text
stringlengths 87
880k
| pmid
stringlengths 1
8
| accession_id
stringlengths 9
10
| license
stringclasses 2
values | last_updated
stringlengths 19
19
| retracted
stringclasses 2
values | citation
stringlengths 22
94
| decoded_as
stringclasses 2
values | journal
stringlengths 3
48
| year
int32 1.95k
2.02k
| doi
stringlengths 3
61
| oa_subset
stringclasses 1
value |
|---|---|---|---|---|---|---|---|---|---|---|---|
==== Front
J Ethnobiol EthnomedJournal of Ethnobiology and Ethnomedicine1746-4269BioMed Central London 1746-4269-1-101627091310.1186/1746-4269-1-10ResearchEnvironmental perception of gatherers of the crab 'caranguejo-uçá' (Ucides cordatus, Decapoda, Brachyura) affecting their collection attitudes Alves Rômulo RN [email protected] Alberto K [email protected]ández Malva IM [email protected] Departamento de Biologia, Universidade Estadual da Paraíba and Programa de Pós-Graduação em Ciências Biológicas (Zoologia), Departamento de Sistemática e Ecologia, Universidade Federal da Paraíba, 58051-900 João Pessoa, PB, Brazil2 Departamento de Sistemática e Ecologia, Universidade Federal da Paraíba, 58051-900 João Pessoa, PB, Brazil, PB, Brasil2005 3 11 2005 1 10 10 26 8 2005 3 11 2005 Copyright © 2005 Alves et al; licensee BioMed Central Ltd.2005Alves et al; licensee BioMed Central Ltd.This is an Open Access article distributed under the terms of the Creative Commons Attribution License (), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
==== Body
Introduction
Mangrove forests are productive ecosystems found along the coastal zones of Brazil, providing several profitable resources such as timber, medicinal products, natural dye, fish, crustaceans, and molluscs. For the littoral dwellers of Northeast Brazil, the Brachyura crabs are a major economic resource. The main species they commercialize are the 'goiamum' (Cardisoma guanhumi), 'siris' (Callinectes spp), and the land crab 'caranguejo-uçá' (Ucides cordatus). The land crab is the most exploited species, and of most relevance for people living in the surrounding mangrove areas in the State of Paraiba [1-3].
U. cordatus lives in an individual burrow ca. 1 m deep, situated under mangrove trees. Adult crabs have few predators, notably the crab-eating racoon (Procyon cancrivorous), monkeys, and hawks [4]. Despite this, a high predation pressure on U. cordatus is exerted by humans who harvest this species for food [5]. In the Northeast Region of Brazil, the exploitation of U. cordatus holds particular socio-economic importance since it involves many local residents, who benefit from both direct and indirect employment [2,3]. Crab gatherers have observed that the natural stock of U. cordatus has decreased alarmingly since 1998, when an unexpected crab mass-mortality event occurred in the mangrove habitats of the Paraiban littoral [3]. The subsequent low crab abundance created social problems in the surroundings of those mangrove areas and seriously affected the economic welfare of poor people who depend upon crab gathering for their livelihoods.
The need for research on the exploitation of mangrove ecosystems and on U. cordatus in particular was emphasized by Maneschy [6], who also suggested the need to study the socio-economics of crab collecting, which has recently experienced an increase in production demand. The life of gatherers is intimately linked to ecological processes and cycles, and their daily involvement with the exploitation of other natural resources will likely help them to develop harvesting strategies for maximizing the crab catch efficiency. An understanding of the ecology of U. cordatus by local gatherers is an important component of the process of exploitation [2,6].
In recent years, researchers have emphasized the importance of traditional knowledge amongst fishermen. They have also emphasized the potential role that traditional fishing practices can play in the development and implementation of sustainable fishery management in the modern world [7-9]. Human communities which rely directly on their natural resources for subsistence, often have a detailed understanding of their local environment [10-12]. The economic, social, and cultural activities of such people often depend upon local environmental goods and services [13].
Ecologists and environmental managers have generally disregarded the possibilities of learning from the traditional human communities [14]. However, a recent acknowledgment of their relevance has led to an intensification of studies on traditional knowledge [14-18]. In Brazil alone, Diegues [19] listed 868 relevant publications on traditional human populations, of which nearly 80% were published over the last 20 years, and mainly in the last decade.
Traditional knowledge may help in the establishment of management plans aimed at the sustainable exploitation of natural resources [2,3]. Nordi [20] observed that the government environmental organization controlling the capture of U. cordatus does not consider local human knowledge of the ecology of the species, a fact that possibly explains the poor effectiveness of the regulations governing the exploitation of this resource. U. cordatus individuals are caught manually or by the use of some tools which allow easier access to them. In most of Brazilian States professional crab gatherers are male [21].
During normal harvest procedures gatherers select crabs with respect to both sex and size. In particular, male crabs are preferred due to their higher flesh yield [22]. Gatherers have developed the ability for distinguishing the sex of crabs as inferred from the track the animal leaves close to burrow openings, as well as from the size of burrow entrances. This perception is important since it directly influences the capture process since large male crabs are preferred due to their higher commercial value. The aim of the present work is to evaluate the ability of gatherers to discriminate the sex and size of crabs, and the importance of this ability on the development of successful harvest strategies. It is also intended here to evaluate the implications of this perception for establishing measures aimed at the conservation and management of U. cordatus.
Methods
Study areas
The research was carried out in the mangrove forests of the Mamanguape river, Northeast Brazil, located between lat 06°43' and 06°51' S, and long 35°07' and 34°54' W. The estuary studied here, is nearly 24 km long from east to west, and 2.5 km wide nearest to the river mouth (Figs 1 and 2).
Figure 1 Map showing the estuary of the River Mamanguape.
Figure 2 Aerial view of River Mamanguape estuary (Photo: João Carlos).
All the area of the influence of the Mamanguape river is within the permanent protected area (APA, in Portuguese) of 'Barra do Rio Mamanguape' of IBAMA (the Brazilian Institute for the Environment and Natural Resources), formed also by the estuaries of the rivers Miriri and Estivas, and covering a total of 14,460 ha. In that estuary there is a large extension of mangrove forest, islands, and islets (the latter formed by sandy-clay-loam banks), and barrier reefs in the rivers mouths together form an elongated sand bar. The whole area is located in the municipalities of Rio Tinto, Marcação, and Lucena, in the State of Paraíba (Figure 2).
The estuary of the APA includes ca. 6,000 ha of a reasonably well preserved mangrove forest, the largest of its kind in the State of Paraíba. The common trees are Rhizophora mangle, Avicennia germinans, A. schaueriana, Laguncularia racemosa and Conocarpus erectus. The tallest R. mangle trees reach 25 m and are up to 60 cm DBH (diameter at breast height); A. germinans trees are taller than 30 m and have up to 65 cm DBH [1].
Despite the fact that the Mamanguape mangrove forest is relatively well preserved, some sites appear to be affected by anthropogenic activities, mainly derived from sugar cane cultivation. Watanabe [23] reported water contamination in one of the estuary tributaries from sugar cane plantations. According to local fishermen, a decrease in fish production has been observed due to an increase in the amount of agrochemicals used in plantations along the river banks.
Procedures
The research was performed between November and October 2002. Before starting the field work, a meeting was held with crab gatherers to inform them about the goals of our study, and research methods, and to propose their participation in our investigations.
In order to respect indigenous intellectual property rights, we followed an informal research protocol: before the survey, we briefly explained the nature of the research and specific objectives to each interviewee [28]. We were precluded from adopting a formal approach using interview consent forms, owing to the poor level of organization among the community of gatherers, and high illiteracy rates [2,3].
We selected ten gatherers typical of the community of gatherers, with each having at least 20 years of experience and being from the middle income range. The perception and discrimination ability of crab gatherers was analysed through direct observations, as a 'non-member participant observer' [24], complemented by open interviews and informal conversations [25-27]. Each gatherer was interviewed individually, there was no time limit for the interviews and they lasted between 1 and 3 hours each. We joined the gatherers (one at a time) in their daily activities. During the course of the study they captured 210 crabs. Morphological measurements were taken using venier callipers, accurate to 1 mm. Crab carapaces were measured with respect to: (a) height, taken dorsoventrally from its central point; (b) length, taken from the sagittal plain on the dorsal surface of the animal; and (c) width, taken transversally from the first pair of pereiopods. The carapace width corresponds to the largest body dimension (Fig. 3). The burrow opening of each caught animal was simultaneously measured with respect to height (the longest axis) and to width (the shortest axis).
Figure 3 Morphometric measurements taken from the carapace of Ucides cordatus (Photo: Guy Nishida).
The gatherer was asked to guess the sex the animal he was going to catch. His perception was then evaluated after a careful examination of the crab external sexual characteristics, following Mota Alves [29]. The possible association between what gatherer expectation and observed sex classification was evaluated using a chi-square test in a 2 × 2 contingency table with Yates correction [30].
Results
The carapaces length ranged from 3.2 cm to 5.8 cm, with a mean of 4.27 cm (± 0.424); the width ranged from 4.1 cm to 6.7 cm with a mean of 5.53 cm (± 0.537); and the height ranged from 2.6 cm to 6.2 cm with a mean of 3.52 cm (± 0.480). The height of the burrow openings ranged from 3.4 cm to 7.3 cm with a mean of 5.72 cm (± 0.694), and the width ranged from 3.2 cm to 7.1 cm with a mean of 5.21 cm (± 0.693). The data for both carapace size and burrow entrance size were normally distributed.
The correlation analysis performed on the burrows opening dimensions and carapace dimensions showed positive and significant correlations (p < 0.05) (Table 1).
Table 1 Pearson's correlation (r values) between the burrow opening dimensions and the Ucides cordatus carapace dimensions (n = 210; * significant at p < 0.05).
Burrow opening dimensions (cm) Carapace dimensions (cm)
Carapace height Carapace length Carapace width
Burrow opening height 0.37* 0.58* 0.59*
Burrow opening width 0.40* 0.60* 0.62*
The variables with the most biological significance with respect to crab behaviour of going into and coming out of the burrows, are the width of the burrow opening and the crab carapace length. The regression analysis performed with these two variables generated the equation: burrow opening width = 1.02 + 0.98 × crab length; and showed that each centimetre increase of the carapace length corresponds to 0.98 cm increase of the burrow opening width (r = 0.60; p < 0.05).
The crab gatherers reported that they are able to distinguish male from female animals through the tracks their limbs leave on soil at the burrow entrance (Fig. 4). The females do not have virtually any hairs in their legs (Fig. 5), and leave thin deep tracks at the burrow entrance, in opposition to the hairy legs of males (Fig. 6), leaving relatively wide and shallow tracks. Figure 7 shows a land-crab gatherer during harvesting activities.
Figure 4 Ucides cordatus specimen entering the burrow (Photo: Guy Nishida).
Figure 5 Legs of Ucides cordatus females virtually without hairs (Photo: Guy Nishida).
Figure 6 Hairy legs of Ucides cordatus males (Photo: Guy Nishida).
Figure 7 Gatherer of the land crab 'caranguejo uçá' (Ucides cordatus) in the mangrove habitat the River Mamanguape estuary.
The data obtained in this study showed a highly successful perception of crab sex amongst gatherers, i.e. 75.2% of the 210 individuals caught were successfully identified with respect to sex, of which 45.2% were males and 30% were females (Table 2). The chi-square test showed a positive and highly significant association (χ2 = 53.27; df = 1; p < 0.01) between the animal sex expected by the crab gatherer, and the sex recorded following examination of the animals. The 24.8% of mis-identifications comprised of 6.7% of the males and 18.1% of the females they caught. Consequently, mis-identifications are nearly three times more likely in the case of female individuals compared to males (Table 2).
Table 2 Percentage success of sex perception of Ucides cordatus crabs by local gatherers.
Sex expected by the gatherer Sex observed Gatherer hits (%) Gatherer misses (%)
Male Female
Male 95 (hit) 38 (miss) 45.2 18.1
Female 14 (miss) 63 (hit) 30.0 6.7
Total 109 (51.90%) 101 (48.10%) 75.2 24.8
Discussion
The proportionality between the crab's carapaces and the burrow openings had already been reported by other authors [31-33]. Their results were confirmed by this present study.
The correlation analysis showed that the crab morphometric values of length and width, as well as burrow opening dimensions were the main variables which had a significant association. The following conclusions can be reached from our results: carapace height was weakly correlated with burrow opening height (r = 0.37) and with burrow opening width (r = 0.40), probably because U. cordatus individuals enter the burrow by flexing their chelipeds frontally to the body, in a way that the longest axis of the burrow opening (its height) corresponds to the carapace height (Figure 3). This lateral movement of young and adult male and female crabs was also observed by Costa [33]. Geraldes and Calventi [34] reported that the 'caranguejo-uçá' enters the burrow laterally, positioning first either the largest or the smallest claw. They also observed that the burrow opening was a little larger (0.34 cm) than the crab carapace length.
The gatherers' observation that the larger the burrow the bigger the crab (as was confirmed by the present study), is important with regard to maximising the harvest obtained by the gatherers.
With respect to the gatherers perception of crab sex, Maneschy [6] also reported that in Pará State, North of Brazil, they distinguish the crabs' sex through their tracks left in front of the burrow openings, but that sometimes mistakes are made. According to crab gatherers of River Mamanguape area, some other factors affect their ability to successfully discriminate the sex of crabs, including the tide dynamics and sediment compaction. Areas washed away by high tides or with hardened sediments, are difficult places to recognize the crabs tracks. Gatherers are also aware of the sex and size of the crabs they target by direct observation of the burrow entrances. Therefore the catches are selective, since large male specimens have a higher commercial value.
Based on the catch selectivity of U. cordatus individuals, where both smaller sized and female individuals are not collected but returned to their habitat, Ivo and Gesteira [22] stated that this crustacean can be sustainably harvested. However, despite this conclusion, depletions of natural stocks of U. cordatus have been reported in the States of Paraíba [2,3], Espírito Santo [35], and Pará [36]. It is therefore important to consider several other factors, besides exploitation through human harvesting, that have contributed to reduce natural populations of U. cordatus, the most important of which is habitat degradation. In Brazil, like in many other tropical countries, mangrove forests have been widely cleared through anthropogenic activities
Another important factor is the precarious social conditions within which the crab gatherers' families survive [2,3], a factor that is associated with the unpredictability of both the capture success of the animal and market demand, and often forces gatherers to harvest crabs in a non-selective approach [37]. All interviewees had a low income of about US$ 100 per month (roughly equal to the minimal wage in Brazil). Crab gathering is the most important means of subsistence for the families of the crab gatherers, despite their supplemental income from small-scale agriculture activity, and the extraction of other natural resources. Crab gatherers are economically marginal groups, extremely poor and largely ignored by other artisanal fishermen. Despite this, the animals they capture are an important component of regional cuisine, as well as being sold commercially in urban centers.
The focus on crab harvesting as providing the economic foundation of these communities may lead them to an economic dependence on buyers, who are therefore in a position to exploit the gatherers. This would mask their perception about their harmful actions on the environment. In the case of common resources, a subject discussed by Burke [38], we agree that '... simply because resource users are not aware of the collective environmental costs of resource use does not make them irrational. It simply means their resource use follows a rationale other than the logic of the commons – possibly the logic of consuming more of a thing for which they have a preference'.
At the present time and despite prohibitions imposed by the government concerning the minimum size of crabs allowed to be caught, small sized and female specimens are also being collected. Therefore, the size-selective harvesting is practised only at sites where large crabs are abundant. This present situation may lead to an over-exploitation of crab populations in mangrove forests of the State of Paraíba, especially at sites presently experiencing strong anthropocentric pressure. A similar situation has been observed in the mangrove ecosystems of the States of Rio de Janeiro and São Paulo, where U. cordatus is considered to be under risk of extinction [39,40].
As observed by Bergallo [39] the decrease in the rate at which the individuals of some species can meet each other, is a clear sign of the reduction in population size. This observation corroborates the need to carry out conservation programs, especially considering that U. cordatus is a valuable economic resource. The strong (and perhaps unsustainable) predation pressure exerted on this species through human exploitation can be interpreted as a consequence of (1) the lack of an efficient environmental management policy, (2) the lack of studies and actions regulating the commercialization of its catch, and (3) the lack of alternative socio-economic proposals aimed at improving the welfare of local human communities directly involved with the exploitation of U. cordatus.
The implementation of a successful management strategy fundamentally requires the involvement of the main stakeholders, who must be made aware of the need for the conservation of the natural resource as a guarantee for its sustainable exploitation [41]. In this sense, the gatherers are critical stakeholders in the process of establishing management plans, which would ensure the sustainable exploitation of the resource. In Brazil, management of coastal fisheries is usually imposed on local communities by the federal government, without taking the either traditional knowledge or the reality of their socioeconomic conditions into consideration – with negative consequences for management plans.
The knowledge derived from traditional fishermen may contribute to strategic ecological managements [42], particularly in research areas with scarce or non-existent data [43,44]. Ethnoecological studies may also help in promoting dialogue and cooperation between fishers and scientists. The scientific literature illustrates the significance and value of traditional knowledge in Brazil. Alves and Nishida [2], for example, reported on the influence of tide variations on the ecdysis of the mangrove crab Ucides cordatus in natural environments, based on information obtained from gatherers. Marques [45] also reported the gathering of relevant local information on the diet of an estuarine fish (Arius herzbergii, Ariidae).
Cultural values and perceptions arising from more experienced elder people need to be taken into account before any decision about their livelihoods can be considered by the relevant governmental authorities. Their contributions are essentially concerned with the sustainable management of local resource, and such contributions unfortunately have not yet been given satisfactory representation in the decision making process.
Conclusion
The results obtained here make evident that gatherers of Ucides cordatus have developed an intimate knowledge of the natural history of this species. Thus, they have developed skills that led them to become efficient harvesters. The unique nature of this local knowledge demonstrates the need for considering these factors in the implementation of management plans of coastal mangrove ecosystems. Gatherers' knowledge can provide a useful basis for understanding local crab stocks and their population dynamics. This kind of information may be used for establishing extractive reserves as well as for delimiting the harvest season and for establishing protected areas where animal species reproduce, aiming to maintain their natural stocks.
In the State of Paraíba the law prohibits that females of any size and males smaller than 4.5 cm of carapace length be captured. We believe that the re-examination of the existing laws of crab harvesting and other crustaceans are urgently needed for the preservation of natural stocks of these animals. The gatherer's environmental perception and its consequent influence on capture efficiency, are the main factors to be considered when making decisions about sustainable exploitation of mangrove ecosystems resources. Such decisions would only be effectively applied if the local knowledge held by crab gatherers as well as their socioeconomic conditions do not continue to be ignored by governmental authorities.
Acknowledgements
We are grateful for the cooperation of gatherers of 'caranguejo-uçá' in Paraíba State, since our research was founded on their knowledge. We are also grateful to CAPES (Coordenadoria de Aperfeiçoamento de Pessoal de Ensino Superior), WWF and USAID for financial supports, and to Dr Breno Grisi for translating this work into English and Toby Gardner for English comments and suggestions.
==== Refs
Palludo D Klonowski VS Barra de Mamanguape – PB: Estudo do impacto do uso de madeira de manguezal pela população extrativista e da possibilidade de reflorestamento e manejo dos recursos madeireiros Série Cadernos da Reserva da Biosfera 1999 16 Mata Atlântica, MAB, UNESCO 7 54
Alves RRN Nishida AK A ecdise do caranguejo-uçá, Ucides cordatus (Crustacea, Decapoda, Brachyura) na visão dos caranguejeiros Interciencia 2002 27 110 117
Alves RRN Nishida AK Aspectos socioeconômicos e formas de percepção ambiental dos catadores de caranguejo-uçá Ucides cordatus cordatus (L. 1763) (Decapoda, Brachyura) do estuário do rio Mamanguape Interciencia 2003 28 36 43
Koch V Epibenthic production and energy flow in the Caeté mangrove estuary, North Brazil PhD thesis 1999 University of Bremen
Diele K Life history and population structure of the exploited mangrove crab Ucides cordatus (L.) (Decapoda: Brachyura) in the Caeté estuary Zentrum fur Marine tropenökologie 2000 Center for Tropical Marine Ecology. Bremen
Maneschy MC A mulher está se afastando da pesca? Continuidade e mudança do papel da mulher na manutenção doméstica entre famílias de pescadores no litoral do Pará Boletim do Museu Paraense Emílio Goeldi, Antropologia 1995 11 145 166
Diegues AC Sea tenure, traditional knowledge and managementamong brazilian artisanal fishermen 2002 NUPAUB. Universidade de São Paulo
Ruddle K Systems of knowledge: Dialogue, relationships and process Environment, Development and Sustainability 2000 2 277 304 10.1023/A:1011470209408
Cordell JC Remaking the waters: the signficance of sea tenure-based protected areas Third Conference on Property rights, economics and environment 2000 International Center for research on Environmental issues, Aix-en-Provence, France
Berlin B Ethnobiological Classification Principles of Categorization of Plants and Animals in Tradicional Societes 1992 Princeton University Press
Gadgil M Berkes F Folke C Indigenous knowledge for biodiversity conservation Ambio 1993 22 151 156
Berkes F Sacred Ecology – Traditional Ecological Knowledge and Resource Management 1999 Taylor & Francis, Philadelphia
Diegues AC Os pescadores artesanais e a questão ambiental Povos e Mares 1995 NUPAUB, São Paulo, Brasil
Poizat G Baran E Fishermen's knowledge as background information in tropical fish ecology: a quantitative comparison with fish sampling results Environmental Biology of fishes 1997 50 435 449 10.1023/A:1007317423165
Johannes RE Johannes RE Fishing and traditional knowledge Traditional Ecological Knowledge: a Collection of Essays 1989 IUCN, Gland 39 42
Pajaro MG Vincent ACJ Seahorse conservation in Central Philippines: a community-based approach Sea Wind 1996 10 7 12
Pajaro MG Vincent ACJ Buhat DY Perante NC The role of seahorse fishers in conservation and management Proceedings of the 1st International Symposium in Marine Conservation Hong Kong 1997 118 126
Skaptadóttir UD Women coping with change in an iceland fishing community: a case study Women's Studies International Forum 2000 23 311 321 10.1016/S0277-5395(00)00089-3
Diegues CA Arruda RSV Silva VCF Figols FAB Andrade D Biodiversidade e comunidades tradicionais no Brasil: Os saberes tradicionais e a Biodiversidade no Brasil 1999 NUPAUB-USP, PROBIO-MMA, CNPq, São Paulo
Nordi N A captura do caranguejo-uçá (Ucides cordatus) durante o evento reprodutivo da espécie: o ponto de vista dos caranguejeiros Revista Nordestina de Biologia 1994 9 41 47
Vergara Filho WL Pereira Filho O As mulheres do caranguejo Ecologia e Desenvolvimento 1995 5 34 36
Ivo CTC Gesteira TCV Sinopse das observações sobre a bioecologia e pesca do caranguejo-uçá, Ucides cordatus cordatus (Linnaeus, 1763), capturado em estuários de sua área de ocorrência no Brasil Boletim Técnico Científico CEPENE 1999 7 9 51
Watanabe T Oliveira RB Sassi R Melo GN Moura GF Gadelha CL Machado VMN Evidences of contamination caused by sugar-cane monoculture and associated industrial activities in water bodies of the State of Paraíba, Northeast Brazil Acta Limnologica Brasiliensia 1994 5 85 101
Sttebins RA Fitting in: the researcher as learner and participant Quality and Quantity 1987 21 103 108
Huntington HP Using traditional ecological knowledge in science: methods and applications Ecological Applications 2000 10 1270 1274
Mello LG Antropologia cultural Editora Vozes, Rio de Janeiro 1995
Chizzoti A Pesquisa em ciências humanas e sociais 2000 Cortez editora, São Paulo
Silvano R Begossi A Local knowledge on a cosmopolitan fish Ethnoecology of Pomatomus saltatrix (Pomatomidae) in Brazil and Australia Fisheries Research 2005 71 43 59 10.1016/j.fishres.2004.07.007
Mota Alves MI Sobre a reprodução do caranguejo-uçá, Ucides cordatus (Linnaeus), em mangues do Estado do Ceará (Brasil) Arquivos de Ciências do Mar 1975 15 85 91
Zar JH Biostatistical Analysis 1984 Prentice-Hall, Englewood Cliffs. NJ, USA
Alcântara-Filho P Contribuição ao estudo da biologia e ecologia do caranguejo-uçá, Ucides cordatus (L. 1763) (Crustacea, Decapoda, Brachyura), no manguezal do rio Ceará (Brasil) Arquivos de Ciências do Mar 1978 18 1 41
Blankensteyn A Cunha Filho D Freire AS Distribuição, estoques pesqueiros e conteúdo protéico do caranguejo do mangue Ucides cordatus (L. 1763) (Brachyura Ocypodidae) nos manguezais da Baía das Laranjeiras e adjacências, Paraná, Brasil Arquivos de Biologia e Tecnologia 1997 40 331 349
Costa RS Fisioecologia do caranguejo-uçá, Ucides cordatus (Linnaeus, 1763) -Crustáceo, Decápode – do Nordeste brasileiro PhD Thesis 1972 , USP-University of São Paulo, São Paulo State, Brazil
Geraldes MG Calventi IB Estudios experimentales para el mantenimiento en cautiverio del cangrejo Ucides cordatus Ciencia Interamericana 1983 23 41 53
Nunes AGA Os argonautas do mangue: uma etnografia visual dos caranguejeiros do Município de Vitória – ES MSc Dissertation 1998 UNICAMP-Campinas University, Campinas, São Paulo State, Brazil
Gondim M Araújo FB Redução dos tamanhos dos caranguejos (Ucides cordatus L.) capturados nos manguezais de Maracanã, zona do Salgado Paraense [abstract] Anais do 3° Congresso de Ecologia do Brasil 1996 260
Nordi N Os catadores de caranguejo-uçá (Ucides cordatus) da região de Várzea Nova (PB): Uma abordagem ecológica e social PhD Thesis 1992 UFSCar-Federal University of São Carlos, São Carlos, São Paulo State, Brazil
Burke BE Hardin Revisited: A Critical Look at Perception and the Logic of the Commons Human Ecology 29 449 476 10.1023/A:1013145905257
Bergallo HG Rocha CFD Van Sluys M Alves MAS A fauna ameaçada de extinção do Estado do Rio de Janeiro 2000 EdUERJ, Rio de Janeiro
Governo do Estado de São Paulo Fauna ameaçada no Estado de São Paulo 1998 Série PROBIO/SP. Secretaria do Meio Ambiente
Rodrigues AMT Branco EJ Saccardo AS Blankensteyn A A explotação do caranguejo Ucides cordatus (Decapoda: Ocypodidae) e o processo de gestão participativa para normatização da atividade na região Sudeste- Sul do Brasil Boletim do Instituto de Pesca 2000 26 63 78
Seixas C Begossi A Ethnozoology of caiçaras from Aventureiro, Ilha Grande Journal of Ethnobiology 2001 21 107 135
Ruddle K A Guide to the Literature on Traditional Community-Based Fishery Management in the Asia-Pacific Tropics 1994 Rome: FAO Fisheries Circular Number 869, FIPP/C869
Johannes RE The case for data-less marine resource management: examples from tropical nearshore finfisheries Trend in Ecology and Evolution 1998 13 243 246 10.1016/S0169-5347(98)01384-6
Marques JGW Aspectos ecológicos na etnoictiologia dos pescadores do complexo estuarino-lagunar de Mundaú-Manguaba, Alagoas PhD Thesis 1991 Universidade Estadual de Campinas, São Paulo
|
16270913
|
PMC1289291
|
CC BY
|
2021-01-04 16:40:12
|
no
|
J Ethnobiol Ethnomed. 2005 Nov 3; 1:10
|
utf-8
|
J Ethnobiol Ethnomed
| 2,005 |
10.1186/1746-4269-1-10
|
oa_comm
|
==== Front
Reprod Biol EndocrinolReproductive biology and endocrinology : RB&E1477-7827BioMed Central London 1477-7827-3-561623617910.1186/1477-7827-3-56ReviewHuman trophoblast function during the implantation process Staun-Ram Elsebeth [email protected] Eliezer [email protected] Laboratory for Research in Reproductive Sciences, Department of Obstetrics and Gynecology, Ha'Emek Medical Center, 18101, Afula, Israel2 Rappaport Faculty of Medicine, Technion-Israel Institute of Technology, Haifa, Israel2005 20 10 2005 3 56 56 1 8 2005 20 10 2005 Copyright © 2005 Staun-Ram and Shalev; licensee BioMed Central Ltd.2005Staun-Ram and Shalev; licensee BioMed Central Ltd.This is an Open Access article distributed under the terms of the Creative Commons Attribution License (), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
The implantation process involves complex and synchronized molecular and cellular events between the uterus and the implanting embryo. These events are regulated by paracrine and autocrine factors. Trophoblast invasion and migration through the uterine wall is mediated by molecular and cellular interactions, controlled by the trophoblast and the maternal microenvironment. This review is focused on the molecular constituents of the human trophoblast, their actions and interactions, including interrelations with the uterine endometrium.
==== Body
1. Introduction
Successful implantation depends on synchronization between the developmental stages of the embryo itself and the complex series of molecular and cellular events that are induced in the pregnant uterus by paracrine and autocrine regulators [1]. Embryos prepare for implantation during their cleavage stage. Successive cleavages must produce sufficient cells by the time the blastocyst is formed to permit the full formation of inner cell mass and trophectodermal cells. The latter are the origin of the cytotrophoblastic cells which ensue to become either the villous cytotrophoblastic cells which will proliferate and differentiate by fusion to form the syncytiotrophoblast, or they will stream out of the syncytiotrophoblast to form mononuclear multilayered invasive extravillous cytotrophoblastic cells [2]. The process of implantation begins six to seven days following fertilization [3] and consists basically of three stages [4]. Apposition is the first stage denoting the initial, still unstable, adhesion of the blastocyst to the uterine wall. At this stage the pinopodes, which are micro protrusions from the apical uterine epithelium surface, inter-digitate with microvilli on the apical syncytiotrophoblast surface of the blastocyst[5] (Figure 1). The stable adhesion, which is the next step, reveals an increased physical contact between the blastocyst and the uterine epithelium, while the embryonic pole is oriented toward the epithelium. The last stage is the invasion process, which starts with the penetration of the syncytiotrophoblasts through the uterine epithelium and followed by infiltration of the mononuclear cytotrophoblasts invading the entire endometrium, the inner third of the myometrium and the uterine vasculature. Villous cytotrophoblast cells at the tip of the anchoring villi proliferate outwards from the underlying basement membrane to form cell columns from which cells migrate into the decidua (interstitial trophoblast cells) and invade the maternal spiral arteries and designate the endovascular trophoblast cells. The latter allows the trophoblasts to be in direct contact with the maternal blood establishing the uteroplacental circulation [4]. Trophoblast invasion and migration is probably controlled by components of the trophoblast itself and maternal microenvironment, through molecular and cellular interaction (Figure 2).
The human implantation process is unique thus no other mammals can provide a true model [2]. Ethical restrictions and limited availability of human placental tissue limits the possibilities of human implantation studies. Consequently most knowledge comes from in vitro experiments using cultured human trophoblasts or cell-lines, mainly obtained from choriocarcinoma. Although the differences in reproductive physiology between species limit the relevance, knowledge on the possible role of various factors in the implantation process comes from knockout-studies in mice. Primates provide a more physiological relevant model, therefore in vivo and in vitro studies in baboons have contributed to our knowledge [6]. All in all, human in vitro and mammal in vitro and in vivo studies have provided important information, which helped in understand at least part from the complex process of implantation. This review focuses on the molecular constituents of the human trophoblast, their actions and interactions, which seem to be crucial for successful implantation.
2. Cell adhesion molecules
2.1 Integrins
During the mid-luteal stage of the menstrual cycle, apparently in response to progesterone, epithelial apical membrane projections named pinopodes appear in the uterus [5]. The biological nature of the pinopodes has not been determined, but their short appearance during the implantation window, interconnecting with microvilli on the apical syncytiotrophoblast surface of the blastocyst, suggests an involvement in embryo implantation [7]. Adherence of the trophectoderm cells of the blastocyst to the pinopode membrane has been suggested to occur through adhesive molecules such as E-cadherin, present in the pinopode epithelial membrane [5]. In vitro studies showed that although no direct contact between trophectoderm and pinopode is seen, blastocysts tends to attach to areas of cultured endometrial epithelium containing pinopodes [5].
Integrins are heterodimeric membrane glycoproteins, composed of α and β subunits, capable of binding to various extracellular matrix (ECM) components and cell adhesion molecules, thereby influencing and mediating adhesion, migration, invasion, cytoskeleton reorganization and cellular signaling [8]. Endometrial integrins are hormone-dependent and vary throughout the menstrual cycle [9]. The integrin repertoire of the endometrium may play an important role for obtaining a successful implantation. According to a timed expression correlating with embryo attachment, the αVβ3 and the α4β1 integrins are considered markers of uterine receptivity [10-12]. αVβ3 has been shown to be highly expressed at the time of embryo attachment, and aberrant expression of αVβ3 is associated with infertility [13,14]. Women with recurrent miscarriages were found to have a lower concentration of α4β1 and α5β1 integrins in the stroma during the implantation window, than women with unexplained infertility [15]. The relevance of integrins alteration, in glandular epithelium or stroma, but not in luminal epithelium is not well understood, since they are not likely to be involved in the early parts of implantation [16]. The trophectoderm also express several integrins, α3, α5, β1, β3, β4 and β5, supposed to be implicated in blastocyst attachment to the endothelial surface [17-19]. Trophoblasts modulate their integrin repertoire during invasion of the stroma and during differentiation, varying in invasive and non-invasive cells [2,16]. Several factors known to be involved in the implantation process may act through mediation of integrins. For example, Insulin-like growth factor (IGF)-I-induced migration of extravillous trophoblast cells (EVT) probably involves internalization of the α5β1-integrin on EVT [19], as well as influencing the αVβ3-integrin pathway [18]. In female mice lacking a functional integrin β1 gene, embryos develop normally to the blastocyst stage but fail to implant properly and die [20]. No other integrins were found in knockout studies to be involved in implantation defects [21], but inactivation of αVβ3 by the disintegrin echistatin significantly reduced implantation sites in mice [22]. Apparently several integrins may have a redundant role in implantation [23].
2.2 MUC1
Uterine cell-surface glycoproteins such as MUC1 are thought to provide a barrier to trophoblast invasiveness, by controlling accessibility of integrin receptors to their ligands [8]. MUC1 increases from the proliferative to the secretory phase in endometrial tissue and then decreases in the late secretory phase [24]. Progesterone combined with estrogen up-regulates MUC1 at the receptive endometrium [24]. Blastocysts were shown to deplete MUC1 locally from the implantation site, whereas during the apposition phase, when the blastocyst apparently must be stationary in position, MUC1 was up-regulated [25]. MUC1 may thereby serve as an anti-adhesive molecule that hinders blastocyst attachment to the uterine epithelium until 3 days after entrance to the uterus [26].
3. Extracellular matrix proteins
The trophoblastic cells are confronted with various matrix proteins and basement membranes, when penetrating the uterine wall. These ECM components, including collagens (Col), fibronectin (FN), laminin (LN), vitronectin (VN), trophin and tastin, influence cell functions by binding to integrins, thereby effecting adhesion, migration, differentiation and spreading [2]. Trophin and Tastin for example can be found in endometrial epithelium as well as in trophoblasts, and may be involved in blastocyst attachment by forming a cell-adhesion molecular complex [26].
The components of the ECM are changed during the menstrual cycle [27]. The changes associated with decidualization, during the mid-secretory phase, are especially relevant for implantation. This includes an increase in hyaluronan, a decrease in collagen VI and a slower production of collagen III and I. During decidualization, stromal fibroblasts differentiate into decidual cells, and a basement membrane, containing laminin, entactin, collagen IV and heparan sulphate proteoglycan, appears around each cell [28].
ECM components affect the behavior and function of trophoblastic cells by affecting matrix metalloproteases (MMPs) and their tissue inhibitors (TIMPs) as shown by Xu et al, 2001 [29]. Trophoblasts, grown in the presence of VN, LN or FN, secreted more MMP-9 than in the presence of Col I, or IV, whereas MMP-2, TIMP-2 and MMP-14 were not affected. TIMP-3 on the contrary was inhibited by the presence of VN. Trophoblast adhesiveness was highest in the presence of Col I and IV compared with the other matrix proteins. Since cytotrophoblast cells also produce LN, FN and VN during the first trimester, these matrix proteins may be part of an autocrine mechanism, regulating MMP expression and cell invasiveness [29].
4. Growth factors
4.1 Epidermal Growth Factor
Epidermal Growth Factor (EGF) is expressed both in decidual and trophoblastic cells [30] and affects implantation in several ways. EGF induces trophoblast invasion [31,32], trophoblast differentiation [33,34] and trophoblast proliferation [35], and is therefore regarded a major regulator of the implantation process. EGF has been shown to increase MMP-2 and MMP-9 activity in trophoblastic cells [32,36,37], and also urokinase-type plasminogen activator (uPA) and plasminogen activator inhibitor-1 (PAI-1) activity [36] in trophoblastic cells, thereby inducing cell invasion. EGF stimulates secretion of human chorionic gonadotrophin (hCG) and human placental lactogen (hPL) from trophoblastic cells [38,33], and has been shown to induce α2 integrin expression in a choriocarcinoma cell-line, an effect found to be related to increased invasiveness [38].
4.2 Heparin-binding EGF-like growth factor
Heparin-binding EGF-like growth factor (HB-EGF) shares a common receptor with EGF and transforming growth factor (TGF)-α. HB-EGF is expressed in stromal and epithelial cells of the uterus and is thought to regulate endometrial proliferation, secretion and decidualization [21]. The expression of HB-EGF is maximal in the mid-secretory phase in uterine epithelial cells; therefore HB-EGF may also be involved in the regulation of blastocyst implantation [21].
4.3 Transforming growth factor β
TGF-β is expressed both in endometrial and trophoblastic cells [2]. TGF-β was shown to inhibit trophoblast proliferation and invasion apparently by stimulating TIMP secretion and decreasing MMP activation through down-regulation of plasminogen activators [39]. In another study TGF-β was found to inhibit trophoblast invasion by reducing MMP-9 and uPA secretion, but did not affect TIMP levels or cell proliferation [40]. Elevated TGF-β activity has been reported in the plasma of pre-eclamptic mothers [41] and may be implicated in the impaired implantation associated with pre-eclampsia [42].
4.4 Insulin-like growth factor binding protein-1
Insulin-like growth factor binding protein-1 (IGFBP-1) is the main secretory product of the decidualized endometrium. IGFBP-1 modulates the metabolic effect of insulin-growth factor (IGF)-I and IGF-II and has been shown to increase the gelatinolytic activity of trophopblasts [43] and trophoblast invasiveness mainly by increasing cell migration [44,45].
5. Cytokines
5.1 Leukaemia inhibitory factor
Leukaemia inhibitory factor (LIF) secreted from the uterus is regarded an important factor in embryo implantation. Female mice lacking a functional LIF gene are fertile, but their blastocysts fail to implant, even though they are viable and can implant when transferred to wild-type recipients [46]. Maximal expression of LIF in endometrial epithelium is during the implantation window [21]. LIF and its receptor are also expressed in pre-implantation embryos [47] and in cytotrophoblasts [48]. LIF inhibits gelatinase activity in cytotrophoblasts, thereby effecting cell invasiveness [49]. LIF may also be involved in immune tolerance through regulation of HLA-G, a class I MHC molecule specifically expressed by invasive cytotrophoblast cells [50].
5.2 Interleukin-1
Interleukin-1 (IL-1) is present at the feto-maternal interphase; trophoblastic cells and decidualized stromal cells produce IL-1, and the IL-1 receptor is present in endometrial epithelial cells as well as in trophoblasts [2]. IL-1 may be one of the first signals of the blastocyst acting upon the endometrium, since in vitro IL-1 increases endometrial secretion of prostaglandin E2, LIF and of integrin β3 subunit expression [51]. In mice IL-1 receptor antagonist given prior to implantation significantly reduces the number of implanted embryos, indicating a role for IL-1 in embryo implantation [52]. IL-1 can stimulate MMP-9 activity in trophoblasts [53] and expression in endometrial stroma cells [54], thereby inducing trophoblast invasion Among other cytokines also present at the maternal-feto interface are IL-6, which stimulates MMP-2 and MMP-9 activity [55] whereas IL-10 down-regulates MMP-9 and trophoblast invasion [56].
6. Hormones
6.1 Human Chorionic Gonadotropin
Syncytiotrophoblastic cells secrete hormones including progesterone and human chorionic gonadotropin (hCG), which play important roles in the implantation process. hCG affects several processes during pregnancy, besides the well-known maintenance of the corpus luteum, including cell growth and differentiation. The trophoblasts themselves express a truncated and inactive hCG receptor until ninth week of gestation, then switching to the full length receptor, allowing hCG autocrinic regulation of various functions including cell differentiation in the trophoblasts [57]. In trophoblastic neoplasms these receptors have been found to be over-expressed [2]. Positive correlation between hCG level and trophoblast invasion has been shown in ectopic pregnancies [58,59], indicating a possible stimulatory effect on trophoblast invasiveness. hCG was found in vitro to increase invasiveness of a trophoblastic choriocarcinoma cell-line [60] however in vitro studies of primary trophoblast showed the opposite effect [61,62]. Therefore, the influence of hCG on trophoblast in vivo invasion remains unclear. hCG stimulates the cAMP pathway, and forskolin, which directly activates adenylate cyclase and increases cAMP, also stimulates trophoblast invasiveness, both in a trophoblast choriocarcinoma cell-line and in primary trophoblastic cells [32]. This is in agreement with the finding that hCG increases MMP-9, an important key-factor in trophoblast invasion [57]. Lately hCG was shown in vitro to stimulate trophoblast migration through an IGF-II effect [63].
hCG influences several uterine factors, for example increases the expression of COX-2 gene, an enzyme involved in prostaglandin biosynthesis [21], LIF and vascular endothelial growth factor (VEGF) [57], suggesting a role in endometrial vascularization. In the baboon hCG was shown to cause physiological effects on the uterine endometrium in vivo, including an increase in glycodelin expression and secretion by the glandular epithelium, and differentiation of subepithelial stromal fibroblasts characterized by expression of the alpha smooth muscle actin, associated with the initiation of decidualization [64,65]. This suggests that the primate blastocyst signal modulates the uterine environment prior to implantation [64].
Hyperglycosylated hCG (HhCG), also called invasive trophoblast antigen (ITA), is an hCG variant with extra-large O-linked oligosaccharides. HhCG is the predominant form of hCG in early pregnancy and in choriocarcinoma, where aggressive trophoblast invasion takes place [66]. It is produced by poorly differentiated or invasive trophoblast cells and decreases as pregnancy advances [67]. The fact that HhCG is dominant around the time of implantation and in the 3 weeks that follow [68] makes it an interesting molecule that deserves further exploration. Low maternal mid-trimester urine HhCG has been found to predict preeclampsia, which is associated with poor trophoblast invasion [69].
6.2 Progesterone
Progesterone has a crucial role in preparing the uterus for the developing embryo and for obtaining a successful pregnancy. Stromal cells differentiate into decidual cells in respond to progesterone during the decidualization process, characterized by morphological changes and secretion of prolactin [70]. Progesterone is also required for maintenance of the pregnancy by stimulating and maintaining uterine functions, necessary for early embryonic development, implantation, placentation and fetal development. Hormones such as hCG, produced by the trophoblast, maintain this progesterone production in early pregnancy [71]. Progesterone was suggested to affect trophoblast invasiveness through the down-regulation of MMP-9 [72]. Progesterone is known to restrain endometrial breakdown by inhibiting MMPs. This inhibitory effect implies that progesterone impedes the invasion of trophoblast cells into the endometrial tissue. In a recent report progesterone was found to decrease invasion and gelatinase expression in early first trimester trophoblast cells and to increase cell invasion and MMP-2 expression in late trophoblast cells [73]. A differential progesterone receptor (PR) profile was documented with the dominance of PRB in the early trophoblast and dominance of PRA in the late trophoblast [73]. This differential PR profile is compatible with the inverse temporal effect of the hormone on the trophoblast cells. In mice a similar dual role of progesterone in embryo implantation has been reported, when progesterone promoted attachment and invasion of primary trophoblasts, mainly through MMP-2 stimulation, but inhibited invasion of secondary trophoblasts [74].
7. Inflammatory factors
7.1 Cortico-releasing hormone
The implanting embryo suppresses the maternal immune process, thereby preventing rejection. Cortico-releasing hormone (CRH) is produced by both trophoblasts and placental deciduas [75]. In mice, implantation can be highly reduced by anti-CRH antibody, indicating a role in embryo implantation [76]. Fas and its ligand FasL, belongs to the tumor necrosis family (TNF). Fas and Fas-FasL interaction, plays an important role in the regulation of immune tolerance, mainly by inducing apoptosis in cells carrying Fas, including T and B lymphocytes [75]. FasL is expressed on cytotrophoblastc as well as on maternal decidual cells of the placenta [77]. CRH was found in vitro to stimulate FasL expression in extravillous human trophoblasts, thereby enabling them to induce apoptosis of the surrounding activated T lymphocytes [75]. Rats treated with a CRH receptor 1 (CRHR1) antagonist had diminished FasL endometrial expression and reduced number of implantation sites, suggesting that locally produced CRH promotes implantation and maintenance of early pregnancy mainly by killing activated T cells [75]. The CRH-Fas-FasL system is not the sole maternal immunetolerance mechanism preventing embryo rejection. Mice with an inactivating mutation of Fasl gene or CRH and CRHR1-deficient mice [78-80] can be fertile.
7.2 Tumor necrosis factor-α
The pleiotropic cytokine TNF-α and its two receptors are present in the endometrium, as well as in placental trophoblasts [81]. TNF-α was shown in vitro to increase uPA secretion from cytotrophoblasts, thereby properly enhances the degradation of fibronectin during trophoblast penetration of the endometrial ECM [84]. Although, TNF-α did not affect MMP-9 concentration, the up-regulation of uPA increases activation of MMP-9 through the plasminogen activation system, thereby enhancing trophoblast invasiveness [84]. This is consistent with another report describing TNF-α stimulation of MMP-9 gelatinolytic activity, without affecting MMP-9 concentration [53]. At the same time, TNF-α decreases MMP-2 concentration and activity as well as hCG secretion [53,84]. It is therefore suggested, that high TNF-α levels presented during inflammatory responses [85], could be responsible in pathologic processes such as pregnancy loss, preterm delivery and preeclampsia, for abnormal trophoblast endocrine function [84,86]. In a recent study, despite increased MMP-9 expression, TNF-α was found to inhibit in vitro trophoblast migration and invasion. However, the plasminogen activator inhibitor-1 (PAI-1) that blocks the plasminogen activator system, was found to be increased [86]. Thus, TNF-α seems to exert diverse in vitro effects, depending on individual trophoblast preparation, type of cell-line and cytokine concentration. This affects the value of conclusions which can be derived from in vitro studies.
A link between elevated PAI-1 levels detected in plasma and syncytium of preeclamptic women, and elevated TNF-α found in preeclamptic sera, villi and deciduas has been suggested [86]. The elevated TNF-α level could result from local hypoxic conditions developing under reduced maternal and fetal vascular perfusion [87].
7.3 Prostaglandins
Prostaglandins are synthesized from arachidonic acid by phospholipase A2, followed by cyclooxygenase (COX). Prostaglandin E2 (PGE2) is essential for mammalian female reproduction. PGE2 is involved in regulation of decidualization of endometrial stomal cells [88], apparently through stimulation of IL-11, since blocking of IL-11 in PGE2 treated cells reduces decidualization and COX inhibitor reduces IL-11 secretion from these cells [89]. In rats, expression of the PGE receptor EP2 is highly detected at implantation sites in luminal epithelium, peaking on day 6 of pregnancy [90]. Phospholipase A2 and PGE2 receptor were described to be up-regulated in human endometrium during the window of implantation [91]. Experiments in mice have shown that prostaglandins are essential for the correct timing of implantation [92]. Reduced levels of COX-2 or COX-2 ligands cause deferred implantation and reduced litter size, and prostaglandin treatment resumes on-time implantation [92]. COX-1, COX-2 and prostaglandin E synthase (PGES), which catalyses COX products to PGE2, are highly expressed in mice at the blastocyst stage [93].
8. Extracellular degrading matrix proteinases
8.1 MMPs
The matrix metalloproteinases (MMPs) are a family of zinc-containing endopeptidases capable of degrading all components of the ECM, both interstitial matrix and basement membrane. MMPs are thought to play an important role in tumor progression and metastasis [94]. Twenty-six mammalian and twenty-two human MMPs are known so far [28,95]. The vertebrate MMPs each have distinct but often overlapping substrate specificities. Human MMPs fall into five classes according to primary structure and substrate specificity: collagenases, gelatinases, stromelysins, membrane-type and nonclassified MMPs [96]. MMPs regulate cell behavior in numerous ways, including cell-matrix and cell-cell interactions and the release, activation or inactivation of autocrine or paracrine signaling molecules and cell surface receptors. ECM-degradation permits cellular invasion to take place and influence processes such as cell shape, movement, cytoskeleton machinery and matrix-derived signals [95]. MMPs are regulated at several levels, including transcriptional, secretion, activation, inhibition and degradation. Transcriptional regulation is cell, tissue and MMP-specific and includes several cytokines and growth factors [97]. MMPs are produced as proenzymes, requiring removal of the propeptide domain for activation. The extracellular activation of most MMPs can be initiated by activated MMPs or by several serine proteinases, including uPA, plasminogen, thrombin and elastase. MMP-2 is activated at the cell surface through a multi-step pathway involving membrane type (MT) – MMPs and TIMP-2 [98]. MMPs are inhibited by α2-macroglobulin, in tissue fluids, and in tissue by TIMPs, which bind to MMPs in a 1:1 stoichiometric fashion [99].
In vitro studies suggest that successful implantation and placentation result from the balance between secretion of MMPs from the trophoblast and their inhibition by TIMPs [100,101]. The gelatinases, MMP-2 and -9, degrade Collagen IV, the main component of the basement membrane, and are therefore regarded as key enzymes in the implantation process, enabling the invasion of the trophoblast cells through the decidua and into the maternal vasculature. Several in vitro studies have found the gelatinases to be required for trophoblast invasion [102-105,32].
Gelatinases, mainly MMP-2, are secreted from the embryo already at the blastocyst stage [106-108]. Lately we and others have shown, that MMP-2 and MMP-9 have a differential expression throughout the first trimester, with MMP-2 being the main gelatinase in early first trimester (6–8 w) and MMP-9 being dominant in late first trimester (9–12 w) [32,104]. MMPs may have other important actions in the implantation process, besides ECM degradation, including regulation of bioactivity of growth factors, cytokines and angiogenic factors [27]. This includes MMP-2 or -9 activation of TGFβ [109] or release of IGF by degradation of IGFBPs [110]. Another role may be modulation of angiogenic factors such as endothelin-1, a vasoconstrictor [111], or angiostatin, an angiogenic inhibitor [112]. Interestingly, knockout mice, deficient in MMP-2 or MMP-9, are fertile and only mild effects have been reported: MMP-2 deficient mice have a subtle delay in their growth [113] and MMP-9 deficient mice show a decreased litter size and an increase in percentage of infertile mice [114].
MT1-MMP deficient mice die before puberty; therefore no conclusions can be made on reproductive capacity [115]. In MMP-7 null-mice both MMP-3 and MMP-10 were up-regulated in the uterus, whereas in MMP-3 deficient mice MMP-7 and MMP-11 were up-regulated, thereby indicating the presence of a compensation mechanism [116].
The MMP gene promoter contains several cis-regulatory elements, often acting synergistically, with varying importance and effect, depending upon cell-type and inducer. AP-1 sites give several MMP genes in various cell types the ability to be induced by phorbol esthers and act in some cases synergistically with adjacent Ets-binding sites [117,2,118]. AP-1 was found to be necessary, but not sufficient for transactivation of the MMP-9 gene in human trophoblasts, and antisense against Jun and Fos transcription factors, belonging to the AP-1 complex, was found to inhibit MMP-9 gelatinolytic activity in trophoblasts [118]. The importance of the Ets site was shown by the study of knockout mice for ets2 transcription factor, which resulted in deficient MMP-9 expression and early embryonic lethality [119]. Lately, it has been shown that EGF induction of MMP-9 as well as TIMP-1 in an extravillous trophoblastic cell-line is through the MAPK and PI3K pathways [37].
Recently ADAM (a disintegrin and metalloproteinase, adamalysin)-19 has been detected during early pregnancy in the endometrium and the placenta of the rhesus monkey, and may therefore also be involved in trophoblast invasion and degradation of the ECM [120].
8.2 TIMPs
Tissue inhibitors of matrix metalloproteinases (TIMPs) are the main inhibitors of MMPs in tissue, physiologically controlling their activity. The four known TIMP proteins (TIMP 1–4) inhibit MMPs in a 1:1 stoichiometric fashion, by interaction of mainly the C-terminal with the MMP catalytic site. Individual TIMPs differ in their ability to inhibit various MMPs, and in their gene regulation and tissue-specific patterns of gene expression [121]. TIMP-1 and 2 exhibit inhibitory activity against the active forms of all MMPs, TIMP-1 preferentially binding MMP-9 in both active and latent form [122], and TIMP-2 preferentially binding active or latent MMP-2 [123]. TIMP-2 in addition has an important role in activation of MMP-2, together with MT1-MMP.
TIMPs (1–3) are produced by trophoblastic and decidual tissues throughout gestation [124,125,100]. TIMP-4 is secreted from mouse blastocyst, and the addition of specific TIMP-4 antibody increases the expression and activity of MMP-2 and MMP-9 [126]. In addition this group also found TIMP-4 to be expressed in a malignant choriocarcinoma human cell-line (JEG-3) [127]. Lately TIMP-1 and -3 and to a lesser extent TIMP-2 were detected in pre-implantation human embryos, indicating that MMP and TIMP genes are among the first genes to be expressed in the developing embryo, preparing for implantation [128]. Growth factors and cytokines known to effect trophoblast invasiveness may act by up or down-regulation of TIMPs. TGF-β for example inhibits trophoblast invasion by up-regulating TIMP-1 and PAI-1 and down-regulating uPA [129]. TIMPs may have additional roles besides MMP inhibition, including increasing cell proliferation [130,131] and embryo development [132].
8.3 Serine proteases
The plasminogen activator system includes the urokinase-type plasminogen activator (uPA), the tissue-type plasminogen activator (tPA), the PA inhibitors PAI-1 and PAI-2 and the cell surface uPA receptor. The PA system converts plasminogen into the active serine protease plasmin, which can degrade ECM. The activity of the PA system is balanced by the inhibitors (PAI-1 and -2) [133]. Besides directly degrading ECM, the PA system has an indirect effect, through proteolytic activation of MMPs. Both uPA and plasmin are reported in the uterus [134] and in trophoblasts [135]. Studies in mice and rats suggest a role for the PA system in the implantation process [136]. A recent report found that Adrenomedullin (ADM), a polypeptide belonging to calcitonin gene-related peptide superfamily, enhances in vitro trophoblast proliferation and invasion [137]. Both ADM binding sites and ADM are present in the trophoblasts, the latter is most abundant in first trimester placenta [138]. The report showed that ADM decreased PAI-1 expression and increased MMP-2 activity, thereby enhancing the downstream reaction of cell invasion [137].
9. Endovascular invasion
The development of a placental vascular network is essential for the growth and maintenance of the developing embryo. Several factors are involved in this angiogenic process, including VEGF (vascular endothelial growth factor), PDGF (platelet-derived growth factor) and PAF (platelet-activating factor) [26].
VEGF induces angiogenesis and increases permeabilization of blood vessels. VEGF and its receptors are expressed in both the endometrium and in trophoblastic cells [1]. Mouse embryos with functional inactivation of one VEGF allele die on day 11–12 of pregnancy and show several malfunctions on the vascular system [139]. VEGF mRNA expression can be detected already in the blastocyst, enabling the implanting embryo to induce angiogenesis at the implantation site by binding to endometrial receptors [1]. VEGF expression is up-regulated in placental tissues by hypoxia, associated with early placental development, whereas another angiogenic factor PIGF (Placental growth factor) is down-regulated [140]. Surprisingly placental VEGF was reduced in pre-eclamptic pregnancies, despite the prolonged hypoxic condition associated with pre-eclampsia [141]. PIGF, on the other hand, was decreased in pre-eclampsia, as expected [142]. TGF-β and TNF-α, two pro-angiogenic factors present in the uterus, increased VEGF expression in a trophoblast cell-line [143]. This is especially interesting in light of the elevated TGF-β level, thought to be involved in pre-eclampsia [42], which may be an attempt to raise the low VEGF level.
ICAM (intercellular adhesion molecule), VCAM (vascular cell adhesion molecule) and PECAM (platelet endothelial cell adhesion molecule) are endothelial-cell adhesion molecules, playing an important role in endothelial activation and are elevated in the maternal circulation during pregnancy. Co-culture of endothelial cells and trophoblasts lead to an increased expression of ICAM, VCAM and E-selectin, indicating that factors released from the trophoblasts activate endothelial cells [108]. In pre-eclamptic pregnancies the expression of these adhesion molecules in the maternal circulation is further increased [108,144]. Elevated expression of E-selectin, as found in pre-eclampsia and in placental tissues cultured under hypoxia-reoxygenation conditions, may be mediated by the cytokine tumor necrosis factor-alpha, since anti-TNFα antibody reduced this activation [145]. Elevated VCAM in maternal blood together with elevated hyperhomocyst(e)inemia, a preeclampsia risk factor, were found to be strongly associated with an increased risk of preeclampsia [146]. This is in contrast to the result of an immunocytochemical study of placental bed biopsies from normal and pre-eclamptic women, which found no difference in ICAM, PECAM, VCAM and E-selectin expression between the two groups, suggesting that these molecules are not implicated in the etiology of pre-eclampsia [147]. In another study MCAM (melanoma cell adhesion molecule) expression was reduced in trophoblasts in placentas from pre-eclamptic women compared to normal, as detected by immunohistochemistry [148]. VCAM is lower in term than in first trimester placenta, indicating importance in the developmental stage. The reduced VCAM expression found in pregnancies complicated with fetal growth restriction further supports this concept [149]. In a comparison study of first trimester serum between normal pregnant women and women with pregnancy-induced hypertension (PIH) ICAM and E-selectin, but not VCAM or P-selectin, were significant higher in women who developed PIH late in gestation, suggesting that these factors can serve as effective indicators of the onset of PIH [150].
10. Conclusion
Ethical restrictions and limited availability of human tissue confined our studies on human implantation. It must be appreciated that most knowledge comes from in vitro experiments using cultured human trophoblasts or cell-lines and in vivo studies from other species. All the same, it is clear today that the implantation process depends on appropriate timing and is regulated by various factors of both maternal and embryonic origin. The success of this process is a result of complex interactions between these factors and comprehension of the process demands further characterization of these interactions.
New technologies that allow the profiling of tissues at the genomic, transcriptomic and proteomic levels are becoming available will probably bring more information and hopefully will help to shed more light on the implantation process. Further understanding of the process will enable new strategies in treating implantation failure both in natural and in assisted reproduction.
Figure 1 A schematic representation of a blastocyst approaching the receptive endometrium, defined by the integrin profile and appearance of pinopodes. Early signaling between the blastocyst and the endometrium precedes the attachment.
Figure 2 A schematic representation of an implanting blastocyst, highlighting interactions between trophoblastic and endometrial cells, including integrins, growth factors, cytokines, hormones and proteases.
==== Refs
Krüssel JS Bielfeld P Polan ML Simón C Regulation of embryonic Implantation European Journal of Obstetrics and Gynecology and Reproductive Biology 2003 110 2 9 12932861 10.1016/S0301-2115(03)00167-2
Bischof P Campana A Molecular mediators of implantation Bailliere's Clinical Obstetrics and Gynaecology 2000 14 801 814 10.1053/beog.2000.0120
Vigano P Mangioni S Pompei F Chiodo I Maternal-conceptus Cross Talk – A Review Placenta 2003 24 S56 S61 14559031 10.1016/S0143-4004(03)00137-1
Norwitz ER Schust DJ Fisher SJ Implantation and the survival of early pregnancy N Engl J Med 2001 345 1400 1408 11794174 10.1056/NEJMra000763
Lopata A Bentin-Ley U Enders A "Pinopodes" and Implantation Reviews in Endocrine & metabolic Disorders 2002 3 77 86 12007284 10.1023/A:1015455709833
Carver JC Martin K Spyropoulou I Barlow D Sargent I Mardon H An in-vitro model for stromal invasion during implantation of the human blastocyst Human Reproduction 2003 18 283 290 12571163 10.1093/humrep/deg072
Nacas G Pinopodes as markers of endometrial receptivity in clinical practice Hum Reprod 1999 14 99 106 10690805
Burghardt RC Johnson GA Jarger LA Ka H Garlow JE Spencer TE Bazer FW Integrins and Extracellular Matrix Proteins at the Maternal-Fetal Interface in Domestic Animals Cell Tissues Organs 2002 172 202 217 10.1159/000066969
Lessey BA Endometrial integrins Endocrinologist 1995 5 214 221
Lessey BA Castelbaum AJ Buck CA Lei Y Yowell CW Sun J Further characterization of endometrial integrins during the menstrual cycle and in pregnancy Fertil Steril 1994 62 497 506 8062944
Lessey BA Endometrial integrins and the establishment of uterine receptivity Hum Reprod 1998 13 247 258 discussion 259–261 9755427
Nardo LG Nikas G Makrigiannakis A Sinatra F Nardo F Synchronous expression of pinopodes and alpha v beta 3 and alpha 4 beta 1 integrins in the endometrial surface epithelium of normally menstruating women during the implantation window J Reprod Med 2003 48 355 361 12815909
Lessey BA Castelbaum AJ Sawin SW Sun J Integrin as markers of uterine receptivity in women with primary unexplained infertility Fertil Steril 1995 63 535 542 7851583
Lessey BA Ilesanmi AO Lessey MA Riben M Harris JF Chwalisz K Luminal and glandular endometrial epithelium express integrins differentially throughout the menstrual cycle: implications for implantation, contraception and infertility Am J Reprod Immunol 1996 35 195 204 8962647
Skrzypczak J Mikolajczyk M Szymanowski K Endomatrial receptivity: expression of alpha3beta1, alpha4beta1 and alphaVbeta1 endometrial integrins in women with impaired fertility Reprod Biol 2001 1 85 94 14666169
Bowen JA Hunt JS The role of Integrins in Reproduction Proc Soc Exp Biol Med 2000 223 331 343 10721002 10.1046/j.1525-1373.2000.22348.x
Campbell S Swann HR Seif MW Kimber SJ Aplin JD Cell adhesion molecules on the oocyte and preimplantation human embryo Hum Reprod 1995 10 1571 1578 7593541
Kabir-Salmani M Shiokawa S Akimoto Y Sakai K Nagamatsu S Sakai S Nakamura Y Lofti A Kawakami H Iwashita M Alphavbeta3 integrin signaling pathway is involved in insulin-like growth factor I-stimulated human extravillous trophoblast cell migration Endocrinology 2003 144 1620 1630 12639947 10.1210/en.2002-220886
Kabir-Salmani M Shiokawa S Akimoto Y Sakai K Iwashita M The role of alpha(5)beta(1)-integrin in the IGF-I-induced migration of extravillous trophoblast cells during the process of implantation Mol Hum Reprod 2004 10 91 97 14742693 10.1093/molehr/gah014
Stephens LE Sutherland AE Klimanskaya IV Andrieux A Meneses J Pedersen RA Damsky CH Deletion of beta 1 integrin in mice results in inner cell mass failure and periimplantation lethality Genes Dev 1995 9 1883 1895 7544312
Simón C Martin JC Pellicer A Paracerine regulators of implantation Bailliere's Clinical Obstetrics and Gynaecology 2000 14 815 826 10.1053/beog.2000.0121
Illera MJ Cullinan E Gui Y Yuan L Beyler SA Lessey BA Blockade of the α5β3 integrin adversely affects implantation in the mouse Biol Reprod 2000 62 1285 1290 10775178 10.1095/biolreprod62.5.1285
Aplin JD Kimber SJ Trophoblast-uterine interactions at implantation Reproductive Biology and Endocrinology 2004 2 48 15236654 10.1186/1477-7827-2-48
Dominguez F Pellicer A Simon C Paracrine dialogue in implantation Molecular and Cellular Endocrinology 2002 186 175 181 11900893 10.1016/S0303-7207(01)00659-1
Meseguer M Aplin JD Caballero-Campo P O'Connor JE Martin JC Remohi J Pellicer A Simon C Human endometrial mucin MUC1 is up-regulated by progesterone and down-regulated in vitro by the human blastocyst Biol Reprod 2001 64 590 601 11159362 10.1095/biolreprod64.2.590
Hill JA Maternal-Embryonic Cross-Talk Ann N Y Acad Sci 2001 943 17 25 11594538
Salamonsen LA Shuster S Stern R Deposition of hyaluronan in human endometrium across the menstrual cycle: implications for implantation and menstruation Cell Tissue Res 2001 306 335 340 11702245 10.1007/s004410100452
Salamonsen LA Nie G Proteinases at the Endometrial-Trophoblast Interface: Their Role in Implantation Reviews in Endocrine & Metabolic Disorders 2002 3 133 143 12007290 10.1023/A:1015407012559
Xu P Wang Y Piao Y Bai S Xiao Z Jia Y Luo S Zhuang L Effects of Matrix Proteins on the Expression of Matrix Metalloproteinase-2, -9. and -14 and Tissue Inhibitors of Metalloproteinases in Human Cytotrophoblast Cells During the First Trimester Biology of Reproduction 2001 65 240 246 11420245 10.1095/biolreprod65.1.240
Hofmann GE Scott RT JrBergh PA Deligdisch L Immunohistochemical localization of epidermal growth factor in human endometrium, decidua, and placenta J Clin Endocrinol Metab 1991 73 882 887 1890159
Bass KE Morrish D Roth I Bhardwaj D Taylor R Zhou Y Fisher SJ Human Cytotrophoblast Invasion Is Up-regulated by Epidermal Growth Factor: Evidence That Paracrine Factors Modify This Process Developmental Biology 1994 164 550 561 8045351 10.1006/dbio.1994.1223
Staun-Ram E Goldman S Gabarin D Shalev E Expression and importance of matrix metalloproteinases 2 and 9 (MMP-2 and -9) in human trophoblast invasion Reproductive Biology and Endocrinology 2004 2 59 15294019 10.1186/1477-7827-2-59
Maruo T Matsuo H Otani T Mochizuki M Role of epidermal growth factor (EGF) and its receptor in the development of the human placenta Reprod Fertil Dev 1995 7 1465 1470 8743148 10.1071/RD9951465
Dakour J Li H Chen H Morrish DW EGF promotes development of a differentiated trophoblast phenotype having c-myc and junB proto-oncogene activation Placenta 1999 20 119 126 9950153 10.1053/plac.1998.0336
Li RH Zhuang LZ The effect of growth factors on human normal placental cytotrophoblast cell proliferation Hum Reprod 1997 12 830 834 9159451 10.1093/humrep/12.4.830
Anteby EY Greenfield C Natanson-Yaron S Goldman-Wohl D Hamani Y Khudyak V Ariel I Yagel S Vascular endothelial growth factor, epidermal growth factor and fibroblast growth factor-4 and -10 stimulate trophoblast plasminogen activator system and metalloproteinase-9 Mol Hum Reprod 2004 10 229 235 14996996 10.1093/molehr/gah031
Qiu Q Yang M Tsang BK Gruslin A EGF-induced trophoblast secretion of MMP-9 and TIMP-1 involves activation of both PI3K and MAPK signaling pathways Reproduction 2004 128 355 363 15333786 10.1530/rep.1.00234
Nakatsuji Y Nishio Y Tani N Adachi K Ohmichi M Hisamoto K Morishige K Kurachi H Tasaka K Murata Y Matsuura N Epidermal growth factor enhances invasive activity of BeWo choriocarcinoma cells by inducing alpha2 integrin expression Endocr J 2003 50 703 714 14709841 10.1507/endocrj.50.703
Graham CH Connelly I MacDougall JR Kerbal RS Stetler-Stevenson WG Lala PK Resistence of malignant trophoblast cells to both the anti-proliferative and anti-invasive effects of transforming growth factor beta Experimental Cell Research 1994 103 1641 1650
Lash GE Otun HA Innes BA Bulmer JN Searle RF Robson SC Inhibition of Trophoblast Cell Invasion by TGFB1, 2, and 3 Is Associated with a Decrease in Active Proteases Biol Reprod 2005, Apr 27 15858216
Djurovic S Schjetlein R Wisloff F Haugen G Husby H Berg K Plasma concentrations of Lp(a) lipoprotein and TGF-bera1 are altered in preeclampsia Clin Genet 1997 52 371 376 9520129
Caniggia I Winter J Lye SJ Post M Oxygen and placental development during the first trimester: implications for the pathophysiology of pre-eclampsia Placenta 2000 21 S25 30 10831118 10.1053/plac.1999.0522
Bischof P Meisser A Campana A Involvement of trophoblast in embryo implantation: regulation by paracrine factors J Reprod Immunol 1998 39 167 177 9786460 10.1016/S0165-0378(98)00020-5
Hamilton GS Lysiak JJ Han VK Lala PK Autocrine-paracrine regulation of human trophoblast invasiveness by insulin-like growth factor (IGF)-II and IGF-binding protein (IGFBP)-1 Exp Cell Res 1998 244 147 156 9770358 10.1006/excr.1998.4195
Irving JA Lala PK Functional role of cell surface integrins on human trophoblast cell migration: regulation by TGF-beta, IGF-II and IGFBP-1 Exp Cell Res 1995 217 419 427 7535237 10.1006/excr.1995.1105
Stewart CL Kaspar P Brunet LJ Bhatt H Gadi I Kontgen F Abbondanzo SJ Blastocyst implantation depends on maternal expression of leukaemia inhibitory factor Nature 1992 359 76 79 1522892 10.1038/359076a0
Chen HF Shew JY Ho HN Hsu WL Yang YS Expression of leukemia inhibitory factor and its receptor in preimplantation embryos Fertil Steril 1999 72 709 713 10.1016/S0015-0282(99)00306-4
Sharkey AM King A Clark DE Burrows TD Jokhi PP Charnock-Jones DS Loke YW Smith SK Localization of leukemia inhibitory factor and its receptor in human placenta throughout pregnancy Biology of reproduction 1999 60 355 364 9916002 10.1095/biolreprod60.2.355
Bischof P Haenggeli L Campana A Effect of leukemia inhibitory factor on human cytotrophoblast differentiation along the invasive pathway American Journal of Reproductive Immunology 1995 34 225 230 8579759
Bamberger AM Jenatschke S Schulte HM Loning T Bamberger MC Leukemia inhibitory factor (LIF) stimulates the human HLA-G promoter in JEG3 chriocarcinoma cells Endocrinol Metab 2000 85 3932 3936 10.1210/jc.85.10.3932
Vigano P Mangioni S Pompei F Chiodo I Maternal-conceptus Cross talk – A review Placenta 2003 24 56 61 10.1016/S0143-4004(03)00137-1
Simón C Frances A Piquette GN el Danasouri I Zurawski G Dang W Polan ML Embryonic implantation in mice is blocked by interleukin-1 receptor antagonist Endocrinology 1994 134 521 528 8299552 10.1210/en.134.2.521
Meisser A Chardonnens D Campana A Bischof P Effects of tumour necrosis factor alpha, interleukin-I alpha, macrophage colony stimulating factor and transforming growth factor beta on trophoblastic matrix metalloproteinases Molecular Human Reproduction 1999 5 252 260 10333360 10.1093/molehr/5.3.252
Huang HY Wen Y Irvine JC Kruessel JS Soong YK Polan ML Cytokine mediated regulation of TIMP-1, TIMP-3 and 92-kDa Type IV collagenase messenger RNA expression in human endometrial stroma cells J Clin Endiocrinology 1998 83 1721 1729 10.1210/jc.83.5.1721
Meisser A Cameo P Islami D Campana A Bischof P Effects of interleukin-6 (IL-6) on cytotrophoblastic cells Molecular Human Reproduction 1999 5 1055 1058 10541568 10.1093/molehr/5.11.1055
Roth I Fisher SJ IL-10 is an autocrine inhibitor of human placental cytotrophoblast MMP-9 production and invasion Developmental Biology 1999 205 194 204 9882507 10.1006/dbio.1998.9122
Licht P Russu V Wildt L On the role of human chorionic gonadotrophin (hCG) in the embryo-endometrial microenviroment: implications for differentiation and implantation Semin Reprod Med 2001 19 37 47 11394202 10.1055/s-2001-13909
Oktay K Brzyski RG Miller EB Krugman D Association of serum beta-hCG levels with myosalpingeal invasion and viable trophoblast mass in tubal pregnancy Obstet Gynecol 1994 84 803 806 7524002
Klein M Graf A Kiss H The relation between depth of trophoblast invasion and beta-hCG levels in tubal pregnancies Archives of Gynecology and Obstetrics 1995 256 85 88 7541981
Zygmunt M Hahn D Munstedt K Bischof P Lang U Invasion of cytotrophoblastic JEG-3 cells is stimulated by hCG in vitro Placenta 1998 19 587 593 9859862 10.1016/S0143-4004(98)90019-4
Yagel S Geva TE Solomon H Shimonovitz S Reich R Finci-Yeheskel Z Mayer M Milwidsky A High levels of human chorionic gonadotropin retard first trimester trophoblast invasion in vitro by decreasing urokinase plasminogen activator and collagenase activities J Clin Endocrinol Metab 1993 77 1506 1511 8263134 10.1210/jc.77.6.1506
Milwidsky A Finci-Yeheskel Z Yagel S Mayer M Gonadotropin-mediated inhibition of proteolytic enzymes produced by human trophoblast in culture J Clin Endocrinol Metab 1993 76 1101 1105 7684389 10.1210/jc.76.5.1101
Zygmunt M McKinnon T Herr F Lala PK Han VK HCG increases trophoblast migration in vitro via the insulin-like growth factor-II/mannose-6 phosphate receptor Mol Hum Reprod 2005 11 261 267 15749784 10.1093/molehr/gah160
Fazleabas AT Donnelly KM Srinivasan S Fortman JD Miller JB Modulation of the baboon (Papio anubis] uterine endometrium by chorionic gonadotrophin during the period of uterine receptivity Proc Natl Acad Sci USA 1999 96 2543 2548 10051679 10.1073/pnas.96.5.2543
Srisuparp S Strakova Z Fazleabas AT The role of chorionic gonadotrophin (CG) in blastocyst implantation Arch Med Res 2001 32 627 634 11750740 10.1016/S0188-4409(01)00330-7
Cole LA Khanlian SA Sutton JM Davies S Stephens ND Hyperglycosylated hCG (invasive trophoblast antigen, ITA) a key antigen for early pregnancy detection Clin Biochem 2003 36 647 655 14636881 10.1016/S0009-9120(03)00108-5
Palomaki GE Knight GJ Roberson MM Cunningham GC Lee JE Strom CM Pandian R Invasive trophoblast antigen (hyperglycosylated human chorionic gonadotropin) in second-trimester maternal urine as a marker for down syndrome: preliminary results of an observational study on fresh samples Clin Chem 2004 50 182 189 14709646 10.1373/clinchem.2003.023986
Cole LA Sutton JM Higgins TN Cembrowski GS Between-method variation in human chorionic gonadotropin test results Clin Chem 2004 50 874 882 15016725 10.1373/clinchem.2003.026989
Bahado-Singh RO Oz AU Kingston JM Shahabi S Hsu CD Cole L The role of hyperglycosylated hCG in trophoblast invasion and the prediction of subsequent pre-eclampsia Prenat Diagn 2002 22 478 481 12116306 10.1002/pd.329
Dunn CL Kelly RW Critchley HO Decidualization of the human endometrial stromal cell: an enigmatic transformation Reprod Biomed Online 2003 7 151 161 14567882
Spencer TE Bazer FW Conceptus signals for establishment and maintenance of pregnancy Reproductive Biology and Endocrinology 2004 2 49 15236653 10.1186/1477-7827-2-49
Shimonovitz S Hurwitz A Hochner-Celnikier D Dushnik M Anteby E Yagel S Expression of gelatinase B by trophoblast cells: down-regulation by progesterone American Journal of Obstetrics and Gynecology 1998 1788 457 461 9539508 10.1016/S0002-9378(98)70420-X
Goldman S Shalev E Difference in Progesterone Receptor Isoforms Ratio, Between Early and Late First Trimester Human Trophoblast, Is Associated with Differential Cell Invasion and Matrix Metalloproteinase2 (MMP2) Expression Biol Reprod 2005 Aug 31
Dai B Cao Y Liu W Yang Y Chen D Duan E Dual roles of progesterone in embryo implantation in mouse Endocrine 2003 21 123 132 12897374 10.1385/ENDO:21:2:123
Makrigiannakis A Zoumakis E Kalantaridou S Coutifaris C Margioris AN Coukos G Rice KC Gravanis A Chrousos GP Cortoco-releasing hormone promotes blastocyst implantation and early maternal tolerance Nature Immunology 2001 2 1018 1023 11590404 10.1038/ni719
Athanassakis I Farmakiotis V Aifantis I Gravanis A Vassiliadis S Expression of corticotrophin-releasing hormone in the mouse uterus: participation in embryo implantation J Endocrino 1999 163 221 227 10.1677/joe.0.1630221
Bamberger AM Shulte HM Thuneke I Erdmann I Bamberger CM Asa SL Expression of the apoptosis-inducing Fas ligand (FasL) in human first and third trimester placenta and choriocarcinoma cells J Clin Endocrinol Metab 1997 82 3173 3175 9284765 10.1210/jc.82.9.3173
Karray S Kress C Cuvellier S Hue-Beauvais C Damotte D Babinet C Levi-Strauss M Complete loss of Fas ligand gene causes massive lymphoproliferation and early death, indicating a residual activity of gld allele J Immunol 2004 172 2118 2125 14764677
Muglia L Jacobson L Dikkes P Majzoub JA Corticotropin-releasing hormone deficiency reveals major fetal but not adult glucocorticoid need Nature 1995 373 427 432 7830793 10.1038/373427a0
Smith GW Aubry JM Dellu F Contarino A Bilezikjian LM Gold LH Chen R Marchuk Y Hauser C Bently CA Sawchenko PE Koob GF Vale W Lee KF Corticotropin releasing factor receptor 1-deficient mice display decreased anxiety, impaired stress response, and aberrant neuroendocrine development Neuron 1998 20 1093 1102 9655498 10.1016/S0896-6273(00)80491-2
Hunt JS Expression and regulation of the tumour necrosis factor-alpha gene in the female reproductive tract Reprod Fertil Dev 1993 5 141 153 8265799 10.1071/RD9930141
Tabibzadeh S Zupi E Babaknia A Liu A Marconi D Romanini C Site and menstrual cycle-dependent expression of proteins of the tumour necrosis factor (TNF) receptor family, and BCL-2 oncoprotein and phase-specific production Hum Reprod 1995 10 277 286 7539446
Yang Y Yelavarthi KK Chen HL Pace JL Tettanova PF Hunt JS Molecular biochemical and functional characteristics of tumor necrosis factor alpha produced by human placental cytotrophoblastic cells J Immuno 1993 150 5614 5624
Monzón-Bordonaba F Vadillo-Ortega F Feinberg RF Modulation of trophoblast function by tumor necrosis factor-alpha: a role in pregnancy establishment and maintenance? Am J Obstet Gyneco 2002 187 1574 1580 10.1067/mob.2002.128028
Hillier SL Witkin SS Khron MA Watts DH Kiviat NB Eschenbach DA The relationship of amniotic fluid cytokines and preterm delivery, amniotic fluid infection, histologic chorioamnionitis, and chorioamnion infection Obstet Gyneco 1993 81 941 948
Bauer S Pollheimer J Hartmann J Husslein P Aplin JD Knofler M Tumor necrosis factor-alpha inhibits trophoblast migration through elevation of plasminogen activator inhibitor-1 in first-trimester villous explant cultures Clin Endocrinol Metab 2004 89 812 822 10.1210/jc.2003-031351
Benyo DF Miles TM Conrad KP Hypoxia stimulates cytokine production by villous explants from the human placenta J Clin Endocrinol Metab 1997 82 1582 1588 9141553 10.1210/jc.82.5.1582
Frank GR Brar AK Cedars MI Handwerger S Prostaglandin E2 enhances human endometrial stromal cell differentiation Endocrinology 1994 134 258 263 7506205 10.1210/en.134.1.258
Dimitriadis E Stoikos C Baca M Fairlie WD McCoubrie JE Salamonsen LA Relaxin and prostaglandin E(2) regulate interleukin 11 during human endometrial stromal cell decidualization J Clin Endocrinol Metab 2005 90 3458 3465 15784719 10.1210/jc.2004-1014
Shi JJ Ma XH Diao HL Ni H Xu LB Zhu H Yang ZM Differential expression of prostaglandin E receptor subtype EP2 in rat uterus during early pregnancy Histol Histopathol 2005 20 1021 1028 16136483
Kao LC Tulac S Lobo S Imani B Yang JP Germeyer A Osteen K Taylor RN Lessey BA Giudice LC Global gene profiling in human endometrium during the window of implantation Endocrinology 2002 143 2119 2138 12021176 10.1210/en.143.6.2119
Dey SK Fatty link to fertility Nature 2005 435 34 35 15875005 10.1038/435034a
Tan HN Liu Y Diao HL Yang ZM Cyclooxygenases and prostaglandin E synthases in preimplantation mouse embryos Zygote 2005 13 103 108 16128405 10.1017/S0967199405003187
Westermarck J Kahari VM Regulation of matrix metalloproteinase expression in tumor invasion FASEB J 1999 13 781 972 10224222
Sternlicht MD Werb Z How Matrix Metalloproteinases Regulate Cell Behavior Annu Rev Cell Dev Biol 2001 17 463 516 11687497 10.1146/annurev.cellbio.17.1.463
Hidalgo M Eckhardt SG Development of Matrix Metalloproteinase Inhibitors in Cancer Therapy J Natl Cancer Inst 2001 93 178 193 11158186 10.1093/jnci/93.3.178
Fini ME Cook JR Mohan R Brinckerhoff CE Parks WC, Mecham RP Regulation of matrix metalloproteinase gene expression Matrix Metalloproteinases 1998 New York: Academic 299 356
Strongin AY Collier I Bannikov G Marmer BL Grant GA Goldberg GI Mechanism of cell surface activation of 72-kDa type IV collagenase. Isolation of the activated form of the membrane metalloproteinase J Biol Chem 1995 270 5331 5338 7890645 10.1074/jbc.270.10.5331
Goldman S Shalev E The role of matrix metalloproteinases in human endometrial and ovarien cycles European Journal of Obstetrics & Gynecology and Reproductive Biology 2003 111 109 121 14597237 10.1016/S0301-2115(03)00341-5
Niu R Okamoto T Iwase K Nomura S Mizutani Quantitative Analysis of Matrix Metalloproteinase-2 and -9 and their Tissue Inhibitors-1 and -2 in Human Placenta through gestation Life Sciences 2000 66 1127 1137 10737363 10.1016/S0024-3205(00)00416-1
Dey SK Lim H Das SK Reese J Paria BC Daikoku T Wang H Molecular Cues to Implantation Endocrine Reviews 2004 25 341 373 15180948 10.1210/er.2003-0020
Librach CL Werb Z Fitzgerald ML Chiu K Corwin NM Esteves RA Grobelny D Galardy R Damsky CH 92-kDa type IV collagenase mediates invasion of human cytotrophoblasts J Cell Biol 1991 113 437 449 1849141 10.1083/jcb.113.2.437
Bischof P Matelli M Campana A Itoh Y Ogata Y Nagase H Importance of metalloproteinases in human trophoblast invasion Early Pregnancy Biology and Medicine 1995 1 263 269
Xu P Wang Y Zhu S Luo S Piao Y Zhuang L Expression of Matrix Metalloproteinase-2, -9 and -14, Tissue Inhibitors of Metalloproteinse-1, and Matrix Proteins in Human Placenta During the First trimester Biol Reprod 2000 62 988 994 10727268 10.1095/biolreprod62.4.988
Isaka K Usuda S Ito H Sagawa Y Nakamura H Nishi H Suzuki Y Li YF Takayama M Expression and Activity of Matrix Metalloproteinase 2 and 9 in Human Trophoblasts Placenta 2003 24 53 64 12495660 10.1053/plac.2002.0867
Puistola U Ronnberg L Martikainen H Turpeenniemi-Hujanen T The human embryo produces basement membrane collagen (type IV collagen)-degrading protease activity Hum Reprod 1989 4 309 311 2715307
Turpeenniemi-Hujanen T Feinberg RF Kauppila A Puistola U Laminin in the human embryo implantation: analog to the invasion by malignant cells Fertil Steril 1992 58 105 113 1378028
Wang Y Zhang Y Lewis DF Gu Y Li H Granger DN Alexander JS Protease chymotrypsin mediates the endothelial expression of P- and E-selectin, but not ICAM and VCAM, induced by placental trophoblasts from pre-eclamptic pregnancies Placenta 2003 24 851 861 13129682 10.1016/S0143-4004(03)00132-2
Yu Q Stamenkovic I Cell surface-localized matrix metalloproteinases-9 proteolytically activates THF-beta and promotes tumor invasion and angiogenesis Genes Dev 2000 14 163 176 10652271
Martin DC Fowlkes JL Babic B Khokha R Insulin-like growth factor II signaling in neoplastic proliferation is blocked by transgenic expression of the metalloproteinase inhibitor TIMP-1 J Cell Biol 1999 146 881 892 10459021 10.1083/jcb.146.4.881
Fernandez-Patron C Radomski MW Davidge ST Vascular matrix metalloproteinase-2 cleaves big endothelin-1 yielding a noval vasoconstrictor Circ Res 1999 85 906 911 10559137
Patterson BC sang QA Angiostatin-converting enzyme activities of human matrilysin (MMP-7) and gelatinase B/type IV collagenase (MMP-9) J Biol Chem 1997 272 28823 28825 9360944 10.1074/jbc.272.46.28823
Itoh T Ikeda T Gomi H Nakao S Suzuki T Itohara S Unaltered secretion of beta-amaloid precursor protein in gelatinase A (matrix metalloproteinase 2)-deficient mice J Biol Chem 1997 272 22389 22392 9278386 10.1074/jbc.272.36.22389
Dubois B Arnold B Opdenakker G Gelatinase B deficiency impairs reproduction J Clin Invest 2000 106 627 628 10974013
Holmbeck K Bianco P Caterina J Yamada S Kromer M Kuznetsov SA Mankani M Robey PG Poole AR Pidoux I Ward JM Birkedal-Hansen H MT1-MMP-deficient mice develop dwarfism, osteopenia, arthritis, and connective tissue disease due to inadequate collagen tuerover Cell 1999 99 81 92 10520996 10.1016/S0092-8674(00)80064-1
Rudolph-Owen LA Hulboy DL Wilson CL Mudgett J Matrisian LM Coordinate expression of matrix metalloproteinase family members in the uterus of normal, matrilysin-deficient, and stromelysin 1-deficient mice Endocrinol 1997 138 4902 4911 10.1210/en.138.11.4902
Pendas AM Balbin M Llana E Jimenez MG Lopez-Otin C Structural analysis and promoter characterization of the human collagenase-3 gene (MMP13) Genomics 1997 40 222 33 9119388 10.1006/geno.1996.4554
Bischof P Truong K Campana A Regulation of Trophoblastic Gelatinases by Proto-oncogenes Placenta 2003 24 155 163 12566242 10.1053/plac.2002.0890
Yamamoto H Flannery ML Kupriyanov S Pearce J McKercher SR Henkel GW Maki RA Werb Z Oshima RG Defective trophoblast function in mice with a targeted mutation of Ets2 Genes Dev 1998 12 1315 1326 9573048
Wang HZ Zhao YG Wang HM Yang Q Lin HY Sang QX Zhu C Expression of adamalysin 19/ADAM19 in the endometrium and placenta of rhesus monkey (Macaca mulatta) during early pregnancy Mol Hum Reprod 2005 11 429 435 15901844 10.1093/molehr/gah183
Baker AH Edwards DR Murphy G Metalloproteinase inhibitors: biological actions and therapeutic opportunities J Cell Sci 2002 115 3719 3727 12235282 10.1242/jcs.00063
Goldberg GI Strongin A Collier IE Genrich LT Marmer BL Interaction of 92 kDa type IV collagenase with the tissue inhibitor of metalloproteinases prevents dimerization, complex formation with interstitial collagenase, and activation of the pro-enzyme with stromelysin J Biol Chem 1992 267 4583 4591 1311314
Stetler-Stevenson WG Krutzsch HC Liotta LA Tissue inhibitor of metalloproteinase-2 (TIMP-2), a new member of metalloproteinase inhibitor family J Biol Chem 1989 264 17374 17378 2793861
Hurskainen T Hoyhtya M Tuuttila A Oikarinen A Autio-Harmainen H mRNA expression of TIMP-1, -2 and -3 and 92-kDa type IV collagenase in early human placenta and decidual membrane as studied by in situ hybridization J Histochem Cytochem 1996 44 1379 1388 8985130
Ruck P Marzusch K Horny H-P Diet J Kaiserling E The distribution of tissue inhibitor of metalloproteinases-2 (TIMP-2) in the human placenta Placenta 1996 17 263 266 8761971 10.1016/S0143-4004(96)90047-8
Zhang J Zhao YG Cao YJ Sang QX Duan EK Expression and implications of tissue inhibitor of metalloproteinase-4 in mouse embryo Mol Hum Reprod 2003 9 143 149 12606590 10.1093/molehr/gag019
Zhang J Cao YJ Zhao YG Sang QX Duan EK Expression of matrix metalloproteinase-26 and tissue inhibitor of metalloproteinase-4 in human normal cytotrophoblast cells and a choriocarcinoma cell line, JEG-3 Mol Hum Reprod 2002 8 659 666 12087081 10.1093/molehr/8.7.659
Wang H Wen Y Mooney S Li H Behr B Polan ML Matrix metalloproteinase and tissue inhibitor of matrix metalloproteinase expression in human preimplantation embryos Fertility and Sterility 2003 80 736 742 14505747 10.1016/S0015-0282(03)00782-9
Chakraborty C Gleeson LM McKinnon T Lala PK Regulation of human trophoblast migration and invasiveness Can J Physiol Pharmacol 2002 80 116 124 11934254 10.1139/y02-016
Hayakawa T Yamashita K Uchijima M Iwata K Growth-promoting activity of tissue inhibitor of metalloproteinases (TIMP-1) for a wide range of cells FEBS Letters 1992 298 29 32 1544418 10.1016/0014-5793(92)80015-9
Hayakawa T Yamashita K Ohuchi E Shinagawa A Cell Growth promoting activity of tissue inhibitor of metalloproteinase-2 (TIMP-1) J Cell Sci 1994 107 2372 2379
Satoh T Kobayashi K Yamashita S Kikuchi M Sendai Y Hoshi H Tissue inhibitor of metalloproteinases (TIMP-1) produced by granolosa and oviduct cells enhances in vitro development of bovine embryos Biol Reprod 1994 50 835 844 8199264 10.1095/biolreprod50.4.835
Vassalli JD Sappino AP Berlin D The plasminogen activator/plasmin system J Clin Invest 1991 88 1067 1072 1833420
Nordengren J Pilka R Noskova V Ehinger A Domanski H Andersson C Hoyer-Hansen G Hansson SR Casslen B Differential localization and expression of urokinase plasminogen activator (uPA), its receptor (uPAR), and its inhibitor (PAI-1) mRNA and protein in endometrial tissue during the menstrual cycle Mol Hum Reprod 2004 10 655 663 15243126 10.1093/molehr/gah081
Floridon C Nielsen O Holund B Sunde L Westergaard JG Thomsen SG Teisner B Localization and significance of urokinase plasminogen activator and its receptor in placental tissue from intrauterine, ectopic and molar pregnancies Placenta 1999 20 711 721 10527826 10.1053/plac.1999.0425
Aflalo ED Sod-Moriah UA Potashnik G Har-Vardi I Differences in the implantation rates of rat embryos developed in vivo and in vitro: possible role for plasminogen activators Fertil Steril 2004 81 780 785 15019809 10.1016/j.fertnstert.2003.10.014
Zhang X Green KE Yallampalli C Dong YL Adrenomedullin Enhances Invasion by Trophoblast Cell Lines Biol Reprod 2005
Moriyama T Otani T Matyo T Expression of adrenomedullin by human placental cytotrophoblasts and choriocarcinoma JAr cells J Clin Endocrinol Metab 2001 86 3958 3961 11502838 10.1210/jc.86.8.3958
Carmeliet P Ferreira V Breier G Pollefeyt S Kieckens L Gertsenstein M Fahrig M Vandenhoeck A Harpal K Eberhardt C Declercq C Pawling J Moons L Collen D Risau W Nagy A Abnormal blood vessel development and lethality in embryos lacking a single VEGF allele Nature 1996 380 435 439 8602241 10.1038/380435a0
Shore VH Wang CL Torry RJ Caudle MR Torry DS Vascular endothelial growth factor, placenta growth factor and their receptors in isolated human trophoblast Placenta 1997 18 657 665 9364601 10.1016/S0143-4004(97)90007-2
Cooper JC Sharkey AM Charnock-Jones DS Palmer CR Smith SK VEGF mRNA levels in placentae from pregnancies complicated by pre-eclampsia Br J Obstet Gynaecol 1996 103 1191 1196 8968234
Torry DS Wang HS Wang TH Caudle MR Torry RJ Preeclampsia is associated with reduced serum levels of placenta growth factor Am J Obstet Gynecol 1998 179 1539 1544 9855593 10.1016/S0002-9378(98)70021-3
Chung IB Yelian FD Zaher FM Gonik B Evans MI Diamond MP Svinarich DM Expression and Regulation of Vascular Endothelial Growth Factor in a First Trimester Trophoblast Cell Line Placenta 2000 21 320 324 10833366 10.1053/plac.1999.0481
Llurba E Gratacos E Martin-Gallan P Cabero L Dominquez C A comprehensive study of oxidative stress and antioxidant status in preeclampsia and normal pregnancy Free Radic Biol Med 2004 37 557 570 15256227 10.1016/j.freeradbiomed.2004.04.035
Hung TH Charnock-Jones DS Skepper JN Burton GJ Secretion of tumor necrosis factor-alpha from human placental tissues induced by hypoxia-reoxygenation causes endothelial cell activation in vitro: a potential mediator of the inflammatory response in preeclampsia Am J Pathol 2004 164 1049 1061 14982858
Vadachkoria S Sanchez SE Qiu C Muy-Rivera M Malinow MR Williams MA Hyperhomocyst(e)inemia and elevated soluble vascular cell adhesion molecule-1 concentrations are associated with an increased risk of preeclampsia Gynecol Obstet Invest 2004 58 133 139 15205565 10.1159/000079067
Tziotis J Malamitsi-Puchner A Vlaschos G Creatsas G Michalas S Adhesion molecules expression in the placental bed of pregnancies with pre-eclampsia BJOG 2002 109 197 201 11888102
Liu Q Yan X Li Y Zhang Y Zhao X Shen Y Pre-eclampsia is associated with the failure of melanoma cell adhesion molecule (MCAM/CD146) expression by intermediate trophoblast Lab Invest 2004 84 221 228 14688802 10.1038/labinvest.3700033
Rajashekhar G Loganath A Roy AC Wong YC Expression and secretion of the vascular cell adhesion molecule-1 in human placenta and its decrease in fetal growth restriction J Soc Gynecol Investig 2003 10 352 360 12969778 10.1016/S1071-5576(03)00121-7
Matsubara K Abe E Ochi H Kusanagi Y Ito M Changes in serum concentrations of tumor necrosis factor alpha and adhesive molecules in normal pregnant women and those with pregnancy-induced hypertension J Obstet Gynaecol Res 2003 29 422 426 14641694 10.1111/j.1341-8076.2003.00141.x
|
16236179
|
PMC1289292
|
CC BY
|
2021-01-04 16:37:13
|
no
|
Reprod Biol Endocrinol. 2005 Oct 20; 3:56
|
utf-8
|
Reprod Biol Endocrinol
| 2,005 |
10.1186/1477-7827-3-56
|
oa_comm
|
==== Front
RetrovirologyRetrovirology1742-4690BioMed Central London 1742-4690-2-641624204510.1186/1742-4690-2-64ResearchSilencing of human T-cell leukemia virus type I gene transcription by epigenetic mechanisms Taniguchi Yuko [email protected] Kisato [email protected] Jun-ichirou [email protected] Michiyuki [email protected] Nancy [email protected] Akihiko [email protected] Masao [email protected] Laboratory of Virus Immunology, Institute for Virus Research, Kyoto University, Kyoto 606-8507, Japan2 Department of Hematology, Kumamoto University School of Medicine, Kumamoto 860-8556, Japan3 Laboratory of Infection and Prevention, Institute for Virus Research, Kyoto University, Kyoto 606-8507, Japan4 Department of Epidemiology, Harvard School of Public Health, Boston, Massachusetts 02115, USA5 Department of Laboratory Medicine, Faculty of Medicine, University of Miyazaki, Miyazaki 889-1692, Japan2005 22 10 2005 2 64 64 31 8 2005 22 10 2005 Copyright © 2005 Taniguchi et al; licensee BioMed Central Ltd.2005Taniguchi et al; licensee BioMed Central Ltd.This is an Open Access article distributed under the terms of the Creative Commons Attribution License (), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
Background
Human T-cell leukemia virus type I (HTLV-I) causes adult T-cell leukemia (ATL) after a long latent period. Among accessory genes encoded by HTLV-I, the tax gene is thought to play a central role in oncogenesis. However, Tax expression is disrupted by several mechanims including genetic changes of the tax gene, deletion/hypermethylation of 5'-LTR. To clarify the role of epigenetic changes, we analyzed DNA methylation and histone modification in the whole HTLV-I provirus genome.
Results
The gag, pol and env genes of HTLV-I provirus were more methylated than pX region, whereas methylation of 5'-LTR was variable and 3'-LTR was not methylated at all. In ATL cell lines, complete DNA methylation of 5'-LTR was associated with transcriptional silencing of viral genes. HTLV-I provirus was more methylated in primary ATL cells than in carrier state, indicating the association with disease progression. In seroconvertors, DNA methylation was already observed in internal sequences of provirus just after seroconversion. Taken together, it is speculated that DNA methylation first occurs in the gag, pol and env regions and then extends in the 5' and 3' directions in vivo, and when 5'-LTR becomes methylated, viral transcription is silenced. Analysis of histone modification in the HTLV-I provirus showed that the methylated provirus was associated with hypoacetylation. However, the tax gene transcript could not be detected in fresh ATL cells regardless of hyperacetylated histone H3 in 5'-LTR. The transcription rapidly recovered after in vitro culture in such ATL cells.
Conclusion
These results showed that epigenetic changes of provirus facilitated ATL cells to evade host immune system by suppressing viral gene transcription. In addition, this study shows the presence of another reversible mechanism that suppresses the tax gene transcription without DNA methylation and hypoacetylated histone.
==== Body
Background
Human T-cell leukemia virus type I (HTLV-I) is associated with a neoplastic disease, adult T-cell leukemia (ATL), and inflammatory diseases, such as HTLV-I-associated myelopathy (HAM)/tropical spastic paraparesis (TSP) and HTLV-I-associated uveitis [1,2]. Among HTLV-I carriers, a part of infected individuals develop ATL after a long latent period. During the leukemogenesis by HTLV-I, Tax protein is considered to play a critical role through its pleiotropic actions, which include transactivation of NF-κB, CREB and SRF pathways, transrepression of lck, p18 and DNA polymerase β gene transcriptions, and functional inactivation of p53 and MAD1 [3-6]. Through these actions, Tax induces the proliferation of HTLV-I infected cells and inhibits their apoptosis, resulting in an increase in the number of infected cells. However, since Tax protein is the major target of cytotoxic T-lymphocytes (CTLs) in vivo, the expression also has a negative effect on the survival of ATL cells [7-9]. In some ATL cells, tax gene expression is inactivated by genetic and epigenetic changes, which include deletion, insertion or mutation of the tax gene, and DNA methylation or deletion of 5'-LTR [10-13]. Such inactivation of Tax expression is considered to allow ATL cells to escape from the host immune system.
DNA methylation of retroviruses is regarded as a host defense mechanism for inactivating retrovirus expression [14]. However, it is also recognized as a mechanism for virus-infected cells to escape from the host immune system and establish the latent state. In contrast, human immunodeficiency virus (HIV) is resistant to silencing in vivo. It is because HIV is frequently integrated into active transcriptional unit in vivo [15]. These findings coincide with the fact that HIV vigorously replicates in vivo. On the other hand, DNA methylation accumulated in HTLV-I 5'-LTR has been shown to silence viral gene transcription in leukemic cells [12,13]. In addition, the frequency of integration of HTLV-I provirus into transcriptional units was equivalent to that calculated based on random integration [16], which also increased the silencing. It remains unclear where and when DNA methylation occurs within the HTLV-I provirus genome.
In this study, we analyzed DNA methylation and histone modification in the whole HTLV-I provirus, and observed the progressive accumulation of DNA methylation. In addition, another reversible mechanism silenced viral gene transcription regardless of hyperacetylated promoter region.
Results
Analyses of DNA methylation of HTLV-I provirus
To reveal DNA methylation status within the HTLV-I provirus, we analyzed the DNA methylation by sodium bisulfite sequencing and combined bisulfite restriction analysis (COBRA). Initially, DNA methylation in 5'-LTR, gag, pol, env, pX and 3'-LTR was identified by sodium bisulfite sequencing. In an ATL case (Fig. 1A), the internal regions of the HTLV-I provirus, including gag, pol and env, were heavily methylated. On the other hand, 5'-LTR and pX were partially methylated, and 3'-LTR was not methylated at all. In an ATL cell line, ATL-48T (Fig. 1A), the internal sequences of the HTLV-I provirus were partially methylated, whereas both LTRs were not methylated. Since the analyses by sodium bisulfite sequencing were time-consuming, we established the COBRA method to detect and analyze DNA methylation in a large number of samples, and then compared the results obtained with the two methods. After amplification of sodium bisulfite treated DNAs with each primer sets, the products were digested with TaqI or AccII, which contain one (TaqI) or two (AccII) CpG site(s) within the recognition sequences. When CpG site is methylated, the products retain CpG site, resulting in digestion by these enzymes. On the other hand, CpG is converted to UG when it is unmethylated. Therefore, PCR products are resistant to restriction enzymes (Fig. 1B). With the COBRA method, the extent of DNA methylation was quantified in eight CpG sites throughout the HTLV-I provirus: 5'-LTR (620 according to the numbering by Seiki et al. [17]), gag (1753), pol (2988, 4187 and 5151), env (6113), pX (7258) and 3'-LTR (8342) (Fig. 1C). The extent of DNA methylation detected by the COBRA method was well correlated with that obtained by sodium bisulfite sequencing in both cases studied, as shown in Fig. 1A and 1C.
Figure 1 DNA methylation of the HTLV-I provirus assessed by sodium bisulfite sequencing and COBRA. A. DNA methylation in the HTLV-I provirus was analyzed by sodium bisulfite sequencing in a case of acute ATL and a tax gene-expressing cell line, ATL-48T. Eight DNA regions, which were represented as bars in A, were amplified with sodium bisulfite treated DNA. The PCR products were subcloned into plasmid DNA, and then the sequences of each clone were determined for at least ten clones of each region. Arrowheads indicate the CpG sites that were target sites for COBRA. Closed circle indicates methylated CpG, and open circle means unmethylated CpG. The number of integrated provirus has been shown in parenthesis. B. Representative data of COBRA has been shown. PCR products, which were amplified with sodium bisulfite treated DNAs, were digested with TaqI or AccII. The extent of methylation in each CpG site was measured as described in Methods, and presented as percentages of methylated CpG. The number in parenthesis represents the position of cytidine residue in analyzed CpG site by COBRA according to Seiki et al. [41]. C. DNA methylation studied by COBRA at eight points in the provirus as shown by arrowheads. Each bar represented the extent of DNA methylation at the points shown by arrowhead. The analyses by COBRA were performed three times independently, and the extents of DNA methylation are shown by the mean ± SD. The number in parenthesis shows the position of cytidine residue of CpG site analyzed by COBRA.
DNA methylation throughout the HTLV-I provirus in HTLV-l-transformed and ATL cell lines
Using the COBRA method, we analyzed the DNA methylation throughout the whole HTLV-I provirus of the cell lines (Fig. 2B and 2C). In addition, we also analyzed the tax gene transcription by RT-PCR (Fig. 2A) and the number of integrated HTLV-I proviruses in each cell lines by Southern blot method. Among the tax gene-expressing cell lines (ATL-35T, MT-2, Sez627, MT-4, ATL-55T, MT-1 ATL-48T and ATL-2) (Fig. 2A), internal sequences from gag to pX were variably methylated. However, 5'-LTR was not methylated or partially methylated, while 3'-LTR was not methylated in all cell lines (Fig. 2B). In ATL-43T and TL-Oml, which did not show tax gene transcription (Fig. 2A), 5'-LTR and the internal sequences were heavily methylated (Fig. 2C), indicating the close correlation between the extents of DNA methylation of the provirus, particularly 5'-LTR, and tax gene transcription. As previously reported, the treatment by 5-aza-deoxy-cytidine can recover the tax gene expression of these cell lines, indicating that the latent state by DNA methylation of 5'-LTR is reversible [13].
Figure 2 DNA methylation in ATL cell lines, HTLV-I carriers and ATL cases. The tax gene transcription in ATL cell lines was studied by RT-PCR (A), and the expression of GAPDH gene has been used as a control. DNA methylation throughout the HTLV-I provirus was studied by COBRA in tax gene-expressing (B) and non-expressing cell lines (C). Furthermore, DNA methylation was also analyzed in 20 carriers and 20 ATL cases by COBRA, and representative patterns of DNA methylation are shown in D. The number of HTLV-I provirus has been analyzed by Southern blot method, and shown in the parenthesis (B, C and D). Each bar indicates the extent of DNA methylation that was calculated by COBRA.
Among cell lines, HTLV-I provirus tends to be not so methylated in cell lines with higher copy number of provirus (Fig. 2). The finding that cell lines with higher integrated provirus number contain hypomethylated provirus is speculated to reflect the higher transcription of viral genes.
DNA methylation of the HTLV-I provirus in ATL and HTLV-I carrier states
Next, we analyzed the DNA methylation of the whole HTLV-I provirus in ATL patients and HTLV-I carriers. Although 5'-LTR is frequently deleted in ATL cells [10], we omitted such ATL cases lacking 5'-LTR in this study. In Fig. 2D, we showed the representative pattern of DNA methylation of whole HTLV-I provirus in carriers and ATL patients. In ATL samples, the gag, pol and env regions were heavily methylated, whereas 5'-LTR was not methylated or partially methylated (Fig. 2D and 3A). On the other hand, 5'-LTR was scarcely methylated and the gag, pol and env regions seemed to be less methylated in HTLV-I carriers (Fig. 2D and 3A). We compared DNA methylation of these different eight regions between 20 carriers and 20 ATL cases (Fig. 3A). These differences in DNA methylation were statistically significant in the gag, pol and env regions between the ATL cases and HTLV-I carriers by the Mann-Whitney's U-test. These data suggested that DNA methylation initially occurred in the gag, pol, and env regions, and that DNA methylation of the provirus accumulated during disease progression from the carrier state to the leukemic stage. The frequency of DNA methylation of 5'-LTR did not differ between carriers and ATL patients. However, the extent of DNA methylation among methylation-positive cases was higher in ATL cases than in carriers (p = 0.001). Among ATL cases, the relationship between DNA methylation of 5'-LTR and tax gene transcription was analyzed (Fig. 3B), and the transcript was detected in six cases. In four cases with relative higher amount of tax gene transcripts (Case 1, 9, 12, 20), 5'-LTR was not methylated or slightly methylated. This finding suggests that higher expression of tax gene is associated with unmethyalted or slightly methylated 5'LTR, however, other mechanism(s) silences the tax gene transcription in ATL cells. There is no statistical correlation between the tax gene transcription and DNA methylation of 5'-LTR
Figure 3 Comparison of the DNA methylation in carriers and ATL cases. A. DNA methylation at eight different regions in the HTLV-I provirus was compared between carriers (C) and ATL cases (A). DNA methylation was quantified by COBRA in 20 carriers and 20 ATL cases. Each sample was analyzed three times by COBRA at each site, and circles indicate mean values of DNA methylation. The differences of DNA methylation are statistically significant in the gag, pol and env regions by the Mann-Whitney's U-test. Horizontal bars represent median of DNA methylation in each group. B. The relation between tax gene transcription and DNA methylation of 5'-LTR in the fresh ATL cells has been shown. DNA methylation of 5'-LTR was quantified by COBRA assay and the tax gene transcripts were detected by RT-PCR.
DNA methylation of HTLV-I provirus after seroconversion
The analyses of DNA methylation suggest that it first occurs around the gag, pol and env regions, and then progresses in both the 5' and 3' regions. To study the changes in DNA methylation after infection, we analyzed sequential DNA samples from seroconverters. As shown in Fig. 4A, DNA methylation already existed in the gag, pol and env regions at the seroconversion. In seroconverter 1, DNA methylation was slightly increased at 4 and 13 years after the seroconversion. Increase of DNA methylation at pol region (4187) is statistically significant 13 years later in seroconverter 1 (p = 0.02, by a Student's t-test). On the other hand, there was little change in the DNA methylation in seroconverter 2, although the HTLV-I provirus was already heavily methylated at the seroconversion. When DNA methylation of seroconverters was compared with that in carriers (Fig. 3A), provirus of carriers was more methylated in carriers than that of seroconverters (p < 0.01 by a Student's t-test) except for pol2 in seroconverter 2, and pX region. It suggests that DNA methylation of provirus accumulates during a latent period after seroconversion.
Figure 4 Sequential analyses of the DNA methylation in seroconverters and a cell line. DNA methylation was analyzed by COBRA in sequential samples from seroconverters (A) and in a cell line, ATL-21C, (B) cultured in vitro for more than 9 years. DNA methylation was analysed by COBRA three times, and each bar indicates mean ± SD.
We established an HTLV-I-transformed cell line, ATL-21C, and cultured for over 9 years in vitro, and analyzed the DNA methylation of the HTLV-I provirus. Slight DNA methylation was detected in the pol, env and pX regions at 6 months after culture, however, it did not increase after 9 years. This indicates that the DNA methylation of HTLV-I provirus did not change after long-term in vitro culture (Fig. 4B). On the other hand, the p16 gene in this cell line was not methylated at 6 months after culture, but heavily methylated after 9 years (data not shown). A comparison with the data from the seroconverters suggests that DNA methylation of the HTLV-I provirus tends to accumulate in vivo.
Association with DNA methylation in the neighboring host genome
It is possible that the HTLV-I provirus integrated into the heterochromatin or hypermethylated regions tends to be silenced [18], and that such HTLV-I-infected cells are selected in vivo. Therefore, we analyzed the DNA methylation of the host genome around the integration sites of the HTLV-I provirus. We first determined the integration sites of the HTLV-I provirus in ATL cells, and then analyzed the DNA methylation of genomic DNAs around the integration sites in both ATL cells and normal PBMCs from a non-carrier donor. When genomic DNAs neighboring integration sites were heavily methylated (Fig. 5), 5'-LTR was not methylated in three cases (acute ATL 1, 2 and 21) while they were methylated in two cases (acute ATL 3 and chronic ATL 1). In acute ATL 22, both genomic DNA and 5'-LTR were not so methylated. Thus, DNA methylation in the neighboring genomic regions was not correlated with the methylation status of the provirus among these cases.
Figure 5 DNA methylation of provirus is not associated with methylated CpG sites in the genome. Integration sites of HTLV-I provirus in leukemic cells have been determined by inverse PCR, and then DNA methylation in genome has been analyzed by sodium bisulfite sequencing. DNA methylation of 5'-LTR was also analyzed by sodium bisulfite sequencing method. Vertical bars represent CpG sites. Open circle indicates unmethylated CpG site, and closed one means methylated CpG site. N: normal PBMCs from non-carrier donor.
Histone modification of the HTLV-I provirus
It has been demonstrated that DNA methylation of 5'-LTR is associated with histone deacetylation and silencing of viral gene transcription in cell lines [13]. When ATL-43T, in which tax gene transcription was silenced by hypermethylation of 5'-LTR, was compared with a tax gene-expressing cell line, ATL-48T, a difference was found in the acetylation of histone H3 in 5'-LTR (Fig. 6A and 6B). The histone H3 of 5'-LTR was hypoacetylated in ATL-43T compared with ATL-48T, whereas there were no differences in pX or 3'-LTR among these cell lines. Since the number of HTLV-I provirus in ATL-43T and -48T is one and two copies respectively, and acetylation of histone H3 in pX and 3'-LTR was similar in both cell lines, the number of provirus was thought to have no influence on the results of ChIP assay in 5'-LTR.
Figure 6 Histone modifications in ATL cell lines. Acetylation of histone was analyzed in tax gene-expressing (ATL-48T) and non-expressing (ATL-43T) cell lines by ChIP assays with anti-acetyl-Histone H3 or H4 (A and B) at four different regions (for 5'-LTR, env, pX and 3'-LTR) of the provirus. Representative data has been shown in A. W.C.E.: whole cell extract. ChIP assay was performed three times and quantified as described in Methods. Values are means ± SD(B). *:p < 0.002.
However, the tax gene transcription is silenced in about 20% of ATL cases despite no or partial methylation of 5'-LTR (Fig. 3B) [13], suggesting that there is aother mechanism(s) for suppressing viral gene transcription. To address this question, we studied the histone modification of 5'-LTR in fresh ATL cells with or without tax gene transcription. In a case with tax gene expression, 5'-LTR was not methylated and histone H3 was hyperacetylated (Fig. 7A, Case 1). On the other hand, in Case 2 with heavily methylated 5'-LTR, histone H3 was hypoacetylated in 5'-LTR, which was consistent with the lack of detection of tax gene transcription in this case. However, in Case 3, tax gene transcription could not be detected regardless of 5'-LTR hyperacetylation. After in vitro culture, such cells showed tax gene transcription within one hour (Fig. 7B). Although both Cases 1 and 3 exhibited hyperacetylation of 5'-LTR, tax gene transcription was silenced in Case 3.
Figure 7 DNA methylation and histone modifications in fresh ATL cases. A. The relationships among DNA methylation, tax gene expression and histone modification in 5'-LTR were analyzed in three ATL cases. Cases 1 and 3 have one copy of the complete HTLV-I provirus, while Case 2 has a defective provirus that lacks part of the pol gene. DNA methylation was analyzed by COBRA. The tax gene transcripts could be detected in Case 1, but not in Cases 2 or 3, by RT-PCR. ChIP assays were also performed using primers for 5'-LTR to analyze acetylation of histone H3 (Ac-H3) and H4 (Ac-H4). W.C.E.: whole cell extract. B. Recovery of tax gene expression ex vivo. The PBMCs isolated from Case 3 were immediately cultured ex vivo for several hours and tested the transcription of tax mRNA by RT-PCR.
Discussion
DNA methylation is regarded as a host defense mechanism for inactivating transportable elements such as retroviruses to inhibit viral transcription and the generation of new viruses. On the other hand, it also renders the provirus into a latent state, resulting in the establishment of latent infection. However, it remained unclear how and when the provirus was methylated, and whether DNA methylation changed in vivo.
Tax has the remarkable potency to promote the proliferation of infected cells [3], however, it is also a major target of CTL in vivo [8]. Therefore, HTLV-I controls tax gene expression by own viral proteins, Rex [19], p30 [20,21] and HBZ [22]. In the leukemic cells, several mechanisms have been identified to suppress or abolish Tax expression, including genetic changes of tax gene, deletion of 5'-LTR, and DNA methylation of 5'-LTR. In this study, DNA methylation was shown to occur in internal provirus sequences, such as the gag, pol and env regions, and then extend to 5' (5'-LTR) and 3' (pX) regions. Since DNA methylation of 5'-LTR is associated with tax gene transcription, the finding that 5'-LTR was more highly methylated in ATL cells than in carriers, among cases with methylated 5'-LTR, suggests that such HTLV-I-infected cells and ATL cells with the methylated provirus, which produce lower amounts of viral proteins, are selected in vivo by the host immune system. In this regard, HTLV-I is quite different from another human retrovirus, HIV-1. HIV-1 vectors were resistant to gene silencing in vivo [23,24]. It is noteworthy that the number of CpG sites in the U3 region of HIV-1 LTR (9 sites in LTR of NL43) is much fewer than that of HTLV-I (47 sites in LTR of ATK). This is consistent to the previous report that transcriptional suppression was not associated with DNA methylation of HIV-1 provirus [25]. In addition, HIV-1 provirus is frequently integrated within transcriptional units, which encode the genes that are transcribed in T-cells [15,26]. In such regions, it is possible that HIV-1 tends to escape from transcriptional silencing that is observed in the heterochromatin region such as alphoid repetitive sequences [18]. These data suggest that HIV-1 is more resistant to gene silencing than HTLV-I. Alternatively, it is possible that HTLV-I takes advantage of susceptibility to DNA methylation to escape from the host immune system.
This study shows that 3'-LTR is unmethylated in carriers and ATL cells while 5'-LTR is methylated in about half of cases. In HTLV-I, HTLV-I bZIP (HBZ) gene is encoded by minus strand of provirus [22,27]. We observed that HBZ gene was transcribed in all ATL cells, suggesting that HBZ gene play a critical role in growth of HTLV-I infected cells and ATL cells (submitted for publication). The finding that 3'-LTR is unmethylated in all ATL cases and carriers suggests that HBZ gene transcription is important for proliferation of ATL and HTLV-I infected cells.
Why does DNA methylation occur from the internal sequences of the HTLV-I provirus? Since CpG island is recognized as DNA region that is susceptible to DNA methylation, we analyzed HTLV-I provirus by the criterion by Takai and Jones [28]. CpG islands are present throughout the provirus in 5'-LTR-gag (1–1360), pol (3876–4509), env (5648–6166), env-pX (6446–7561), and pX-3'-LTR (8212–9045) regions. Therefore, the presence of CpG island could not explain why DNA methylation occurred in the internal region of HTLV-I provirus. Among tumor-suppressor genes, which are transcriptionally silenced by DNA methylation, the exon regions are first methylated, and then DNA methylation progresses to the promoter region [29]. When the promoter region is heavily methylated, the transcription of the corresponding gene is silenced. Since 5'-LTR is the promoter/enhancer for viral gene transcription, there might be a similar scenario between the exon/promoter and DNA methylation in both virus and tumor-suppressor genes. Thus, it is possible that gene coding regions are first methylated and DNA methylation spreads to the promoter region of provirus, 5'-LTR.
Transcriptional silencing of tax gene in spite of hyperacetylated histone H3 is recognized as another mechanism to suppress the viral gene transcription in addition to DNA methylation. The prompt recovery of tax gene expression after in vitro culture suggests the presence of an inhibitory factor(s) that binds to 5'-LTR, and suppresses the viral gene transcription in vivo. It is noteworthy that this phenotype is very similar to that of a mouse T-cell line transfected with an HTLV-I LTR-derived reporter plasmid [30]. In that study, a green fluorescent protein-fused Tax (Gax) gene was transfected into a mouse T-cell line, EL-4, and the transduced cells were then injected into Tax-immunized and non-immunized mice. Although Tax-induced cytotoxic T-cells suppressed the expression of the Gax gene in vivo, its expression was shown to recover within three hours when the transduced cells were transferred to in vitro culture. This phenotype resembles that observed in Case 3 in Fig. 7. Considering that Tax is the major target of CTL in vivo, and at the same time, confers growth advantages on the infected cells, such reversible suppression of tax gene expression is thought to be suitable for the survival of HTLV-I infected cells, and ATL cells. In this regard, potentiation of anti-Tax immunity might protect against the development of ATL when combined with possible therapeutics to induce Tax expression [31]. For this purpose, the mechanism for silencing viral transcription regardless of histone H3 hyperacetylation should be studied.
In general, gene silencing is associated with several different mechanisms. DNA methylation in the promoter region silences the gene transcription, whereas gene silencing is often not associated with DNA methylation [32,33]. In such situations, methylation of H3K9 is linked with loss of transcriptions [34]. It is possible that silencing of viral gene transcription renders proviral DNA vulnerable to methylation. Once proviral DNA is methylated, such silencing would be fixed unless such cells are treated with demethylating agents such as 5-aza-deoxy-cytidine.
DNA methylation of the HTLV-I provirus did not accumulate in a cell line that was cultured in vitro for more than 9 years. The finding that the p16 gene was heavily methylated in this cell line excluded the possibility that hypermethylation did not occur in this cell line due to aberrant methylation machinery. Among the seroconverters, the provirus was heavily methylated in internal regions such as gag, pol and env. Taken together, DNA methylation in the provirus is considered to reflect the selection in vivo. Since the growth of in vitro HTLV-I-transformed cell lines depends on Tax expression, cells with suppressed expression of the tax gene do not have the growth advantage in vitro. However, the immune system exerts selection of the infected cells with suppressed tax gene expression in vivo.
Recently, both 5'- and 3'-LTR have been reported to be transcriptionally active, and transcriptional factors and Tax bind equally to both [35]. 3'-LTR may activate the transcription of cellular genes, which are located in the downstream of integration sites. In addition, unmethylated 3'-LTR is critical for transcription of the HBZ gene. Since 5'-LTR is a promoter/enhancer for viral gene transcription, selective methylation of 5'-LTR is considered to silence the transcription of viral genes.
Conclusion
We have demonstrated how DNA methylation of HTLV-I provirus occurred, and how it suppressed viral gene transcription. When 5'-LTR was heavily methylated, viral transcription was silenced, which is thought to reflect the immune system selection in vivo. In addition, mechanisms other than DNA methylation suppresses viral gene transcription regardless of histone H3 hyperacetylation. The mechanism of such suppression requires further investigation.
Methods
Cells
HTLV-I-associated cell lines (MT-1, MT-2, MT-4, ATL-2, TL-Oml and Sez627) were cultured in RPMI1640 medium supplemented with 10% fetal bovine serum and penicillin/streptomycin. For interleukin-2-dependent cell lines (ATL-43T, 48T and 55T), 100 U/ml of recombinant interleukin-2 (Shionogi, Osaka) was added to the medium. Peripheral blood mononuclear cells (PBMC) or lymph node cells were isolated from HTLV-I carriers and ATL patients after informed consent was obtained. The polyclonal integration of HTLV-I provirus in carriers has been shown by inverse PCR [36], and provirus load was determined by real-time PCR as reported previously [37].
Sodium bisulfite treatment of genomic DNA
Sodium bisulfite treatment was performed as described previously [29]. Briefly, 1–3 μg of genomic DNA was denatured in 0.3 N NaOH at 37°C for 15 min, and 1 μg of salmon sperm DNA was added to each sample as a carrier. Sodium bisulfite (pH 5.0) and hydroquinone were added to each sample to final concentrations of 3 M and 0.05 mM, respectively. The reaction was performed at 55°C for 16 h and the samples were then desalted using the Wizard DNA Clean-Up System (Promega, Madison, WI). Finally, samples were desulfonated in 0.3 N NaOH at 37°C for 15 min.
Sequencing of sodium bisulfite-treated genomic DNA
The sodium bisulfite-treated DNA (200–500 ng) was used as a template for PCR amplification of eight HTLV-I provirus regions. The PCR reactions were performed using FastStart Taq DNA Polymerase (Roche, Mannheim, Germany). The PCR primer pairs and annealing temperatures are shown in Table 1. The amplified PCR products were purified and subcloned into pGEM-T Easy vectors (Promega). For each region, at least 10 clones were sequenced using Big Dye Terminator v3.1 Cycle Sequencing Kit (Applied BioSystems, Foster City, CA) and ABI3100 autosequencer (Applied Biosystems).
Table 1 Primer sets for COBRA and ChIP assay
Site in HTLV-Ia Forward primer Reverse primer Anneal (°C) Enzyme for COBRA
COBRA 620 1st 5'-TTTGGAGTTTATTTAGATTTAG-3' 5'-CCAATAATAAACRACCAACCC-3' 45 TaqI
(5'-LTR) 2nd 5'-GTTTTGTTTGATTTTGTTTGT-3' 5'-AAAAAAATTTAACCCATTACC-3' 49
1753 1st 5'-GGGAGTGTTAAAGATTTTTTTTGGG-3' 5'-ACTCCAATAACCTACTTTCCC-3' 55 TaqI
(gag) 2nd 5'-TTTATTTTTTAAGGTTTGGAGGAG-3' 5'-TTAAAAATCCAAATCTAACAAACCC-3' 55
2988 1st 5'-GTTAAAAAGGTTAATGGAATTTGG-3' 5'-CCTCTAAAAATAATAATAAATCCTC-3' 52 TaqI
(pol) 2nd 5'-GGGTTTTTTGATTTGTTTAGTTTG-3' 5'-AAACTTACTAAAAAAATATCATCC-3' 51
4187 1st 5'-GGGTGAAATTGTGTAGTTTTGTAGG-3' 5'-CCTATTTTCAAACGAATCTACCTCC-3' 57 AccII
(pol) 2nd 5'-GTGATTAGTAGGGTATTTGTGAGAG-3' 5'-ATTATCACAAAAATCATTCCCCC-3' 52
5151 1st 5'-GGTATTATTTTAAGTTTTTTGG-3' 5'-CTCCAATTATAAAAATACAACAAC-3' 46 TaqI
(pol) 2nd 5'-GTTAGTGGAAAGGATTATAGGAGG-3' 5'-AACTTACCCATAATATTAAAAATC-3' 51
6113 1st 5'-GGATTTATTGTTTTGATTTTTAG-3' 5'-CTTTACATAATCCTCCTTACTCCC-3' 51 TaqI
(env) 2nd 5'-GGATTTATTGTTTTGATTTTTAG-3' 5'-CCCAAAACAAAAAATCAAAACC-3' 53
7258 1st 5'-GAGGTGGYGTTTTTTTTTTTGG-3' 5'-CCTTAAAAATCTTAAAAATTCTC-3' 47 TaqI
(pX) 2nd 5'-AAGGATAGTAAATYGTTAAGTATAG-3' 5'-CCCAAATAATCTAATACTCTAAAC-3' 50
8342 1st 5'-YGATGGTAYGTTTATGATTTTYGGG-3' 5'-ACCCCCTCCTAAACTATCTCC-3' 57 TaqI
(3'-LTR) 2nd 5'-YGATGGTAYGTTTATGATTTTYGGG-3' 5'-AACTCCTACTAATTTATTAAACC-3' 52
5'-LTRb 5'-GCTTTGCCTGACCCTGCTTGC-3' 5'-AAGATTTGGCCCATTGCCTAGGG-3' 63
env 5'-TGCCAGCCTCTCCACTTGGCACG-3' 5'-ATGGAGCCGGTAATCCCGCCAGC-3' 64
pX 5'-AAGGATAGCAAACCGTCAAGCACAG-3' 5'-CCCAGGTGATCTGATGCTCTGGAC-3' 63
3'-LTR 5'-CCCCTCATTTCTACTCTCACACGGC-3' 5'-TGGGTGGTTCTTGGTGGCTTCCC-3' 64
a Nucleotide position corresponding to that of ATK. This number means the cytidine of CpG sites analysed.
bFor ChIP assay, we used primers to amplify the indicated regions.
Combined bisulfite restriction analysis (COBRA)
For COBRA, eight different regions of HTLV-I provirus were amplified with sodium bisulfite treated genomic DNAs using each primer sets as shown in Table 1. The nested PCR reactions were performed using FastStart Taq DNA Polymerase (Roche) with the following condition: 5 minutes at 95°C for denaturation, 40 cycles of 30 sec at 95°C, 30 sec at each annealing temperature (Table 1), 30 sec at 72°C, and 2 min at 72°C for final extension. The PCR products were digested for at least 4 hrs with an appropriate restriction enzyme (TaqI or AccII) that had a single recognition site within each product [38]. When CpG site within amplified region was methylated, it was resistant to sodium bisulfite treatment, resulting in digestion by these enzymes. On the other hand, since unmethylated CpG was converted to UG by sodium bisulfite treatment, these enzymes could not digest the amplified DNAs. The digested PCR products were separated in a 3% Nusieve 3:1 agarose (BMA, Rockland, ME) gel. The intensity of each fragment was determined using ATTO Densitograph Ver. 4.0 (ATTO, Tokyo, Japan), and the extent of DNA methylation was calculated as follows: % methylation = 100 × (digested PCR products/undigested+ digested PCR products).
Southern blot analyses
To determine the number of integrated HTLV-I provirus, we performed Southern blot method using HTLV-I probe as described previously [10]. In brief, 5 μg of DNA were digested with EcoRI, separated by electrophoresis in a 0.7% agarose gel, and transferred to nylon membrane (Hybond N+, Amersham Biosciences, Piscataway, NJ). The membrane was hybridized to the alkaline phospatase labeled pX probes. 0.9 kb PCR product of HTLV-I pX region derived from HTLV-I clone λ23-3 was used as probe [39]. DNA probe was labeled, and hybridized to the membrane with Gene Images AlkPhos Direct Labelling and Detection system (Amersham Biosciences).
Inverse-long PCR
To check the HTLV-I integration in PBMCs of carriers, we analyzed the genomic DNAs from carriers by inverse-long PCR method as described previously [36]. In brief, genomic DNA was digested with EcoRI, and then ligated with T4 DNA ligase. Circularized DNA was digested with MluI that cut the provirus at pX region to prevent amplification of provirus itself. Then, treated genomic DNA was amplified with primers as follows: Long-IPCR-F: 5'-TGCCTGACCCTGCTTGCTCAACTCTACGTCTTTG-3', Long-IPCR-R 5'-AGTCTGGGCCCTGACCTTTTCAGACTTCTGTTTC-3'. PCR condition was as follows: 2 min at 98°C for denaturation, 5 cycles (30 sec at 98°C, 10 min at 64°C), followed by 35cycles (30 sec at 94°C, 10 min at 64°C) and 15 min at 72°C for final extension. The PCR products were subcloned into plasmid DNA and their sequences were determined.
DNA methylation in neighboring regions of HTLV-I integration sites
The integration sites of HTLV-I provirus has been determined by inverse long PCR, and DNA methylation of genomic DNAs neighboring integration sites was determined in both ATL cells and PBMCs. The nested PCR reactions were performed using FastStart Taq DNA Polymerase (Roche) with the following condition: 5 minutes at 95°C for denaturation, 40 cycles of 30 sec at 95°C, 30 sec at each annealing temperature (Table 2), 30 sec at 72°C, and 2 min at 72°C for final extension.
Table 2 Primer sets and annealing temperatures for genome specific PCR
Case Locus Forward primer Reverse primer Anneal (°C)
Primers for case Acute ATL 1 5q11.1 1st 5'-TTTGGAGAGGGAATTTTATATTG-3' 5'-ACCCCCTCCTAAACTATCTCC-3' 55
2nd 5'-GGAGTGTAGAGATGTAGTTTTGG-3' 5'-ACCCCCTCCTAAACTATCTCC-3' 50
Acute ATL 2 8p23.1 1st 5'-GAGAAATTTGTGTTGATTTTATTAG-3' 5'-ACCCCCTCCTAAACTATCTCC-3' 47
2nd 5'-TTAGTGGTAGATTAAGTTAAAG-3' 5'-ACCCCCTCCTAAACTATCTCC-3' 45
Acute ATL 3 1q31.1 1st 5'-GGTAGAAATTATAGGTTTTTGTAGG-3' 5'-ACCCCCTCCTAAACTATCTCC-3' 51
2nd 5'-GTTATTTGTGAAGTAAGATGTTTTG-3' 5'-ACCCCCTCCTAAACTATCTCC-3' 53
Acute ATL 21 15q24.3 1st 5'-GAGGTGGATTTTTATTTTATTG-3' 5'-ACCCCCTCCTAAACTATCTCC-3' 52
2nd 5'-GGTTTTTGATTATATTTGGGGAG-3' 5'-ACCCCCTCCTAAACTATCTCC-3' 54
Acute ATL 22 19q13.11 1st 5'-GTTAGTTGTTAGAGAGTTTTTTGG-3' 5'-ACCCCCTCCTAAACTATCTCC-3' 52
2nd 5'-AAGATTATTTAGTTTTTTGGGG-3' 5'-ACCCCCTCCTAAACTATCTCC-3' 54
Chronic ATL 1 1p22.1 1st 5'-GGGTTTGAAGTTTTTTTTGTAGG-3' 5'-ACCCCCTCCTAAACTATCTCC-3' 53
2nd 5'-AAGATTATTTAGTTTTTTGGGG-3' 5'-ACCCCCTCCTAAACTATCTCC-3' (5'-LTR U3) 50
Primers for human genome 5q11.1 1st 5'-TTTGGAGAGGGAATTTTATATTG-3' 5'-CCCAAACTAATCTTCAACTCC-3' 52
2nd 5'-GGAGTGTAGAGATGTAGTTTTGG-3' 5'-CCACCATAAAAAACCCTCCC-3' 54
8p23.1 1st 5'-GAGAAATTTGTGTTGATTTTATTAG-3' 5'-AATATCACTATAACAATAACCAC-3' 46
2nd 5'-TTAGTGGTAGATTAAGTTAAAG-3' 5'-CTCTCAACAAATTCCATCTTTCC-3' 49
1q31.1 1st 5'-GGTAGAAATTATAGGTTTTTGTAGG-3' 5'-CACCATTAAACAAACTAAATTCTC-3' 51
2nd 5'-GTTATTTGTGAAGTAAGATGTTTTG-3' 5'-CACATAAAAAAACCCACACAATC-3' 53
15q24.3 1st 5'-GAGGTGGATTTTTATTTTATTG-3' 5'-ATCTACCTAAAAAACCCACCC-3' 52
2nd 5'-GGTTTTTGATTATATTTGGGGAG-3' 5'-AAAAACCCACCCAAACAAACC-3' 57
19q13.11 1st 5'-GTTAGTTGTTAGAGAGTTTTTTGG-3' 5'-CAACTCCCTAAACCCTCCTCC-3' 52
2nd 5'-GTTTTTTGGTTAAGGTTATGGG-3' 5'-CTCCTACCACGAACCTACTCC-3' 54
1p22.1 1st 5'-GGGTTTGAAGTTTTTTTTGTAGG-3' 5'-CAACAAAAACAATAAACAAAACC-3' 54
2nd 5'-AAGATTATTTAGTTTTTTGGGG-3' 5'-CTTTACACCAATAAATTTAATACC-3' 50
RT-PCR
Total RNA was isolated from PBMCs or lymph node cells using TRIzol Reagent (Invitrogen, Carlsbad, CA) and RT-PCR was performed using RNA LA PCR Kit (AMV) Ver. 1.1 (Takara Bio Inc., Otsu, Japan) according to the manufacturer's protocol. The tax and GAPDH gene transcripts were amplified using the following primers: RPX2 5'-CCGGCGCTGCTCTCATCCCGGT-3' and RPX5 5'-GGCCGAACATAGTCCCCCAGAG-3' (for tax), GAPDH1 5'-ATGGGGAAGGTGAAGGTCGGAGTC-3' and GAPDH1a 5'-CCATGCCAGTGAGCTTCCCGTTC-3' (for GAPDH) under following conditions: 2 minutes at 95°C for denaturation, 35 cycles of 30 sec at 95°C, 30 sec at 62°C, 30 sec at 72°C (for tax), 25 cycles of 30 sec at 95°C, 30 sec at 55°C, 30 sec at 72°C (for GAPDH) and 2 min at 72°C for final extension.
Chromatin immunoprecipitation (ChIP) assay
ChIP assays were performed as described previously [40]. Briefly, ATL cell lines and fresh ATL cells from ATL patients (5 × l05 cells/antibody) were fixed with formaldehyde and then sonicated to obtain soluble chromatin. The chromatin solutions were immunoprecipitated with anti-acetyl-Histone H3 or anti-acetyl-Histone H4 (Upstate Biotechnology), or normal rabbit IgG, overnight at 4°C, and the immunoprecipitates were then collected with 50% protein A and G-Sepharose slurry preabsorbed with 0.1 mg/ml sonicated salmon sperm DNA. The resulting purified DNAs were subjected to PCR reactions using primer sets specific for 5'-LTR, env, pX and 3'-LTR. The sequences of the primers are shown in Table 1. To distinguish 5' and 3'-LTR, we used primers specific for gag and R region of LTR for amplification of 5'-LTR, and primers for pX region and U3 region were used for amplification of 3'-LTR. The PCR reactions were performed using FastStart Taq DNA Polymerase (Roche) with the following condition: 5 minutes at 95°C, 35 or 37 cycles of 30 sec at 95°C, 30 sec at each annealing temperature (Table 1), 30 sec at 72°C, and 2 min at 72°C. The PCR products were electrophoresed in an agarose gel and the results were analyzed using ATTO Densitograph Ver. 4.0. Values were calculated as the signal intensity of each sample normalized by that of the whole cell extract.
Statistical analyses
Statistical analyses were performed using the Mann-Whitney's U-test and Student's t-test.
Competing interests
The author(s) declare that they have no competing interests.
Authors' contributions
YT conceived this project and carriers out most of experiments in Figs. 1, 2, 3, 5 and 6. KN established COBRA assay and performed experiments in Figs. 1 and 2. JY performed experiments in Fig. 7. MM established most of HTLV-I transformed cell lines, and analyzed experiments in Fig. 4. AO and NM provided sequential DNA samples from seroconverters, and analyzed the data. M.Matsuoka directed and supervised the experiments and interpretations All authors read and approved the final manuscript.
Acknowledgements
We thank Shinjiro Hino for valuable suggestions.
==== Refs
Takatsuki K Yamaguchi K Matsuoka M Takatsuki K ATL and HTLV-I-related diseases Adult T-cell leukemia 1994 New York: Oxford University Press 1 27
Matsuoka M Human T-cell leukemia virus type I and adult T-cell leukemia Oncogene 2003 22 5131 5140 12910250 10.1038/sj.onc.1206551
Yoshida M Multiple viral strategies of HTLV-1 for dysregulation of cell growth control Annu Rev Immunol 2001 19 475 496 11244044 10.1146/annurev.immunol.19.1.475
Franchini G Fukumoto R Fullen JR T-cell control by human T-cell leukemia/lymphoma virus type 1 Int J Hematol 2003 78 280 296 14686485
Jin DY Spencer F Jeang KT Human T cell leukemia virus type 1 oncoprotein Tax targets the human mitotic checkpoint protein MAD1 Cell 1998 93 81 91 9546394 10.1016/S0092-8674(00)81148-4
Jin DY Giordano V Kibler KV Nakano H Jeang KT Role of adapter function in oncoprotein-mediated activation of NF-kappaB. Human T-cell leukemia virus type I Tax interacts directly with IkappaB kinase gamma J Biol Chem 1999 274 17402 17405 10364167 10.1074/jbc.274.25.17402
Jacobson S Shida H McFarlin DE Fauci AS Koenig S Circulating CD8+ cytotoxic T lymphocytes specific for HTLV-I pX in patients with HTLV-I associated neurological disease Nature 1990 348 245 248 2146511 10.1038/348245a0
Bangham CR Human T-lymphotropic virus type 1 (HTLV-1): persistence and immune control Int J Hematol 2003 78 297 303 14686486
Kannagi M Ohashi T Harashima N Hanabuchi S Hasegawa A Immunological risks of adult T-cell leukemia at primary HTLV-I infection Trends Microbiol 2004 12 346 352 15223062 10.1016/j.tim.2004.05.005
Tamiya S Matsuoka M Etoh K Watanabe T Kamihira S Yamaguchi K Takatsuki K Two types of defective human T-lymphotropic virus type I provirus in adult T-cell leukemia Blood 1996 88 3065 3073 8874205
Furukawa Y Kubota R Tara M Izumo S Osame M Existence of escape mutant in HTLV-I tax during the development of adult T-cell leukemia Blood 2001 97 987 993 11159527 10.1182/blood.V97.4.987
Koiwa T Hamano-Usami A Ishida T Okayama A Yamaguchi K Kamihira S Watanabe T 5'-long terminal repeat-selective CpG methylation of latent human T-cell leukemia virus type 1 provirus in vitro and in vivo J Virol 2002 76 9389 9397 12186921 10.1128/JVI.76.18.9389-9397.2002
Takeda S Maeda M Morikawa S Taniguchi Y Yasunaga J Nosaka K Tanaka Y Matsuoka M Genetic and epigenetic inactivation of tax gene in adult T-cell leukemia cells Int J Cancer 2004 109 559 567 14991578 10.1002/ijc.20007
Verma M Viral genes and methylation Ann N Y Acad Sci 2003 983 170 180 12724222
Han Y Lassen K Monie D Sedaghat AR Shimoji S Liu X Pierson TC Margolick JB Siliciano RF Siliciano JD Resting CD4+ T cells from human immunodeficiency virus type 1 (HIV-1)-infected individuals carry integrated HIV-1 genomes within actively transcribed host genes J Virol 2004 78 6122 6133 15163705 10.1128/JVI.78.12.6122-6133.2004
Doi K Wu X Taniguchi Y Yasunaga JI Satou Y Okayama A Nosaka K Matsuoka M Preferential selection of human T-cell leukemia virus type I (HTLV-I) provirus integration sites in leukemic versus carrier states Blood 2005 15840694
Seiki M Eddy R Shows TB Yoshida M Nonspecific integration of the HTLV provirus genome into adult T-cell leukaemia cells Nature 1984 309 640 642 6328324 10.1038/309640a0
Jordan A Bisgrove D Verdin E HIV reproducibly establishes a latent infection after acute infection of T cells in vitro Embo J 2003 22 1868 1877 12682019 10.1093/emboj/cdg188
Seiki M Inoue J Hidaka M Yoshida M Two cis-acting elements responsible for posttranscriptional trans-regulation of gene expression of human T-cell leukemia virus type I Proc Natl Acad Sci U S A 1988 85 7124 7128 3174625
Zhang W Nisbet JW Albrecht B Ding W Kashanchi F Bartoe JT Lairmore MD Human T-lymphotropic virus type 1 p30(II) regulates gene transcription by binding CREB binding protein/p300 J Virol 2001 75 9885 9895 11559821 10.1128/JVI.75.20.9885-9895.2001
Nicot C Dundr M Johnson JM Fullen JR Alonzo N Fukumoto R Princler GL Derse D Misteli T Franchini G HTLV-1-encoded p30II is a post-transcriptional negative regulator of viral replication Nat Med 2004 10 197 201 14730358 10.1038/nm984
Gaudray G Gachon F Basbous J Biard-Piechaczyk M Devaux C Mesnard JM The complementary strand of the human T-cell leukemia virus type 1 RNA genome encodes a bZIP transcription factor that down-regulates viral transcription J Virol 2002 76 12813 12822 12438606 10.1128/JVI.76.24.12813-12822.2002
Miyoshi H Smith KA Mosier DE Verma IM Torbett BE Transduction of human CD34+ cells that mediate long-term engraftment of NOD/SCID mice by HIV vectors Science 1999 283 682 686 9924027 10.1126/science.283.5402.682
Gatlin J Padgett A Melkus MW Kelly PF Garcia JV Long-term engraftment of nonobese diabetic/severe combined immunodeficient mice with human CD34+ cells transduced by a self-inactivating human immunodeficiency virus type 1 vector Hum Gene Ther 2001 12 1079 1089 11399229 10.1089/104303401750214294
Pion M Jordan A Biancotto A Dequiedt F Gondois-Rey F Rondeau S Vigne R Hejnar J Verdin E Hirsch I Transcriptional suppression of in vitro-integrated human immunodeficiency virus type 1 does not correlate with proviral DNA methylation J Virol 2003 77 4025 4032 12634362 10.1128/JVI.77.7.4025-4032.2003
Schroder AR Shinn P Chen H Berry C Ecker JR Bushman F HIV-1 integration in the human genome favors active genes and local hotspots Cell 2002 110 521 529 12202041 10.1016/S0092-8674(02)00864-4
Larocca D Chao LA Seto MH Brunck TK Human T-cell leukemia virus minus strand transcription in infected T-cells Biochem Biophys Res Commun 1989 163 1006 1013 2476979 10.1016/0006-291X(89)92322-X
Takai D Jones PA Comprehensive analysis of CpG islands in human chromosomes 21 and 22 Proc Natl Acad Sci U S A 2002 99 3740 3745 11891299 10.1073/pnas.052410099
Nosaka K Maeda M Tamiya S Sakai T Mitsuya H Matsuoka M Increasing methylation of the CDKN2A gene is associated with the progression of adult T-cell leukemia Cancer Res 2000 60 1043 1048 10706122
Furuta RA Sugiura K Kawakita S Inada T Ikehara S Matsuda T Fujisawa J Mouse model for the equilibration interaction between the host immune system and human T-cell leukemia virus type 1 gene expression J Virol 2002 76 2703 2713 11861837 10.1128/JVI.76.6.2703-2713.2002
Hanabuchi S Ohashi T Koya Y Kato H Hasegawa A Takemura F Masuda T Kannagi M Regression of human T-cell leukemia virus type I (HTLV-I)-associated lymphomas in a rat model: peptide-induced T-cell immunity J Natl Cancer Inst 2001 93 1775 1783 11734593 10.1093/jnci/93.23.1775
Soengas MS Capodieci P Polsky D Mora J Esteller M Opitz-Araya X McCombie R Herman JG Gerald WL Lazebnik YA Inactivation of the apoptosis effector Apaf-1 in malignant melanoma Nature 2001 409 207 211 11196646 10.1038/35051606
Yasunaga J Taniguchi Y Nosaka K Yoshida M Satou Y Sakai T Mitsuya H Matsuoka M Identification of aberrantly methylated genes in association with adult T-cell leukemia Cancer Res 2004 64 6002 6009 15342380 10.1158/0008-5472.CAN-04-1422
Bachman KE Park BH Rhee I Rajagopalan H Herman JG Baylin SB Kinzler KW Vogelstein B Histone modifications and silencing prior to DNA methylation of a tumor suppressor gene Cancer Cell 2003 3 89 95 12559178 10.1016/S1535-6108(02)00234-9
Lemasson I Polakowski NJ Laybourn PJ Nyborg JK Transcription regulatory complexes bind the human T-cell leukemia virus 5' and 3' long terminal repeats to control gene expression Mol Cell Biol 2004 24 6117 6126 15226416 10.1128/MCB.24.14.6117-6126.2004
Etoh K Tamiya S Yamaguchi K Okayama A Tsubouchi H Ideta T Mueller N Takatsuki K Matsuoka M Persistent clonal proliferation of human T-lymphotropic virus type I-infected cells in vivo Cancer Res 1997 57 4862 4867 9354450
Yasunaga J Sakai T Nosaka K Etoh K Tamiya S Koga S Mita S Uchino M Mitsuya H Matsuoka M Impaired production of naive T lymphocytes in human T-cell leukemia virus type I-infected individuals: its implications in the immunodeficient state Blood 2001 97 3177 3183 11342446 10.1182/blood.V97.10.3177
Xiong Z Laird PW COBRA: a sensitive and quantitative DNA methylation assay Nucleic Acids Res 1997 25 2532 2534 9171110 10.1093/nar/25.12.2532
Clarke MF Gelmann EP Reitz MS Jr Homology of human T-cell leukaemia virus envelope gene with class I HLA gene Nature 1983 305 60 62 6888550 10.1038/305060a0
Hino S Fan J Taguwa S Akasaka K Matsuoka M Sea urchin insulator protects lentiviral vector from silencing by maintaining active chromatin structure Gene Ther 2004 11 819 828 14985785 10.1038/sj.gt.3302227
Seiki M Hattori S Hirayama Y Yoshida M Human adult T-cell leukemia virus: complete nucleotide sequence of the provirus genome integrated in leukemia cell DNA Proc Natl Acad Sci U S A 1983 80 3618 3622 6304725
|
16242045
|
PMC1289293
|
CC BY
|
2021-01-04 16:36:39
|
no
|
Retrovirology. 2005 Oct 22; 2:64
|
utf-8
|
Retrovirology
| 2,005 |
10.1186/1742-4690-2-64
|
oa_comm
|
==== Front
J Transl MedJournal of Translational Medicine1479-5876BioMed Central London 1479-5876-3-391625577710.1186/1479-5876-3-39ResearchSynergistic inhibition of human melanoma proliferation by combination treatment with B-Raf inhibitor BAY43-9006 and mTOR inhibitor Rapamycin Molhoek Kerrington R [email protected] David L [email protected] Craig L [email protected] Department of Surgery, Division of Surgical Oncology, University of Virginia School of Medicine, Charlottesville, VA, USA2 Center for Cell Signaling, University of Virginia Health System, Charlottesville, VA, USA2005 28 10 2005 3 39 39 12 8 2005 28 10 2005 Copyright © 2005 Molhoek et al; licensee BioMed Central Ltd.2005Molhoek et al; licensee BioMed Central Ltd.This is an Open Access article distributed under the terms of the Creative Commons Attribution License (), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
Background
Targeted inhibition of protein kinases is now acknowledged as an effective approach for cancer therapy. However, targeted therapies probably have limited success because cancer cells have alternate pathways for survival and proliferation thereby avoiding inhibition. We tested the hypothesis that combination of targeted agents would be more effective than single agents in arresting melanoma cell proliferation.
Methods
We evaluated whether BAY43-9006, an inhibitor of the B-Raf kinase, and rapamycin, an inhibitor of the mTOR kinase, would inhibit serum-stimulated proliferation of human melanoma cell lines, either alone or in combination. Proliferation was measured by quantitating melanoma cell numbers with a luciferase for ATP. Phosphorylation of proteins downstream of targeted kinase(s) was assayed by immunoblots. Statistical significance was determined with the Student-T test. Isobologram analysis was performed to distinguish additive versus synergistic effects of combinations of drugs.
Results
Serum-stimulated proliferation of multiple human melanoma cell lines was inhibited by BAY43-9006 and by rapamycin. Melanoma cells containing the B-Raf mutation V599E were more sensitive than cells with wild-type B-raf to 10 nM doses of both BAY43-9006 and rapamycin. Regardless of B-Raf mutational status, the combination of low dose rapamycin and BAY43-9006 synergistically inhibited melanoma cell proliferation. As expected, rapamycin inhibited the phosphorylation of mTOR substrates, p70S6K and 4EBP1, and BAY43-9006 inhibited phosphorylation of ERK, which is dependent on B-Raf activity. We also observed unexpected rapamycin inhibition of the phosphorylation of ERK, as well as BAY43-9006 inhibition of the phosphorylation of mTOR substrates, p70S6K and 4EBP1.
Conclusion
There was synergistic inhibition of melanoma cell proliferation by the combination of rapamycin and BAY 43-9006, and unexpected inhibition of two signaling pathways by agents thought to target only one of those pathways. These results indicate that combinations of inhibitors of mTOR and of the B-raf signaling pathways may be more effective as a treatment for melanoma than use of either agent alone.
B-RafmTORmelanomaBAY43-9006rapamycin
==== Body
Background
In human cancers, mutant oncogenes are frequently associated with disease progression [1]. Thus, there is a need for development of effective therapies that can slow progression of solid tumors by blocking the action of those oncogenes. Cancer therapy has undergone a paradigm shift based on the therapeutic effectiveness of imatinib mesylate (Gleevec). This drug was designed as a specific inhibitor of the BCR-ABL oncogene protein tyrosine kinase, known to be responsible for chronic myeloid leukemia (CML) cells [2]. The therapeutic effectiveness of Gleevec and relative absence of detrimental side-effects has made it a model for the development of an array of new therapeutic agents targeted to inhibit signal transduction enzymes, especially protein kinases.
The recent discovery that 60–70% of human melanomas have activating mutations in B-Raf (with 80% of these mutations caused by a single substitution V599E) make this protein kinase an especially promising target for inhibition [3,4]. Indeed, lead compounds have been produced and tested, and currently are working their way through clinical trials. One example is BAY43-9006 (aka: sorafenib, N-(3-trifluoromethyl-4-chlorophenyl)-N'-(4-(2-methylcarbamoyl pyridin-4-yl)oxyphenyl)urea), an investigational compound, currently in phase II and III clinical trials, designed to inhibit both B-Raf and C-Raf kinases [5,6]. B-Raf is a component of a cell signaling pathway which includes the upstream activator of Raf, called Ras, and the direct substrate of Raf, called MEK1/2 and the MEK substrate ERK1/2 [7]. B-Raf phosphorylates MEK1 and MEK2 on Ser217 and Ser221, which activates it to dual phosphorylate ERKs, at Thr202/Tyr204 for human ERK1 and Thr185/Tyr187 for human ERK2 [8,9]. Mutations in RAF which cause constitutive activation of Raf kinase are thought to promote events leading to carcinogenesis. Pre-clinical and early phase I studies have suggested that BAY 43-9006 may be of therapeutic value not only in human tumors containing ras gene mutations, but also in tumors over-expressing growth factor receptors that activate the Ras/ERK pathway [10]. However, these studies have not addressed the effects of BAY 43-9006 in combination with any other kinase inhibitors.
Another molecular pathway commonly mutated (30–60%) in melanomas involves loss of the PTEN tumor suppressor gene, which can lead to constitutive activation of the mTOR kinase signaling pathway [11-13]. Inhibition of mTOR kinase is feasible with the macrolide natural product rapamycin (aka: sirolimus, RAPA, Rapamune, AY-22989, and NSC-226080). Rapamycin is an FDA-approved agent used as immunosuppressive therapy post organ transplant [14,15]. More recent clinical application of rapamycin has been with coated stents to suppress the neo-intima formation during restenosis in response to balloon angioplasty [16]. The action of rapamycin is understood to involve binding to the receptor protein FKBP12; this drug-protein complex binds to the mTOR protein kinase and interferes with phosphorylation of two well-recognized downstream targets, p70S6K (p70 ribosomal S6 kinase) and 4EBP1 (aka: 4E binding protein 1, eukaryotic translation initiation factor 4E binding protein, and PHAS-1) [17]. An appreciation of the potent inhibition of cell growth and protein synthesis, as well as cell cycle arrest, imposed by rapamycin led to testing of its derivatives, in particular CCI-779, as cytostatic agents, especially for various cancers refractory to other forms of cancer chemotherapy [18-20]. Pharmacokinetic analysis revealed that CCI-779 was progressively converted into rapamycin, its main metabolite, beginning as early as 15 minutes after infusion of the drug [20], therefore, we used rapamycin in our studies.
Our interest is in combining targeted agents for these pathways in an effort to determine if such treatments will be effective in the treatment of melanoma. We hypothesized that the combination of multiple targeted therapeutic agents would result in enhanced inhibition of melanoma cell proliferation compared to either drug alone, because of synergy between effects on two pathways. Here we show that serum-stimulated melanoma cell proliferation is inhibited by either rapamycin or BAY43-9006, with B-Raf V599E mutants showing an increased sensitivity to each drug at 10 nM compared to melanoma cells with wild-type B-Raf. Each of these drugs inhibited the serum-stimulated phosphorylation of known Raf and mTOR substrates. What was unexpected was that each of the drugs inhibited phosphorylation in both the Raf and mTOR pathways, suggesting there was interdependence or cross-talk between these pathways in melanoma cells. Furthermore, the combination of rapamycin with BAY43-9006 was synergistic compared to either drug alone at inhibiting proliferation of wild-type B-Raf and V599E mutant B-Raf melanoma cell lines.
Methods
Cell Culture
Melanoma cell lines used in this study were derived from tumors from patients at the University of Virginia (VMM5A, VMM18, and VMM39), as described previously [21,22]. All of the cell lines were cultured in RPMI 1640 medium supplemented with 5% fetal bovine serum, 2 mM L-glutamine, penicillin (100 units/ml), and streptomycin (100 μg/ml) at 37°C in 5% CO2, unless otherwise indicated. As a control, cells were incubated in Dulbecco's Phosphate buffered saline (PBS). VMM39 is a representative cell line from human melanomas known to contain a wild-type B-Raf gene and VMM18 and VMM5A both contain the V599E B-Raf activating mutation [23]. Other human melanoma cell lines listed in Table 1 include VMM12, a malignant melanoma cell line derived from tumors from a patient at the University of Virginia which is known to contain the V599E B-Raf activating mutation [23]. DM122 is a melanoma cell line derived from tumors from a patient at Duke University, and is known to contain a wild-type B-Raf gene (data not shown). DM6 and DM331 are melanoma cell lines derived from tumors obtained from patients at Duke University, however, their B-Raf status remains to be determined.
Table 1 Melanoma cell proliferation is stimulated by serum.
Cell Line
VMM18 2.0
DM6 2.3
VMM39 2.1
VMM5A 2.0
VMM12 2.4
DM331 1.7
DM122 2.0
Reagents and Inhibitors
The MEK1/2 inhibitor U0126 (Catalog # 662005) and BAY43-9006 (Catalog # 553011) were purchased from Calbiochem, and stock solutions were made in DMSO. Rapamycin (R-5000) was purchased from LC Laboratories and a stock solution was made in DMSO.
Cell Proliferation Assays
Melanoma cells (25,000 cells per well) were plated in 96-well plates in RPMI plus either 5% FBS or 0.5% FBS, and cell numbers were assayed at time 0 and after 4, 8, 16, 24, 48, and 72 hours using Cell Titer 96 Aqueous (Promega Catalog# G3580; Madison, WI), according to the instructions provided by the manufacturer. Serum-dependent rates of growth were calculated using the slope of the lines from the growth curves, as shown in Figure 1A for VMM18. For experiments to examine the effects of the signal transduction inhibitors on serum-dependent melanoma cell proliferation, melanoma cells (1,000 cells per well) were plated in triplicate in a 96-well plates with 5% fetal bovine serum and allowed to adhere overnight. After 12–16 h, the cells were washed and treated with inhibitors as indicated for one hour. Cells were then washed and grown in RPMI medium with 5% FBS for 48 hours. Cell numbers were assayed with Cell Titer-Glo (Promega Catalog # G7571) according to the instructions provided by the manufacturer. The triplicate values were all within 5% and the mean values were calculated and plotted with error bars representing the standard deviation of triplicate samples from 3 independent experiments.
Figure 1 A, Growth curves of VMM18 melanoma cells in 5% and 0.5% serum-containing media. VMM18 melanoma cells were cultured in media containing either 5% or 0.5% serum and were assayed in triplicate at times 0, 4, 8, 16, 24, 48, and 72 hours using Cell Titer 96 Aqueous (Promega; Madison, WI) according to the directions supplied by the manufacturer. The absorbance at 490 nm (OD 490) measures the quantity of formazan product from the MTS-based assay and is directly proportional to the number of live cells present. The R2 values for the linear regression lines determined using Microsoft Excel are listed above each line. The solid dark line with the squares represents the data collected from VMM18 melanoma cells grown in media with 5% serum. The thin line with the circles represents the data from VMM18 melanoma cells grown in media containing 0.5% serum. B, Western blot analysis of 4EBP1 from melanoma cell lines grown in 5.0% FBS and 0.5% FBS. Phosphorylation of 4EBP1 was assayed by its reduced migration in SDS-PAGE and the proteins were detected by immunoblotting. VMM5A, VMM18, and VMM39 cells were grown in media as indicated. C, Western blot analysis of ERK from melanoma cell lines grown in 5.0% FBS and 0.5% FBS. The dual phosphorylation of ERK was analyzed by phosophosite-specific immunoblotting in VMM5A, VMM18, and VMM39 melanoma cells cultured as described (upper panel). The total amount of ERK protein was determined by immunoblotting with a separate antibody (lower panel). The relative phosphorylation of ERK was quantitated by densitometry analysis using Image Quant 5.2 software and the values are given below the top panel.
Western Blot Analysis
For analysis of proteins in Figure 1B and 1C, VMM5A, VMM18, and VMM39 melanoma cells were plated in Petri dishes and incubated for 24 hours in either RPMI medium plus 5% FBS or 0.5% FBS. After 24 hours, the cells were harvested and lysed as described for analysis of proteins in Figures 4 and 5. For analysis of the proteins in Figures 4 and 5, VMM18 melanoma cells were plated in petri dishes, treated with drugs or not for one hour, washed, and incubated overnight in RPMI medium plus 5% FBS. The next day, cells were rinsed with room temperature PBS, frozen by placing the dish on a mixture of acetone and dry ice. Cells were lysed in one ml of ice-cold 5% trichloroacetic acid for 10 minutes, scraped from the dish using a Costar cell lifter and the slurry was transferred to a 1.5-ml microcentrifuge tube and centrifuged for 10 minutes at 10,000 × g. The supernatant was discarded, and the pellet was washed twice with cold acetone to extract away the trichloroacetic acid and the proteins resuspended in resolubilization buffer (20 mM Tris, 23 mM glycine, 1 mM EDTA, and 10 mM β-glycerophosphate). Protein yields were determined by BCA analysis. Proteins were resuspended in SDS-containing sample buffer, heated for 10 min at 100°C, and 10 ng/lane was resolved by SDS-PAGE and transferred to Immobilon-P (Millipore). Membranes were blocked in 1% BSA in 50 mM Tris-Cl (pH 7.5), 0.9% NaCl, 0.05% Tween 20, and 0.01% antifoam A. Membranes were probed with antibodies listed below. Proteins were detected with Pierce SuperSignal West Pico Chemiluminescent substrate (#34080) as recommended by the manufacturer and used to expose to Kodak BioMax film. Films exposed within the linear response range were scanned and used for densitometry analysis by Image Quant 5.2.
Figure 4 Phosphorylation of mTOR targets in VMM18 melanoma cells treated with rapamycin, BAY43-9006, or U0126. The phosphorylation of p70S6K and 4EBP1 was assayed by migration in SDS-PAGE and the proteins detected by immunoblotting. Control VMM18 cells grown in PBS without serum were analyzed in lane 1 and VMM18 melanoma cells in lanes 2–6 were grown 24 hours in media that contained 5% serum. Cells not treated with drugs (lanes 1 and 2) were compared to cells pre-treated for one hour with rapamycin (lane 3), BAY43-9006 (lane 4), a combination of rapamycin and BAY43-9006 (lane 5), or U0126 (lane 6). GAPDH was immunoblotted as a loading control (bottom panel).
Figure 5 Phosphorylation of ERK in VMM18 melanoma cells treated with rapamycin, BAY43-9006, or U0126. The phosphorylation of ERK was analyzed by phosophosite-specific immunoblotting in VMM18 melanoma cells cultured and treated as described in Figure 4. The total amount of ERK protein was determined by immunoblotting with a separate antibody. The relative phosphorylation of ERK was quantitated by densitometry analysis using Image Quant 5.2 software and the values are given below the top panel.
Antibodies
Anti-p70S6 Kinase, clone SB20 Antibody (Catalog # 05-781, used at 1:8000) was purchased from Upstate. 4E-BP1 Antibody (Catalog #9452, used at 1:500) was purchased from Cell Signaling. GAPDH Antibody (Catalog # MAB374, used at 1:500) was purchased from Chemicon International. Anti-phospho MAP Kinase (ERK1/2), clone 12D4 antibody (Catalog # 05-481, used at 1:500) was purchased from Upstate. Anti-MAP Kinase 2/ERK2 antibody (Catalog #06-333, used at 1:500) was also purchased from Upstate. Phospho-MEK1/2 (Ser217/221) Antibody (Catalog#2354S, used at 1:1000) was purchased from Cell Signaling. Anti-Mouse IgG, peroxidase-linked species-specific whole antibody from sheep, secondary antibody (Catalog #NA931, used at 1:5000) was purchased from Amersham Biosciences. Anti-rabbit IgG, peroxidase-linked species-specific whole secondary antibody from donkey (Catalog # NA934, used at 1:5000) was also purchased from Amersham Biosciences.
Isobologram Analysis
To assess whether a combination dose of rapamycin and BAY43-9006 is synergistic or simply additive, a focused isobologram method was used as described previously [24]. An IC70 was selected, and these doses of each drug alone were plotted as the ordinate and abscissa in a Cartesian log-log plot. The straight line connecting these IC70 values is the locus of points (dose pairs) that produce a simply additive combination. In an isobologram, the IC70 dose pairs for two drugs together which fall on the line indicate an additive effect. Points above this line indicate antagonism, and points below the line indicate synergism.
Human Subjects
All of the research involving human subjects was approved by the University of Virginia's IRB (Human Investigation Committee) in accordance with assurances filed with and approved by the Department of Health and Human Services. Informed consent was obtained from all of the study participants.
Results
Proliferation of melanoma cells expressing wild-type and V599E mutant B-Raf
We examined the serum-dependent proliferation of various human melanoma cell lines. Figure 1 A is a growth curve for the VMM18 cell line, which is representative of the growth curves generated for each of the cell lines from a collection of human melanomas (Table 1). Cell proliferation was determined by the number of cells at 0, 4, 8, 16, 24, 48, and 72 hours, quantitated using the Cell Titer 96 Aqueous (Promega Corporation, Madison, WI) assay which measures reduction of MTS (a novel tetrazolium compound). These human melanoma cell lines proliferated even in limiting serum (0.5%). However, all showed higher rates of proliferation ~ 2-fold in the presence of 5% serum.
We could detect activation of the mTOR and ERK signaling pathways in proliferating melanoma cells. Shown in Figure 1 B is a Western blot detecting the phosphorylation of the mTOR substrate, 4EBP1, from 3 different melanoma cell lines grown in the presence of either 5% or 0.5% serum. The phosphorylation of 4EBP1 has previously been demonstrated to retard migration in SDS-PAGE [25], seen as the upper band in the doublet in the even numbered lanes (Figure 1B). Shown in Figure 1 C is a Western blot detecting both the dual phosphorylated (activated) form of ERK, as well as total ERK protein in three different melanoma cell lines grown in 5% or 0.5% serum. The quantitation of the relative phosphorylation of ERK relative to total ERK is shown between the blots, demonstrating about a two-fold increase in phosphorylation. The phosphorylation of ERK paralleled the relative increase in proliferation for each of these cell lines.
BAY43-9006 and rapamycin inhibit proliferation of melanoma cells
We examined the serum-dependent proliferation of multiple human melanoma cell lines and the effects of inhibition of B-Raf by BAY43-9006 and of mTOR by rapamycin. Melanoma cell lines were tested for proliferation after treatment with a single dose of BAY43-9006 or rapamycin using Cell Titer-Glo (Promega Corporation, Madison, WI), a luminescence-based ATP cell viability assay. Cells were exposed to different doses of either drug for one hour. Then, the media was changed and the cells were cultured for two days in the presence of serum. We found that micromolar concentrations were cytotoxic, because cell numbers decreased after two days, whereas nanomolar concentrations were growth inhibitory. Melanoma cells showed dose-dependent inhibition with 0.01 nM to 100 nM of BAY43-9006 (Figure 2A), or rapamycin (Figure 2B). Proliferation of the cells was inhibited in either 5% or 0.5% serum.
Figure 2 Inhibition of various melanoma cell lines by BAY43-9006 and rapamycin. A, The serum-stimulated proliferation of VMM18, VMM5A, and VMM39 melanoma cells was assayed in triplicate by Cell Titer Glo (Promega; Madison, WI). Inhibition was determined as the difference between the number of cells in the control without any added drug compared to the number of cells following treatment with an inhibitor. Cells were treated for one hour with doses of BAY43-9006 ranging from 0.01 nM to 100 nM and were assayed for serum-stimulated proliferation after 48 hours. The mean value for triplicate samples from three independent experiments is plotted relative to the control (treated with vehicle, DMSO). Error bars are present at each point on the graph. The Student T test was performed and the bracket and asterisk represents a statistically significant difference between VMM39 (wt) and VMM5A (V599E) and VMM18 (V599E) cells at the 10 nM concentration with a p value < 0.002. B, The serum-stimulated proliferation of VMM18, VMM5A, and VMM39 melanoma cells were assayed as described above except melanoma cells were treated with rapamycin instead of BAY43-9006 with doses ranging from 0.01 nM to 100 nM. The mean value for triplicate samples from three independent experiments is plotted. Error bars are present at each point on the graph. The Student T test was performed and the bracket and asterisk represents a statistically significant difference between VMM39 (wt) and VMM5A (V599E) and VMM18 (V599E) cells at the 10 nM concentration with a p value < 0.002.
Among the melanoma cell lines, there was a significant difference in the amount of inhibition at 10 nM BAY43-9006 or rapamycin. We observed that melanoma cell lines that contain the V599E mutation in B-Raf (VMM5A and VMM18) were more sensitive to BAY43-9006 and to rapamycin, compared to cell lines with wild-type B-Raf (VMM39). This difference in growth inhibition was observed in two additional cell lines, one wild-type (DM122) and one V599E (VMM12). Therefore, nanomolar concentrations of either BAY43-9006 or rapamycin inhibit the proliferation of melanoma cells, whether or not they have mutated B-Raf.
Combining Rapamycin with BAY43-9006 synergistically inhibits serum-dependent proliferation of melanoma cells
Melanoma cell proliferation was inhibited by either BAY43-9006 or rapamycin over the 0.01 – 100 nM concentration range. A combination of the two drugs was markedly more effective than either drug alone at inhibiting serum-stimulated melanoma cell proliferation. For example, 0.01 nM of each drug together was more effective at inhibiting melanoma cell proliferation than 1 nM of either drug alone. To assess synergism versus additivity quantitatively, we used a focused isobologram method (Figure 3). Treatment of three melanoma cell lines (VMM39, VMM5A, and VMM18) with rapamycin alone induced a 70% growth inhibition (IC70) from approximately 10 nM (VMM39) to 2 nM (VMM5A and VMM18). These were plotted on the ordinate. The IC70 concentration for BAY43-9006 alone was in the range of approximately 5 to 10 nM, in different cell lines, and these were plotted on the abscissa. Compared to the single agents, the IC70 for the dose pairs (rapamycin and BAY43-9006 together) falls below the line, for each of these melanoma cell lines, indicating that the combination is synergistic (Figure 3). Furthermore, VMM18, which contains the V599E substitution, was more sensitive to the combination treatment than melanoma cell lines with wild-type B-Raf, consistent with the enhanced sensitivity at the 10 nM dose of each agent (Figure 2 above). However, all melanoma cell lines tested displayed synergistic inhibition of proliferation, indicating that these drugs were more effective in combination than alone.
Figure 3 Isobologram analyses of inhibition of melanoma cell proliferation. A, VMM39 melanoma cells, B, VMM5A melanoma cells, and C, VMM18 melanoma cells. The straight line connecting the IC70 points (additivity line) is the locus of all dose pairs that should give the same effect. A dose pair (data point) below the line is synergistic, a dose pair on the line is additive, and a dose pair above the line is antagonistic. Note: all graphs are on a log-scale.
Rapamycin and BAY43-9006 inhibit phosphorylation of proteins in the mTOR signaling pathway in melanoma cells
Melanoma cells were treated with rapamycin and BAY 43-9006, either singly or in combination, for one hour, and protein phosphorylation was examined by Western blot analysis 24 hours later. Rapamycin is an inhibitor of mTOR kinase and reduces phosphorylation of its substrates, p70S6K and 4EBP1 [17]. BAY 43-9006 is a chemical inhibitor of B-Raf kinase and reduces phosphorylation of MEK and ERK [26,27]. VMM18 melanoma cells grown in the presence of 5% serum had enhanced phosphorylation of p70S6K and 4EBP1 (Figure 4, lane 2) relative to cells grown in the absence of serum (Figure 4, PBS, lane 1). The phosphorylation of p70S6K and 4EBP1 retards migration in SDS-PAGE. Antibodies to these proteins were used to show all the protein and therefore enable evaluation of the fraction phosphorylated under different conditions. Treatment of VMM18 melanoma cells with a 10 nM dose of rapamycin inhibited the serum-stimulated phosphorylation of p70S6K and 4EBP1 (Figure 4, lane 3). Parallel treatment of VMM18 melanoma cells with a 10 nM dose of BAY43-9006 unexpectedly inhibited serum-stimulated phosphorylation of p70S6K and 4EBP1 (Figure 4, lane 4). There is not a well-documented requirement of Raf-MEK-ERK activity for the phosphorylation of mTOR substrates p70S6K and 4EBP1. Combination treatment with a 10 nM dose of rapamycin plus a 10 nM dose of BAY43-9006 blocked phosphorylation of p70S6K and 4EBP1 as effectively as either drug alone (Figure 4, lane 5). Thus, even though cell proliferation was suppressed more effectively by this combination of drugs, this was not reflected in a detectable further decrease in phosphorylation of the mTOR target proteins p70S6K and 4EBP1. As an additional control, we treated VMM18 melanoma cells with U0126, a MEK inhibitor, which blocked serum-stimulated phosphorylation of both p70S6K and 4EBP1 (Figure 4, lane 6). This result showed that MEK/ERK activities contribute to phosphorylation of p70S6K and 4EBP1.
We noted that total 4EBP1 in cells treated with a combination of rapamycin plus BAY43-9006, or with U0126, was lower relative to untreated cells or cells treated with either rapamycin or BAY43-9006 alone. Equal recovery of other proteins from the cells was demonstrated by immunoblotting both for p70S6K and for GAPDH, used as a loading control. We do not understand the basis for the reduced recovery of 4EBP1, but it did not seem to depend simply on the phosphorylation state because phosphorylation was blocked with the single drug treatments, without change in the level of the 4EBP1 protein.
Rapamycin and BAY43-9006 inhibit phosphorylation of proteins in the B-Raf-MEK-ERK signaling pathway in melanoma cells
In VMM18 melanoma cells, the dual phosphorylation (Tyr/Thr) of ERK was 9-fold higher in cells grown in 5% serum relative to cells grown in the absence of serum (Figure 5, lanes 2 versus 1). There also was an increased level in the dual phosphorylation (Ser217/221) of MEK (not shown). Treatment of VMM18 melanoma cells with a 10 nM dose of BAY 43-9006 produced a 75% decrease in the dual phosphorylation of ERK (Figure 5, lane 4) and reduced the phosphorylation of MEK below detection levels (not shown). These results were consistent with the inhibition of B-Raf by BAY43-9006. On the other hand, when VMM18 melanoma cells were treated with a 10 nM dose of rapamycin, the dual phosphorylation of ERK was reduced by about half (Figure 5, lane 3). Our interpretation of this result is that mTOR activity is required to maintain the phosphorylation of ERK in melanoma cells. Combination treatment of a 10 nM dose of rapamycin plus a 10 nM dose of BAY 43-9006 reduced the phosphorylation of ERK to a level even below that seen in cells grown in the absence of serum (Figure 5, lane 5). This inhibition of ERK phosphorylation by combination of rapamycin and BAY43-9006 was as effective as inhibition of MEK by the U0126 compound (Figure 5, compare lanes 5 and 6).
Discussion
New cancer treatments involve directly targeting enzymes essential for the growth and proliferation of cancer cells. The mTOR pathway regulates cell growth, and the Raf/MEK/ERK pathway is critical for cell proliferation. Activating mutations in B-Raf have been found in 60–70% of human melanomas, making B-Raf a potential target for small molecule inhibitors as therapy [3,4,11,28]. Indeed, new drugs such as BAY43-9006 have been developed as selective inhibitors of B-Raf and are currently in Phase II clinical trials [6]. Inhibition of mTOR by rapamycin has been standard treatment for immunosuppression following organ transplant [15], and the rapamycin derivative CCI-779 is now being clinically tested as a cancer chemotherapy [18-20]. Thus, B-Raf and mTOR are acknowledged targets for anti-proliferative therapy [29].
Current knowledge suggests that B-Raf and mTOR protein kinases operate in separate signaling pathways. The B-Raf kinase is activated by GTP-Ras in response to growth factors and phosphorylates MEK, which in turn activates ERK to phosphorylate downstream targets such as kinases and transcription factors that promote cell division [30]. The mTOR kinase responds to both nutrient and growth factor signals to activate p70S6K and 4EBP1 to increase protein translation as part of a cell growth response [31]. Increase in cell growth (size) is a pre-requisite for cell proliferation. Because the B-Raf and mTOR pathways are thought to operate in parallel, we hypothesized that combined inhibition of these kinases would be effective in blocking cell growth and cell proliferation. Though our results with multiple melanoma cell lines support that hypothesis, they also gave some unexpected results.
Human tumors deficient in PTEN have activated Akt, and are especially sensitive to mTOR inhibitors [13]. However, pharmacogenomic profiling indicates that melanomas are not, in general, PTEN deficient and therefore would be unresponsive to mTOR inhibitors. Results from a phase II trial using CCI-779 alone showed only one response among 33 observed patients [32]. These data suggest that CCI-779 is not sufficiently active in melanoma as a single agent. However, our data show that melanoma cell proliferation is effectively inhibited in vitro by low doses of rapamycin. Together, these findings argue against use of CCI-779 as a single agent, but support investigation of mTOR inhibitors as part of combination therapy for treatment of patients with malignant melanoma.
With regards to B-Raf, recent structural studies have shown that BAY43-9006 interacts with an inactive conformation of B-Raf [33]. In biochemical assays, the kinase activity of V599E B-Raf is less sensitive to inhibition by BAY43-9006 than wild-type B-Raf, suggesting that melanomas with the B-Raf V599E mutation might be resistant to the effects of this drug [33]. However, in the present study, proliferation of the human melanoma cells was inhibited by BAY43-9006, and at a dose of 10 nM, the cells that contained mutated B-Raf V599E were more sensitive than cells with wild-type B-Raf. In clinical studies with BAY43-9006 plus chemotherapy, objective tumor regressions were more common in patients who had wild-type B-raf [34]. The findings of the current report support continued investigation of BAY43-9006 for treatment of patients with melanoma, and suggest that clinical effects observed may be due to some effects that are independent of B-raf kinase activity.
We found that multiple human melanoma cell lines proliferated in culture at different relative rates in the absence of serum and that the addition of serum to the medium doubled the rate of proliferation. Thus, we could use the consistent serum response to compare cell growth and proliferation with a variety of melanoma cell lines. At concentrations in the nanomolar range, we observed dose-dependent inhibition of cell proliferation by either rapamycin or BAY43-9006. In every cell line examined, combination of BAY43-9006 and rapamycin produced synergistic inhibition of cell proliferation compared to either drug alone. This suggests that administration of a combination of an mTOR inhibitor (rapamycin or CCI-779) and BAY43-9006 could be an especially effective approach to therapy of melanoma.
Our results indicate that rapamycin and BAY43-9006 inhibit their cognate targets in melanoma cells (mTOR and B-Raf respectively), as well as downstream effectors thought to be in other pathways, providing evidence for cellular cross-talk between the different signaling pathways studied. Specifically, we found that BAY43-9006 inhibited serum-stimulated phosphorylation of p70S6K and 4EBP1, and rapamycin blocked serum-stimulated phosphorylation of ERK. Previously published results have suggested interdependence between mTOR and Raf-MEK-ERK signaling [35-40]. In vascular smooth muscle cells under hyperglycemic conditions (25 mM versus 5 mM glucose), inhibition of PI3K with LY294002 or inhibition of mTOR by rapamycin reduced the level of ERK Tyr-phosphorylation [35]. In cardiomyocytes, PKC-dependent activation of mTOR and p70S6K was inhibited by U0126, implicating a requirement for MEK [36]. Rapamycin inhibited the FGF-2 induced proliferation of two different small cell lung cancer lines (SCLC), whereas PD098059 inhibited one and not the other [37]. Combination of rapamycin and PD098059 was not tested. In proximal tubular epithelial cells, insulin-activated phosphorylation of 4EBP1 could be inhibited by PD098059, suggesting a requirement for MAPK [38]. Another report shows that following hypertonic stress, HEK 293 cells show increase in protein synthesis, and simultaneous inhibition of both mTOR and ERK was required to prevent de novo translation [39].
Since there appears to be cross-talk between mTOR and Raf-MEK-ERK pathways, it might be expected that combination therapy with rapamycin and BAY43-9006 might simply be additive. To our knowledge, the effects of combining inhibitors of these two pathways on proliferation of melanoma cells had not previously been examined. However, studies are in development for such combination therapies in human clinical trials, sponsored by the Clinical Trials Evaluation Program (CTEP) of the NIH. In the present study, we found that the combination of inhibitors synergistic for inhibition of melanoma cell proliferation.
Cancer cells may be dependent on particular oncogenes for cell growth, which renders them sensitive to drugs that inhibit these protein targets. Under these circumstances, single chemical inhibitors are efficacious, such as Gleevec inhibition of BCR-ABL in CML [2]. However, in a number of different cancers, single drug targeted therapy is only effective in about half of the patients [41]. These cancer cells utilize either alternate pathways or compensatory mechanisms to evade inhibition. Under these circumstances, combination therapy that inhibits different pathways may be especially effective. Our results show synergistic inhibition of cell proliferation with drugs against different pathways. Further, we exposed effects on pathways not thought to be targeted by agents currently used in the clinic. Because a combination of rapamycin and BAY43-9006 is more effective at inhibiting melanoma cell proliferation than either drug alone, further studies of this combination in animal models and clinical trials deserve to be examined.
Competing interests
The author(s) declare that they have no competing interests.
Acknowledgements
We thank the members of the Slingluff laboratory and Drs. John Lawrence, Christopher Thomas and Yukiko Misawa for helpful discussions. We also thank Dr. Tom Sturgill for the use of his luminometer in our cell proliferation assays. This work was partly supported by grant CA77584 to Dr. David L. Brautigan from USPHS NCI and partly supported by funds from the Harrison Foundation to the University of Virginia Cancer Center and Dr. Craig L. Slingluff, Jr. Support was also provided by the NIH/NCI grant R01 CA57653 (to CLS) and the University of Virginia Cancer Center Support Grant (NIH/NCI P30 CA44579).
==== Refs
Rodenhuis S ras and human tumors Semin Cancer Biol 1992 3 241 247 1421168
Shah NP Sawyers CL Mechanisms of resistance to STI571 in Philadelphia chromosome-associated leukemias Oncogene 2003 22 7389 7395 14576846 10.1038/sj.onc.1206942
Davies H Bignell GR Cox C Stephens P Edkins S Clegg S Teague J Woffendin H Garnett MJ Bottomley W Davis N Dicks E Ewing R Floyd Y Gray K Hall S Hawes R Hughes J Kosmidou V Menzies A Mould C Parker A Stevens C Watt S Hooper S Wilson R Jayatilake H Gusterson BA Cooper C Shipley J Hargrave D Pritchard-Jones K Maitland N Chenevix-Trench G Riggins GJ Bigner DD Palmieri G Cossu A Flanagan A Nicholson A Ho JW Leung SY Yuen ST Weber BL Seigler HF Darrow TL Paterson H Marais R Marshall CJ Wooster R Stratton MR Futreal PA Mutations of the BRAF gene in human cancer Nature 2002 417 949 954 12068308 10.1038/nature00766
Pollock PM Meltzer PS A genome-based strategy uncovers frequent BRAF mutations in melanoma Cancer Cell 2002 2 5 7 12150818 10.1016/S1535-6108(02)00089-2
Tuveson DA Weber BL Herlyn M BRAF as a potential therapeutic target in melanoma and other malignancies Cancer Cell 2003 4 95 98 12957284 10.1016/S1535-6108(03)00189-2
Ahmad T Eisen T Kinase inhibition with BAY 43-9006 in renal cell carcinoma Clin Cancer Res 2004 10 6388S 92S 15448036 10.1158/1078-0432.CCR-040028
Roux PP Blenis J ERK and p38 MAPK-activated protein kinases: a family of protein kinases with diverse biological functions Microbiol Mol Biol Rev 2004 68 320 344 15187187 10.1128/MMBR.68.2.320-344.2004
Hartkamp J Troppmair J Rapp UR The JNK/SAPK activator mixed lineage kinase 3 (MLK3) transforms NIH 3T3 cells in a MEK-dependent fashion Cancer Res 1999 59 2195 2202 10232608
Burack WR Sturgill TW The activating dual phosphorylation of MAPK by MEK is nonprocessive Biochemistry 1997 36 5929 5933 9166761 10.1021/bi970535d
Hotte SJ Hirte HW BAY 43-9006: early clinical data in patients with advanced solid malignancies Curr Pharm Des 2002 8 2249 2253 12369852 10.2174/1381612023393053
Krasilnikov M Ivanov VN Dong J Ronai Z ERK and PI3K negatively regulate STAT-transcriptional activities in human melanoma cells: implications towards sensitization to apoptosis Oncogene 2003 22 4092 4101 12821943 10.1038/sj.onc.1206598
Sansal I Sellers WR The biology and clinical relevance of the PTEN tumor suppressor pathway J Clin Oncol 2004 22 2954 2963 15254063 10.1200/JCO.2004.02.141
Xu G Zhang W Bertram P Zheng XF McLeod H Pharmacogenomic profiling of the PI3K/PTEN-AKT-mTOR pathway in common human tumors Int J Oncol 2004 24 893 900 15010827
Morris RE Wu J Shorthouse R A study of the contrasting effects of cyclosporine, FK 506, and rapamycin on the suppression of allograft rejection Transplant Proc 1990 22 1638 1641 1697111
Chiaramonte S Dissegna D Ronco C Monitoring of immunosuppressive therapy in renal transplanted patients Contrib Nephrol 2005 146 73 86 15567922
Regar E Sianos G Serruys PW Stent development and local drug delivery Br Med Bull 2001 59 227 248 11756213 10.1093/bmb/59.1.227
Chan S Targeting the mammalian target of rapamycin (mTOR): a new approach to treating cancer Br J Cancer 2004 91 1420 1424 15365568 10.1038/sj.bjc.6602162
Atkins MB Hidalgo M Stadler WM Logan TF Dutcher JP Hudes GR Park Y Liou SH Marshall B Boni JP Dukart G Sherman ML Randomized phase II study of multiple dose levels of CCI-779, a novel mammalian target of rapamycin kinase inhibitor, in patients with advanced refractory renal cell carcinoma J Clin Oncol 2004 22 909 918 14990647 10.1200/JCO.2004.08.185
Peralba JM DeGraffenried L Friedrichs W Fulcher L Grunwald V Weiss G Hidalgo M Pharmacodynamic Evaluation of CCI-779, an Inhibitor of mTOR, in Cancer Patients Clin Cancer Res 2003 9 2887 2892 12912932
Raymond E Alexandre J Faivre S Vera K Materman E Boni J Leister C Korth-Bradley J Hanauske A Armand JP Safety and pharmacokinetics of escalated doses of weekly intravenous infusion of CCI-779, a novel mTOR inhibitor, in patients with cancer J Clin Oncol 2004 22 2336 2347 15136596 10.1200/JCO.2004.08.116
Brinckerhoff LH Kalashnikov VV Thompson LW Yamshchikov GV Pierce RA Galavotti HS Engelhard VH Slingluff CLJ Terminal modifications inhibit proteolytic degradation of an immunogenic MART-1(27-35) peptide: implications for peptide vaccines Int J Cancer 1999 83 326 334 10495424 10.1002/(SICI)1097-0215(19991029)83:3<326::AID-IJC7>3.0.CO;2-X
Kittlesen DJ Thompson LW Gulden PH Skipper JC Colella TA Shabanowitz J Hunt DF Engelhard VH Slingluff CLJ Human melanoma patients recognize an HLA-A1-restricted CTL epitope from tyrosinase containing two cysteine residues: implications for tumor vaccine development J Immunol 1998 160 2099 2106 9498746
Huntington JT Shields JM Der CJ Wyatt CA Benbow U Slingluff CLJ Brinckerhoff CE Overexpression of collagenase 1 (MMP-1) is mediated by the ERK pathway in invasive melanoma cells: role of BRAF mutation and fibroblast growth factor signaling J Biol Chem 2004 279 33168 33176 15184373 10.1074/jbc.M405102200
Tallarida RJ Drug synergism: its detection and applications J Pharmacol Exp Ther 2001 298 865 872 11504778
Lehman JA Calvo V Gomez-Cambronero J Mechanism of ribosomal p70S6 kinase activation by granulocyte macrophage colony-stimulating factor in neutrophils: cooperation of a MEK-related, THR421/SER424 kinase and a rapamycin-sensitive, m-TOR-related THR389 kinase J Biol Chem 2003 278 28130 28138 12740386 10.1074/jbc.M300376200
Hilger RA Kredtke S Scheulen ME Seeber S Strumberg D Correlation of ERK-phosphorylation and toxicities in patients treated with the Raf kinase inhibitor BAY 43-9006 Int J Clin Pharmacol Ther 2004 42 648 649 15598034
Strumberg D Richly H Hilger RA Schleucher N Korfee S Tewes M Faghih M Brendel E Voliotis D Haase CG Schwartz B Awada A Voigtmann R Scheulen ME Seeber S Phase I Clinical and Pharmacokinetic Study of the Novel Raf Kinase and Vascular Endothelial Growth Factor Receptor Inhibitor BAY 43-9006 in Patients With Advanced Refractory Solid Tumors J Clin Oncol 2004 15613696
Erhardt P Schremser EJ Cooper GM B-Raf inhibits programmed cell death downstream of cytochrome c release from mitochondria by activating the MEK/Erk pathway Mol Cell Biol 1999 19 5308 5315 10409722
Dancey J Sausville EA Issues and progress with protein kinase inhibitors for cancer treatment Nat Rev Drug Discov 2003 2 296 313 12669029 10.1038/nrd1066
Smalley KS A pivotal role for ERK in the oncogenic behaviour of malignant melanoma? Int J Cancer 2003 104 527 532 12594806 10.1002/ijc.10978
Bjornsti MA Houghton PJ The TOR pathway: a target for cancer therapy Nat Rev Cancer 2004 4 335 348 15122205 10.1038/nrc1362
Margolin KA Longmate J Baratta T Synold T Weber J Gajewski T Quirt I Christensen S Doroshow JH CCI-779 in metastatic melanoma: A phase II trial of the California Cancer Consortium. Journal of Clinical Oncology, 2004 ASCO Annual Meeting Proceedings (Post-Meeting Edition) 2004 Vol 22, No 14S (July 15 Supplement) Abstract No. 7523
Wan PT Garnett MJ Roe SM Lee S Niculescu-Duvaz D Good VM Jones CM Marshall CJ Springer CJ Barford D Marais R Mechanism of activation of the RAF-ERK signaling pathway by oncogenic mutations of B-RAF Cell 2004 116 855 867 15035987 10.1016/S0092-8674(04)00215-6
Flaherty KT BMSLTDLRJSBLCDWBODPJ Phase I/II trial of BAY 43-9006, carboplatin (C) and paclitaxel (P) demonstrates preliminary antitumor activity in the expansion cohort of patients with metastatic melanoma. J Clin Oncol 2004 22 7507
Campbell M Allen WE Sawyer C Vanhaesebroeck B Trimble ER Glucose-potentiated chemotaxis in human vascular smooth muscle is dependent on cross-talk between the PI3K and MAPK signaling pathways Circ Res 2004 95 380 388 15242975 10.1161/01.RES.0000138019.82184.5d
Iijima Y Laser M Shiraishi H Willey CD Sundaravadivel B Xu L McDermott PJ Kuppuswamy D c-Raf/MEK/ERK pathway controls protein kinase C-mediated p70S6K activation in adult cardiac muscle cells J Biol Chem 2002 277 23065 23075 11940578 10.1074/jbc.M200328200
Pardo OE Arcaro A Salerno G Tetley TD Valovka T Gout I Seckl MJ Novel cross talk between MEK and S6K2 in FGF-2 induced proliferation of SCLC cells Oncogene 2001 20 7658 7667 11753643 10.1038/sj.onc.1204994
Bhandari BK Feliers D Duraisamy S Stewart JL Gingras AC Abboud HE Choudhury GG Sonenberg N Kasinath BS Insulin regulation of protein translation repressor 4E-BP1, an eIF4E-binding protein, in renal epithelial cells Kidney Int 2001 59 866 875 11231341 10.1046/j.1523-1755.2001.059003866.x
Naegele S Morley SJ Molecular cross-talk between MEK1/2 and mTOR signaling during recovery of 293 cells from hypertonic stress J Biol Chem 2004 279 46023 46034 15292274 10.1074/jbc.M404945200
Lehman JA Gomez-Cambronero J Molecular crosstalk between p70S6k and MAPK cell signaling pathways Biochem Biophys Res Commun 2002 293 463 469 12054624 10.1016/S0006-291X(02)00238-3
Leyland-Jones B Trastuzumab: hopes and realities Lancet Oncol 2002 3 137 144 11902499 10.1016/S1470-2045(02)00676-9
|
16255777
|
PMC1289294
|
CC BY
|
2021-01-04 16:39:28
|
no
|
J Transl Med. 2005 Oct 28; 3:39
|
utf-8
|
J Transl Med
| 2,005 |
10.1186/1479-5876-3-39
|
oa_comm
|
==== Front
Nutr Metab (Lond)Nutrition & Metabolism1743-7075BioMed Central London 1743-7075-2-261621611910.1186/1743-7075-2-26ResearchFat intake and composition of fatty acids in serum phospholipids in a randomized, controlled, Mediterranean dietary intervention study on patients with rheumatoid arthritis Hagfors Linda [email protected] Ingela [email protected]öldstam Lars [email protected] Gunnar [email protected] Department of Food and Nutrition, Umeå University, SE-901 87 Umeå, Sweden2 Department of Clinical Chemistry, Kalmar County Hospital, SE-391 85 Kalmar, Sweden3 Department of Medicine, Visby Hospital, SE-621 84 Visby, Sweden2005 10 10 2005 2 26 26 24 11 2004 10 10 2005 Copyright © 2005 Hagfors et al; licensee BioMed Central Ltd.2005Hagfors et al; licensee BioMed Central Ltd.This is an Open Access article distributed under the terms of the Creative Commons Attribution License (), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
Background
We have previously reported that rheumatoid arthritis patients, who adopted a modified Cretan Mediterranean diet, obtained a reduction in disease activity and an improvement in physical function and vitality. This shift in diet is likely to result in an altered intake of fatty acids. Therefore, the objective of the present study was to examine the dietary intake of fatty acids, as well as the fatty acid profile in serum phospholipids, during the dietary intervention study presented earlier.
Results
From baseline to the end of the study, changes in the reported consumption of various food groups were observed in the Mediterranean diet group. The change in diet resulted in a number of differences between the Mediterranean diet group and the control diet group regarding the fatty acid intake. For instance, a lower ratio of n-6 to n-3 fatty acids was observed in the Mediterranean diet group, both assessed by diet history interviews (dietary intake) and measured in serum phospholipids. Moreover, the patients in the Mediterranean diet group that showed a moderate or better clinical improvement during the study (diet responders), had a higher reported intake of n-3 fatty acids and a lower ratio of n-6 to n-3 fatty acids compared to the patients with minor or no improvement. Also the fatty acid profile in serum phospholipids differed in part between the diet responders and the diet non-responders.
Conclusion
The changes in the fatty acid profile, indicated both by dietary assessments and through fatty acids in s-phospholipids may, at least in part, explain the beneficial effects of the Cretan Mediterranean diet that we have presented earlier.
==== Body
Background
We have previously reported that a shift to a modified Cretan Mediterranean diet decreased the disease activity and improved the physical function and vitality of Swedish rheumatoid arthritis (RA) patients [1]. Among the characteristics of the experimental diet used in this study, was a relatively high consumption of fish, olive oil and canola (rapeseed) oil, as well as a low intake of other fats and of red meat. A change from a typical Swedish diet to a Cretan Mediterranean diet is likely to result in a changed intake of fatty acids, for example an increased intake of n-3 fatty acids. The impact of dietary fatty acids on rheumatoid inflammation has been investigated in a number of in-vitro and animal studies, as well as in randomized, placebo-controlled trials [2]. The fatty acids most thoroughly studied in relation to RA, are the n-3 fatty acids eicosapentaenoic acid (EPA; 20:5n-3) and docosahexaenoic acid (DHA; 22:6n-3). At least thirteen randomized controlled trials, ranging from 12 to 52 weeks, have shown that supplementation with fish oil (which contains EPA and DHA) resulted in beneficial effects on RA symptoms [3,4]. The most commonly observed benefit is a decreased number of tender joints [5], but improvements in morning stiffness, number of swollen joints, pain index, physicians global assessment and grip strength have also been reported [4]. In addition, a meta-analysis, as well as a mega-analysis (in which the original data sets were analysed), confirmed the efficacy of fish oils in reducing RA symptoms [6]. However, only the reduction in the tender joint count and in morning stiffness was shown to be statistically significant using these approaches.
In supplementation studies, olive oil has sometimes been used as a placebo oil. In some of these studies improvements were reported also in the group of RA patients treated with olive oil [7-9]. The exact mechanism for this effect has been uncertain. However, in a recent study, a compound in extra-virgin olive oil, called oleocanthal, was shown to inhibit the activity of cyclooxygenase enzymes [10].
The relationship between n-3 fatty acids/fish intake, olive oil and RA, has also been observed in case control studies. In these studies a high consumption of fish [11] and olive oil [12,13] was associated with decreased risk of developing RA.
In view of the results from earlier studies, the beneficial effects observed in the randomized controlled trial in which the Cretan Mediterranean diet was tested [1], may be caused by an increse in n-3 fatty acids in relation to n-6 fatty acids and/or an increased intake of olive oil. The aim of the present study was to examine the consumption of food items with relevance to the fat intake, the dietary intake of fatty acids, as well as the fatty acid profile in serum phospholipids, in the subjects participating in this study.
Results
Reported dietary intake
To estimate the reported consumption of foods with relevance to the fatty acid intake, a self-administrated questionnaire was used. At baseline the Cretan Mediterranean diet (MD) group (n = 26) and the control diet (CD) group (n = 25) reported a similar intake frequency of all the food groups investigated (Table 1). By the end of the study the MD group had decreased their intake of several food groups such as meat, processed meat (including cured meat, sausage, pâté or the like), sweets and high fat dairy products, while the intake of fish, shellfish, poultry and cheese with less than 17% fat, had increased compared to baseline. The CD group also decreased their intake of meat, from baseline to the end of the study, but apart from that, no differences were seen in that group regarding the intake frequencies of foods with relevance to the fatty acid intake.
Table 1 Comparison of reported food consumption frequencies (servings per month) between the Mediterranean Diet (MD) group and the Control Diet (CD) group.
MD group (n = 26) CD group (n = 25) P-value*
Time (weeks after dietary shift) Time (weeks after start of study)
0 3 12 0 3 12
Fish 6 (6-6) 16 (6–16) 16 (16-16)§ 6 (2–6) 6 (6–16) 6 (2–6) <0.001
Shellfish 2 (0–2) 2 (0–6) 4 (2–6)‡ 2 (0–2) 2 (0–2) 2 (0–2) <0.001
Meat, minced meat or offal from pig, cattle or sheep 16 (6–16) 0 (0-0) 0 (0-0)§ 16 (6–16) 6 (6–16) 6 (6–16)† <0.001
Poultry 2 (2–6) 16 (6–16) 11 (6–16)§ 2 (2–6) 2 (2–6) 2 (2–6) <0.001
Processed meat¶(not on bread) 6 (2–6) 0 (0-0) 0 (0-0)§ 6 (2–6) 6 (2–6) 2 (2–6) <0.001
Processed meat¶on bread 11 (6–26) 0 (0-0) 0 (0-0)§ 16 (2–16) 6 (4–16) 6 (2–16) <0.001
Cheese with more than 17% fat 16 (14–26) 0 (0-0) 0 (0-0)§ 26 (11–70) 16 (4–26) 16 (4–26) 0.002
Cheese with less than 17% fat 6 (0–9) 16 (0–26) 16 (6–26)‡ 0 (0–4) 0 (0–6) 0 (0–11) 0.005
Dairy products with 27% fat or more 2 (0–6) 0 (0-0) 0 (0-0)‡ 2 (0–6) 2 (0–6) 6 (2–6) <0.001
Dairy products with 10–15% fat 2 (0–6) 0 (0-0) 0 (0–2)‡ 2 (0–6) 0 (0–6) 0 (0–6) 0.013
Ice cream 2 (0–6) 0 (0–0.5) 0 (0–2)‡ 2 (0–4) 2 (0–2) 2 (0–2) 0.015
Sweets (including chocolate) 6 (2–16) 0 (0–2) 0 (0–2)§ 6 (1–6) 6 (0–6) 6 (2–6) <0.001
Buns, cookies or cakes 6 (5–16) 0 (0–2) 0 (0–2)§ 6 (2–16) 6 (2–16) 6 (2–16) <0.001
Nuts 0 (0–2) 0 (0–2) 0 (0–3) 0 (0–2) 0 (0–2) 0 (0–2) 0.704
Seeds 0 (0-0) 0 (0–2) 0 (0–2) 0 (0-0) 0 (0-0) 0 (0-0) 0.016
Data are presented as medians (25th–75th percentiles). *The p-values refer to differences between the MD- and the CD group concerning the change from baseline to week 12. Differences between the groups were analyzed with the Mann-Whitney U test. Statistically significant change from baseline to week 12: †p < 0.05; ‡p < 0.01; §p < 0.001. Within-group differences in week 12 compared to baseline were evaluated by the Wilcoxon signed ranks test. ¶Including cured meat, sausage, pâté or the like.
Regarding the reported food choices, significantly more MD subjects than CD subjects changed their choice of milk and fermented milk to milk and fermented milk with a lower fat content, or excluded these foods totally (p = 0.003 and p = 0.038 respectively). In addition, more MD subjects than CD subjects used olive oil and canola oil, as well as the liquid margarine and half-fat margarine supplied during the study period (p < 0.001 for all comparisons).
The nutrient intake during the second half of the study was estimated by means of diet history interviews (conducted with 17 individuals in each group). Two persons in the CD group and one in the MD group were identified as under-reporters, based on the individual Goldberg cut-off [14] for the food intake level (FIL). In addition, one CD subject was identified as an over-reporter. Since the results at the group level were unchanged, whether we excluded under- and over-reporters or not, the results of all the diet history interviews are presented (Table 2). One exception was however, the percentage of energy (E%) derived from polyunsaturated fatty acids (PUFAs), which was not significantly different between the groups when under- and over-reporters were excluded.
Table 2 Comparison of average daily intake (excluding supplements) of energy, fat and specific fatty acids between the Mediterranean Diet (MD) group and the Control Diet (CD) group. The dietary intake is based on the diet history interviews performed between study weeks seven and twelve.
MD group (n = 17) CD group (n = 17) P-value*
Energy (MJ) 8.8 ± 1.6 9.8 ± 3.2 p = 0.242
Fat (g) 60.4 ± 21.9 89.3 ± 33.1 p = 0.005
Total saturated fatty acids (g) 18.3 ± 8.2 40.5 ± 18.3 p < 0.001
Total monounsaturated fatty acids (g) 25.3 ± 10.5 31.9 ± 11.1 p = 0.088
Total polyunsaturated fatty acids (g) 11.8 ± 3.9 10.6 ± 3.5 p = 0.381
Total n-6 fatty acids (g)† 7.9 ± 2.5 8.2 ± 2.7 p = 0.743
Total n-3 fatty acids (g)‡ 3.1 ± 1.3 2.0 ± 0.9 p = 0.008
Ratio n-6:n-3 2.7 ± 0.6 4.4 ± 0.9 p < 0.001
Fat (E%§) 25.0 ± 5.3 33.7 ± 5.6 p < 0.001
Total saturated fatty acids (E%) 7.5 ± 2.4 15.0 ± 3.7 p < 0.001
Total monounsaturated fatty acids (E%) 10.5 ± 2.8 12.2 ± 2.2 p = 0.067
Total polyunsaturated fatty acids (E%) 5.0 ± 1.1 4.1 ± 1.1 p = 0.028¶
Fatty acids (g):
4:0–10:0 0.77(0.32–1.59) 3.51(1.72–4.79) p < 0.001
12:0 0.44(0.28–0.93) 1.67(0.93–2.56) p < 0.001
14:0 1.81(1.07–2.90) 4.68(2.79–6.19) p < 0.001
16:0 9.46(7.29–13.29) 18.54(14.98–26.00) p < 0.001
18:0 2.80(1.99–4.34) 6.85(5.94–9.97) p < 0.001
20:0 0.14(0.10–0.19) 0.19(0.16–0.38) p = 0.016
16:1n-7 0.95(0.70–1.22) 1.25(1.05–1.84) p = 0.014
18:1n-9 20.50(14.98–28.05) 27.55(21.52–35.67) p = 0.049
18:2n-6 7.42(5.80–9.46) 7.90(5.78–9.77) p = 0.892
18:3n-3 1.79(1.23–2.36) 1.42(1.09–2.09) p = 0.454
20:4n-6 0.08(0.05–0.10) 0.08(0.04–0.13) p = 0.708
20:5n-3 0.35(0.23–0.59) 0.11(0.06–0.18) p < 0.001
22:5, n-3 and n-6 0.07(0.05–0.11) 0.02(0.01–0.05) p = 0.001
22:6n-3 0.73(0.44–1.03) 0.21(0.12–0.30) p < 0.001
Data are presented as mean ± SD for normally distributed variables and as medians (25th–75th percentiles) for variables with skew distributions. *The P-values refer to the difference between diet- and control group. Differences between groups were analyzed by the Students t-test for independent samples for normally distributed variables and by the Mann-Whitney U test for variables with skew distributions; †sum of 18:2n-6 and 20:4n-6; ‡sum of 18:3n-3, 20:5n-3 and 22:6n-3; §E% = percent of total energy. ¶Difference between groups regarding E% polyunsaturated fatty acids was not significant (p = 0.101) when under- and over-reporters were excluded.
According to the results from the diet history interviews, the MD group had a lower intake of fat and saturated fatty acids (SFAs), compared to the CD group (Table 2). Although there was no significant difference between the groups regarding the total intake of monounsaturated fatty acids (MUFAs), the intake of both oleic acid (OA; 18:1n-9) and palmitoleic acid (16:1n-7) was higher in the CD group. Regarding PUFAs, the reported intake of n-6 fatty acids did not differ significantly between the groups, while the intake of the long-chain n-3 fatty acids was higher in the MD group. In this group the total intake of the n-3 fatty acids (18:3n-3, 20:5n-3 and 22:6n-3) was 3.1 g/day (of which EPA+DHA constituted 1.2 g/day), versus 2.0 g/day (EPA+DHA = 0.4 g/day) in the CD group. This difference in turn resulted in a significantly lower ratio of ingested n-6 to n-3 fatty acids in the MD group. Conversely, there was no significant difference between the groups regarding the absolute intake (g/day), or in E%, of α-linolenic acid (α-LNA; 18:3n-3) and linoleic acid (LA; 18:2n-6). In the MD group, the E% α-LNA was 0.80 and the E% LA was 3.3, versus 0.62 E% α-LNA and 3.1 E% LA acid in the CD group. However, as a consequence of the difference between the groups regarding the total fat intake, both α-LNA, LA and OA contributed to a higher proportion of the fat intake in the MD group (mean intake of α-LNA, LA and OA expressed as g/100 g total fat was 3.2, 13.3 and 37.4 in the MD group, and 1.9, 9.4 and 32.7 in the CD group, p < 0.001, p < 0.001 and p = 0.012, respectively).
Fatty acid composition in serum phospholipids
At baseline there were no significant differences between the groups regarding the composition of fatty acids in serum phospholipids (Table 3). From baseline to the end of the study the percentage of the saturated fatty acid arachidic acid (20:0) increased slightly in the MD group. Increased relative amounts were also observed in the MD group regarding the n-3 fatty acids EPA and DHA, and for the total percentage of n-3 fatty acids. As regards n-6 fatty acids, there was a decrease in the relative amount of dihomo-γ-linolenic acid (20:3n-6) and in the total percentage of n-6 fatty acids in the MD group. The differences in the proportions of n-6 and n-3 fatty acids in this group, resulted in a decrease in the n-6:n-3 ratio. In the CD group there were no significant changes in any of the fatty acids analyzed. When the correlation between the reported intakes of individual fatty acids and their relative amounts in s-phospholipids was calculated, both the relative intake of EPA (g/100 g total fat) and the absolute intake of EPA (g/day) were significantly correlated to the percentage of EPA in s-phospholipids (Table 4). Also the relative and absolute intakes of DHA were related to the percentage of EPA in s-phospholipids. Regarding α-LNA there was a significant correlation between the absolute intake of α-LNA (g/day) and the level of α-LNA in s-phospholipids.
Table 3 Comparison of the fatty acid composition (per cent of total fatty acids) in serum phospholipids between the Mediterranean Diet (MD) group and the Control Diet (CD) group.
MD group (n = 26) CD group (n = 25) P-value*
Time (weeks after dietary shift) Time (weeks after start of study)
0 3 12 0 3 12
16:0 28.3 (23.6–31.0) 29.0 (24.6–32.2) 27.8 (25.3–31.3) 28.9 (27.2–31.7) 29.4 (27.1–32.3) 29.8 (26.6–32.1) 0.797
18:0 14.5 (13.1–15.2) 13.7 (12.2–15.3) 13.7 (12.4–15.7) 15.0 (13.2–17.1) 14.6 (13.3–16.5) 15.0 (13.3–17.0) 0.584
20:0 0.0 (0.0–0.1) 0.0 (0.0–0.2) 0.0 (0.0–0.3)† 0.0 (0.0–0.2) 0.0 (0.0–0.2) 0.0 (0.0–0.2) 0.100
24:0 1.4 (1.0–2.2) 1.4 (1.0–1.9) 1.6 (1.1–2.0) 1.4 (1.0–1.6) 1.1 (0.8–1.7) 1.3 (0.8–1.8) 0.959
Total saturated fatty acids 44.0 (39.1–46.8) 44.7 (37.8–48.3) 43.9 (39.6–49.4) 45.6 (41.4–49.9) 45.6 (41.1–49.1) 46.0 (42.7–49.9) 0.705
16:1n-7 0.8 (0.6–0.9) 0.7 (0.6–0.8) 0.8 (0.6–0.9) 0.7 (0.6–1.1) 0.8 (0.6–0.9) 0.8 (0.6–1.0) 0.339
18:1n-7 1.7 (1.3–1.8) 2.0 (1.4–2.2) 1.8 (1.6–2.0) 1.7 (1.3–1.9) 1.7 (1.5–2.1) 1.8 (1.5–2.0) 0.557
18:1n-9 9.9 (8.2–10.9) 9.7 (8.3–11.2) 9.4 (7.3–10.6) 10.2 (8.3–11.1) 10.6 (8.9–11.5) 10.9 (8.3–11.5) 0.087
22:1n-11 0.6 (0.2–1.6) 0.6 (0.2–1.5) 0.8 (0.8–1.8) 0.6 (0.0–1.3) 0.3 (0.0–1.0) 0.6 (0.0–1.1) 0.356
Total monounsaturated fatty acids 13.1 (11.4–13.8) 12.8 (11.2–14.6) 12.6 (11.3–13.8) 13.3 (11.8–14.4) 13.6 (11.6–14.7) 13.9 (11.4–15.0) 0.207
18:2n-6 16.7 (11.5–19.9) 16.5 (10.7–18.6) 14.4 (11.0–16.9) 16.0 (12.5–19.2) 16.7 (12.8–18.1) 17.0 (13.5–19.5) 0.044
20:2n-6 0.4 (0.2–0.7) 0.3 (0.2–0.5) 0.4 (0.2–0.6) 0.3 (0.2–0.5) 0.3 (0.2–0.5) 0.4 (0.3–0.6) 0.458
20:3n-6 2.6 (1.8–3.2) 2.5 (1.7–3.0) 2.2 (1.6–2.6)‡ 2.1 (1.5–3.1) 2.2 (1.7–2.8) 2.4 (1.8–3.1) 0.003
20:4n-6 6.9 (5.0–7.7) 6.8 (4.8–8.2) 6.6 (4.1–8.0) 5.9 (4.1–9.8) 6.6 (4.2–8.5) 6.5 (4.8–8.7) 0.144
Total n-6 fatty acids 25.2 (18.9–30.7) 26.8 (17.5–30.0) 23.8 (17.6–28.3)† 24.4 (18.4–32.2) 25.1 (19.3–29.9) 25.2 (21.5–30.5) 0.008
18:3n-3 0.3 (0.0–0.4) 0.2 (0.0–0.4) 0.4 (0.0–0.4) 0.3 (0.2–0.4) 0.3 (0.1–0.5) 0.3 (0.0–0.5) 0.395
20:5n-3 1.4 (1.1–1.8) 1.8 (1.1–2.8) 2.1 (1.4–3.3)§ 1.1 (1.0–1.5) 1.1 (1.0–1.7) 1.1 (0.9–1.4) <0.001
22:5n-3 0.7 (0.2–1.7) 0.6 (0.2–1.4) 0.8 (0.2–1.7) 0.5 (0.2–1.3) 0.6 (0.2–1.0) 0.6 (0.3–1.0) 0.395
22:6n-3 3.9 (2.2–4.8) 4.2 (3.1–6.6) 5.0 (2.5–7.2)§ 3.0 (1.5–4.3) 3.5 (1.7–4.2) 3.0 (2.2–4.5) <0.001
Total n-3 fatty acids 6.1 (5.2–7.4) 7.3 (5.3–10.0) 8.6 (6.0–11.7)§ 5.2 (4.3–7.0) 4.9 (3.7–6.4) 5.2 (3.8–7.0) 0.001
n-6:n-3 ratio 4.2 (3.0–5.6) 3.1 (2.7–3.7) 2.5 (2.1–3.5)‡ 4.6 (3.6–5.6) 4.4 (3.4–5.6) 4.9 (4.1–5.6) 0.002
Total polyunsaturated fatty acids 32.7 (25.9–38.4) 33.5 (24.4–40.1) 32.4 (22.7–39.0) 29.3 (22.3–37.5) 31.0 (24.0–36.6) 30.9 (25.7–37.4) 0.072
Data are presented as medians (25th–75th percentiles). *The p-values refer to differences between the MD- and the CD group concerning the change from baseline to week 12. Differences between the groups were analyzed with the Mann-Whitney U test. Statistically significant change from baseline to week 12: †p < 0.05, ‡p < 0.01, §p < 0.001. Within-group differences in week 12 compared to baseline were evaluated by the Wilcoxon signed ranks test.
Table 4 Spearman's correlation coefficients between the reported intake of fatty acids or foods and fatty acids in serum phospholipids.
Fatty acids in s-phospholipids
Dietary fatty acids/foods EPA DHA α-LNA total n-3
EPA g/day 0.58* 0.24 0.04 0.43†
EPA g/100 g total fat 0.58* 0.32 0.04 0.49*
DHA g/day 0.60* 0.19 -0.05 0.40†
DHA g/100 g total fat 0.57* 0.26 -0.09 0.45*
α-LNA g/day 0.13 -0.09 0.38† 0.03
α-LNA g/100 g total fat 0.24 0.11 0.29 0.20
Change in fish intake‡ 0.48* 0.46* 0.14 0.51*
Change in shellfish intake‡ 0.33† 0.48* 0.03 0.34†
*p < 0.01; †p < 0.05; ‡change from baseline to week 12 in relation to change in percentage of fatty acids in s-phospholipids.
Furthermore, changes in the reported consumption frequencies of fish and shellfish were positively related to changes in EPA, DHA and the total percentage of n-3 fatty acids in s-phospholipids (Table 4).
Differences between diet responders and diet non-responders
When the MD group was divided into diet responders (n = 15) and diet non-responders (n = 11), based on their clinical improvement during the study, the diet responders had a higher reported relative intake (fatty acids expressed in relation to the total fat intake) of EPA, docosapentaenoic acid (22:5, which in the food composition database used includes both n-3 and n-6 isomers) and DHA compared to the diet non-responders (p = 0.032, p = 0.040 and p = 0.011 respectively). This resulted in a lower intake ratio of n-6 to n-3 fatty acids in the group of diet-responders. The median n-6 to n-3 ratio was 2.2 in the diet responders and 3.0 in the diet non-responders (p = 0.032). Also the intake of the MUFA palmitoleic acid (16:1n-7), in relation to the total fat intake, was higher in the diet responders (p = 0.015), while the E% of total MUFAs was higher in the diet non-responders (p = 0.048). However, when the intake of total MUFAs was expressed in relation to the total fat intake there was no significant difference between the groups (p = 0.380).
Regarding the fatty acid composition in serum phospholipids there were some differences between the diet responders and the diet non-responders at baseline. The diet responders had a higher relative amount of lignoceric acid (24:0) and lower relative amounts of arachidic acid (20:0), vaccenic acid (18:1n-7) and α-LNA (p = 0.042, p = 0.022, p = 0.007 and p = 0.024 respectively). The change in fatty acid composition from baseline to the end of the study also differed between the groups. The relative amounts of the SFAs palmitic acid (16:0) and stearic acid (18:0), as well as the total SFAs, increased in the diet non-responders, while these fatty acids decreased in the diet responders (p = 0.040, p = 0.033 and p = 0.007 respectively for differences between the groups). On the other hand, the n-3 fatty acid docosapentaenoic acid (22:5n-3) increased in the diet responders while a slight decrease was observed in the diet non-responders (p = 0.048). Regarding the other n-3 fatty acids in serum phospholipids, no significant differences were found between the groups.
No significant differences between the groups were observed concerning the reported consumption of food groups. However, from baseline to the end of the study there was a tendency towards a greater increase in the fish consumption (p = 0.059), as well as a greater decrease in the consumption of dairy products with 10–15% fat (p = 0.076), in the group of diet responders compared to the non-responders.
Discussion
Dietary intake of fat and specific fatty acids
In this study we used a questionnaire to estimate the reported consumption of foods with relevance to the fatty acid intake, and diet history interviews to assess the intake of fat and specific fatty acids. In the MD group, the changes in the reported consumption of food items were in line with the advice given to the patients, i.e. to replace meat, and processed meat with fish and poultry, to use olive oil, canola oil or margarine based on canola oil, and to choose low fat dairy products. This was also reflected in the reported intake of fat assessed by the diet history interviews. For instance, choosing low fat dairy products instead of higher fat alternatives is likely to result in a decreased intake of both total fat and of SFAs. Thus, both assessments of dietary intake indicated that the MD group changed their dietary intake in the direction aimed at.
The traditional Cretan Mediterranean diet is well known for its high content of MUFAs from olive oil. According to a study by Cleland at al, the incorporation of EPA into neutrophil membranes after fish-oil supplementation was higher in healthy subjects eating a diet high in OA and low in LA, compared to subjects on a diet high in LA [15]. Thus, when the dietary intake of OA is high in relation to LA, this may be of advantage to the n-3 fatty acids in the competition between n-6 and n-3 fatty acids. In this study we also used canola oil which, similar to olive oil, is rich in the MUFA OA. Nevertheless, there was a tendency towards a lower intake of MUFAs in the MD group (p = 0.088). This was probably a result of the considerably lower intake of total fat in this group, since the relative amount of OA was still significantly higher than in the CD group.
Among the most important characteristics of the experimental diet was the amount of n-3 fatty acids and the relation between n-6 and n-3 fatty acids. The original Cretan Mediterranean diet was rich in n-3 fatty acids from both animal sources and plant sources [16]. In the present study, the MD group was instructed to use canola oil and margarine based on canola oil, which are good sources of α-LNA. This, together with an increased intake of fish and shellfish, resulted a significantly lower ratio of n-6:n-3 fatty acids in the MD group (2.7 versus 4.4). The n-6:n-3 ratio has been reported to be 1–2 in the traditional diet of Greece, while the ratio in many Western countries is around 15–17 [17]. However, in the Swedish nationwide dietary survey "Riksmaten 1997–98" [18], where the patterns of food and nutrient intake in Sweden were studied on a nationwide basis, the ratio n-6:n-3 was reported to be around 5. This is not far from the ratio in our CD group, while the MD group obtained a ratio closer to that of the traditional Greek diet. However, it is important to remember, that when data from different countries and different studies are compared, different fatty acids may have been used to calculate these ratios.
A modified Cretan Mediterranean diet, similar to our experimental diet, has been tested by de Lorgeril et al [19,20], in a study on secondary prevention of coronary heart disease. In that study the aim was to reduce the intake of LA to 4 E% and the intake of α-LNA was to exceed 0.6 E%. After eight weeks the experimental group in their study had an LA intake of 3.6 E% and an α-LNA intake of 0.76 E%. In the present study, the E% of LA was even lower and the E% of α-LNA was almost the same in the MD group as in the study by de Lorgeril et al. However, the intake of LA and α-LNA in our CD group was not very different from the MD group. This may be explained by the relatively high use of canola oil and margarine based on canola oil in Sweden. Although we encouraged our MD group to increase the intake of these α-LNA sources, these foods are also important sources of n-3 fatty acids in the Swedish population in general [18]. Thus, regarding essential fatty acids, even the intake of the CD group was close to that of the experimental group in the study by de Lorgeril et al [20].
In the present study, the reported total intake of n-3 fatty acids in the MD group was 3.1 g/day and of these 1.2 g were EPA + DHA. Most studies of dietary supplementation with fish oil have used between 1 and 7.1 g EPA + DHA per day [2]. On the basis of these studies, a daily intake of 3–6 g long-chain n-3 fatty acids (usually EPA + DHA) has been recommended in order to improve RA symptoms [5,21]. Hence, although the MD group reported an increased consumption of fish, the intake of long chain n-3 PUFAs only reached the lower range of the amounts used in studies of fish oil supplementation. However, in these studies n-3 fatty acids are usually added to the diet, regardless of the intake of other fatty acids. In a few supplementation studies the amount of competitor n-6 fatty acids has been considered as well [3,22,23]. Volker et al [3] investigated the effect of fish oil supplementation (40 mg/kg body weight/day) in RA subjects with a background diet of <10 g n-6 fatty acids. After 15 weeks, the experimental group achieved substantial incorporation of EPA in plasma and monocyte phospholipids, and significant improvements in six of nine outcome measures. Furthermore, in a study by Adam et al [22] the effect of fish oil supplementation (30 mg/kg body weight/day) in two groups on different diets was studied. One group were on a diet low in arachidonic acid (AA; 20:4n-6; less than 90 mg/day) and the other group ate a normal western diet. Patients from both groups were randomized to receive either fish oil capsules or placebo capsules for three months, in a double-blind, crossover manner. The results of the study showed that patients on the AA-low diet improved more than the other group regarding tender and swollen joints. As the authors conclude, this indicates that the ratio of AA:EPA is decisive for the clinical effectiveness of fish oil supplementation. In the present study both groups reported a daily intake of less than 10 g of n-6 fatty acids, and less than 90 mg AA. Based on the studies mentioned above, the low background intake of n-6 fatty acids may have amplified the effect of the Mediterranean diet.
Fatty acid composition in serum phospholipids
The dietary intake of fat is often difficult to assess by means of dietary assessment methods. Therefore, there is a need for objective markers of fatty acid intake to be used as a complement to dietary assessment methods [24,25]. In the present study, we measured fatty acids in serum phospholipids, which are influenced by the dietary intake of the past few days [26]. The changes we found regarding the fatty acid profile were in agreement with the reported nutrient intake, i.e. an increase in the relative amount of n-3 fatty acids. We also observed a decrease in n-6 fatty acids in serum phospholipids. However, since an increased relative amount of one fatty acid cause a decrease in the percentage of another, this is not necessarily a consequence of a decreased intake of n-6 fatty acids. Somewhat surprisingly, the saturated fatty acid arachidic acid (20:0) increased marginally, but significantly, in the MD group. One possible reason for this could be an increased consumption of canola oil. Although the amount of arachidic acid is low in canola oil, this oil contains more arachidic acid than other oil, margarine or butter.
The use of fatty acid patterns in serum phospholipids as biomarkers of the dietary intake is, limited by the fact that not only diet affects the levels of fatty acids in biological specimens [24]. For instance, most fatty acids can be synthesized by humans and, within the body, fatty acids can also be converted to other fatty acids by means of desaturation and elongation. Furthermore, the pattern of plasma fatty acids in phospholipids has been reported to be altered in RA patients compared to healthy subjects [27]. In this study, the strongest correlation between the intake of fatty acids and the corresponding biomarkers, was observed regarding EPA (rs = 0.58 for both absolute and relative intakes). Similar, as well as both stronger and weaker correlations have been reported by others, possibly due to variations in the quality of the dietary data [28-30]. In other studies the percentage of DHA in phospholipids has been correlated to the intake of this fatty acid [28,29], while in the present study, the intake of DHA was only correlated to the percentage of EPA in phospholipids. The reason for this is probably that the dietary sources of EPA and DHA are the same. Also a significant correlation between α-LNA in serum phospholipds and the absolute intake of this fatty acid (rs = 0.38) was found in the present study. This result is partly in line with a study by Sasaki et al [31], while others have reported only weak correlation regarding α-LNA [29,30].
Sometimes, the relative fatty acid levels reflect the intake of specific foods, especially when the food item is the major source of the fatty acid in question. In this study changes in the reported consumption of fish and shellfish were related to changes in the percentage of both EPA, DHA and total n-3 fatty acids, with Spearman's correlation ranging from 0.46 to 0.51 for fish and from 0.33 to 0.48 for shellfish. Regarding fish intake, this is in agreement with other studies [28,32,33]. The correlation between shellfish intake and long-chain n-3 fatty acids may also be related to the consumption of fish. Although the reported intake of shellfish increased significantly in the MD group, the level of shellfish consumption was not very high in this study. Thus, compared to the intake of fish, shellfish was probably not a major source of n-3 fatty acids. Furthermore, there was a significant correlation between the change in fish intake and the change in shellfish intake (data not shown), indicating that the individuals reporting an increased intake of shellfish are likely to be the same individuals as those with an increased fish intake.
Can the altered intake of fatty acids explain the beneficial effects on RA disease activity?
To analyse the possible connection between the altered intake of fatty acids in the MD group and the improvement regarding disease activity, we divided the MD group into diet responders and diet non-responders, based on the individual change in disease activity during the study. It is important to remember that only 15 (diet responders) and 11 (diet non-responders) patients were compared in this analysis. Hence, the power to detect differences between diet responders and diet non-responders was fairly low. Still, when these two groups were compared differences in both the reported intake of fatty acids and the fatty acid composition in serum phospholipids were found. Overall, the results point towards a more favourable fatty acid intake in the group of diet responders, at least regarding PUFAs, which in turn suggests a better compliance to the experimental diet in this group. Thus, the differences found between diet responders and diet non-responders indicate that the fatty acid intake contributed to the positive effect of this diet on RA disease activity.
Conclusion
In conclusion, the changes in the reported consumption of food items in the MD group were in line with the advice given during the study. As a result, the total fat intake was lower in the MD group compared to the CD group, and in the MD group a lower percentage of the energy intake was derived from SFAs. The MD group also had a lower intake ratio of n-6:n-3 fatty acids. A corresponding change in the relation between n-6 and n-3 fatty acids was observed in s-phospholipids. Furthermore, the MD patients who were characterised as diet responders, had a higher reported intake of long chain n-3 fatty acids and a lower ratio of n-6 to n-3 fatty acids compared to the diet non-responders. Also the fatty acid profile in serum phospholipids differed in part between the diet responders and the diet non-responders. These findings point towards a better compliance to the experimental diet in the group of diet responders.
Altogether, the changes in the fatty acid profile indicated by three methods, a questionnaire, diet history interviews and fatty acids in s-phospholipids may, at least in part, explain the beneficial effects on the clinical measurements demonstrated earlier [1]. Thus, the results of this study support previous studies indicating the importance of the fatty acid intake in patients with RA.
Methods
Patients, study design and the experimental diet
The study was a randomised, parallel, dietary intervention study over three months. In total, 56 patients with RA were included in the study. Of these, 26 MD subjects (21 women and 5 men, mean age 58 years) and 25 CD subjects (20 women and 5 men, mean age 59 years) completed the study. The patients, study design and the dietary intervention have been described in detail elsewhere [1,34]. In brief, the patients were randomized to either a modified Cretan Mediterranean diet group or a control diet group, by means of block randomization stratified for sex. At baseline the two groups were equal except for the disease duration and the body mass index (BMI). The MD group had a significantly higher BMI and a longer disease duration compared to the CD group (p = 0.024 and 0.047, respectively).
The experimental diet used in the present study was based on the Cretan Mediterranean diet previously tested by de Lorgeril et al, in a secondary prevention study of coronary heart disease [19]. However, some modifications of the diet were done in order to suit Swedish food habits. We instructed our MD group to eat a large amount of vegetables, fruit, pulses, cereals, fish (particularly fish with a high content of ω-3 fatty acids) and nuts and seeds with a high content of α-LNA. The intake of meat (such as pork, beef, lamb or mutton) and processed meat (including cured meat, sausage, pâté or the like) were to be replaced by poultry, fish or vegetarian dishes. Both olive oil and canola oil were used in salad dressings and for food preparation. The MD group was also informed to use two types of margarine based on canola oil: a liquid margarine (80% fat) for food preparation and half-fat margarine (40% fat) to use on bread. In addition, the MD group was advised to replace high fat dairy products with low fat products. In the present study, no recommendations were given regarding alcohol consumption. To compensate for the antioxidants in wine, we advised the MD group to drink green or black tea.
To promote good compliance with the Mediterranean diet some food items were supplied free to the MD group, namely: frozen vegetables, tea, olive oil, canola oil and the liquid and half-fat margarine based on canola oil. Olive oil and canola oil, were supplied by Karlshamns AB, vegetables by Nestlé Sweden AB and margarine and tea by Van den Bergh Foods AB.
The CD subjects were instructed to adhere to their ordinary diet. If the subjects of the study used any dietary supplementation (e.g. fish oils, vitamins, minerals, etc.) prior to the study this was recorded. All such supplementation had to be kept unchanged during the study.
Dietary assessments
To assess the dietary intake both a self-administered questionnaire and diet history interviews were used. The questionnaire was completed by the patients of both groups at baseline, in week three and twelve. This questionnaire included both open and closed questions, mainly concerning food choices, and it was designed to investigate compliance with the Mediterranean diet. Regarding questions on frequencies of food intake, the subjects should state their average intake of various food items by marking one of six alternatives ranging from "rarely or never" to "two or more times per day". To enable comparisons of the consumption between the two groups, as well as consumption at different points in time, the food frequencies were converted to average consumption per month. For example, if a consumption of 3–5 times per week was marked this would be converted to 16 times per month.
Diet history interviews, covering the dietary intake during the past month, were carried out to obtain more detailed data on the energy and nutrient intake during the intervention period. The interviews were performed between study weeks seven and twelve and were conducted with 34 patients from both the MD group and the CD group (15 women and 2 men from each group). The only selection criterion for taking part in the diet history interviews was that the subjects were included in the study on February 15th 1999, or later. Fore more details regarding the questionnaire and the diet history interviews see reference 34.
The nutritional analysis package MATs 4_03e was used to calculate the estimated intake of energy and nutrients based on the diet history interviews. This program is based on the Swedish National Food Administration's food composition database, PC-kost (version 2_97). If composite food items and supplements not listed in the database were reported, the nutrient content was entered manually. The energy and nutrient intake was calculated both including and excluding dietary supplements. Still, when dietary supplements were included, the total intake of n-3 fatty acids increased marginally in the MD group and not at all in the CD group. Therefore, only the dietary intake without supplements will be presented. However, when the correlation between the reported dietary intake of fatty acids and their relative content in s-phospholipids were analyzed, the results from the diet history interviews including supplements were used.
Validation of the diet history interviews
A validation of the diet history interviews by means of the doubly labelled water method (performed for nine subjects) and biological markers of the intake of protein, sodium and potassium has been presented elsewhere [35]. To identify under- and over-reporters we used the individual Goldberg cut-offs for the FILs [14]. These cut-offs are based on the physical activity level for each subject assessed by a three-day activity registration, which has been described earlier [35]. The FIL is the individuals reported energy intake divided by the estimated basal metabolic rate. The basal metabolic rate was estimated based on body weight, age group and sex, according to standard equations [36].
Categorization of diet responders and diet non-responders
In this intervention study the disease activity was assessed by means of a combined index called disease activity score from 28 joints (DAS28) [37]. The patients in the MD group who had a moderate or better clinical improvement from baseline to the end of the study, which is defined as a decrease of >0.6 in DAS28, were categorized as diet responders and the remaining MD patients as diet non-responders.
Determination of the composition of fatty acids in serum phospholipids
Sampling
Blood samples for the analysis of fatty acids in serum phospholipids were taken at baseline and in weeks three and twelve. After a one-night fast, samples were collected in tubes without additives. The blood was cooled in ice water for 30 min before centrifugation at 1500 g for 15 min at +4°C. The serum was separated and stored at -70°C until analysis.
Folch extraction
Serum (1 ml) was extracted using methanol (2 ml) and chloroform (4 ml). Potassium chloride, 0,88%; (2 ml with the water content of the sample excluded) was added and the mixture was shaken thoroughly for 30 seconds before being allowed to settle. The aqueous layer was removed, one fourth of the volume of the organic phase of methanol-saline (1:1, v/v) was added and the washing procedure was repeated. The bottom layer containing the purified lipids was collected and the solvent was evaporated in a gentle stream of nitrogen at +30°C [38].
Separation of lipid classes using thin layer chromatography (TLC)
The purified lipids were solved in 40 μl chloroform containing 0,05% (v:v) of antioxidant butylated hydroxytoluene (BHT, 10% in ethanol (w:v)) and separated on silica gel plates 60, F254 (MERCK) by TLC, using petroleumether: diethyl ether: acetic acid (81:18:1). The different lipid classes were visualized using UV-light at 254 nm. The standards used were L-α-Phosphatidylcholine and L-α-Phosphatidylethanolamine, Larodan Fine Chemicals. The phospholipid spots were transferred into screw-capped glass tubes for preparation of fatty acid methyl esters [39].
Transesterification using sodium methoxide
The silica gel containing the lipids was dissolved in sodium-dried diethyl ether (0,5 ml) and methyl acetate (20 μl). Sodium methoxide, 0,5 M, in dry methanol (20 μl) was added. After 15 minutes at room temperature the reaction was stopped by the addition of acetic acid (2 μl). The solvent was evaporated in a gentle stream of nitrogen at +30°C. Iso-hexane with BHT (0,05%; 1 ml) was added and the mixture was centrifuged at 1500 g for 2 minutes. The supernatant layer was removed using a Pasteur pipette into a sample tube [40].
GC analysis
The composition of the fatty acids was determined using gas chromatography. Of the purified methylated sample 1 μl was injected into a HP 5890 series II gas chromatograph suited with a FID and an HP-FFAP capillary column (30 m × 0,25 mm × 0,25 μm). The He carrier gas flow was 13 ml/min. The detector temperature was 260°C. The injector split 1:50 at 220°C. A temperature programme was used with an initial temperature of 160°C held for 5 minutes, raised from 160 to 220°C at a rate of 2°C/min and 220°C held for 30 minutes.
The fatty acid methyl esters were identified by comparison with the retention times of the standards: GLC-68D and GLC-68A, Nu-Chek-Prep and Qualimix Fish 89 – 5540, Larodan Fine Chemicals.
The individual fatty acids were expressed as the percentage of total fatty acids (relative amount). The relative amounts were quantified by integrating the area under the peak and dividing the results by the total area of all fatty acids.
Statistical methods
The statistical analyses were performed using SPSS for Windows version 11.0.1. The differences in fat intake between groups were analyzed using the Student's t-test for independent samples, when the variables were normally distributed. For fatty acids with skewed distributions the Mann-Whitney U test was used. As regards the reported consumption of food items and the composition of fatty acids in serum phospholipids, the Mann-Whitney U-test was performed to test differences between groups at baseline. For these variables, within group differences from baseline to week 12 were evaluated by means of the Wilcoxon signed ranks test.
The Spearman's rank correlation was used to evaluate the association between the reported dietary intake of fatty acids (assessed by diet history interviews including supplements) and their relative content in s-phospholipids. When calculating the correlation, the mean value of the results from week six and twelve were used regarding fatty acids in s-phospholipids, since the diet history interviews were performed between study weeks seven and twelve.
Spearman's rank correlation was also used to evaluate the association between the reported consumption of fish (estimated by means of the questionnaire) and the long chain n-3 fatty acids in s-phospholipids. For these assessments the change from baseline to week 12 was used.
All results were considered statistically significant at a two-tailed p-value of <0.05.
List of abbreviations
AA, arachidonic acid; BMI, body mass index; CD, control diet; DAS 28, disease activity score from 28 joints; DHA, docosahexaenoic acid; EPA, eicosapentaenoic acid, E%, percentage of the energy intake; FIL, food intake level; LA, linoleic acid; α-LNA, α-Linolenic acid; MD, Mediterranean diet; MUFA, monounsaturated fatty acids; OA, oleic acid; PUFA, polyunsaturated fatty acids; RA, rheumatoid arthritis; SFA, saturated fatty acids
Competing interests
The author(s) declare that they have no competing interests.
Authors' contributions
LH participated in the conception and design, acquisition of data, data analysis, interpretation of data and the writing of the manuscript. IN participated in the data analysis, the critical revision of the manuscript and the writing of parts of the paper. LS participated in the conception and design, acquisition of data, interpretation of data and the critical revision of the manuscript. GJ participated in the conception and design, data analysis, interpretation of data, and the critical revision of the manuscript. All authors have read and approved the final manuscript.
Acknowledgements
We would like to express our gratitude to our collaborators, Lena Martinsson, Eva Wolke, Lena Henningsson, Marianne Olsson, Ann-Louise Karlsson, Mona Bäckström and Gunnel Gustavsson for administrative assistance, information, advice and support to the patients concerning their diet, help with the clinical examinations, and in the handling of blood samples for chemical analyses.
We would like to thank Maria Bengtsson and Jenny Wannstedt for performing diet history interviews, Petra Rydén for her help with the processing of food questionnaires, and Christel Larsson, Ulla Johansson, Magdalena Rosell, Agneta Hörnell, Anette Jonsäll and Ylva Mattsson-Sydner for valuable discussions.
We also wish to express our gratitude to all the participants.
The study was supported by grants from the Faculty of Social Sciences of Umeå University, the Swedish Foundation for Health Care Sciences and Allergy Research, the Health Research Council in the Southeast of Sweden, the Swedish Rheumatism Association, the Swedish Nutrition Foundation, the JC Kempe Memorial Scholarship Fund, the "Borgerskapet i Umeå" Fund and the "Uppsala Hemsysterskola" Fund.
==== Refs
Sköldstam L Hagfors L Johansson G An experimental study of a Mediterranean diet intervention for patients with rheumatoid arthritis Ann Rheum Dis 2003 62 208 214 12594104 10.1136/ard.62.3.208
Calder PC Zurier RB Polyunsaturated fatty acids and rheumatoid arthritis Curr Opin Clin Nutr Metab Care 2001 4 115 121 11224655
Volker D Fitzgerald P Major G Efficacy of fish oil concentrate in the treatment of rheumatoid arthritis J Rheumatol 2000 27 2343 2346 11036827
James MJ Cleland LG Dietary n-3 fatty acids and therapy for rheumatoid arthritis Semin Arthritis Rheum 1997 27 85 97 9355207 10.1016/S0049-0172(97)80009-1
Kremer JM n-3 Fatty acid supplements in rheumatoid arthritis Am J Clin Nutr 2000 349 351
Fortin PR Lew RA Liang MH Wright EA Beckett LA Chalmers TC Sperling RI Validation of a meta-analysis: the effects of fish oil in rheumatoid arthritis J Clin Epidemiol 1995 48 1379 1390 7490601 10.1016/0895-4356(95)00028-3
Cleland LG French JK Betts WH Murphy GA Elliott MJ Clinical and biochemical effects of dietary fish oil supplements in rheumatoid arthritis J Rheumatol 1988 15 1471 1475 2849682
Kremer JM Lawrence DA Jubiz W DiGiacomo R Rynes R Bartholomew LE Sherman M Dietary fish oil and olive oil supplementation in patients with rheumatoid arthritis. Clinical and immunologic effects Arthritis Rheum 1990 33 810 820 2363736
Brzeski M Madhok R Capell HA Evening primrose oil in patients with rheumatoid arthritis and side-effects of non-steroidal anti-inflammatory drugs Br J Rheumatol 1991 30 370 372 1913008
Beauchamp GK Keast RSJ Morel D Lin J Pika J Han Q Lee C-H Smith AB Breslin PAS Phytochemistry: ibuprofen-like activity in extra-virgin olive oil Nature 2005 437 45 6 16136122 10.1038/437045a
Shapiro JA Koepsell TD Voigt LF Dugowson CE Kestin M Nelson JL Diet and rheumatoid arthritis in women: a possible protective effect of fish consumption Epidemiology 1996 7 256 263 8728438
Linos A Kaklamanis E Kontomerkos A Koumantaki Y Gazi S Vaiopoulos G Tsokos GC Kaklamanis PH The effect of olive oil and fish consumption on rheumatoid arthritis – a case control study Scand J Rheumatol 1991 20 419 426 1771399
Linos A Kaklamani VG Kaklamani E Koumantaki Y Giziaki E Papazoglou S Mantzoros C Dietary factors in relation to rheumatoid arthritis: a role for olive oil and cooked vegetables? Am J Clin Nutr 1999 70 1077 1082 10584053
Black AE Critical evaluation of energy intake using the Goldberg cut-off for energy intake:basal metabolic rate. A practical guide to its calculation, use and limitations Int J Obes 2000 24 1119 1130 10.1038/sj.ijo.0801376
Cleland LG James MJ Neumann MA D'Angelo M Gibson RA Linoleate inhibits EPA incorporation from dietary fish-oil supplements in human subjects Am J Clin Nutr 1992 55 395 399 1310374
Simopoulos AP The Mediterranean diets: What is so special about the diet of Greece? The scientific evidence J Nutr 2001 3065 3073
Simopoulos AP The importance of the ratio of omega-6/omega-3 essential fatty acids Biomed Pharmacother 2002 56 365 379 12442909 10.1016/S0753-3322(02)00253-6
Becker W Pearson M Riksmaten 1997–98: Kostvanor och näringsintag i Sverige Metod-och resultatanalys Uppsala: Livsmedelsverket 2002
de Lorgeril M Renaud S Mamelle N Salen P Martin J-L Monjaud I Guidollet J Touboul P Delaye J Mediterranean alpha-linolenic acid-rich diet in secondary prevention of coronary heart disease Lancet 1994 343 1454 1459 7911176 10.1016/S0140-6736(94)92580-1
de Lorgeril M Salen P Modified Cretan Mediterranean diet in the prevention of coronary heart disease and cancer World Rev Nutr Diet 2000 87 1 23 10929524
Cleland LG James MJ Fish oil and rheumatoid arthritis: antiinflammatory and collateral health benefits J Rheumatol 2000 27 2305 2307 11036821
Adam O Beringer C Kless T Lemmen C Adam A Wiseman M Adam P Klimmek R Forth W Anti-inflammatory effects of a low arachidonic acid diet and fish oil in patients with rheumatoid arthritis Rheumatol Int 2003 23 27 36 12548439
Sundrarjun T Komindr S Archararit N Dahlan W Puchaiwatananon O Angthararak S Udomsuppayakul U Chuncharunee S Effects of n-3 fatty acids on serum interleukin-6, tumour necrosis factor-alpha and soluble tumour necrosis factor receptor p55 in active rheumatoid arthritis J Int Med Res 2004 32 443 454 15458276
Arab L Biomarkers of fat and fatty acid intake J Nutr 2003 925 932
Bates CJ Thurnham DI Bingham SA Margetts BM Nelson M Margetts BM, Nelson M Biochemical markers of nutrient intake Design concepts in nutritional epidemiology 1997 2 Oxford: Oxford University Press 170 240
Kohlmeier L Biomarkers of fatty acid exposure and breast cancer risk Am J Clin Nutr 1997 1548 1556
Navarro E Esteve M Olivé A Klaassen J Cabré E Tena X Fernández-Bañares F Pastor C Gassull MA Abnormal fatty acid pattern in rheumatoid arthritis. A rationale for treatment with marine and botanical lipids J Rheumatol 2000 27 298 303 10685788
Andersen LF Solvoll K Drevon CA Very-long-chain n-3 fatty acids as biomarkers for intake of fish and n-3 fatty acid concentrates Am J Clin Nutr 1996 64 305 311 8780338
Kobayashi M Sasaki S Kawabata T Hasegawa K Akabane M Tsugane S Single measurement of serum phospholipid fatty acid as a biomarker of specific fatty acid intake in middle-aged Japanese men Eur J Clin Nutr 2001 55 643 650 11477462 10.1038/sj.ejcn.1601194
Ma J Folsom AR Shahar E Eckfeldt JH Plasma fatty acid composition as an indicator of habitual dietary fat intake in middle-aged adults. The Atherosclerosis Risk in Communities (ARIC) Study Investigators Am J Clin Nutr 1995 62 564 571 7661118
Sasaki S Ushio F Amano K Morihara M Todoriki O Uehara Y Toyooka E Serum biomarker-based validation of a self-administered diet history questionnaire for Japanese subjects J Nutr Sci Vitaminol (Tokyo) 2000 46 285 296 11227800
Bjerve KS Brubakk AM Fougner KJ Johnsen H Midthjell K Vik T Omega-3 fatty acids: essential fatty acids with important biological effects, and serum phospholipid fatty acids as markers of dietary omega 3-fatty acid intake Am J Clin Nutr 1993 801 805
Amiano P Dorronsoro M de Renobales M Ruiz de Gordoa JC Irigoien I Very-long-chain omega-3 fatty acids as markers for habitual fish intake in a population consuming mainly lean fish: the EPIC cohort of Gipuzkoa. European Prospective Investigation into Cancer and Nutrition Eur J Clin Nutr 2001 55 827 832 11593343 10.1038/sj.ejcn.1601242
Hagfors L Leanderson P Sköldstam L Andersson J Johansson G Antioxidant intake, plasma antioxidants and oxidative stress in a randomized, controlled, parallel, Mediterranean dietary intervention study on patients with rheumatoid arthritis Nutrition Journal 2003 2 5 12952549 10.1186/1475-2891-2-5
Hagfors L Westerterp K Sköldstam L Johansson G Validity of reported energy expenditure and reported intake of energy, protein, sodium and potassium in rheumatoid arthritis patients in a dietary intervention study Eur J Clin Nutr 2005 59 238 245 15483633 10.1038/sj.ejcn.1602064
Department of Health Dietary reference values for food energy and nutrients for the United Kingdom Report on health and social subjects 41 1991 London: HMSO
van Riel PLCM van Gestel AM Clinical outcome measures in rheumatoid arthritis Ann Rheum Dis 2000 28 31 10.1136/ard.59.suppl_1.i28
Folch J Lees M Stanley GHS A simple method for the isolation and purification of total lipids from animal tissues J Biol Chem 1957 226 497 509 13428781
Luostarinen R Boberg M Saldeen T Fatty acid composition in total phospholipids of human coronary arteries in sudden cardiac death Atherosclerosis 1993 99 187 193 8503947 10.1016/0021-9150(93)90021-L
Christie WW A simple procedure for rapid transmethylation of glycerolipids and cholesteryl esters J Lipid Res 1982 23 1072 1075 6897259
|
16216119
|
PMC1289295
|
CC BY
|
2021-01-04 16:37:46
|
no
|
Nutr Metab (Lond). 2005 Oct 10; 2:26
|
utf-8
|
Nutr Metab (Lond)
| 2,005 |
10.1186/1743-7075-2-26
|
oa_comm
|
==== Front
PLoS Comput BiolPLoS Comput. BiolpcbiplcbploscompPLoS Computational Biology1553-734X1553-7358Public Library of Science San Francisco, USA 1632276110.1371/journal.pcbi.001005205-PLCB-MI-0254plcb-01-06-09Message from ISCBBioinformatics Education—Perspectives and Challenges Message from ISCBRanganathan Shoba Shoba Ranganathan is Professor and Chair of Bioinformatics, Macquarie University, Sydney, Australia; Adjunct Professor at the National University of Singapore, Singapore; and Organizer of the Workshop on Education in Bioinformatics, ISCB Special Interest Group. E-mail: [email protected]
[email protected]
11 2005 25 11 2005 1 6 e52Copyright: © 2005 Shoba Ranganathan.2005This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are properly credited. Citation:Ranganathan S (2005) Bioinformatics education—Perspectives and challenges. PLoS Comput Biol 1(6): e52.
==== Body
Education in bioinformatics has undergone a sea change, from informal workshops and training courses to structured certificate, diploma, and degree programs—spanning casual self-enriching courses all the way to doctorate programs. The evolution of curriculum, instructional methodologies, and initiatives supporting the dissemination of bioinformatics is presented here.
Building on the early applications of informatics (computer science) to the field of biology, bioinformatics research entails input from the diverse disciplines of mathematics and statistics, physics and chemistry, and medicine and pharmacology. Providing education in bioinformatics is challenging from this multidisciplinary perspective, and represents short- and long-term efforts directed at casual and dedicated learners in academic and industrial environments. This is an NP-hard problem.
Training in bioinformatics remains the oldest and most important rapid induction approach to learning bioinformatics skills. Both formal (short-term courses) and informal training (on-demand “how-to” procedures) have remained the mainstay of on-the-job programs.
After almost a decade of short-term training, and retraining students, faculty, and scientists in discrete aspects of bioinformatics, the impetus to formalize bioinformatics education came in 1998 from Russ Altman [1], with a wish list of topics for an ideal bioinformatics educational program at the masters and PhD levels. Given the multidisciplinary nature of bioinformatics and the need for designing cross-faculty courses, by 2001 only a handful of universities had successfully commenced formal education in bioinformatics, with others waiting and watching.
The Workshop on Education in Bioinformatics (WEB) 2001 (http://surya.bic.nus.edu.sg/web01/) was launched at the 2001 International Conference on Intelligent Systems for Molecular Biology (ISMB) as a satellite meeting, to provide for the first time a forum for bioinformatics educators to meet, discuss, and exchange ideas and suggestions. WEB addresses fundamental educational and pedagogical issues to determine the nature, extent, and content of, and delivery tools available for, bioinformatics degree and training programs, and to provide focus points and suggestions for improvement of nascent degree programs.
WEB has served as the single annual meeting for education in bioinformatics, with plenary, oral, and poster presentations. WEB has, over the past four years, witnessed the evolution of bioinformatics education through talks and posters. Typically, presentations have sequentially progressed from a masters degree to a PhD [2,3] or from a minor to a complete program [4]. Several WEB attendees themselves started educational programs [5] and are now active members of the ISCB Education Committee, where a list of currently available educational programs is maintained [6]. The new educational initiatives showcased include the S* Life Science Informatics Alliance in WEB 2001 and the Worldwide Universities Network in WEB 2002. Special sessions on industry needs (WEB 2002) and pedagogy (WEB 2003 and WEB 2004) have also been included in the program to provide perspective and depth.
Bioinformatics education became mainstream with the ISMB 2005 conference, featuring a joint education workshop integrating WEB with the efforts of the ISCB's Education Committee (referred to as WEB 2005 here) on the opening day of the conference.
How well do the new graduates from these educational programs fit the jobs available to them? How many areas should a bioinformatics graduate be an expert in? Apart from a deep understanding of algorithms, programming, and life sciences, solving problems in genomics, proteomics, and/or medical informatics appears to be the current requirement. With the blurring boundaries between bioinformatics and new areas of endeavor such as forensics and biodiversity/ecoinformatics, where should we, as educators, draw the line? With industry's annual fluctuating demands for specific in-depth knowledge, how can we create a bioinformatics world?
Some of these issues will be discussed in the 2006 WEB meeting. But here are four take home messages from the earlier workshops. First, with the growing demand for computational biologists, there is a persistent and continuing demand for bioinformatics education at all levels—formal and informal, face-to-face and distance learning, and short-term training and rigorous long-term academic programs. Clearly, “one size does not fit all” [7], since the trend is to develop new and innovative educational programs addressing niche areas within bioinformatics. Second, there is still no single bioinformatics education program that can satisfy today's wish list of essential ingredients [8,9]. But by enabling the student cohort in self-learning, with a problem-based approach [10], the ontology of education (“to draw forth”) is realized. Third, percolation of basic bioinformatics down to the undergraduate level, especially for all life science majors [11] and optionally for physical and computer science majors [12] would develop the multidisciplinary skills required for bioinformatics. Fourth, “faster, further, and more” summarizes the components that would go into curricula of the future. ISCB members, educators, and bioinformatics professionals are encouraged to participate in WEB, and contribute their valuable ideas and expertise to this bioinformatics educational initiative.
Abbreviations
ISMBIntelligent Systems for Molecular Biology
WEBWorkshop on Education in Bioinformatics
==== Refs
References
Altman RB 1998 A curriculum for bioinformatics: The time is ripe Bioinformatics 14 549 550 9841111
Yang UC 2001 A master degree bioinformatics program in Taiwan [abstract]. Workshop on Education in Bioinformatics; 2001 26 July; Copenhagen, Denmark. In Silico Biol Available: http://surya.bic.nus.edu.sg/web01/abstracts/yang-poster.html . Accessed 3 November 2005.
Yang UC 2002 A Ph.D degree bioinformatics program in Taiwan [abstract]. Workshop on Education in Bioinformatics; 2002 8 August; Edmonton, Canada. In Silico Biol Available: http://surya.bic.nus.edu.sg/web02/abstracts/yang.html . Accessed 3 November 2005.
Hughey RP 2003 Bioinformatics degree programs at UCSC [abstract]. Workshop on Education in Bioinformatics; 2003 28 August; Brisbane, Australia. In Silico Biol Available: http://surya.bic.nus.edu.sg/web03/abstracts/hughey.html . Accessed 3 November 2005.
Jennings SF 2003 An integrated program in bioinformatics [abstract]. Workshop on Education in Bioinformatics; 2003 28 August; Brisbane, Australia. In Silico Biol Available: http://surya.bic.nus.edu.sg/web03/abstracts/jennings.html . Accessed 3 November 2005.
International Society for Computational Biology 2005 ISCB listing of degree/certificate programs worldwide La Jolla (California) International Society for Computational Biology Available: http://iscb.org/univ_programs/program_board.php . Accessed 31 October 2005.
Tholesson M 2004 Tholesson, M. teaching bioinformatics in Uppsala—One size does not fit all [abstract]. Workshop on Education in Bioinformatics; 2004 3 July; Glasgow, United Kingdom. In Silico Biol Available: http://surya.bic.nus.edu.sg/web04/abstracts/thollesson.html . Accessed 3 November 2005.
Mount DW 2002 Just how much does a bioinformatics specialist need to know [abstract]? Workshop on Education in Bioinformatics; 2002 8 August; Edmonton, Canada. In Silico Biol Available: http://surya.bic.nus.edu.sg/web02/abstracts/mount.html . Accessed 3 November 2005.
Ranganathan S 2005 Bioinformatics education today! ISCB education session/WEB05 (Workshop on Education in Bioinformatics 2005) [workshop]. 15th International Conference on Intelligent Systems for Molecular Biology; 2005 25–29 June; Detroit, United States of America Available: http://www.iscb.org/ismb2005/workshops.html# . Accessed 3 November 2005.
Choo KH Tong JC Tan TW Ranganathan S 2004 Application of problem based learning pedagogy in global online bioinformatics education [abstract]. Workshop on Education in Bioinformatics; 2004 3 July; Glasgow, United Kingdom. In Silico Biol Available: http://surya.bic.nus.edu.sg/web04/abstracts/tan.html . Accessed 3 November 2005.
Bellgard M 2003 Critical analysis of ICT/bioinformatics learning: Objectives for students in life sciences [abstract]. Workshop on Education in Bioinformatics; 2003 28 June; Brisbane, Australia. In Silico Biol Available: http://surya.bic.nus.edu.sg/web03/abstracts/bellgard.html . Accessed 3 November 2005.
Isaev A 2003 Bioinformatics for students with mathematical background [abstract]. Workshop on Education in Bioinformatics; 2003 28 June; Brisbane, Australia. In Silico Biol Available: http://surya.bic.nus.edu.sg/web03/abstracts/isaev.html . Accessed 3 November 2005.
|
16322761
|
PMC1289384
|
CC BY
|
2021-01-05 09:18:23
|
no
|
PLoS Comput Biol. 2005 Nov 25; 1(6):e52
|
utf-8
|
PLoS Comput Biol
| 2,005 |
10.1371/journal.pcbi.0010052
|
oa_comm
|
==== Front
PLoS Comput BiolPLoS Comput. BiolpcbiplcbploscompPLoS Computational Biology1553-734X1553-7358Public Library of Science San Francisco, USA 1632276210.1371/journal.pcbi.001005905-PLCB-PV-0225plcb-01-06-06PerspectivesScience PolicyInterdisciplinary ResearchNIH Roadmap Interdisciplinary Research Initiatives NIH RoadmapHuerta Michael F *Farber Gregory K Wilder Elizabeth L Kleinman Dushanka V Grady Patricia A Schwartz David A Tabak Lawrence A *To whom correspondence should be addressed. E-mail: [email protected] F. Huerta is Associate Director, National Institute of Mental Health and co-leader of the National Institutes of Health Roadmap Interdisciplinary Research Consortia Initiative; Gregory K. Farber is Program Director, National Center for Research Resources and co-leader of the National Institutes of Health Roadmap Interdisciplinary Research Consortia Initiative; Elizabeth L. Wilder is Program Director, National Institute of Diabetes and Digestive and Kidney Diseases and Coordinator of the National Institutes of Health Roadmap Interdisciplinary Research Initiatives; Dushanka V. Kleinman is Assistant Director for Roadmap Coordination, Office of the Director; Patricia A. Grady is Director, National Institute of Nursing Research and co-leader of the National Institutes of Health Roadmap Interdisciplinary Research Initiatives; David A. Schwartz is Director, National Institute of Environmental Health Sciences and co-leader of the National Institutes of Health Roadmap Interdisciplinary Research Initiatives; Lawrence A. Tabak is Director, National Institute of Dental and Craniofacial Research and co-leader of the National Institutes of Health Roadmap Interdisciplinary Research Initiatives, National Institutes of Health, Bethesda, Maryland, United States of America.
11 2005 25 11 2005 1 6 e59Copyright: © 2005 Huerta et al.2005This is an open-access article distributed under the terms of the Creative Commons Public Domain declaration, which stipulates that, once placed in the public domain, this work may be freely reproduced, distributed, transmitted, modified, built upon, or otherwise used by anyone for any lawful purpose. Citation:Huerta MF, Farber GK, Wilder EL, Kleinman DV, Grady PA, et al. (2005) NIH roadmap interdisciplinary research initiatives. PLoS Comput Biol 1(6): e59.
==== Body
A recent Perspective [1] suggested that the National Institutes of Health (NIH) Roadmap interdisciplinary research initiatives (http://nihroadmap.nih.gov/interdisciplinary/) embrace interdisciplinary teams without recognizing the importance of interdisciplinary individuals, and that these initiatives recognize something other than scientific questions as drivers of interdisciplinary research. As representatives of the implementation work group for these initiatives, we welcome this opportunity to clarify the approach that the interdisciplinary portion of the NIH Roadmap is taking to support science conducted beyond the borders of individual disciplines.
The NIH Roadmap interdisciplinary research work group issued seven requests for applications (RFAs). Recognizing that interdisciplinary individuals require knowledge of multiple disciplines, the work group emphasized individual training in four of these RFAs. Together, the training initiatives represent a coherent trans-NIH effort to not only provide training opportunities to individuals, but in the process, nurture and cultivate interest and investment in interdisciplinary research at the organizations to which those grants were awarded. The driving forces behind which disciplines are intertwined and drawn upon in these training activities are the questions that those participating are interested in pursuing (and which are, of course, relevant to the mission of the NIH).
Support provided under two other RFAs is also aimed at individuals—in these cases, to facilitate the development and enhancement of methods that work across disciplines contributing to behavioral, social, and biomedical research. Again, the scientific questions that interest those being funded will determine the nature of this interdisciplinary activity.
Finally, one RFA encouraged projects comprising teams of investigators from different disciplines to begin to develop new ways of thinking about, and addressing, significant research problems in research relevant to health and illness. Grants awarded under this RFA were specifically meant to facilitate thoughtful planning for what were hoped to become long-term interdisciplinary research projects. Again, these efforts are driven by the scientific questions that interest the participating investigators.
Very recently, the National Academies published “Facilitating Interdisciplinary Research” [2], a report that considered several aspects of this topic from many perspectives. We are delighted that the conceptualization and implementation of the NIH Roadmap interdisciplinary research initiatives conform well with the findings and recommendations presented in that report, including the importance of both individuals and teams in conducting such research.
It is clear that interdisciplinary research will play an increasingly significant role in improving human health. In recognition of this, the NIH is proceeding deliberately and rapidly to enhance its ability to advance this important and exciting paradigm. The Roadmap initiatives described above, which support both individual and team efforts and which are driven by the scientific questions that are of greatest interest to investigators and trainees, represent a major thrust in this direction.
Abbreviations
NIHNational Institutes of Health
RFArequests for application
==== Refs
References
Eddy S 2005 “Antedisciplinary” science PLoS Comput Biol 1 e6. DOI: 10.1371/journal.pcbi.0010006 16103907
National Academy of Sciences 2005 Facilitating interdisciplinary research Washington, DC National Academies Press 306 p.
|
16322762
|
PMC1289386
|
CC0
|
2021-01-05 09:18:23
|
no
|
PLoS Comput Biol. 2005 Nov 25; 1(6):e59
|
utf-8
|
PLoS Comput Biol
| 2,005 |
10.1371/journal.pcbi.0010059
|
oa_comm
|
==== Front
PLoS Comput BiolPLoS Comput. BiolpcbiplcbploscompPLoS Computational Biology1553-734X1553-7358Public Library of Science San Francisco, USA 1632276310.1371/journal.pcbi.001006105-PLCB-RV-0239R1plcb-01-06-10ReviewBioinformatics - Computational BiologyEvolutionGenetics/EvolutionVirologyNoneQuasispecies Made Simple ReviewBull J. J Meyers Lauren Ancel Lachmann Michael **To whom correspondence should be addressed. E-mail: [email protected]. J. Bull is at the Institute for Cellular and Molecular Biology, Section of Integrative Biology, University of Texas, Austin, Texas, United States of America. Lauren Ancel Meyers is at the Institute for Cellular and Molecular Biology, Section of Integrative Biology, University of Texas, Austin, Texas, United States of America, and at the Santa Fe Institute, Santa Fe, New Mexico, United States of America. Michael Lachmann is at the Max Planck Institute for Evolutionary Anthropology, Leipzig, Germany.
11 2005 25 11 2005 1 6 e61Copyright: © 2005 Bull et al.2005This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are properly credited.Quasispecies are clouds of genotypes that appear in a population at mutation–selection balance. This concept has recently attracted the attention of virologists, because many RNA viruses appear to generate high levels of genetic variation that may enhance the evolution of drug resistance and immune escape. The literature on these important evolutionary processes is, however, quite challenging. Here we use simple models to link mutation–selection balance theory to the most novel property of quasispecies: the error threshold—a mutation rate below which populations equilibrate in a traditional mutation–selection balance and above which the population experiences an error catastrophe, that is, the loss of the favored genotype through frequent deleterious mutations. These models show that a single fitness landscape may contain multiple, hierarchically organized error thresholds and that an error threshold is affected by the extent of back mutation and redundancy in the genotype-to-phenotype map. Importantly, an error threshold is distinct from an extinction threshold, which is the complete loss of the population through lethal mutations. Based on this framework, we argue that the lethal mutagenesis of a viral infection by mutation-inducing drugs is not a true error catastophe, but is an extinction catastrophe.
Citation:Bull JJ, Meyers LA, Lachmann M (2005) Quasispecies made simple. PLoS Comput Biol 1(6): e61.
==== Body
Introduction
The concept of a mutation–selection balance is one of the oldest and most fundamental pillars of population genetics: natural selection increases the frequency of fit variants while mutations introduce unfit variants, giving rise to an equilibrium distribution balanced between these two effects. Mutation–selection balance has been invoked to explain the persistence of undesirable genes, for example, those underlying inbreeding depression, genetic diseases, and even senescence. Despite the long history of the concept, some of its consequences were only realized in 1971, when Manfred Eigen studied mutation–selection dynamics in long genomes [1]. He found that populations do not necessarily attain classic mutation–selection balances in which the wild-type allele is most common, but rather attain an equilibrium with an abundant assemblage of mutant genotypes and a rare wild-type. He and Peter Schuster later called this collection of genotypes at equilibrium a quasispecies [2]. This concept offered not only an intuitive extension of the mutation–selection theory based on simple one- or two-locus systems, but also a novel insight into the impact of mutation rate on evolutionary dynamics. In particular, Eigen found that there are states in which a trivial boost in the mutation rate can lead to a fundamental change in the composition of genotypes in the population. This change, a phase transition in physics terms, is called the error catastrophe.
The error catastrophe has been applied liberally as a metaphor for complications of high mutation rates, as likely plagued primordial life [1] and currently challenges extant viruses with RNA genomes [3]. The error-catastrophe model inspired treatments to extirpate viral populations by mutation enhancement [4,5], and the model has been generalized to explain the attraction of populations to mutationally robust regions of fitness landscapes [6]. The error catastrophe has imparted a mystique to the quasispecies concept, and much of the literature on RNA virus evolution now uses quasispecies as an enriched synonym for a high mutation rate. An excellent and short review of the topic and its relationship to population genetics theory is provided by Wilke [7].
Eigen's insights were developed in the context of genomes with many loci, each of which suffered mutation. Appropriately, the quasispecies has since been considered in this large-genome context. Yet many of its concepts are easily illustrated in the much simpler case of few genotypes, which is our approach here. Our results are not new, per se, but our models should convey quasispecies and error-catastrophe concepts to a broad audience and correct some common misunderstandings.
The Simplest Quasispecies
Our basic model has the fewest number of genotypes needed to demonstrate a quasispecies and an error threshold: two [8]. Genotype A
1 has fitness w
1, and of those w
1 offspring a fraction 1 − μ1 retain the A
1 genotype (Figure 1). Its mutants are converted into the other genotype, A
2, which has the lower fitness w
2. A
2 reproduces its genotype with fidelity 1 − μ2, and all of its mutants die.
Figure 1 Model of Two Genotypes with Forward Mutation
Each genotype Ai has its own fitness wi and mutational loss μi. Mutation is asymmetric, so that A
1 gives rise to A
2, but not vice versa.
Quasispecies concepts address equilibia, that is, the final distributions of genotypes in populations that have evolved to a stable state. In the quasispecies model, mutation and natural selection steer the population toward the equilibrium distribution, regardless of the initial distribution of genotypes. If the population does not start at the equilibrium, then mutation and natural selection steer it toward equilibrium in the quasispecies model. Mutations introduce new types with various fitnesses while natural selection causes more fit variants to increase in frequency at the expense of less fit variants. Ultimately, a population reaches a genotype distribution at which these two forces exactly cancel each other, leaving the genotype distribution unchanged. This stable assemblage of genotypes has been called both a mutation–selection balance and a quasispecies. Some authors prefer to reserve the quasispecies concept for the mutation–selection balance extreme in which the wild-type is rare [9]. This preference may stem from the property that, under a high mutation rate, the quasispecies distribution as a whole rather than a single genotype is the target of selection [2], as illustrated below.
It is essential to distinguish the process of evolution toward the mutation–selection balance from an error catastrophe, as the two are completely different. We thus begin by briefly illustrating the dynamic approach to mutation–selection balance. For any combination of mutation rates and fitnesses in our model, a population will evolve until it reaches a unique equilibrium distribution of A
1 and A
2. Figure 2 illustrates the evolution of two populations—one that initially consisted entirely of A
1 and another that consisted almost entirely of A
2. Both eventually reach the same equilibrium proportions because the mutation rates and fitnesses are the same for both cases. This end-state is the mutation–selection balance, whereby mutation continually creates A
2 from A
1 and natural selection continually purges A
2 in favor of A
1; it is also a quasispecies where A
1 is the wild-type and A
2 is the mutational “cloud” that persists with it. In a strict sense, equilibrium is reached only after infinitely many generations, but the deviation from equilibrium quickly becomes insignificant.
Figure 2 Dynamic Approach to the Quasispecies Equilibrium, for the Genotypes Illustrated in Figure 1
Drawn for w
1 = 1.5, w
2 = 1.0, μ1 = 0.7, and μ2 = 0.6. The upper, black curve represents a different starting condition than the lower, gray curve, but they both equilibrate to the same value.
If a mutation rate or fitness is changed, the equilibrium distribution changes. Thus, the analogous figure for a different equilibrium would have the curves converging to a different value on the vertical axis. In particular, if we continually lower the fitness (w
1) or the mutation-free fraction of offspring (1 − μ1) of A
1 relative to that of A
2, then the equilibrium distribution will move to progressively lower values on the vertical axis, until it lacks A
1 entirely. This brings us to the error catastrophe.
Error Catastrophe: The “Loss” of A
1 at Equilibrium
Error thresholds and error catastrophes are properties of mutation–selection equilibria, and an error catastrophe is a specific type of change in the equilibria. Consider again the genotypes and fitnesses illustrated in Figure 1. The product w
1(1 − μ1) is the number of A
1 offspring of an A
1 parent; we will refer to this product as the replacement rate of the genotype—the product of fitness times the mutation-free fraction of offspring. For mathematical simplicity, we assume that this number is an absolute quantity greater than one and that it does not change with population density. Hence, the population is forever expanding in unlimited space (this assumption is convenient but not necessary). For convenience, we will continue to treat A
1 and A
2 as single genotypes, but later in the paper, we will explain how they can be treated as sets of genotypes and how additional types can be included (e.g., A
3, A
4, ...).
Since A
1 has a higher fitness than A
2, it follows that, for small mutation rates μ1, the replacement rate of A
1 is higher than the replacement rate of A
2, or w
1 (1 − μ1) > w
2 (1 − μ2). As long as the replacement rate of A
1 exceeds that of A
2, both genotypes will be maintained at equilibrium: A
1 because of its higher replacement rate, and A
2 because of mutation from A
1. If the mutation rate of A
1 increases such that the mutation-free fraction of A
1 drops relative to that of A
2, however, the replacement rate inequality will eventually reverse, yielding a simple error catastrophe: A
1 is no longer maintained. The error threshold is the point at which both genotypes have identical replacement rates. Beyond this point, A
2 has the higher replacement rate, and A
1 is absent at all equilibria—the lack of back mutations ensures that it is not recreated from A
2.
The nature of this simple error threshold and error catastrophe is easily visualized (Figure 3). The left part of the figure shows A
1 and A
2 replacement rate functions approaching each other as the mutation rate increases toward the error threshold. The functions cross at the error threshold, giving rise to an error catastrophe (complete loss of A
1) for higher mutation rates. The right part of the figure illustrates that, despite the suggestive terminology (threshold and catastrophe), an error catastrophe is the culmination of a gradual decrease in the equilibrium frequency of A
1 rather than a dramatic one. An apt analogy is the geological transition from mountains to plains: starting from the mountains, the plains are approached gradually until they are reached, at which point the plains continue in the absence of any trace of mountains. (Note, however, that a true phase transition does occur at the error threshold, as has been shown by Leuthausser [10,11].) The equilibrium proportion of A
1 approaches zero as the mutation rate approaches the error threshold. Just before the threshold, the population is dominated by the “cloud” of A
2 mutants around a tiny fraction of A
1, and thus fulfills the standard criteria for a quasispecies.
Figure 3 The Error Threshold and Error Catastrophe
Left: replacement rates of the two genotypes in Figure 1, w
1(1 − μ1) for A
1 (solid black line extending to dashes) and w
2(1 − μ2) for A
2 (gray line). Here we let μ1 = kμ and μ2 = μ, with k > 1 so the two replacement rates can be plotted as lines in the same plane. Since w
1 > w
2, the constraint on k ensures a higher replacement rate of A
1 than of A
2 at low mutation rates and the reverse at higher mutation rates. The point at which the lines intersect is the error threshold, beyond which A
1 is absent, hence the use of dashes for this part of its replacement rate function. If replacement rate drops below unity, the population goes extinct, so the functions are not extended below one on the vertical axis.
Right: the frequency of A
1 declines as the mutation rate increases until the error threshold is reached. At higher mutation rates, only A
2 is present. The decline in the frequency of A
1 with μ toward the error threshold may be linear (as shown here), concave, or convex, depending on parameter values. (Right side drawn for w
1 = 3.0, w
2 = 2, and k = 2.)
An error threshold exists because deleterious mutations impact some genotypes more than others (reviewed in [7]): in this model the threshold is breached only if A
1 experiences a potentially much greater mutational loss than A
2. Beyond the error threshold this deleterious effect of mutation is retarded: by replacing a mutation-sensitive genotype with a mutation-robust genotype, the error catastrophe reduces the speed at which the replacement rate drops in response to increases in the mutation rate [7]. This outcome can be seen in the left side of Figure 3: the A
1 replacement rate function is steeper than the A
2 replacement rate function, so increases in the mutation rate have less impact after the error threshold. (Not only do the functions depict genotype replacement rate, but the uppermost functions also give the mean replacement rate in a population at equilibrium.) Below we will specifically describe how such mutational robustness may be achieved; we have depicted an error catastrophe between two genotypes, but the logic is easily extended to larger sets of mutationally connected genotypes. Again, we emphasize that an error catastrophe is not a process per se; rather it is a change in the equilibrium distribution of genotypes arising under a relatively high mutation rate. If the mutation rate of a population is increased from below to above an error threshold, however, then there will be a dynamical loss of A
1 as the population approaches the new equilibrium.
Three implications of this analysis are significant. First, an error catastrophe is not equivalent to a population extinction. Second, fitness landscapes may contain multiple error thresholds. Third, the error threshold is affected by finite population size, the inclusion of back mutation from the mutant to the wild-type, and recombination. We now discuss each of these in detail.
Error catastrophes versus population extinction.
High mutation rates can cause extinction, but not as an error catastrophe (Figure 3). An error catastrophe takes place when one genotype's replacement rate exceeds another's and thus displaces it at the equilibrium—a replacement of one self-sustaining genotype by another self-sustaining genotype. The population cannot maintain the genotype with the higher fitness, A
1, since its high mutation rate causes it to have a lower replacement rate than A
2. Thus, the error threshold refers to the loss of the highest fitness genotype in favor of other genotypes with lower fitness but greater mutational robustness.
In contrast, extinction occurs at the point that the best genotype cannot reproduce itself to maintain a minimum population size (wi[1 − μi] < 1 for all genotypes). The most trivial example of the distinction between an error threshold and an extinction threshold is the case in which only one genotype exists, and all mutations are lethal. No error catastrophe is possible because no other genotypes exist, but extinction is obviously possible. Our model is one genotype more complicated: mutant offspring of A
1 survive as A
2, but mutants of A
2 are inviable. Thus, there is not an additional error threshold beyond A
2, because the model does not allow a third inferior genotype. An A
2 population, however, will go extinct if its net growth rate drops below one (the horizontal axis in Figure 3). While an error threshold depends on the relative replacement rates of the genotypes, the extinction threshold depends on the absolute replacement rates of the genotypes.
There are several possible relationships between error thresholds and extinction thresholds. The genotype with highest fitness in this model (A
1) can disappear through either an error catastrophe or a population extinction, but A
2 can disappear only through extinction. Depending on the values of w
1, w
2, and μ2, the rate of mutation μ1 at which A
1 would be lost to A
2 in an error catastrophe may be less than or greater than the value of μ1 at which the population would go extinct. If the error threshold lies beyond the extinction value, then as μ1 increases, the A
1 + A
2 population will go extinct of its own accord before a true error catastrophe can happen, because the population has disappeared before the mutation rate rises high enough for the error catastrophe (Figure 4, right). If the opposite is true, then A
1 will be displaced by A
2 in an error catastrophe before the replacement rate of either type drops below one (Figure 4, left).
Figure 4 Extinction versus Error Catastrophe
Left: the error threshold is crossed at a lower mutation rate (μ) than the extinction threshold.
Right: the extinction threshold is crossed before the error threshold.
Otherwise as in Figure 3.
Multiple error thresholds.
The model implies that multiple error thresholds may exist in a fitness landscape. For example, a fitness landscape may include multiple Ai with associated wi and μi (Figure 5). With an appropriate hierarchy of mutation rates, error thresholds could operate sequentially, so that as mutation rates increase, the population drops from one Ai to the next, then to the next, and so on [12]. It is also possible for an error catastrophe to jump over an intermediate peak [13]. For example, if the mutation rate at which the error threshold is crossed between A
1 and A
2 exceeds the mutation rate for the error threshold between A
2 to A
3, then a transition will occur from A
1 to A
3, skipping the intermediate in which A
1 has disappeared but A
2 remains (see below). The interplay between extinction catastrophes and error catastrophes can likewise add interesting complexity to the model.
Figure 5 Model of Several Genotypes with Forward Mutation
Each genotype Ai has its own fitness wi and mutational loss μi so that error thresholds can exist between each pair of genotypes.
Implications of finite population size, back mutation, and recombination.
Our model assumes deterministic dynamics and ignores the stochastic effects of finite population sizes. We have argued that populations will equilibrate either to a combination of A
1 and A
2 or to A
2 exclusively. This equilibration strictly requires an infinite amount of time in an infinite population. In a finite population, as long as the effective population size times the mutation rate is significantly bigger than one (Nμ ≫ 1), the finite population behaves similarly to the infinite population, except that the error threshold is not as sharp. In particular, the lack of back mutations ensures that random events will eventually cause the loss of genotype A
1 from the population, so that only genotype A
2 remains. If A
2 represents multiple types of mutants, then this random loss of A
1 is Muller's ratchet [14,15]. Beyond the error threshold, this transition is simply much faster, because selection favors it. The threshold is also altered by the introduction of back mutations from A
2 to A
1, as described in Box 1.
Box 1. Back Mutation
Our model of two genotypes ignored back mutation, hence A
1 could not be regenerated once lost from the population (see Figure 1). The absence of back mutations is often assumed in models of population genetics (e.g., the “infinite alleles” and “infinite sites” models of population genetics [42,43]). This assumption offers analytical convenience and is perhaps an approximate interpretation of asexual genomes, in which an “allele” can be considered the entire suite of mutations in a genome; a new mutation, which can occur anywhere in the genome, is unlikely to recreate a former allelic state. Nonetheless, since back mutations can occur in many systems, it is useful to study their effect on the error threshold. In particular, they present a special challenge to the concept of an error threshold: when back mutations occur, the wild-type genome is never strictly lost in a deterministic sense. What, then, is the counterpart to a strict error threshold? To approach this question, we return to the simple model in Figure 6.
Figure 6 Model of Two Genotypes with Back Mutation, μ3
As in Figure 1, except that some of the mutations away from A
2 recreate genotype A
1. The total mutation rate of A
2 is still μ2, so the back mutation rate cannot exceed this value (i.e., μ3 ≤ μ2).
At this point we use simple matrix algebra to explore the problem. Assuming discrete time steps, let ni be the number of individuals with genotype Ai; mutation precedes reproduction, so the fitness wi accrues to an individual after its genotype experiences any mutation (the model is just as easily constructed with mutation occurring after reproduction). Two equations describe the system:
The system can then be described with the aid of the transition matrix:
so that
. The only new parameter with respect to our earlier model is μ3, which allows for back mutation and cannot exceed μ2. If we set μ3 = 0, then these equations describe our earlier model exactly (in Figure 1). To evaluate the long term behavior of the system, we consider n
1(t) and n
2(t) as t becomes arbitrarily large; that is, we calculate (n
1, n
2) Mt for large t. The standard method for evaluating long-term behavior in such a linear system is to consider the eigenvalues of the system, found as the values of λ satisfying
To establish a baseline, we first assume no back mutation (μ3 = 0). Equation 4 then yields the solutions
These two eigenvalues are the respective replacement rates of A
1 and A
2 described in Figure 3. Importantly, λ1 is larger for some mutation rates, but λ2 is larger for others, and it is the intersection of the eigenvalues that gives rise to the error threshold. This simple picture appears to change fundamentally with even a small level of back mutation. When μ3 > 0, the two eigenvalues become
where x = w
1(1 − μ1) and y = w
2(1 − μ2).
Whereas a transition from λ1 to λ2 as the dominant eigenvalue occurred at x = y in the absence of back mutation, now λ1 remains superior at all values of x and y. As μ3 approaches arbitrarily close to zero, the eigenvalues approach each other but do not cross. Thus, there is no strict error threshold unless μ3 equals zero exactly. Yet as μ3 approaches zero, the system appears more and more like that with no back mutation (Figure 7). So the mathematical discontinuity between systems with and without back mutation does not obviously translate into a meaningful biological distinction, but large amounts of back mutation do change the picture.
Figure 7 Eigenvalues When a Small Level of Back Mutation Occurs
In contrast to the case of no back mutation in Figure 3, the eigenvalues here do not cross but change slope as they approach each other. The dark curves (upper) correspond to the maximum eigenvalue (λ1) and the light curves (lower) to the smaller eigenvalue. The thick, outer curves are drawn for a back mutation rate of μ3 = 0.05, and the thin, inner curves for μ3 = 0.001. The effect of increasing back mutation is clearly seen as increasing the minimum distance between the eigenvalue functions. Other parameters are the same as for the right side of Figure 3: w
1 = 3.0, w
2 = 2, and k = 2.
When Nμ ≪ 1, the above description does not hold. Mutations occur so rarely, that the population is very likely to fix or lose each mutation before the next one arises. In this case, it no longer makes sense to talk about a quasispecies—there is no mutation–selection balance, but instead a fixed type, the wild-type, with possibly some recent mutants. The wild-type essentially undergoes a random walk in genotype space, in which it tends to “hang out” more often in areas with high fitness [16]. Sella and Hirsh [16] showed that population size rather than mutation rate is the critical randomizing factor: the smaller the population size, the less the population is able to maintain genotypes with high fitness.
We ignored the effect of recombination, as does most of the literature on the error catastrophe. The effect of recombination on the quasispecies can be very complicated. A surprising result is that recombination can make a genotype more vulnerable to an error catastrophe, causing the error threshold to occur at a lower mutation rate [17]. Although a full explanation of this effect does not seem to have been provided, one contributing factor is that, close to the error threshold, the quasispecies contains mainly mutants, and very few individuals of type A
1. Therefore, most individuals of type A
1 will mate with those of type A
2. If our model is interpreted as a single-locus model, recombination would make no difference. But if A
2 is a collection of genotypes with mutations at possibly many loci and A
1 is the mutation-free class, then recombination could decrease the rate at which A
1 parents have A
1 offspring and thus cause the error threshold to occur earlier [17]. Recombination might also operate in the reverse direction, however, if A
2 genotypes frequently recombine to create A
1 offspring. A greater exploration of this problem is warranted.
Survival of the Flattest
Error thresholds require that some genotypes or phenotypes are more sensitive to mutation than others. Biologically, many phenotypes have the property that they can be produced by multiple, mutationally related genotypes, and mutational robustness is known to vary among protein and nucleic acid sequences [13,18–22]. The set of genotypes for a particular phenotype is called its neutral network [23]. Since mutations within a neutral network preserve the phenotype, this genetic redundancy can reduce the phenotypic mutation rate without altering the underlying genetic mutation rate. As discussed above, a lower rate of phenotypic mutations (e.g., A
1 to A
2) leads to a higher net replacement rate of the phenotype.
Here we illustrate how the breadth of a neutral network—the number of and mutational connectivity among genotypes contained within it—affects the competitiveness of the phenotype. In particular, “survival of the flattest” refers to the ability of inferior phenotypes with large neutral networks to displace superior phenotypes with smaller neutral networks. This phenomenon was recently demonstrated in digital life simulations and simulated RNA molecules [6,24].
We explore the impact of neutrality on the error threshold with a simple extension of our previous model, this time allowing two genotypes to have the phenotype A
2 (Figure 8). This creates a minimal neutral network of two genotypes. In the absence of back mutation (μ3 = 0), the error threshold is unaffected by the “network” and is the same as with a single A
2 genotype: A
2 here has exactly the same properties as the A
2 in Figure 1, so the A
1/A
2 error threshold is the same as in Figure 1. Yet the error threshold does change when back mutation is allowed from the second A
2 genotype (
) to the first (A
2). Back mutation provides a mutational “connectivity” among both genotypes in the neutral network and renders the network more competitive against A
1. This connectivity serves to bolster the frequency of the left A
2 genotype at the expense of the right A
2. Since only the right peak experiences mutations off the neutral network (at rate μ2), this shift increases the overall replacement rate of A
2, allowing A
2 to more easily supplant A
1. Higher rates of back mutation increase this effect (Figure 9, top) and also retard the extinction of A
2. (As we discuss later, allowing an infinite succession of A
2 genotypes without back mutation has the same effect on the error threshold as abolishing the forward mutation rate of a single A
2 genotype, since the population will never mutate off the A
2 neutral network.)
Figure 8 Model of Three Genotypes with a Simple Neutral Network
Both A
2 genotypes have the same fitness and the same forward mutation rates μ2. The error threshold with μ3 = 0 is the same as in Figure 1. As μ3 increases, the A
2 network becomes more competitive against A
1, and thus the error threshold value of μ1 decreases. Equations for this system can be derived and analyzed with similar methods as in Box 1.
Figure 9 Network Size and Mutational Connectivity Affects Replacement Rate
Top: effect of back mutation (μ3) between the pair of A
2 genotypes in Figure 8. The vertical axis gives the replacement rate of the A
2 network. Since all A
2 genotypes are mutationally interconnected, they ultimately grow at the same rate in the limit. In the absence of back mutation, the neutral network has the replacement rate w
2 (1 − μ2), and replacement rate increases as back mutation is increased up to its maximum of μ3,max = 1 − μ2. (The replacement rate function is
obtained as the eigenvalue associated with A
2 in the transition matrix.)
Bottom: the replacement rate of the neutral network also increases with the number of genotypes in it. Figure 8 shows the network with two genotypes; third and additional genotypes would be added so that each is mutationally balanced: each additional genotype loses μ2 of its progeny to become the genotype to its right and loses μ3 of its progeny to become the genotype to its left, but it receives μ2 mutants from the genotype to its left and receives μ3 mutants from the genotype to its right. The benefit of increasing the network size is shown for three different combinations of forward and back mutation rates.
This model can also be extended to demonstrate the impact of neutral network size on the error threshold. In this case, we add A
2 genotypes into the middle of the neutral network, each with forward mutation rate μ2 and back mutation rate μ3. Figure 9 (bottom) shows that as the number of genotypes increases, the replacement rate of the A
2 network also increases, thus shifting the error threshold in favor of A
2 and buffering A
2 against extinction.
In summary, the error threshold decreases as the fitness landscape around the inferior phenotype broadens, either via the extension of the neutral network to include additional genotypes or via a net decrease in mutation off the network. The error threshold above which a particular wild-type is lost will depend on its fitness relative to the fitness and breadth of the inferior network. The broader and more mutationally connected the inferior neutral network, the greater its tolerance for mutation and the more likely an error catastrophe that tips the population in its favor. Nonetheless, a sufficiently high fitness can compensate for a high rate of mutational loss.
These trade-offs also apply to more complex fitness landscapes consisting of multiple neutral networks with differing fitnesses. In this case, the sequence of error thresholds becomes difficult to intuit and has not previously been addressed quantitatively. Figure 10 illustrates a succession of error catastrophes in a single landscape with varying widths and heights of neutral networks, for a genome with 40 loci. In general, the narrower networks are skipped as the error catastrophes move to the flatter networks. In this particular model, there is no back mutation. Error thresholds exist because the genomic mutation rate is proportional to the number of unmutated loci, hence the phenotypes of lower fitness experience lower overall mutation rates.
Figure 10 Multiple Error Thresholds on a Binary Sequence of Length 40
Each allele has two states—wild-type and mutant. The per base mutation rate from wild-type to mutant is identical for all loci. Back mutation from mutant to wild-type is not allowed. As mutant alleles fix in the population, the proportion of the sequence that can still mutate decreases. Multiple error thresholds exist because the effective mutation rate per genome drops as mutations accumulate. The model is otherwise as in Figure 1 and Box 1, and as specified below.
Top: the fitness landscape. The x-axis shows the mutation distance from the “perfect” genotype consisting of all wild-type (zero mutant) alleles. The y-axis shows the fitness.
Lower left: equilibrium distributions of genotypes in the population for seven mutation rates from 0.0001 to 0.045. In all plots the x-axis indicates the number of mutations in the genome and the y-axis indicates frequency in the steady-state population; the shortest bars merely indicate the presence of genotypes at some (low) frequency, and the absence of a bar indicates absence of that genotype from the equilibrium population. Plots 1–5 show equilibria before the first error catastrophe, that is, for mutation rates that are less than the first error threshold. Plot 6 shows an equilibrium after the first error catastrophe and before the second error catastrophe, and plot 7 shows an equilibrium after the second catastrophe. The vertical dashed green lines indicate the stable networks that can attract error catastrophes. These correspond to the first (wild-type), third (5–14 mutations), and fourth (15–31 mutations) networks, starting from the left. The second network, with 1–4 mutations (represented by a red line), is never stable as the fittest type in the population, and is skipped.
Lower right: replacement rates of the various fitness plateaus. Numbers correspond to panels in the figure to the left. The x-axis shows the mutation rate; the y-axis shows the growth rate (with a log scale). One can see two error thresholds. The red dashed line represents the replacement rate of the second network, with 1–4 mutations, and shows why this plateau never stabilizes—it is too narrow relative to its fitness advantage.
Comparison to Eigen's and Other Models
For those familiar with Eigen's original quasispecies model ([1], a simplified version appears on p. 480, Table 8) and more recent versions applied to nucleotides [2,8,25], we relate it to our simpler model using our notation. In those models, a genome consists of L loci with a per locus mutation rate of ν. A single mutant-free, wild-type genotype A
1 has highest fitness w
1, and all other genotypes with one or more mutations fall into the same neutral network in which each genotype has the same fitness w
2. Eigen's classic error threshold result is that, depending on L and w
1/w
2, there is a specific per base mutation rate ν beyond which the wild-type is lost or maintained only through back mutation and the w
2 neutral network prevails, as in our Figure 10 (the most-fit genotype is maintained if back mutation is allowed, as in Box 1).
A second class of models is more challenging to fit into the current framework [15,26]. In those models, a linear series of genotypes is connected by forward mutation but not back mutation (
) -->
, and all genotypes have the same mutation rate. An error threshold exists in those models but not in ours. One critical difference is that those models include infinitely many genotypes. Thus, if our A
1/A
2 model is restricted to a single forward mutation rate for all genotypes, an error threshold appears only when infinitely many A
2 genotypes exist. In this extreme, the A
2 network no longer experiences any mutational loss, and hence its mutation rate effectively disappears. The existence of an error threshold in those models also requires that the fitnesses of all of the genotypes be greater than some positive value [15]. Thus, those models also appear to be consistent with our framework.
RNA Viruses, Quasispecies, and Lethal Mutagenesis
The error catastrophe has found its greatest appeal in work on RNA viruses, whose genomes have high mutation rates. One appeal of the quasispecies concept in this work is that it can potentially explain the high levels of sequence variation observed in RNA virus samples: the high mutation rate of RNA genomes should lead to a mutation–selection balance with lots of variation. Yet despite the frequent reference to quasispecies concepts in the literature, relatively few studies have demonstrated that observed variation is truly a quasispecies equilibrium [9]. An early study of the RNA phage Qβ reported that sequence variation in a population was high but approximately stable over time around a consensus sequence, supporting the basic concept of a quasispecies [27]. There are now countless studies revealing population variation in RNA viruses of eukaryotes, but one cannot easily determine whether the variation stems primarily from mutation–selection balance or from the simple alternative of selection for different genotypes within the spatially heterogeneous environment of the host [9]. Part of the difficulty is that quasispecies models consider mutation only as a degenerative process reducing replacement rate. They do not address mutation as an exploratory process for finding new fitness peaks—a process that must obviously be considered when interpreting empirical data. Quasispecies models also tend to ignore population structure and migration.
Further support for quasispecies behavior was provided by an empirical study of the dynamical approach toward quasispecies equilibrium. Burch and Chao [28] obtained a high-fitness isolate of the RNA phage φ6. Upon propagation in the same environmental conditions, however, the fitness of the viral lineage declined to the level of culture from which the isolate was originally obtained. Their interpretation, with which we agree, was that the initial isolate was a somewhat rare, high-fitness type from a quasispecies distribution, and that it re-evolved the equilibrium quasispecies distribution on further passage. It thus seems clear that some RNA virus populations satisfy the basic tenets of quasispecies theory, even if based on relatively few studies.
Quasispecies error thresholds have been invoked in an important medical application, the extinction of RNA virus populations by lethal mutagenesis—elevating mutation rates with base analog drugs to the point that the virus population disappears. In 2005, an entire issue of Virus Research was devoted to this topic (volume 107, issue 2). Base analogs are incorporated into the viral genome as it is being copied from the parent strand. In general, incorporation of a base analog can have two types of effects: chain termination, resulting in immediate death of the progeny strand, or mutation, whereby the base analog is later miscopied when the genome gets replicated. The latter mechanism underlies the antiviral activity of the drug ribavirin for some RNA viruses [4,29,30]. It is commonly argued that the extinction of the viral population by mutagenesis is an example of an error catastrophe (e.g., [4,5,30–32]). Others have argued against this interpretation [7], and our models illustrate clearly that extinction is distinct from an error catastrophe, and hence that lethal mutagenesis does not necessarily produce an error catastrophe.
From a practical perspective, it does not matter whether lethal mutagenesis is referred to as an error catastrophe or extinction catastrophe. There may be a benefit, however, to distinguishing between these two concepts and, more generally, to clarifying the concept of a quasispecies. For example, consider a protocol of sublethal mutagenesis, whereby a virus mutation rate is artificially elevated close to an extinction threshold but the virus population persists. Conventional wisdom suggests that such a strategy should be disastrous from a host health perspective, because the increased mutation rate may accelerate viral adaptation. Yet two components of the quasispecies framework suggest that a permanently elevated mutation rate may be deleterious to the virus even when extinction is not achieved. First, a higher mutation rate reduces mean fitness around any local fitness peak. Second, a high mutation rate “flattens” the fitness landscape, so that some narrow peaks cannot be attained. Thus, a sustained high mutation rate that does not drive the virus to extinction may enforce a low mean fitness. Holmes [33] has likewise argued that the high mutation rates of RNA viruses may constrain adaptation. Without greater insight into the structure of biological fitness landscapes, we cannot yet assess whether high mutation rates will expedite or thwart adaptation. Evolutionary theory suggests that either outcome is possible.
In light of this potential impact of high mutation rates, it seems paradoxical that a new class of drugs is being touted for its ability to reduce mutation rates, and thereby block the evolution of resistance to other drugs [34]. Yet mutation reduction and lethal mutagenesis may both be feasible strategies for controlling a virus, with the two methods working at opposite ends of the evolutionary spectrum. Mutation suppression may thwart evolutionary exploration of the fitness landscape while mutation enhancement may overwhelm the genome with lethal mutations. There is necessarily a wide zone of mutation rates between these extremes that allow viral persistence.
How to Demonstrate an Error Catastrophe
We have used a variety of simplifications to explain and depict error catastrophes from a mathematical perspective. Yet it may not be obvious how an error catastrophe would manifest itself in a population, especially how it would appear in a sample of genome sequences taken before and after crossing the error threshold. One possible scenario is similar to that of Figure 10: the error catastrophe is the deterministic accumulation of mildly deleterious nucleotides in a genome while preserving sites that are functionally essential. Figure 10 does not assign different fitness effects among loci but is easily extended to that case. The population will suffer the mutational loss of all nucleotides that cannot be maintained under the high mutation rate, while maintaining only those for which the fitness cost of a mutation is sufficiently large. This is similar to Kondrashov's version of Muller's ratchet, in which high fitness alleles ratchet out of the population until either the remaining alleles are sufficiently robust to mutation or all viable alleles have disappeared [35]. The different neutral networks in this process are neighbors in sequence space and differ by one or a few deleterious mutations. Thus the error catastrophe is observed as a population-wide change in nucleotide frequencies at possibly very few sites in the genome.
A second scenario is that of a population in which each individual's genome belongs to either of two “disjoint” neutral networks from very different parts of the fitness landscape. This type of polymorphism could arise if one type was present in the starting population and the other was introduced from a different population through migration. For this scenario, we assume that the population size is sufficiently small relative to the mutational distance between the two neutral networks, so that it is essentially impossible for the population to reach one network from the other via mutation (as modeled in some of the “survival of the flattest” work cited above). If one of the networks is more mutationally robust but of lower fitness than the other, an error catastrophe can cause the loss of the higher-fitness network. This type of error catastrophe might reveal itself in genomic data as the loss of polymorphism through the independent mutational meltdown and subsequent disappearance of one of the original types. Back mutation complicates the previous scenario by continually re-introducing the fitter nucleotides, which are then maintained at only a slightly higher level than neutral nucleotides. Back mutation is not a factor in the disjoint network model because the first network to disappear is exceedingly unlikely to be re-introduced via point mutations from the remaining population.
Given the difficulty of demonstrating the relatively simple properties of quasispecies dynamics and equilibrium properties, it is not surprising that the empirical demonstration of an error threshold/catastrophe has not been achieved. An interesting step in this direction was provided recently in a study that estimated the error threshold mutation rate that might apply to a primitive organism consisting of RNA molecules—a riboorganism [36]. The study concluded that an error catastrophe might be avoided in a substantially larger genome than had previously been thought. This result stemmed in part from the high degree of neutrality inferred from existing functional RNA molecules.
The systems offering the best hope for empirically demonstrating an error threshold are in fact naked RNA molecules evolved by directed evolution: ribozymes (catalytic RNA molecules) and aptamers (binding molecules) evolved in vitro from initial pools of random-sequence molecules. In this work, mutational robustness is recognized as a key property of molecules affecting evolution (e.g., [37]), but the concept has been approached from the perspective of adaptive evolution rather than error catastrophes (the Kun et al. study [36] cited above is unusual in applying the perspective of error catastrophes to ribozymes). A sequence “motif” is the set of bases at specific positions in the RNA molecule essential for function; changes in those bases destroy function but changes at other positions have no major effect. When there exist both short and long sequence motifs that are functionally similar, then, by simple combinatorics, the short motif will attain higher representation in a random sequence library than the long motif will. Thus, short motifs are more likely to be selected at the outset, creating a bias known as the “tyranny of short motifs” [38]. Furthermore, since the length of a motif determines the number of potential mutational targets, short motifs will typically be more mutationally robust than long motifs. In subsequent rounds of mutagenesis and selection, this may create additional bias toward short motifs.
Suppose a long motif has slightly greater functionality (our fitness, wi) than a short motif. By virtue of differences in mutational robustness among the different motifs, there should be an error threshold mutation rate, above which the long motifs will be displaced by the short motifs. If one knows the fitnesses (wi) and the number of critical bases in the long and short motifs (mi), it should be possible to estimate the value of the error threshold by using
as the replacement rate of motif i. The ligase ribozymes seem like an appropriate model for such a calculation, because the class I ligases have a much longer motif (~70 bases) than the class II and class III ligases (~40 bases; [39]). Kinetic parameters of classes II and III were not determined for optimized ligase molecules, so it is not yet possible to estimate realistic fitness differences between the best members of each family, and even if kinetic properties were known, extrapolation to fitness is not trivial. Nonetheless, if all point mutations of critical bases are lethal and others are neutral, the error threshold would be a per base mutation rate of roughly 0.003 if class I had a fitness advantage of 10% over classII/III, a mutation rate of 0.01 for a 50% advantage, and a mutation rate of 0.06 for a 10-fold advantage of class I. An experimental demonstration of an error threshold may thus be feasible in this system.
The ligase ribozymes will permit an easier demonstration of an error threshold than most other empirical systems, because the alternative functional networks have been identified and can be seeded into the experiment. We thus know in advance what to compare. If, instead, we have information about only the wild-type but not the suboptimal type that dominates beyond the error threshold (as would be the case when working with viruses or bacteria), then it becomes far more complicated to identify an error threshold empirically. Even with the ligase ribozymes, if a population was initiated with only the class I molecule, the class II or III molecules might not evolve at high mutation rates because their sequences might be too distant in sequence space to arise in a quasispecies of class I molecules. All three classes arose in the initial experiment because they were selected from a random sequence pool rather than evolved from a single type.
Identification of an error threshold is certainly facilitated by use of a previously characterized genotype–fitness map, but that kind of information is not usually available. In ignorance of the alternative neutral networks for a phenotype, the obvious protocol to use in blindly testing for an error catastrophe is to impose stabilizing selection for the wild-type trait while progressively increasing the mutation rate. One would then test for characteristics of an error catastrophe, such as a population shift in genotype space or increased mutational robustness of the genotypes evolved at high mutation rate. The key difficulty in demonstrating an error catastrophe without a priori knowledge of alternative networks is that there are population genetic processes not considered in the simple quasispecies models that give rise to superficially similar outcomes. Thus, under our suggested experimental protocol, shifts in genotype space may stem from (i) an error catastrophe—the deterministic fixation of deleterious mutations, (ii) stochastic fixation of deleterious and neutral mutations, (iii) the higher mutation rate allowing the population to find higher adaptive peaks, or (iv) adaptive evolution because mutagenesis has altered the selective environment (other than through higher mutation rate). Likewise, increases in mutational robustness may, in fact, suggest an error catastrophe, or may simply be a by-product of a general decline in fitness under high mutation rates. Recent work suggests that genotypes of reduced fitness generally experience a higher fraction of mutations with beneficial effect than genotypes of high fitness [40,41]. This pattern may stem from the local structure of the fitness landscape: a genotype at a local optimum can experience no beneficial mutations, and the further a genotype lies from a peak, the greater the chance that a random mutation improves fitness. Even before sorting out these various interpretations, there are fundamental challenges to demonstrating that a shift in genotype space or in mutational robustness has actually occurred (e.g., Figure 10 shows some of the difficulties).
The theory of quasispecies error thresholds appears sound. The major avenues for new developments are twofold: (i) obtaining empirical support and (ii) incorporating additional realism into the basic models. There are several papers that make progress on this second point, by including recombination [17], stochastic effects [14], and the shape of the mutation–fitness function [15]. We still have much to learn about the relevance of error catastrophes to the natural world, and the effort promises to yield many new insights.
We thank J. F. Crow, C. Wilke, B. Kerr, H. Wichman, and two anonymous reviewers for comments, and in some cases, identifying errors. Discussions with L. Chao and C. Burch helped motivate this study. JJB was supported by National Institutes of Health GM 57756, LAM by National Science Foundation DEB-0445351. ML and LAM thank G. Sella for numerous discussions of the error threshold.
==== Refs
References
Eigen M 1971 Selforganization of matter and the evolution of biological macromolecules Naturwissenschaften 58 465 523 4942363
Eigen M Schuster P 1977 The hypercycle. A principle of natural self-organization. Part A: Emergence of the hypercycle Naturwissenschaften 64 541 565 593400
Eigen M 1993 Viral quasispecies Sci Am 269 42 49 8337597
Anderson JP Daifuku R Loeb LA 2004 Viral error catastrophe by mutagenic nucleosides Annu Rev Microbiol 58 183 205 15487935
Jonsson CB Milligan BG Arterburn JB 2005 Potential importance of error catastrophe to the development of antiviral strategies for hantaviruses Virus Res 107 195 205 15649565
Wilke CO Wang JL Ofria C Lenski RE Adami C 2001 Evolution of digital organisms at high mutation rates leads to survival of the flattest Nature 412 331 3 11460163
Wilke CO 2005 Quasispecies theory in the context of population genetics BMC Evol Biol 5 44 16107214
Swetina J Schuster P 1982 Self-replication with errors. A model for polynucleotide replication Biophys Chem 16 329 345 7159681
Holmes EC Moya A 2002 Is the quasispecies concept relevant to RNA viruses? J Virol 76 460 465 11739715
Leuthausser I 1986 An exact correspondence between Eigen's evolution model and a two-dimensional Ising system J Chem Phys 84 1884 1885
Leuthausser I 1987 Statistical-mechanics of Eigen evolution model J Stat Phys 48 343 360
Tannenbaum E Shakhnovich EI 2004 Solution of the quasispecies model for an arbitrary gene network Phys Rev E Stat Nonlin Soft Matter Phys 70 021903 15447511
Ancel LW Fontana W 2000 Plasticity, evolvability, and modularity in RNA J Exp Zool 288 242 283 11069142
Nowak M Schuster P 1989 Error thresholds of replication in finite populations mutation frequencies and the onset of Muller's ratchet J Theor Biol 137 375 395 2626057
Wagner GP Krall P 1993 What is the difference between models of error thresholds and Muller's ratchet? J Math Biol 32 33 44
Sella G Hirsh AE 2005 The application of statistical physics to evolutionary biology Proc Natl Acad Sci U S A 102 9541 9546 15980155
Boerlijst MC Bonhoeffer S Nowak MA 1996 Viral quasi-species and recombination Proc R Soc Lond B Biol Sci 263 1577 1584
Burch AD Ta J Fane BA 1999 Cross-functional analysis of the Microviridae internal scaffolding protein J Mol Biol 286 95 104 9931252
Huynen MA 1996 Exploring phenotype space through neutral evolution J Mol Evol 43 165 169 8703081
Huynen MA Stadler PF Fontana W 1996 Smoothness within ruggedness: The role of neutrality in adaptation Proc Natl Acad Sci U S A 93 397 401 8552647
Li WH 1997 Molecular evolution Sunderland (Massachusetts) Sinauer Associates 487 p.
Novak CR Fane BA 2004 The functions of the n terminus of the phix174 internal scaffolding protein, a protein encoded in an overlapping reading frame in a two scaffolding protein system J Mol Biol 335 383 390 14659765
Fontana W Schuster P 1987 A computer model of evolutionary optimization Biophys Chem 26 123 147 3607225
Wilke CO 2001 Selection for fitness versus selection for robustness in RNA secondary structure folding Evolution Int J Org Evolution 55 2412 2420
Eigen M McCaskill J 1988 Molecular quasispecies J Phys Chem 92 6881 6891
Charlesworth B 1990 Mutation–selection balance and the evolutionary advantage of sex and recombination Genet Res 55 199 221 2394378
Domingo E Sabo D Taniguchi T Weissmann C 1978 Nucleotide sequence heterogeneity of an RNA phage population Cell 13 735 744 657273
Burch CL Chao L 2000 Evolvability of an RNA virus is determined by its mutational neighbourhood Nature 406 625 628 10949302
Crotty S Andino R 2002 Implications of high RNA virus mutation rates: Lethal mutagenesis and the antiviral drug ribavirin Microbes Infect 4 1301 1307 12443894
Crotty S Cameron CE Andino R 2001 RNA virus error catastrophe: Direct molecular test by using ribavirin Proc Natl Acad Sci U S A 98 6895 6900 11371613
Graci JD Cameron CE 2002 Quasispecies, error catastrophe, and the antiviral activity of ribavirin Virology 298 175 180 12127780
Vignuzzi M Stone JK Andino R 2005 Ribavirin and lethal mutagenesis of poliovirus: Molecular mechanisms, resistance and biological implications Virus Res 107 173 181 15649563
Holmes EC 2003 Error thresholds and the constraints to RNA virus evolution Trends Microbiol 11 543 546 14659685
Cirz RT Chin JK Andes DR de Crecy-Lagard V Craig WA 2005 Inhibition of mutation and combating the evolution of antibiotic resistance PLoS Biol 3 e176. DOI: 10.1371/journal.pbio.0030176 15869329
Kondrashov AS 1994 Muller's ratchet under epistatic selection Genetics 136 1469 1473 8013919
Kun A Santos M Szathmary E 2005 Real ribozymes suggest a relaxed error threshold Nat Genet 37 1008 1011 16127452
Joyce GF 2004 Directed evolution of nucleic acid enzymes Annu Rev Biochem 73 791 836 15189159
Jhaveri SD Hirao I Bell S Uphoff KW Ellington AD 1997 Landscapes for molecular evolution: Lessons from in vitro selection experiments with nucleic acids Annu Rep Comb Chem Mol Divers 1 169 191
Ekland EH Szostak JW Bartel DP 1995 Structurally complex and highly active RNA ligases derived from random RNA sequences Science 269 364 370 7618102
Poon A Otto SP 2000 Compensating for our load of mutations: Freezing the meltdown of small populations Evolution Int J Org Evolution 54 1467 1479
Whitlock MC Otto SP 1999 The panda and the phage: Compensatory mutations and the persistence of small populations Trends Ecol Evol 14 295 296 10407424
Kimura M Crow JF 1964 The number of alleles that can be maintained in a finite population Genetics 49 725 738 14156929
Ewens WJ 1974 A note on the sampling theory for infinite alleles and infinite sites models Theor Popul Biol 6 143 148 4445971
|
16322763
|
PMC1289388
|
CC BY
|
2021-01-05 09:18:23
|
no
|
PLoS Comput Biol. 2005 Nov 25; 1(6):e61
|
utf-8
|
PLoS Comput Biol
| 2,005 |
10.1371/journal.pcbi.0010061
|
oa_comm
|
==== Front
PLoS Comput BiolPLoS Comput. BiolpcbiplcbploscompPLoS Computational Biology1553-734X1553-7358Public Library of Science San Francisco, USA 1632276410.1371/journal.pcbi.001006405-PLCB-RA-0168R1plcb-01-06-05Research ArticleBioinformatics - Computational BiologyGenetics/Functional GenomicsNeuroscienceSystems BiologySaccharomycesCaenorhabditisQuantitative Analysis of Genetic and Neuronal Multi-Perturbation Experiments Quantitative Multi-Perturbation AnalysisKaufman Alon 1*Keinan Alon 2Meilijson Isaac 3Kupiec Martin 4Ruppin Eytan 25*1 Interdisciplinary Center for Neural Computation, Hebrew University, Jerusalem, Israel
2 School of Computer Science, Tel-Aviv University, Tel-Aviv, Israel
3 School of Mathematical Sciences, Tel-Aviv University, Tel-Aviv, Israel
4 Department of Molecular Microbiology and Biotechnology, Tel-Aviv University, Tel-Aviv, Israel
5 School of Medicine, Tel-Aviv University, Tel-Aviv, Israel
Sander Chris EditorMemorial Sloan Kettering Cancer Center, United States of America* To whom correspondence should be addressed. E-mail: [email protected] (AK), [email protected] (ER) 11 2005 25 11 2005 26 10 2005 1 6 e6417 7 2005 26 10 2005 Copyright: © 2005 Kaufman et al.2005This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are properly credited.Perturbation studies, in which functional performance is measured after deletion, mutation, or lesion of elements of a biological system, have been traditionally employed in many fields in biology. The vast majority of these studies have been qualitative and have employed single perturbations, often resulting in little phenotypic effect. Recently, newly emerging experimental techniques have allowed researchers to carry out concomitant multi-perturbations and to uncover the causal functional contributions of system elements. This study presents a rigorous and quantitative multi-perturbation analysis of gene knockout and neuronal ablation experiments. In both cases, a quantification of the elements' contributions, and new insights and predictions, are provided. Multi-perturbation analysis has a potentially wide range of applications and is gradually becoming an essential tool in biology.
Synopsis
Which are the important elements of a system? What are their relative contributions to the performance of the various tasks the system is involved in? These simple and basic questions typically arise when analyzing the workings of any system, and of biological systems in particular. In the latter, the elements may be genes, proteins, cells, or tissues, depending on the level and scope of the analysis. To address these questions in a causal manner, perturbations are required, where the elements are perturbed and the resulting performance function is recorded. This approach has been one of the cornerstones of biological research. However, it has usually been confined to the perturbation of a single element at a time, which may lead to misleading results if the elements of the system functionally interact with each other. This paper addresses these questions by providing a quantitative and rigorous method for the analysis of multi-perturbation experiments, where more than one element may be concomitantly perturbed. The workings of the new method are demonstrated in the analysis of genetic multi-knockout experiments of DNA repair in the yeast Saccharomyces cerevisiae and a neural circuit in the nematode Caenorhabditis elegans accounting for chemotaxis. However, the method is general and can be applied to study many other systems in numerous pertinent biological domains.
Citation:Kaufman A, Keinan A, Meilijson I, Kupiec M, Ruppin E (2005) Quantitative analysis of genetic and neuronal multi-perturbation experiments. PLoS Comput Biol 1(6): e64.
==== Body
Introduction
System identification (localization of function) in biological networks is currently mainly studied in genetics by high-throughput expression profiling and in neuroscience by functional brain imaging. While these techniques have proved to be very useful and productive [1,2], the correlation-based approach they employ does not necessarily identify causal relations. Previous studies have shown that there may be, at times, a weak correlation between the expression of different genes and their role in various cellular functions [3,4]. In a similar vein, the need to add causal perturbation analysis to complement correlation-based ones has also been raised in the neuroscience literature (e.g., [5,6]). To causally deduce the roles played by system elements (genes, proteins, neurons, brain regions, etc.), perturbation studies, in which functional performance is measured after deletion, mutation, or lesion of the different elements, have traditionally been employed. However, the vast majority of these studies have perturbed only one element at a time, often resulting in little phenotypic effect. Hence, in complex biological systems, multiple concomitant perturbations should be employed to reveal the contributions of the different elements to the system's functioning. In genetics, the lack of phenotypic effects may be due to the existence of duplicates, alternative pathways, and functional overlaps [1,7]. To uncover these effects, new experimental techniques are now emerging to carry out the necessary multi-perturbation studies [1,8,9]. Specifically, the recent discovery of RNA interference [10,11] and the rapid recent advances in gene silencing with RNA interference chips [12,13] may advance multi-perturbation technology to our doorstep. In neuroscience, techniques such as transcranial magnetic stimulation allow researchers to induce reversible “virtual lesions,” enabling them to perform multi-lesion experiments for the analysis of cognitive and perceptual tasks in humans [14,15]. A question remains: how can the results of such multi-perturbation experiments be integrated and analyzed, and what knowledge can be extracted from them?
To address this challenge, Keinan et al. [16] developed the multi-perturbation Shapley value analysis (MPA) method, and presented its application to the analysis of a neural model of the lamprey swimming controller and to the analysis of reversible cooling deactivation experiments in cats. Here we expand these results in two fundamental ways. First, we present the first application of the MPA to the analysis of gene knockout experiments and to the analysis of neuronal ablation data. Second, we present a complementary method for the analysis of multi-perturbation data, the functional influence network (FIN) algorithm. In contrast to existing methods for network inference and system identification in biology—which employ methods from machine learning such as Bayesian network inference [17,18], Boolean networks [19], and other reverse engineering methods [20]—both MPA and FIN are based on concepts from game theory. These new methods are the first to utilize fundamental results from game theory to assess the contribution of system elements and functional subsets of such elements to the overall system performance function. For the task of determining the functional contribution of system elements, these game theory tools are more adequate than standard machine learning approaches employing error minimization (e.g., [21]), since they are based on a solid axiomatic framework and provide a unique contribution assignment [16]. Recently, there have been a number of attempts to utilize game theory approaches in neuroscience, but these had a completely different goal of constructing decision-making models [22,23].
The goal of MPA is to define and calculate the contribution (importance) of system elements to a certain function, from a dataset of a series of multi-perturbation experiments. In each such experiment, a different subset of the system elements is concomitantly perturbed (denoting a perturbation configuration), and the system's performance in the studied function is measured. The FIN algorithm analyzes the same multi-perturbation data. It describes the incremental contribution of each subset of elements to the function studied, and produces a compact representation, composed only of the most important subsets. As the full set of all theoretically possible multi-perturbation experiments required for the MPA and FIN computation is usually unavailable, both analyses employ a predictor algorithm to compute the system's performance on the missing multi-perturbation experiments.
In the following sections we describe the MPA and FIN methods and present their application to two different biological systems: the DNA post-replication repair (PRR) pathway in Saccharomyces cerevisiae, and laser ablation studies of Caenorhabditis elegans chemosensory neurons.
Results
Quantitative Multi-Perturbation Analysis
Multi-perturbation Shapley value analysis.
The starting point of MPA [16] is a dataset of multi-perturbation experiments studying a system's performance in a certain function. In each such experiment, a different subset of the system's elements are perturbed concomitantly (denoting a perturbation configuration), and the system's performance following the perturbation is measured. Given this dataset, the goal of MPA is to ascribe to each element its contribution (importance) to the studied function.
The basic observation underlying MPA is that the multi-perturbation setup is essentially equivalent to a coalitional game. A coalitional game is defined by a pair (N, v), where N = {1,…, n} is the set of all players and v(S), for every S ⊆ N, is a real number associating a worth with the coalition S, and v(∅︀) = 0. In the context of multi-perturbations, N denotes the set of all the system's elements, and for each S ⊆ N, v(S) denotes the performance measured under the perturbation configuration in which all the elements in S
are intact and the rest are perturbed.
A payoff profile of a coalitional game is the assignment of a payoff to each of the players. A value is a function that assigns a unique payoff profile to a coalitional game. The function is efficient if the sum of the payoffs assigned to all players is v(N). The definite efficient value in game theory and economics for coalitional games is the Shapley value [24], defined as follows: let the marginal importance of player i to a coalition S, with i ∉ S, be
Then, the Shapley value is defined by the payoff
assigned to player i, for all i ∈ N, where ℜ is the set of all n! orderings of N, and Si(R) is the set of players preceding i in the ordering R. The Shapley value has a clear intuitive interpretation, denoting the average marginal importance of player i to the game. Importantly, it has an axiomatic foundation (see Text S1), which is well suited for the analysis of biological data [16]. The MPA uses the Shapley value as the unique fair measure of each element's contribution (importance) to the function in question.
Obviously, conducting the large number of multi-perturbation experiments (exponential to the size of the system) required for the computation of the Shapley value is most often intractable. In such cases, MPA involves training a predictor using a given subset of multi-perturbation experiments to predict the performance levels of all missing experiments. Given the predicted outcomes of all multi-perturbation experiments, the predicted Shapley value is calculated as the Shapley value based on these outcomes. The accuracy of such an analysis depends on the accuracy of the predictions [16] and is determined using standard cross validation techniques such as leave-one-out [25]. The analysis presented throughout this paper is based on a projection pursuit regression predictor [26]. The accuracy of the predicted contributions is strongly dependent on the accuracy of the predictor used. However, the outcome—the predicted contributions—is more accurate then the individual predictions provided by the predictor because of the fact that the Shapley value is obtained via an averaging over a large number of predictions. Assuming that the predictor is unbiased, prediction errors tend to cancel each other out, resulting in a predicted Shapley value that is unbiased and very similar to the real one. See Protocol S1 for a detailed discussion on the robustness and stability of the predicted contributions produced by the MPA.
Functional influence network.
The FIN algorithm, based on the series of multi-perturbation experiments, begins with the computation of a performance prediction function F, in the form of a multilinear weighted summation over all 2n subsets (summands) of the n elements in the system (see Materials and Methods). The weight of each summand describes the marginal contribution of that subset of elements to the value of F. Given a configuration S of perturbed and intact elements, the goal of F(S) is to accurately describe the experimentally measured performance value of the system in the task studied under that configuration, v(S). After obtaining F, the FIN algorithm proceeds to prune its summands and retains only the most significant ones, to obtain a compact, approximate performance prediction function F˜
(the detailed algorithm is provided in the Materials and Methods). The latter preserves a high percentage of F's original prediction accuracy but aims to provide a compact functional description that may be visualized, if sufficiently compact. Each summand of F and F˜
can be viewed as an integrative functional pathway in the sense that the knockout of any of its elements will zero its contribution. In cases where the full set of all possible multi-perturbation experiments required for the FIN computation is unavailable, the FIN, similar to MPA, uses a predictor algorithm to compute the system's performance in the missing experiments. The accuracy of the resulting FIN description for a given prediction accuracy is part of a broader conceptual issue, that of the relationship between “predictive knowledge” (the prediction accuracy) and “descriptive knowledge” (provided in the case in hand by the FIN output). A comprehensive investigation of this fundamental issue is out of the scope of the current work and will be addressed in a separate paper (preliminary results have been recently presented by Kaufman et al. in the BioPathways Special Interest Group, ISMB 2005).
Gene Knockout Analysis: The Rad6 Pathway
Post-replication repair.
We performed a multi-knockout study of the DNA PRR system of the yeast S. cerevisiae. DNA repair pathways in yeast have been classified genetically into three major repair systems specialized on different types of damage: (1) the excision repair (Rad3) group, which is mainly involved in the repair of UV-irradiated DNA, (2) the recombination repair (Rad52) group, which is mainly involved in repair of damage caused by ionizing radiation and of double-strand breaks in the DNA, and (3) the PRR (Rad6) group of genes. This last pathway is believed to facilitate replication in situations where lesions in the template strand would otherwise cause a stalling of the replication machinery, as occurring following UV radiation.
A key physiological target of the PRR pathway is PCNA, a homotrimeric ring-shaped protein that encircles DNA, functioning as a freely sliding clamp that tethers DNA polymerase to the DNA template. The current hypothesis posits that following the stalling of the replicative DNA polymerases (when lesions are encountered), PCNA is modified, and the replicative polymerase is replaced by trans-lesion polymerases. Ubiquitination of PCNA is carried out by the Rad6 ubiquitin-conjugating enzyme, which is targeted to the stalled replication fork through physical interactions with the Rad18 cofactor [27,28]. During DNA synthesis, Replication factor C (RFC), a heteropentameric protein complex, is necessary for loading PCNA onto double-stranded DNA at the primer–template junction. Recently, several proteins with similarities to Rfc1 (the large subunit of RFC) were found to form RFC-like complexes (RLCs), including Elg1, Rad24, and Ctf18. These RLCs may act similarly to RFC, loading PCNA or PCNA-related complexes that act as clamps for specific DNA polymerases [29]. In addition to the three RLC genes, our study includes RAD18, a gene needed for PCNA modification, and REV3, which encodes an alternative DNA polymerase (ζ).
The analyzed data include a series of multi-knockout experiments carried out in the lab of one of the authors (M. K.), testing the ability of the resulting mutants to resolve the single-stranded gaps created after UV irradiation. Hence, the perturbations were gene knockouts, and the elements were the five genes listed above. The performance under investigation was UV survival, measured by the relative number of colonies that survived compared to the wild-type yeast strain (normalized on a scale from zero to one). The dataset included 21 multi-knockout experiments (see Table S1). Prediction of the full multi-knockout set (i.e., 32 multi-perturbations) was obtained using projection pursuit regression, and explains 79.6% of the data variance via leave-one-out cross validation.
MPA and FIN analysis.
Figure 1 displays the results of MPA of the Rad6 data, leading to a quantification of the causal contribution of each of the Rad6 genes to PRR. The most important genes are RAD18 and REV3, the modifier gene and the DNA polymerase, respectively. All three RLCs play a causal role as well, but their importance differs markedly.
Figure 1 Contributions of Genes in the Rad6 Pathway to PRR Functioning (Normalized Such That Their Sum Equals One)
The multi-knockout data can be utilized to construct a unique weighted multilinear performance prediction function F, which, given any configuration of knocked-out and intact genes, can accurately predict the PRR performance level. However, this function contains 32 (or, more generally, 2n) terms corresponding to all possible knockout configurations and hence is unintelligible and uninformative to the biologist. To extract the relevant information in the data and make it explicit, F is further processed via the FIN algorithm to construct a compact and yet fairly accurate functional prediction function, F˜
. Each of the terms in F˜
can be viewed as a serial functional pathway, whose contribution depends on the intactness of all its component genes. (Obviously, membership in a functional pathway does not necessarily imply that there are direct physical interactions between the elements of the pathway.) F˜
's compactness can be utilized to visualize the PRR functioning via a FIN diagram, as shown in Figure 2.
Figure 2 The FIN Diagram of the PRR Pathway
Visualizes the compact performance function F˜ = e ⋅ (d ⋅ 0.26 + c ⋅ 0.17 + a ⋅ 0.12 + d ⋅ (c ⋅ 0.2 + a ⋅ 0.15) + (c ⋅ a) ⋅ 0.1), where a through e are Boolean variables representing the genes (a = ELG1, b = CTF18 , c = RAD24 , d = REV3, and e = RAD18 ). The investigated genes are represented as binary nodes whose state is determined according to the state of the corresponding genes in a given perturbation configuration, intact or knocked out. The nodes are connected with edges, their weight representing the functional influence between the two endpoint genes (the width of the edge is proportional to its weight). Given a knockout configuration, the expected performance level F˜ can be calculated by summing up the weights on the edges between intact nodes that form a connected component with the function node F˜. For example, in a mutant where both REV3 and ELG1 are knocked-out, the intact nodes are CTF18, RAD24, and RAD18. The edge RAD24–RAD18 is the resulting connected subgraph of F˜ predicting a performance level of 0.17. Observe that there are three main (two-node) pathways leading to F˜
, lines RAD24–RAD18 , ELG1–RAD18, and REV3–RAD18, where RAD18 is an essential gene in all of them. The RLC CTF18 has no significance in the FIN description even though it has a contribution of 4% (see Figure 1); it contributes marginally across many insignificant summands and does not play a significant role in any major one.
The FIN analysis gave rise to two new hypotheses. First, as is evident in Figure 2, both ELG1 and RAD24 play a significant role even without REV3 (edges RAD24–RAD18 and ELG1–RAD18 in Figure 2). Hence, there is probably another polymerase (or perhaps more than one) involved in the PRR process, suggesting that both ELG1 and RAD24 play a role loading this additional DNA polymerase. A good candidate for such an additional polymerase is Pol-η, encoded by the RAD30 gene. This alternative polymerase has been shown to be dependent on Rad6/Rad18 for activity. Second, the REV3–RAD18 pathway (edge REV3–RAD18 in Figure 2) encompasses 26% of the system's repair performance. This suggests that there are some additional DNA polymerase loaders besides those investigated. Alternatively, some of the functionality of the DNA polymerase may be maintained even in the absence of the RLCs.
Analysis of Neuronal Ablations: Chemotaxis in C. elegans
We turn to address the question of function localization in the nervous system, focusing on laser ablation experiments of the C. elegans chemosensory neurons. The behavior studied in these experiments was chemotaxis, in which the nematode directs its movements according to chemical gradients in the environment, moving toward the highest concentration of food or fleeing from toxins. Reanalyzing the data published by Bargmann and Horvitz [30], we compared the qualitative conclusions given in their paper to the quantitative analogs obtained by applying the MPA. The elements studied were eight sensory neuron pairs (out of a total of 16 pairs that form the chemosensory system [31]). In each laser ablation experiment both neurons in a pair were either left intact or perturbed. The performance measures, chemotaxis to various attractants (each composing a distinct functional task), were evaluated by placing the animal on an agar plate with a gradient of an attractant on one side of the plate, and scoring the chemotaxis performance by counting the number of times the animal arrived at the peak of the gradient minus the number of times the animal arrived at the control plug at the opposite side of the plate. The level of chemotaxis performance was evaluated under 31 perturbation configurations, according to the protocol described in [30]. Prediction of the full set of 256 multi-lesions needed to calculate the neurons' contributions was obtained using projection pursuit regression as the predictor. A cross validation leave-one-out procedure shows that the predictor explains 65%–80% of the data variance depending on the attractant type.
Neuronal contribution analysis.
Figure 3 displays the contributions of the different neuron pairs to four different attractants tasks (serotonin, chlorine, cAMP, and biotin). As evident, the ASE pair is the most important to chemotaxis across all attractants, in line with the results of Bargmann and Horvitz [30]. However, MPA additionally shows how the importance of all other neurons varies among the different attractants. The processing of the serotonin task is more distributed (the neuronal contributions are more equally spread) across the network than that of the other tasks. A FIN diagram of the relatively simple and localized cAMP task is provided in Protocol S2. Notably, the ASH pair has a negative, inhibitory, contribution to chemotaxis (chemotaxis will be more successful on average if the ASH pair is ablated). This observation is in line with other more recent experimental assays showing that ASH plays a role in mediating C. elegans avoidance of toxic chemicals [32]. Interestingly, examining the interaction between these two neuron pairs, ASE and ASH, shows a negative interaction in three attractant tasks, serotonin, biotin, and cAMP (the interaction calculation is described in Materials and Methods). In these tasks, ASE's contribution is suppressed by ASH, i.e., the contribution of ASE is lower when ASH is intact than when it is ablated, and similarly, ASH is suppressed by ASE. Thus, these results lead to the prediction that ASE will be an antagonist of avoidance behavior where ASH is likely to be highly activated.
Figure 3 Contributions of the Eight Neuron Pairs to the Different Chemotaxis Attractant Tasks (Normalized Such That Their Sum for Each Attractant Equals One)
Multiple task analysis.
The contributions of the neuron pairs across the different tasks can be summarized in a contribution matrix, where Cij in the matrix denotes the contribution of element j to task i (Table S2). This matrix description permits the utilization of a series of analyses that are not applicable when the results are only summarized in a qualitative manner. Singular value decomposition (SVD), a standard method for dimension reduction previously utilized in various biological applications (e.g., [33]), can be applied to the contribution matrix to reveal both its “neuron space” and its “task space,” identifying similar tasks and similar functional contributions of neurons. Figure 4A presents the results of an SVD of the contribution matrix in the task space. The figure shows four main clusters of the neurons, based on the contributions of different neurons across the attractants (i.e., the column vectors of the contribution matrix in Table S2). The distinct placement of the ASE and ASJ neurons is notable. Neurons participating in each of the clusters {ADF, ASG, ASH} and {ASI, ASK, ADL} have similar functional roles across the investigated attractants. Figure 4B presents the results of SVD in the neuron space. The processing of serotonin chemotaxis is localized very differently than the processing of the other attractants, and the processing of cAMP and biotin is localized in a very similar manner. The similarity between the tasks was already observed by Bargmann and Horvitz [30], although not shown in a rigorous manner. Chemotaxis to chlorine, which was thought to be processed similarly to that of cAMP and biotin [30], is actually processed quite differently.
Figure 4 SVD Analysis of the Contribution Matrix
Uses the two main principal components of the SVD, which together explain 96% of the data's variance.
(A) “Task space,” presenting the projections of the neurons' contribution vectors (column vectors of the contribution matrix) onto the two main principal eigenvectors (PCs) of the task space.
(B) “Neuron space,” presenting the projections of the tasks' contribution vectors (row vectors of the contribution matrix) onto the two main principal eigenvectors (PCs) of the neuron space.
Discussion
This paper presents a multi-perturbation analysis of two different biological systems. The analysis reinforces previously known knowledge in a quantitative manner and leads to new insights. The MPA analysis of the PRR system shows that each of the RLCs has a different magnitude of contribution to the PRR process. The FIN analysis gives rise to the hypotheses that there are additional polymerase loading complexes in yeast and that DNA polymerase ζ encoded by REV3 is probably not the only polymerase involved in PRR. The analysis of C. elegans's chemotaxis provides a more refined picture of the sensory network and rigorously reinforces previous findings.
MPA and FIN are the first methods to our knowledge to harness game theory concepts for the analysis of biological systems. Further work is needed to better adapt these methods to the constraints of biological systems, most notably, the limited depth (i.e., number of concomitantly perturbed elements) of multi-perturbations in biology. However, this is likely to be a very rewarding endeavor, as such multi-perturbation analysis has potentially many applications. The most direct and natural ones are those concerning the analysis of causal perturbation data, e.g., in genetics, using gene silencing with RNA interference. In neuroscience, there is now a new prospect of carrying out experimental perturbation studies using transcranial magnetic stimulation. This technique allows researchers to induce “virtual lesions” in normal subjects performing various cognitive and perceptual tasks [14,15].
Importantly, MPA and FIN are not limited to causal perturbation analysis, where one controls the lesions made. They may well be applied to sets of naturally given multi-perturbations, e.g., by studying the brain localization of cognitive functions from “multi-lesion” data from stroke patients. In summary, multi-perturbation studies are a necessity if one wants to understand the processing of biological networks in a quantitatively causal manner. The methods described in this paper are a harbinger of this new kind of study, offering a novel and rigorous way of making sense out of them.
Materials and Methods
The basic MPA and FIN analysis methods are described at the beginning of the Results. Here we provide a description of the extension of MPA to a two-dimensional interaction analysis and the details of the FIN algorithm.
MPA interaction analysis.
In complex systems, the importance of an element may strongly depend on the state (perturbed or intact) of other elements. A higher order description may be necessary to capture these interactions. Such high-dimensional analysis provides further insights into the network's functional organization.
We focus on the description of two-dimensional interactions. A natural definition of the latter is as follows [16]: let
be the Shapley value of element i in the subgame of all elements without element j, where vN
\{j} is the value function over the set (N\{j}), which denotes the set N without the element j. Intuitively, this is the average marginal importance of element i when element j is perturbed.
Let us now define the coalitional game (M, vM), where M = N\{i, j} ∪ {(i, j)}((i, j) is a new compound element composed of both i and j) and vM
(S), for S ⊆ M, is defined by
where v is the payoff function of the original game with elements N. Then γ(i, j) = γ(i, j)(M, vM), the Shapley value of element (i, j), is the average marginal importance of elements i and j when jointly added to a configuration. The two-dimensional interaction between element i and element j, j ≠ i, is then defined as
which quantifies how much the average marginal importance of the two elements together is larger (or smaller) than the sum of the average marginal importance of each of them when the other is perturbed. Intuitively, this symmetric definition (Ii,j = Ij,i) quantifies the synergistic interaction between elements i and j, denoting how much “the whole is greater than the sum of its parts.” In cases where the whole is smaller than the sum of its parts, i.e., when the two elements exhibit functional overlap or redundancy, the interaction is negative. Based on the genetic interaction nomenclature of Brendel and Haynes [34], an interaction will be defined as “epistatic” if γi,j¯
is zero and γi,j has a positive contribution, i.e., the intactness of j is essential for i's contribution. The Shapley interaction index [35] provides a more general measure for the interaction among players.
A detailed description of the FIN analysis.
The performance prediction function F(S) can be uniquely computed as the sum
, where S denotes the set of intact elements in a given perturbation configuration and the summation goes over all its subsets T [36,37]. The coefficients a(T) of the summands are the dividends, describing the incremental importance of each summand T to the performance being studied. These dividends can be uniquely calculated from the multi-perturbation data (both given and predicted) according to equation 6 (the cardinality of the sets S and T is denoted by corresponding lower-case letters: s = |S| and t = |T|),
The dividend computation is performed in an iterated manner. It begins from the dividend of the null group, and each iteration computes the dividend (incremental contribution) of subsequently larger, subsuming subsets.
To compute a compact and intelligible approximation of F,
F˜
, a greedy heuristic algorithm is employed that retains only the summands T with the largest dividends a(T), while maintaining a predefined level of prediction accuracy. The latter is measured with respect to the performance of the original F (by the normalized mean squared error between F˜
and F over all perturbation configurations). The algorithm first selects statistically significant summands (based on a null hypothesis that the dividend magnitude is zero), and then eliminates those with a low magnitude to obtain F˜
. To visualize F˜
, we construct the FIN diagram. This construction starts with an algebraic simplification, rewriting F˜
to minimize the number of appearances of each element. This is done by combining clauses and placing elements common to a few summands as multipliers of the weighted summation of the corresponding, residual summands. In the DNA PRR investigation (in the Results), for example, this stage results in the function
where a through e are Boolean variables representing the genes, assigned one if the gene is intact and zero if it is knocked out. Based on this simplified representation, we construct the FIN diagram by starting from the function node, F˜
, and connecting it to the variables at the most external level parentheses, assigning weights to the connections according to the corresponding dividend coefficients. This process is recursively repeated by connecting the current leaf nodes on each pathway from the node F˜
to the next level of elements in the remaining parentheses, until the nodes at the most internal parentheses are connected. The resulting FIN diagram depicts the most important functional pathways (interconnected subsets of elements whose contribution depends on the intactness of the other elements in the pathway) and quantifies their relative importance to the function in hand (see Figure 2).
Supporting Information
Protocol S1 MPA Robustness—Partial and Noisy Data
(47 KB PDF)
Click here for additional data file.
Protocol S2 cAMP FIN Diagram
(82 KB PDF)
Click here for additional data file.
Table S1 Multiple Knockout Data
(53 KB PDF)
Click here for additional data file.
Table S2 Chemotaxis Contribution Matrix
(52 KB PDF)
Click here for additional data file.
Text S1 The Axiomatic Basis of the Shapley Value
(50 KB PDF)
Click here for additional data file.
Accession Numbers
The SwissProt (http://www.ebi.ac.uk/swissprot/) accession numbers for the S. cerevisiae proteins discussed in this paper are Ctf18 (P49956), Elg1 (Q12050), Rad18 (P10862), Rad24 (P32641), and Rev3 (P14284).
We thank Nir Yosef, Ya'acov Ritov, Shawn Lockery, and Cori Bargmann for their valuable comments and suggestions. A. Kaufman is supported by the Yeshaya Horowitz Association through the Center of Complexity Science. A. Keinan was supported by the Dan David Prize Scholarship. M. Kupiec was supported by grants from the Recanati Foundation and the Israel Cancer Fund. E. Ruppin's research is supported by the Tauber Fund and the Center of Complexity Science.
Competing interests. A. Keinan, I. Meilijson and E. Ruppin are the authors of a pending PCT patent entitled “A New Method for Multi-Silencing Analysis.”
Author contributions. M. Kupiec conceived and designed the experiments. M. Kupiec performed the experiments. A. Kaufman analyzed the data. A. Kaufman, A. Keinan, I. Meilijson, M. Kupiec, and E. Ruppin contributed reagents/materials/analysis tools. A. Kaufman, A. Keinan, M. Kupiec, and E. Ruppin wrote the paper.
A previous version of this article appeared as an Early Online Release on October 26, 2005 (DOI: 10.1371/journal.pcbi.0010064.eor).
Abbreviations
FINfunctional influence network
MPAmulti-perturbation Shapley value analysis
PRRpost-replication repair
RFCReplication factor C
RLCReplication factor C–like complex
SVDSingular value decomposition
==== Refs
References
Steinmetz LM Davis RW 2004 Maximizing the potential of functional genomics Nat Rev Genet 5 190 201 14970821
Friston K 2002 Beyond phrenology: What can neuroimaging tell us about distributed circuitry? Annu Rev Neurosci 25 221 250 12052909
Winzeler EA Shoemaker D Astromoff A Liang H Anderson K 1999 Functional characterization of the S. cerevisiae genome by gene deletion and parallel analysis Science 285 901 906 10436161
Giaever G Chu AM Ni L Connelly C Riles L 2002 Functional profiling of the Saccharomyces cerevisiae genome Nature 418 387 391 12140549
Kosslyn SM 1999 If neuroimaging is the answer, what is the question? Philos Trans R Soc Lond B Biol Sci 354 1283 1294 10466151
Grobstein P 1990 Strategies for analyzing complex organization in the nervous system: I. lesion experiments Schwartz E Computational neuroscience Cambridge (Massachusetts) MIT Press 19 37
Gu Z Steinmetz LM Gu X Scharfe C Davis RW 2003 Role of duplicate genes in genetic robustness against null mutations Nature 421 63 66 12511954
Carpenter AE Sabatini DM 2004 Systematic genome-wide screens of gene function Nat Rev Genet 5 11 22 14708012
Tong AH Lesage G Bader GD Ding H Xu H 2004 Global mapping of the yeast genetic interaction network Science 303 808 813 14764870
Hammond SM Caudy AA Hannon GJ 2001 Post-transcriptional gene silencing by double-stranded RNA Nat Rev Genet 2 110 119 11253050
Couzin J 2002 Small RNAs make big splash Science 298 2296 2297 12493875
Wheeler DB Bailey SN Guertin DA Carpenter AE Higgins CO 2004 RNAi living-cell microarrays for loss-of-function screens in Drosophila melanogaster cells Nat Methods 1 127 132 15782175
Lehner B Fraser AG 2004 5,000 RNAi experiments on a chip Nat Methods 1 103 104 15782169
Pascual-Leone A Davey NJ Rothwell J Wasserman EM Puri BK 2002 Handbook of transcranial magnetic stimulation Oxford University Press 416 p.
Rafal R 2001 Virtual neurology Nat Neurosci 4 862 864 11528411
Keinan A Sandbank B Hilgetag CC Meilijson I Ruppin E 2004 Fair attribution of functional contribution in artificial and biological networks Neural Comput 16 1887 1915 15265327
Friedman N 2004 Inferring cellular networks using probabilistic graphical models Science 303 799 805 14764868
Pe'er D Regev A Elidan G Friedman N 2001 Inferring subnetworks from perturbed expression profiles Bioinformatics 17 S215 S224 11473012
Ideker TE Thorsson V Karp RM 2000 Discovery of regulatory interactions through perturbation: Inference and experimental design Pac Symp Biocomput 5 305 316
Tegner J Yeung MK Hasty J Collins J 2003 Reverse engineering gene networks: Integrating genetic perturbations with dynamical modeling Proc Natl Acad Sci U S A 100 5944 5949 12730377
Aharonov R Segev L Meilijson I Ruppin E 2003 Localization of function via lesion analysis Neural Comput 15 885 913 12689391
Glimcher PW 2002 Decisions, decisions, decisions: Choosing a biological science of choice Neuron 36 323 332 12383785
Braeutigam S 2005 Neuroeconomics—From neural systems to economic behaviour Brain Res Bull 67 355 360 16216681
Shapley LS 1953 A value for n-person games Kuhn HW Tucker AW Contributions to the theory of games. Volume 2, Annals of mathematics studies 28 Princeton Princeton University Press 307 317
Stone M 1974 Cross-validatory choice and assessment of statistical predictions J R Stat Soc Ser B 36 111 147
Friedman JH Stuetzle W 1981 Projection pursuit regression J Am Stat Assoc 76 817 823
Barbour L Xiao W 2003 Regulation of alternative replication bypass pathways at stalled replication forks and its effects on genome stability: A yeast model Mutat Res 532 137 155 14643434
Johnson RE Prakash S Prakash L 1999 Efficient bypass of a thymine-thymine dimer by yeast DNA polymerase, poleta Science 283 1001 1004 9974380
Ben-Aroya S Koren A Liefshitz B Steinlauf R Kupiec M 2003 ELG1, a yeast gene required for genome stability, forms a complex related to replication factor C Proc Natl Acad Sci U S A 100 9906 9911 12909721
Bargmann CI Horvitz HR 1991 Chemosensory neurons with overlapping functions direct chemotaxis to multiple chemicals in C. elegans
Neuron 7 729 742 1660283
Bargmann CI Mori I 1997 Chemotaxis and thermotaxis Riddle DL Blumenthal T Meyer BJ Priess JR C. elegans II Cold Spring Harbor (New York) Cold Spring Harbor Laboratory Press 717 737
Hilliard MA Bargmann CI Bazzicalupo P 2002
C. elegans responds to chemical repellents by integrating sensory inputs from the head and the tail Curr Biol 12 730 734 12007416
Alter O Brown PO Botstein D 2000 Singular value decomposition for genomewide expression data processing and modeling Proc Natl Acad Sci U S A 97 10101 10106 10963673
Brendel M Haynes RH 1973 Interactions among genes controlling sensitivity to radiation and alkylation in yeast Mol Gen Genet 125 197 216 4359562
Grabisch M Roubens M 1999 An axiomatic approach to the concept of interaction among players in cooperative games Int J Game Theory 28 547 565
Grabisch M Marichal JL Roubens M 2000 Equivalent representations of a set function with applications to game theory and multicriteria decision making Math Oper Res 25 157 178
Harsanyi J 1963 A simplified bargaining model for the n-person cooperative game Int Econ Rev 4 194 220
|
16322764
|
PMC1289391
|
CC BY
|
2021-01-05 09:20:35
|
no
|
PLoS Comput Biol. 2005 Nov 25; 1(6):e64
|
utf-8
|
PLoS Comput Biol
| 2,005 |
10.1371/journal.pcbi.0010064
|
oa_comm
|
==== Front
PLoS Comput BiolPLoS Comput. BiolpcbiplcbploscompPLoS Computational Biology1553-734X1553-7358Public Library of Science San Francisco, USA 1632276510.1371/journal.pcbi.001006505-PLCB-RA-0139R3plcb-01-06-08Research ArticleBioinformatics - Computational BiologyCancer BiologyStatisticsHomo (Human)Allele-Specific Amplification in Cancer Revealed by SNP Array Analysis Allele- and Parent-Specific Copy NumberLaFramboise Thomas 12Weir Barbara A 12Zhao Xiaojun 1Beroukhim Rameen 12Li Cheng 3Harrington David 3Sellers William R 124Meyerson Matthew 125*1 Department of Medical Oncology, Dana-Farber Cancer Institute, Boston, Massachusetts, United States of America
2 The Broad Institute of Harvard and MIT, Cambridge, Massachusetts, United States of America
3 Departments of Biostatistics and Computational Biology, Dana-Farber Cancer Institute and Harvard School of Public Health, Boston, Massachusetts, United States of America
4 Department of Medicine, Harvard Medical School, Boston, Massachusetts, United States of America
5 Department of Pathology, Harvard Medical School, Boston, Massachusetts, United States of America
Bryant Barbara EditorMillennium Pharmaceuticals, United States of America* To whom correspondence should be addressed. E-mail: [email protected] 2005 25 11 2005 28 10 2005 1 6 e6524 6 2005 28 10 2005 Copyright: © 2005 LaFramboise et al.2005This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are properly credited.Amplification, deletion, and loss of heterozygosity of genomic DNA are hallmarks of cancer. In recent years a variety of studies have emerged measuring total chromosomal copy number at increasingly high resolution. Similarly, loss-of-heterozygosity events have been finely mapped using high-throughput genotyping technologies. We have developed a probe-level allele-specific quantitation procedure that extracts both copy number and allelotype information from single nucleotide polymorphism (SNP) array data to arrive at allele-specific copy number across the genome. Our approach applies an expectation-maximization algorithm to a model derived from a novel classification of SNP array probes. This method is the first to our knowledge that is able to (a) determine the generalized genotype of aberrant samples at each SNP site (e.g., CCCCT at an amplified site), and (b) infer the copy number of each parental chromosome across the genome. With this method, we are able to determine not just where amplifications and deletions occur, but also the haplotype of the region being amplified or deleted. The merit of our model and general approach is demonstrated by very precise genotyping of normal samples, and our allele-specific copy number inferences are validated using PCR experiments. Applying our method to a collection of lung cancer samples, we are able to conclude that amplification is essentially monoallelic, as would be expected under the mechanisms currently believed responsible for gene amplification. This suggests that a specific parental chromosome may be targeted for amplification, whether because of germ line or somatic variation. An R software package containing the methods described in this paper is freely available at http://genome.dfci.harvard.edu/~tlaframb/PLASQ.
Synopsis
Human cancer is driven by the acquisition of genomic alterations. These alterations include amplifications and deletions of portions of one or both chromosomes in the cell. The localization of such copy number changes is an important pursuit in cancer genomics research because amplifications frequently harbor cancer-causing oncogenes, while deleted regions often contain tumor-suppressor genes. In this paper the authors present an expectation-maximization-based procedure that, when applied to data from single nucleotide polymorphism arrays, estimates not only total copy number at high resolution across the genome, but also the contribution of each parental chromosome to copy number. Applying this approach to data from over 100 lung cancer samples the authors find that, in essentially all cases, amplification is monoallelic. That is, only one of the two parental chromosomes contributes to the copy number elevation in each amplified region. This phenomenon makes possible the identification of haplotypes, or patterns of single nucleotide polymorphism alleles, that may serve as markers for the tumor-inducing genetic variants being targeted.
Citation:LaFramboise T, Weir BA, Zhao X, Beroukhim R, Li C, et al. (2005) Allele-specific amplification in cancer revealed by SNP array analysis. PLoS Comput Biol 1(6): e65.
==== Body
Introduction
Genomic alterations are believed to be the major underlying cause of cancer [1–3]. These alterations include various types of mutations, translocations, and copy number alterations. The last category involves chromosomal regions with either more than two copies (amplifications), one copy (heterozygous deletions), or zero copies (homozygous deletions) in the cell. Genes contained in amplified regions are natural candidates for cancer-causing oncogenes [4], while those in regions of deletion are potential tumor-suppressor genes [5]. Thus, the localization of these alterations in cell lines and tumor samples is a central aim of cancer research.
In recent years, a variety of array-based technologies have been developed to identify and classify genomic alterations [6–8]. Studies using these technologies typically analyze the raw data to produce estimates of total copy number across the genome [9–11]. However, these studies ignore the individual contributions to copy number from each chromosome. Thus, for example, if a region containing a heterozygous locus undergoes amplification, the question of which allele is being amplified generally remains unanswered. The amplified allele is of interest because it may have been selected for amplification because of its oncogenic effect. Data from array-based platforms have also been employed to identify loss-of-heterozygosity (LOH) events [12,13]. In these studies LOH is typically inferred to have occurred where there is an allelic imbalance in a tumor sample at the same site at which the matched normal sample is heterozygous. A complicating issue (particularly in cancer) is that the imbalance may be due to the amplification of one of the alleles rather than the deletion of the other, and thus LOH may not in fact be present.
Copy number analysis and LOH detection can both be improved by combining copy number measurement with allelotype data. In this paper, we present a probe-level allele-specific quantitation (PLASQ) procedure that infers allele-specific copy numbers (ASCNs) from 100K single nucleotide polymorphism (SNP) array [7] data. Our algorithm yields highly accurate genotypes at the over 100,000 SNP sites. We are also able to infer parent-specific copy numbers (PSCNs) across the genome, making use of the fact that PSCN is locally constant on each chromosome. (PSCNs here mean the copy numbers of each of the two parental chromosomes.) Our results also allow the distinction to be made between true LOH and (false) apparent LOH due to the amplification of a portion of only one of the chromosomes.
The PSCNs of 12 lung cancer samples that we initially analyzed reveal almost exclusively monoallelic amplification of genomic DNA, a result that we subsequently confirm in 89 other lung cell lines and tumors. Monoallelic amplification has previously been noted in the literature on the single gene level [14–16], wherein mutant forms of known oncogenes are amplified, while their wild-type counterparts are left unaltered. To our knowledge, this phenomenon has not previously been described on a genome-wide scale, though proposed mechanisms of amplification such as unequal sister chromatid exchange [17] would suggest monoallelic amplification as the expected result.
In addition, our ASCNs identify the SNP haplotypes being amplified. These haplotypes could conceivably serve as markers for deleterious germ line mutations via linkage disequilibrium. Indeed, the presence of monoallelic amplification makes such linkage studies statistically tractable (see Discussion).
Results
Model Specification and Justification
The 100K SNP array set [7] is a pair of arrays, corresponding to the HindIII and XbaI restriction enzymes, that together are able to interrogate over 100,000 human SNPs. Herein, we shall refer to the pair simply as the 100K SNP array. Its original intended use was to query normal human DNA at specific SNP sites, using a probe set of 40 25-mer oligonucleotide probes to interrogate each SNP. The aim is to identify which of the two alleles—arbitrarily labeled allele A and allele B—occurs in each chromosome at each SNP site. (Note that a diploid normal genome is implicitly assumed, though there are recent reports of copy number variation in normal cells [18,19].) An individual can therefore be genotyped at each SNP as either homozygous AA, homozygous BB, or heterozygous AB.
The design of the array is such that each probe may be classified as either a perfect match (PM; perfectly complementary to one of the target alleles), or a mismatch (MM; identical to a perfect match probe except that the center base is altered so as to be perfectly complementary to neither allele). Further, probes may be subclassified according to whether they are complementary to allele A or allele B, yielding four types of probes: PMA, MMA, PMB, and MMB. A third subclassification is relevant. A probe may either be centered precisely at the SNP site, or may be offset by between one and four bases in either direction. This results in eight types of probes:
. Here the superscripts c and o denote “centered” and “offset,” respectively. Examples of each probe type and their base mismatch properties for a hypothetical SNP are shown in Figure 1. Our model relates a probe's intensity to the number of bases at which it mismatches each of the two allele targets (see below). Note that the eight probe types collapse to five types with respect to affinity for each allele, so that each of the 40 probes in a probe set may be classified as
.
Figure 1 A Hypothetical Example of the Eight Probe Types in the 100K SNP Array [7]
Each probe is a 25-mer designed to be at least partially complementary to a portion of the target fragment. In this diagram, the target contains an A (A allele)/C (B allele) SNP, as shown in brackets. The middle (13th) base of each probe is underlined, and the base corresponding to the SNP site is indicated in bold. The offset probes here are offset by two bases. From the sequences, one can count the number of bases that each probe mismatches each target allele (right columns).
As a first step, we invariant-set normalized [20] all arrays to the same pair (one for the HindIII array and the other for the XbaI array) of baseline arrays using the dChip software (http://www.dchip.org). (Normalization is a standard first step in the analysis of microarray data, and is meant to eliminate unwanted artifacts such as differences in overall array brightness.) Our subsequent analyses are all based on a model that specifies probe intensity as a linear function of the copy numbers of both alleles. The underpinnings of this model are justified by empirical evidence that the signal from oligonucleotide probes is proportional to target quantity up until the point at which the probe becomes saturated [21].
A similar linear model has been well established for use with expression array data [22]. In our model, however, the proportionality parameters depend upon the numbers of bases at which the probe mismatches each target allele. Therefore, we specify the model for (normalized) probe intensity Yk of the kth probe in a fixed SNP's probe set as
Here C
A and C
B are the copy numbers of the A and B alleles, respectively, in the sample being interrogated, and Ak and Bk denote the number of bases (either 0, 1, or 2) at which the kth probe is not perfectly complementary to the A and B targets, respectively. For example, it follows from Figure 1 that the model specifies a PMA probe's intensity as α + β0
C
A + β1
C
B + e. The first term, α, represents background signal, which can arise from optical noise and nonspecific binding [23], and the error e is a normally distributed mean-zero term meant to capture additional sources of variation. Hence the model parameters are α, β0, β1, and β2. These parameters are allowed to be different for forward and reverse strands, and to vary from SNP to SNP, but are assumed to be constant within same-strand portions of probe sets and across different samples in a study. They effectively encode the binding affinities between the probes and targets for each SNP. Finally, our experience indicates that the two-base mismatch signal is essentially indistinguishable from background noise, and hence we set β2 = 0.
From model equation 1 and Figure 1, it directly follows that the background-subtracted mean intensities in a normal sample should depend upon the genotype at the SNP in normal samples according to the inset table in Figure 2. We fit the model to data from nine samples—NA6985, NA6991, NA6993, NA12707, NA12716, NA12717, NA12801, NA12812, and NA12813—that were gathered as part of the International HapMap Project (http://www.hapmap.org). An example of the model fit is illustrated for a specific SNP (rs 2273762) in Figure 2. We estimated values α̂, β̂, and β̂1 for the parameters α, β0, and β1, along with genotyping calls for each sample using an expectation-maximization algorithm [24] (see Materials and Methods). In the figure, it can be seen that each probe classification's mean intensity agrees closely with that assumed by the model (inset table). This is an indication that the model provides a reasonably accurate description of the data.
Figure 2 Average Intensities for Each Probe Type by Sample at a Single SNP (rs 2273762)
The inset table gives the average background-subtracted intensities that would be predicted by our model. The actual background-subtracted mean intensity values (bar graph) in each sample closely agree with what is predicted (inset table).
Genotyping of Normal Samples
We applied our method to the nine samples (see above) that were independently genotyped by centers in the International HapMap Project consortium. Nine different centers were involved in the genotyping of these samples. They employed a variety of platforms, including mass spectroscopy, enzymatic reactions, hybridization, and polymerase chain reaction (PCR)–based techniques. There are approximately 22,000 SNPs that are represented in both the 100K SNP array and the HapMap effort. In the nine samples we studied, a total of 1,198 SNPs were genotyped by two or more different HapMap centers, resulting in 10,782 sample SNP calls. The concordant calls among these multiply genotyped sample SNPs may be treated as being very close to a “gold standard” result, and we used these as a benchmark against which to evaluate the accuracy of our calls. Table 1 summarizes the comparison. The HapMap results have a 98.7% call rate. Among those called, the concordance rate between centers exceeds 99%. Our genotyping algorithm performs quite well, achieving a call rate of 99.27%, and disagreeing with the consensus HapMap genotyping for less than 1% of the calls. The results point to a very high rate of accuracy for our method, and speak well to the suitability of the model.
Table 1 Concordance between Our Model's Calls and Those Made by More Than One Center in the International HapMap Project Effort
A feature of Table 1 that bears further comment is the fact that 16 sample SNPs were called AA by our algorithm and BB by the HapMap consortium. All 16 of these discrepancies occur in either of two SNPs, rs 1323113 or rs 2284867. Close inspection of the raw intensities of the 40 probes at each of these SNPs (data not shown) reveals a strong AA signal for the samples in question. A likely explanation is that the A and B labels were inadvertently switched for these two SNPs when Affymetrix matched its notation to the HapMap effort's alleles.
ASCNs and PSCNs in Cancer DNA Samples
The distinction between ASCN and PSCN may be best understood by considering a hypothetical example of four consecutive SNPs in a genomic region with a total copy number of five. Suppose that the allele A copy numbers for the SNPs are four, zero, five, and one, respectively, leaving allele B copy numbers as one, five, zero, and four. These are what we mean by ASCNs. Taken individually, the ASCNs for the second and third SNPs are noninformative with regard to PSCN, as both the maternal and paternal chromosomes have the same alleles. However, the first and fourth SNPs both indicate that one of the parental chromosomes was amplified to a copy number of four, while the other is unaltered. Thus, we infer PSCNs of four and one for the entire genomic region containing the four SNPs. The ASCN at a SNP site may be viewed as a generalized genotype of the sample.
We initially tested our PLASQ algorithm on a set of 12 lung cancer samples for which we have recently reported total copy number analysis [25], after calibrating the model on 12 normal samples. The cancer samples included one small cell primary tumor, two non-small cell primary tumors, and nine cell lines. Please refer to [25] and the Materials and Methods for additional details. All inferred homozygous deletions are provided in Table 2, while all inferred amplifications with total copy number of at least five are in Table 3. The genome-wide view of inferred PSCN is shown for the H2122 and HCC95 cell lines in Figure 3. The absence of minor chromosome copy numbers (red bars) at high levels on the plot shows that the amplifications are essentially monoallelic.
Table 2 All PLASQ-Inferred Homozygous Deletions, across 12 Lung Cancer Samples
Table 3 All PLASQ-Inferred Amplifications of Total Copy Number of at Least Five, across 12 Lung Cancer Samples
Table 3 Continued
Figure 3 A Depiction of PSCN across the Genome for the Cell Lines H2122 and HCC95
In both graphs green indicates the higher copy number parental chromosome, and red indicates the lower copy number parental chromosome. The total height of each red/green bar indicates the total copy number at the corresponding SNP. Black bars represent homozygous deletions, where total copy number is zero.
All amplicons with total inferred copy number of at least five, throughout all 12 samples, are shown in Figure 4. The most striking feature of this graph is the fact that the vast majority of amplifications exclusively involve only one of the two parental chromosomes. That is, amplification here is monoallelic. Also clear from the figure is the distinction between true LOH (bars with no red portion) and false LOH (bars partly red). We repeated our analysis on 89 other samples (data not shown), on which we similarly obtained the result that amplicons are almost entirely composed of only one of the two parental chromosomes.
Figure 4 PSCNs for All Discovered Amplicons with PLASQ-Inferred Total Copy Numbers of at Least Five
The height of each bar indicates the total copy number for that amplicon. The copy numbers for the parental chromosomes are represented by the red and green portions of each bar. LOH occurs, as indicated, where there is no red portion.
To experimentally validate our PLASQ approach using an independent method, we applied allele-specific real-time PCR. ASCN analysis required changes to the standard copy number analysis by real-time PCR. Standard conditions using Taq polymerase caused the amplification of the target allele, as well as delayed amplification from the other SNP allele. The Stoffel fragment of Taq polymerase, which lacks that enzyme's normal 5′ to 3′ exonuclease activity, increases the specificity of the enzyme for the correct target [26,27]. This consequently increases the amplification delay enough to distinguish the two alleles and calculate accurate copy numbers.
In [25], we used standard real-time PCR to verify the total copy number for “recurrent” amplifications and deletions. We defined an event to be recurrent if it occurred in at least two samples, contained at least four SNPs, and was at least 5 kb in length. The comparison of our PLASQ analysis to both allele-specific and standard real-time PCR is given in Tables 4 and 5 for these recurrent events that occur in our initial 12 samples. PLASQ largely agrees with the PCR measurements for homozygous deletions (Table 4). For amplifications (Table 5), there is strong concordance between our estimates and the allele-specific PCR results. The rounded minor allele estimates differ by at most one copy in all but one case. With regard to major allele copy number inferences in Table 5, our estimates tend to be somewhat low, though they are always at elevated levels where the PCR results are. These discrepancies are likely the result of saturation effects that are well known in oligonucleotide arrays [28]. There is only one case where the total PCR estimate from [25] is lower than the PLASQ total. Here the allele-specific PCR results are in closer agreement with our inferred ASCN, indicating that this is an experimental error in the standard real-time PCR.
Table 4 Comparison of Inferred ASCNs with PCR Results for Deletions
Table 5 Comparison of Inferred ASCNs with PCR Results for Amplifications
One type of discrepancy in Table 5 stands out. In two cases, PLASQ infers an ASCN of one, whereas the experimentally determined copy number was essentially zero. One possible explanation is that our inference is correct and the low PCR estimates are attributable to experimental errors such as suboptimal primer sequences. On the other hand, our ASCN calls are somewhat vulnerable to the inherent noise in hybridization-based intensity measurements. At the single SNP level, deviations of one copy number in either direction may be difficult to detect because of this noise, resulting in slightly inaccurate ASCN calls. However, these inaccuracies are ameliorated in PSCN calls since we may “borrow strength” from neighboring SNPs' raw ASCNs because of the locally constant property of PSCN. Thus, for example, the LOH calls for regions will be very precise even when individual ASCN calls are slightly erroneous.
It is important to note that in all cases, the property of interest—the presence or absence of amplification or deletion in each chromosome—is clearly detectable with our method, as all approaches agree in this regard. Finally, in order to assess the accuracy of our determination of amplicon and deletion boundaries, we compared the results that were determined in [25] using an algorithm implemented in the dChipSNP computational platform [9] to our results. The comparison is shown in Table 6 for the events in Tables 4 and 5. In most cases, our estimated alteration boundaries correspond exactly to those inferred by dChipSNP. Events for which the two approaches differ in their inferences could be due to procedural differences such as varying copy number thresholds used to determine whether or not a gain should be called an amplification.
Table 6 Comparison of PLASQ-Inferred Lesion Boundaries with Those from [25]
Amplification of EGFR Mutant
In order to determine whether amplification could target, in a monoallelic fashion, an activating mutation in one of our samples, we examined sequence data for the EGFR gene. It was shown in [25] that the HCC827 cell line harbors the E746 A750del deletion mutant. This is a known activating mutation [29,30], and our result in Table 5 predicts ASCNs of 11 and two at this locus. It was interesting, therefore, to determine whether the greatly amplified chromosome is the one harboring the mutation. To answer this question, we performed quantitative PCR experiments that are able to differentiate the wild-type copies from the mutant copies (see Materials and Methods). The wild-type allele was found to be unamplified (PCR estimate 0.80), while the total PCR copy number was 39.78. Thus, our method uncovered a targeted amplification of an activating mutant allele over its wild-type counterpart.
Discussion
Many genomic events of interest are easily placed in the context of ASCN and PSCN. LOH at a SNP site occurs where one of the PSCNs is zero. Monoallelic amplification occurs at loci where one parental chromosome has a copy number less than two and the other has a copy number greater than one. We have demonstrated that these events, among others, may be identified though ASCN and PSCN from 100K SNP array data. Examining array data from over 100 lung cancer samples, we have found that amplifications are overwhelmingly monoallelic. Current understanding of the mechanisms behind amplification in tumorigenesis would suggest this as an expected result. For example, Herrick et al. [17] describe mechanisms that would all lead to monoallelic amplification in genes. To our knowledge, however, this phenomenon has not been demonstrated on a genome-wide scale in the literature.
Previous studies have demonstrated monoallelic amplification at specific genes. Hosokawa and Arnold [14] found two tumor cell lines in which a mutant allele of cyclin D1 is amplified but the wild-type copy is not. Zhuang et al. [16] uncovered a similar trend in 16 renal carcinoma tumors heterozygous for a MET mutation, and a study of 26 mouse skin tumors found 16 with a mutant HRAS homolog allele amplified but none with the wild-type allele amplified [15]. Using our procedure, we have uncovered (and validated) an EGFR example in one of our samples. These cases highlight the targeting of one genetic variant for amplification over another at a heterozygous site, presumably in order to give the cell growth advantage. However, further studies involving a larger set of tumors are necessary to uncover multiple instances of the transforming variant being the amplification target. A large number of such cases would provide compelling evidence for the biological significance of allele-specific amplification of genes. In some studies these monoallelic amplifications may be erroneously called LOH because of the allelic imbalance. Our approach was able to determine that, in most cases, the minor allele is not in fact deleted, and thus LOH has not occurred.
ASCN information may be used to identify SNP haplotypes in cancer cell amplicons. This haplotype structure determination has important applications for uncovering candidate oncogenes and tumor suppressor genes. The applications may be understood in the context of a recent study [31] that characterizes the genome as consisting of haplotype blocks—regions with few distinct haplotypes commonly observed in human populations—separated by recombination “hotspots.” Indeed, consider an inherited variant that predisposes a cell toward tumor growth and is selected for amplification. Many SNP sites located in the same haplotype block would be amplified along with the variant. One may determine the haplotype of the amplicon via ASCN. The SNP haplotype in the same block as the gene, therefore, may serve as a marker for the variant through genetic association studies [32]. We point out that, were it not for monoallelic amplification, this endeavor would be far more difficult, for if both parental chromosomes were amplified then both haplotypes would be candidate markers for the deleterious variant. Statistically, the power to detect association would be significantly compromised.
Our method produces, in addition, highly accurate genotype calls in normal cells. Analyzing sample SNPs that were genotyped by at least two independent groups, we had over 99% agreement with their concordant calls. Given the strength of our results, we are now working to apply the model to data from oligonucleotide resequencing arrays [33].
Note that our procedure is does not take into account all types of genomic alterations. For example, it would be somewhat confounded by a translocation event. A translocation would induce a loss of the “local constancy” property of total copy number. Similarly, point mutations are not detectable with our approach, and in fact could adversely affect copy number measurements if they were to occur near 100K SNP sites. Still, we feel that these limitations do not severely impact the applicability of the method.
The structure of our model suggests a very useful extension. A common problem in analyzing the genomic content of tumor cells is that of stromal contamination—the presence of normal cells in the sample. Stromal contamination makes accurate copy number determination difficult because the quantity measured is actually a weighted average of the normal and cancer cell copy numbers. Mathematically, the sample's ASCNs at a fixed SNP site may be expressed as
where p
S is the (unknown) proportion of stroma, C
AS and C
BS are the ASCNs of the stromal cells, and C
AT and C
BT are the (unknown) ASCNs in the tumor. We may treat C
AS and C
BS as known, since a matched normal sample may be genotyped at the SNP. Thus, replacing C
A and C
B in our model with the expressions in equation 2 above gives each probe's intensity as a function of true cancer cell ASCNs and proportion of stromal content. Although beyond the scope of this paper, this is an intriguing bioinformatic approach to a pervasive experimental problem.
In summary, we have presented a procedure, termed PLASQ, that is not only able to localize copy number alterations in cancer cells, but can also identify each chromosome's contribution to these alterations as well as the SNP haplotypes in each event. Our approach has been validated using a variety of independent experimental techniques. We have also described several applications and extensions of our methods, and we have demonstrated that chromosomal amplifications in human lung cancer are monoallelic. Finally, it has come to our attention that, while this work was under review, a pair of papers [34,35] describing methods to infer PSCN from 100K SNP array data was published. The approaches differ from ours, and appear to require matched normal samples.
An R [36] package, downloadable at http://genome.dfci.harvard.edu/~tlaframb/PLASQ, contains procedures and data described in this work.
Materials and Methods
The PLASQ procedure for genotyping normal and aberrant samples (thereby obtaining ASCN and PSCN), beginning with the SNP array .cel files, is outlined in Figure 5. Details of each step are given below and in the Results.
Figure 5 The PLASQ Procedure for Determining ASCN and PSCN from the .cel Files
After normalizing signal intensities from all samples, the model is first fit to the normal samples' data to produce both genotype calls and parameter estimates at each SNP site. The latter are used in the model as applied to the data from the cancer samples. Ordinary least squares fitting produces raw ASCN estimates at each SNP. The corresponding raw total copy number estimates are smoothed using circular binary segmentation. Finally, further processing yields our final ASCN and PSCN inferences (see Materials and Methods). EM algorithm, expectation-maximization algorithm.
DNA samples.
We obtained the Affymetrix .cel files from all lung cancer tumors and cell lines analyzed in [25]. In our analysis, we used the same raw probe-level data that were generated from the experiments in that study. For initial analysis, we selected the cell lines H157, H2087, H2122, H2126, H2882, HCC95, HCC827, HCC1359, and HCC1171, as well as tumors S0177T, S0465T, and S0515T. These 12 samples were chosen because each was found [25] to harbor at least two of the copy number alterations that were considered recurrent. We subsequently applied our approach to the remaining 89 tumors and cell lines in that study. Additionally, the 12 normal samples from that paper were employed in the study. Details about the preparation, hybridization, and image acquisition for all samples may be found in [25], and all .cel files are available at http://research2.dfci.harvard.edu/dfci/snp/. We obtained the HapMap samples' .cel files from the Affymetrix Web site (http://www.affymetrix.com).
Normal sample genotyping.
In this case, for each sample the value of C
A at a SNP is either zero, one, or two. The value of C
B is completely determined by C
A, as C
A + C
B = 2. Thus, we may think of each sample SNP as being in one of three states, corresponding to the AA, AB, and BB genotypes. These states are not known a priori, and neither are the values of α, β0, and β1. We employ an expectation-maximization algorithm [24] at each SNP to infer the genotypes and estimate the parameters. Briefly, we first initialize the probabilities of the three genotypes of each sample using a crude t-test approach. Based on these initial “guesses,” we apply ordinary least squares [37] to our model, finding the maximum likelihood estimates of the parameters α, β0, and β1 (the M step). Next, based upon these estimates, we re-infer the genotype probabilities of each sample using the expected values of the indicator variables for each of the three possible genotypes (the E step). These two steps—maximization and expectation—are iterated until the approximated values of all unknowns converge. The result of this procedure is an estimated probability of each genotype along with parameter estimates. The algorithm's call at each sample SNP is the genotype with the maximum final estimated probability, unless the maximum falls under a user-defined threshold (the default is 99%), in which case a “No Call” is given. We subsequently use the final parameter estimates α̂, β̂, and β̂1 of α, β0, and β1, respectively, in the application of the model to data from cancer cells (see below).
Total copy number in cancer DNA samples.
In an aberrant sample, copy numbers of the A and B alleles are no longer constrained to sum to two at each SNP. After calibrating the model on normal samples as described above, we replace the parameters α, β0, and β1 in our model with their estimates at each SNP. We directly apply least squares estimation to find raw inferences (“raw” because we do not yet exploit local constancy of total copy number) of the A and B copy numbers at each SNP. These rough measures are referred to as the raw ASCNs. While the ASCNs are not locally constant in a sample, their pairwise sums C
A + C
B are. We therefore input the pairwise sums of the raw ASCNs at each SNP into the circular binary segmentation algorithm [38] to infer total copy number. This smoothing algorithm exploits the fact that chromosomal alterations typically occur in segments containing several SNPs. Briefly, circular binary segmentation searches for locally constant sections by recursively splitting chromosomes into candidate subsegments and computing a maximum t-statistic that reflects differences in mean total raw copy number between subsegments. The reference distribution for this statistic, estimated by permutation, is used to decide whether or not to permanently split at each stage. The result is a segmentation of each chromosome in a sample, where the total copy number is deemed constant within each segment. Our raw total copy number of a segment is the mean of the pairwise sums of the raw ASCNs of all SNPs in the segment.
PSCNs and ASCNs.
The circular binary segmentation algorithm divides each sample's genome into segments, each assumed to have the same total copy number. Consider a segment with n SNPs and a raw total copy number T
raw. We infer PSCN for the segment as follows. If n < 4, we consider T
raw to be too noisy due to the small number of observations, and infer PSCNs (1, 1). For n ≥ 4, if T
raw ≤ 0.35, the segment is called a homozygous deletion, giving PSCNs (minor chromosome, major chromosome) = (0, 0). If 0.35 < T
raw ≤ 1.35, we call a heterozygous deletion with PSCNs (0, 1). If T
raw > 1.35, our inferred total copy number T is simply T
raw rounded to the nearest integer (or to two if 1.35 < T
raw ≤ 2.5), and we proceed as follows.
Let A
1, A
2,…, An and B
1, B
2,…, Bn denote the raw ASCNs for the n SNPs in a segment. We consider a SNP i to be homozygous if minimum (Ai, Bi) ≤ 0.5. We must first consider the possibility that one of the parental chromosomes is deleted while the other is amplified, i.e., the SNP may be homozygous either because it was homozygous in the normal cell, or because of LOH. Since the average heterozygosity rate for SNPs on the array is 0.3 [39], the probability of a randomly chosen SNP being homozygous is 0.7. Thus, we model the number of homozygous SNPs in a segment without chromosomal deletion as a binomial (n, 0.7) random variable X. The resulting hypothesis test would reject the null hypothesis of no LOH at the α level if
Making a conservative Bonferroni correction for multiple testing on the total number of segments s, we assume deletion of one chromosome if the null hypothesis is rejected at the α = 0.05/s level. In this case, our inferred PSCNs are (0, T). Otherwise, note that (as discussed in Results) homozygous SNP sites are noninformative with regard to PSCN. Thus, we temporarily ignore those SNPs, leaving m SNPs (m ≤ n) whose raw ASCNs we relabel A
1, A
2,…, Am and B
1, B
2,…, Bm. Our inferred minor chromosome PSCN is then
rounded to the nearest integer. In order to ensure that total copy number is T, the inferred major chromosome PSCN is T − (inferred minor chromosome PSCN).
Once PSCNs are determined, the ASCNs follow immediately from these and the raw ASCNs. The homozygous SNPs (determined as in the paragraph above) are assigned the allele with the larger raw ASCN. Heterozygous SNPs are assigned ASCNs so that the allele with the larger raw ASCN has the copy number of the major parental chromosome.
PCR-based copy number validation.
Relative copy numbers for both alleles of a SNP site were determined by quantitative real-time PCR using both a PRISM 7500 Sequence Detection System (96 well) and a PRISM 7900HT Sequence Detection System (384 well) (Applied Biosystems, Foster City, California, United States). Real-time PCR was performed in 25-μl (96 well) or 12.5-μl (384 well) reactions with 2 ng or 1 ng, respectively, of template DNA. SYBR Green I (Molecular Probes; Eugene, Oregon, United States) and the Stoffel fragment of Taq polymerase (Applied Biosystems) [27] were used for the PCR reaction. The reaction mix used was as described previously [27], with the following exceptions: 3U of Stoffel polymerase, 100 μM dUTP, and 0.5 μM ROX (Invitrogen, Carlsbad, California, United States) were used per reaction. Primers were designed with the help of Primer 3 (http://frodo.wi.mit.edu/cgi-bin/primer3/primer3_www.cgi) and synthesized by Invitrogen. For each SNP site three primers were designed, one common for the region and two designed with the 3′ base of the primer specific for each SNP allele. The common primer plus one of the SNP-specific primers were used for each PCR reaction (0.3 μM each). Primer sequences are available upon request. PCR conditions were as follows: 2 min at 50 °C, 15 min at 95 °C, followed by 47 three-step cycles of (20 s at 95 °C, 20 s at 60 °C, and 30 s at 72 °C). The standard curve method was used to calculate the copy number of each allele of a target SNP site in the tumor DNA sample relative to a reference, the Line-1 repetitive element whose copy number is similar between both normal and cancerous cells. Quantification was based on standard curves from a serial dilution of human normal genomic DNA. The relative target copy number level for each allele of a SNP target site was normalized to normal human genomic DNA, heterozygous for that particular SNP site, as calibrator. Changes in the target allele copy number relative to the Line-1 and the calibrator were determined using the formula (T
target/T
Line-1)/(C
target/C
Line-1), where T
target and T
Line-1 are the DNA quantities from tumor by using the target allele and Line-1, and C
target and C
Line-1 are the DNA quantities from the calibrator by using the target allele and Line-1. The copy number of both alleles for each SNP site was determined in this way.
Real-time PCR was also used to determine the relative copy number of the two EGFR alleles in the HCC827 cell line, which contains the E746 A750del mutation and an amplification of the EGFR region. Real-time PCR was performed with the Stoffel fragment of Taq polymerase using reaction mix and conditions described above. The standard curve method was used to calculate the total copy number of the EGFR gene and the copy number of the wild-type allele in the HCC827 DNA sample normalized to Line-1 and a normal reference DNA. The primer pairs consisted of one common reverse primer, with one forward primer that would bind both EGFR alleles (wild-type and mutated) and one forward primer specific for the wild-type allele. The primer specific for the wild-type EGFR allele was designed so that the 3′ end was located within the DNA deleted by the E746 A750del mutation. Two PCR reactions were performed: one that gave total EGFR copy number (using primer that binds both alleles) and one that gave only wild-type EGFR copy number (using primer specific for wild-type EGFR).
Supporting Information
Accession Numbers
The NCBI Entrez Gene (http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?db=gene) accession numbers for the genes discussed in this paper are cyclin D1 (595), EGFR (1956), HRAS (3265), and MET (4233).
This project was supported by the following grants: Department of Defense grant PC040638 (RB), the Claudia Adams Barr Program in Cancer Research (CL), US National Institute of Allergy and Infectious Diseases grant 2R01 AI052817 (DH), National Cancer Institute grant R01CA109038 (WRS), the Damon-Runyon Cancer Research Foundation (WRS), American Cancer Society grant RSG-03–240–01-MGO (MM), and the Flight Attendant Research Institute (MM). The authors express thanks to Eric Lander for helpful comments, and to the referees, whose careful reading and critique resulted in a much-improved manuscript.
Competing interests. The authors have declared that no competing interests exist.
Author contributions. TL and BAW conceived and designed the experiments. BAW and XZ performed the experiments. TL developed the statistical procedure and analyzed the data. RB contributed reagents/materials/analysis tools. CL, DH, and WRS advised on the project. MM supervised the project. TL and BAW wrote the paper.
A previous version of this article appeared as an Early Online Release on October 28, 2005 (DOI: 10.1371/journal.pcbi.0010065.eor).
Abbreviations
ASCNallele-specific copy number
LOHloss of heterozygosity
MMmismatch
PCRpolymerase chain reaction
PLASQprobe-level allele-specific quantitation
PMperfect match
PSCNparent-specific copy number
SNPsingle nucleotide polymorphism
==== Refs
References
Weinberg RA 1996 How cancer arises Sci Am 275 62 70 8701295
Futreal PA Coin L Marshall M Down T Hubbard T 2004 A census of human cancer genes Nat Rev Cancer 4 177 183 14993899
Weir B Zhao X Meyerson M 2004 Somatic alterations in the human cancer genome Cancer Cell 6 433 438 15542426
Little CD Nau MM Carney DN Gazdar AF Minna JD 1983 Amplification and expression of the c-myc oncogene in human lung cancer cell lines Nature 306 194 196 6646201
Knudson AG 1971 Mutation and cancer: Statistical study of retinoblastoma Proc Natl Acad Sci U S A 68 820 823 5279523
Pollack JR Perou CM Alizadeh AA Eisen MB 1999 Genome-wide analysis of DNA copy-number changes using cDNA microarrays Nat Genet 23 41 46 10471496
Matsuzaki H Dong S Loi H Di X Liu G 2004 Genotyping of over 100,000 SNPs on a pair of oligonucleotide arrays Nat Methods 1 109 111 15782172
Lucito R Healy J Alexander J Reiner A Esposito D 2003 Representational oligonucleotide microarray analysis: A high-resolution method to detect genome copy number variation Genome Res 13 2291 2305 12975311
Zhao X Li C Paez JG Chin K Jänne PA 2004 An integrated view of copy number and allelic alterations in the cancer genome using single nucleotide polymorphism arrays Cancer Res 64 3060 3071 15126342
Huang J Wei W Zhang J Liu G Bignell GR 2004 Whole genome DNA copy number changes identified by high density oligonucleotide arrays Hum Genomics 1 287 299 15588488
Brennan C Zhang Y Leo C Feng B Cauwels C 2004) High resolution global profiling of genomic alterations with long oligonucleotide microarray Cancer Res 64 4744 4748 15256441
Lindblad-Toh K Tannenbaum DM Daly MJ Winchester E Lui WO (2000 Loss-of-heterozygosity analysis of small-cell lung carcinomas using single-nucleotide polymorphism arrays Nat Biotechnol 18 1001 1005 10973224
Wang ZC Lin M Wei LJ Li C Miron A 2004 Loss of heterozygosity and its correlation with expression profiles in subclasses of invasive breast cancers Cancer Res 64 64 71 14729609
Hosokawa Y Arnold A 1998 Mechanism of cyclin D1 (CCND1, PRAD1) overexpression in human cancer cells: Analysis of allele-specific expression Genes Chromosomes Cancer 22 66 71 9591636
Bianchi AB Aldaz CM Conti CJ 1990 Nonrandom duplication of the chromosome bearing a mutated Ha-ras-1 allele in mouse skin tumors Proc Natl Acad Sci U S A 87 6902 6906 1697691
Zhuang Z Park W Pack S Schmidt L Vortmeyer AO 1998 Trisomy 7-harbouring non-random duplication of the mutant MET allele in hereditary papillary renal carcinomas Nat Genet 19 66 69
Herrick J Conti C Teissier T Thierry F Couturier J 2005 Genomic organization of amplified MYC genes suggests distinct mechanisms of amplification in tumorigenesis Cancer Res 65 1174 1179 15735000
Sebat J Lakshmi B Troge J Alexander J Young J 2004 Large-scale copy number polymorphism in the human genome Science 305 525 528 15273396
Iafrate AJ Feuk L Rivera MN Listewnik ML Donahoe PK 2004 Detection of large-scale variation in the human genome Nat Genetics 36 949 951 15286789
Li C Wong W 2001 Model-based analysis of oligonucleotide arrays: Model validation, design issues and standard error application Genome Biol 2 RESEARCH0032 11532216
Bignell GR Huang J Greshock J Watt S Butler A 2004 High resolution analysis of DNA copy number using oligonucleotide microarrays Genome Res 14 287 295 14762065
Li C Wong WH 2001 Model-based analysis of oligonucleotide arrays: Expression index computation and outlier detection Proc Natl Acad Sci U S A 98 31 36 11134512
Irizarry RA Hobbs B Collin F Beaxer-Barclay Y Antonellis K 2003 Exploration, normalization, and summaries of high density oligonucleotide array probe level data Biostatistics 4 249 264 12925520
Dempster AP Laird NM Rubin DB 1977 Maximum likelihood from incomplete data via the EM algorithm J R Stat Soc Ser B 39 1 38
Zhao X Weir BA LaFramboise T Lin M Beroukhim R 2005 Genomic alterations in human lung carcinomas revealed by single nucleotide polymorphism (SNP) array analysis Cancer Res 65 5561 5570 15994928
Lawyer FC Stoffel S Saiki RK Chang SY Landre PA 1993 High-level expression, purification, and enzymatic characterization of full-length Thermus aquaticus DNA polymerase and a truncated form deficient in 5′ to 3′ exonuclease activity PCR Methods Appl 2 275 287 8324500
Germer S Holland M Higuchi R 2000 High-throughput SNP allele frequency determination in pooled DNA samples by kinetic PCR Genome Res 10 258 266 10673283
Naef F Socci ND Magnaso M 2003 A study of accuracy and precision in oligonucleotide arrays: Extracting more signal at large concentrations Bioinformatics 19 178 184 12538237
Paez JG Janne PA Lee JC Tracy S Greulich H 2004 EGFR mutations in lung cancer: Correlation with clinical response to gefitinib therapy Science 304 1497 1500 15118125
Lynch TJ Bell DW Sordella R Gurubhagavatula S Okimoto RA 2004 Activating mutations in the epidermal growth factor receptor underlying responsiveness of non-small-cell lung cancer to gefitinib N Engl J Med 350 2129 2139 15118073
Gabriel SB Schaffner SF Nguyen H Moore JM Roy J 2002 The structure of haplotype blocks in the human genome Science 296 2225 2229 12029063
Lange K 2002 Mathematical and statistical methods for genetic analysis, 2nd ed New York Springer-Verlag 384 p.
Affymetrix 2003 GeneChip CustomSeq resequencing arrays data sheet. Santa Clara (California): Affymetrix Available: http://www.affymetrix.com/support/technical/datasheets/customseq_datasheet.pdf. Accessed 31 October 2005.
Ishikawa S Komura D Tsuji S Nishimura K Yamamoto S 2005 Allelic dosage analysis with genotyping arrays Biochem Biophys Res Commun 333 1309 1314 15982637
Nannaya Y Sanada M Nakazaki K Hosoya N Wang L 2005 A robust algorithm for copy number detection using high-density oligonucleotide single nucleotide polymorphism genotyping arrays Cancer Res 65 6071 6079 16024607
R Development Core Team 2004 R: A language and environment for statistical computing [computer program] Vienna R Foundation for Statistical Computing
Stapleton JH 1995 Linear statistical models New York Wiley 472 p.
Olshen AB Venkatraman ES Lucito R Wigler M 2004 Circular binary segmentation for the analysis of array-based DNA copy number data Biostatistics 5 557 572 15475419
Affymetrix 2004 GeneChip human mapping 100K set data sheet. Santa Clara (California): Affymetrix Available: http://www.affymetrix.com/support/technical/datasheets/100k_datasheet.pdf. Accessed 31 October 2005.
|
16322765
|
PMC1289392
|
CC BY
|
2021-01-05 09:18:23
|
no
|
PLoS Comput Biol. 2005 Nov 25; 1(6):e65
|
utf-8
|
PLoS Comput Biol
| 2,005 |
10.1371/journal.pcbi.0010065
|
oa_comm
|
==== Front
PLoS Comput BiolPLoS Comput. BiolpcbiplcbploscompPLoS Computational Biology1553-734X1553-7358Public Library of Science San Francisco, USA 1632276610.1371/journal.pcbi.001006605-PLCB-RA-0188R1plcb-01-06-07Research ArticleBiochemistryBioinformatics - Computational BiologySystems BiologySaccharomycesRefining Protein Subcellular Localization Refining Protein Subcellular LocalizationScott Michelle S 1Calafell Sara J 1Thomas David Y 2Hallett Michael T 1*1 McGill Center for Bioinformatics, McGill University, Montreal, Quebec, Canada
2 Biochemistry Department, Faculty of Medicine, McGill University, Montreal, Quebec, Canada
Nielsen Henrik EditorTechnical University of Denmark, Denmark* To whom correspondence should be addressed. E-mail: [email protected] 2005 25 11 2005 1 6 e663 8 2005 28 10 2005 Copyright: © 2005 Scott et al.2005This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are properly credited.The study of protein subcellular localization is important to elucidate protein function. Even in well-studied organisms such as yeast, experimental methods have not been able to provide a full coverage of localization. The development of bioinformatic predictors of localization can bridge this gap. We have created a Bayesian network predictor called PSLT2 that considers diverse protein characteristics, including the combinatorial presence of InterPro motifs and protein interaction data. We compared the localization predictions of PSLT2 to high-throughput experimental localization datasets. Disagreements between these methods generally involve proteins that transit through or reside in the secretory pathway. We used our multi-compartmental predictions to refine the localization annotations of yeast proteins primarily by distinguishing between soluble lumenal proteins and soluble proteins peripherally associated with organelles. To our knowledge, this is the first tool to provide this functionality. We used these sub-compartmental predictions to characterize cellular processes on an organellar scale. The integration of diverse protein characteristics and protein interaction data in an appropriate setting can lead to high-quality detailed localization annotations for whole proteomes. This type of resource is instrumental in developing models of whole organelles that provide insight into the extent of interaction and communication between organelles and help define organellar functionality.
Synopsis
Eukaryotic cells are divided into various morphologically and functionally distinct compartments. Proteins must be targeted to the appropriate compartment to ensure proper function. Understanding protein subcellular localization is important to help understand not only the function of individual proteins but also the organization of the cell as a whole. Bioinformatic predictors of localization can provide such information quickly for large numbers of proteins. The authors of this paper have created a localization predictor called PSLT2 that considers the combinatorial presence of protein motifs and domains as well as protein interactions in yeast proteins. PSLT2 can predict the localization of all yeast proteins to nine different compartments: the endoplasmic reticulum, Golgi apparatus, cytosol, nucleus, peroxisome, plasma membrane, lysosome, mitochondrion, and extracellular space. The authors also investigated how to identify and predict proteins that localize to more than one compartment. They compared the localization predictions of PSLT2 to those determined through high-throughput tagging and microscopy experiments for yeast proteins. Disagreements between these methods generally involve proteins that transit through or reside in the secretory pathway.
Citation:Scott MS, Calafell SJ, Thomas DY, Hallett MT (2005) Refining protein subcellular localization. PLoS Comput Biol 1(6): e66.
==== Body
Introduction
Subcellular localizations determine the environments in which proteins operate. As such, subcellular localization influences protein function by controlling access to and availability of all types of molecular interaction partners. Thus, knowledge of protein localization often plays a significant role in characterizing the cellular function of hypothetical and newly discovered proteins. There are several research endeavours that aim to localize whole proteomes by using high-throughput approaches [1–3]. These large datasets provide important information about protein function, and more generally global cellular processes. However, they currently do not achieve 100% coverage of proteomes, and the methodology used can in some cases cause mislocalization of subsets of proteins [4,5]. Complementary methods are necessary to address these problems.
Many efforts have focused on the creation of bioinformatic predictors of localization via different machine learning methods, and using various protein characteristics (reviewed in [6]). Available predictors can be grouped into four general classes based on the protein characteristics considered: amino acid composition and order-based predictors [7,8], sorting signal predictors [9,10], homology-based predictors [11,12], and hybrid methods that use several sources of information to predict localization [5,13,14]. Existing predictors have shortcomings, which can include low coverage, a small number of compartments considered, low predictive accuracy, and misannotation of several classes of proteins that include multi-compartmental proteins, proteins at the boundary of two organelles, and transmembrane proteins.
To address some of these issues, we have created a localization predictor for yeast proteins called PSLT2 (Protein Subcellular Localization Tool 2) that integrates the presence of motifs, domains, and targeting sequences with protein–protein interaction data. We have previously shown that the integration of various motif, domain, and targeting signal information into a probabilistic framework allows for accurate prediction of mammalian protein subcellular localization [5]. PSLT2 is based on similar methodology but uses additional data available for yeast proteins. PSLT2 achieves a prediction accuracy of at least 72% and a coverage of 100% of the yeast proteome. A comparison of PSLT2 predictions to experimentally tagged high-throughput datasets reveals that most disagreements occur when predicting the localization of proteins belonging to the secretory pathway. Such annotation errors can be identified by combining different experimental and bioinformatic methods. Additionally, PSLT2 predictions were used to study the sub-compartmental localization of all yeast proteins, allowing for more specific annotations than presently available. In particular, we investigate how to distinguish between lumenal proteins of organelles and proteins that are associated with these organelles on their cytosolic side. Such distinctions are not widely available in public databases. In fact, many proteins annotated as being inside specific organelles are actually cytosolic proteins that are peripherally associated with the organelle through interactions with organellar membrane proteins or lipids. These sub-compartmental predictions should be extremely useful for the biological community, as they represent a new tool to study cellular processes on an organellar scale. These refined predictions can in turn be used to investigate the extent of interaction and communication between organelles and help define organellar functionality.
Results
Components of the PSLT2 Bayesian Network
Different types of protein characteristics can be used in the prediction of subcellular localization, including the presence of a wide variety of motifs, domains, and targeting sequences that can be identified from the primary sequence. Bayesian networks provide an excellent tool for the integration of such diverse information because they can generate probabilistic models that are well suited to capture the combinatorial aspect of the data [15]. The full PSLT2 Bayesian network (illustrated in Figure 1) is composed of three independent modules (each capable of predicting localization on its own) whose predictions are combined using a naïve Bayes net localization predictor. The motif module is trained to predict localization based on the combinatorial presence of InterPro motifs [16] in proteins. The targeting module uses targeting signals such as signal peptides/anchors as predicted by SignalP [9], mitochondrial targeting peptides predicted by TargetP [10], glycosyl-phosphatidylinositol (GPI) anchors predicted by DGPI (http://129.194.185.165/dgpi/index_en.html), and the presence of transmembrane domains predicted by TMHMM [17]. Finally, the interaction module is trained on protein–protein interaction data from the CORE set of the Database of Interacting Proteins [18]. The modules are described in greater detail in the Materials and Methods section.
Figure 1 Structure of the PSLT2 Bayesian Network
The PSLT2 predictor is composed of three independent modules that can predict localization individually or in combination: the motif, targeting, and interaction modules. Each module can be characterized by the protein information used as input and the localization probabilities (for all compartments [C]) that are generated as output. The motif module accepts combinations of InterPro motifs (M) as input. The targeting module considers the presence of mitochondrial targeting signals (Mi), signal peptides/anchors (Si; S, signal peptide; A, signal anchor; Q, neither), GPI anchors (G), and the number of transmembrane domains (Tm) to predict localization. The interaction module considers the three compartments to which are localized the largest number of interactions partners (see Materials and Methods for more details). The full network (illustrated as the localization module) takes into account the output of all three modules to predict the probability of localization to all compartments (C).
Information used within the targeting module is available for all yeast proteins. However, not all types of information are available for all proteins in the motif and interaction modules. We term a protein as uninformative if it lacks both motif and interaction information. Otherwise, we say it is informative (see Materials and Methods). 83% of yeast proteins are informative.
Statistical Tests of Accuracy
All possible combinations of the three modules can be used to predict localization. To evaluate the contribution of each module, we tested each combination of modules using a 10-fold cross-validation approach. As shown in Table 1, as more information is used by the predictor, its accuracy and coverage increase, suggesting that the different modules provide some complementary information. On one hand, the interaction module alone provides a high prediction accuracy (the highest achieved for a single module) but low coverage. This can be explained by the fact that protein–protein interaction information is a good indicator of localization (proteins must be in close proximity to interact) but protein interaction datasets have high false-negative rates [19]. On the other hand, the targeting module allows 100% coverage (as all proteins can be scanned for the presence of the targeting motifs) but achieves lower predictive accuracy (as these motifs are less informative because they alone are insufficient to distinguish between all compartments). Because the full network achieves the highest accuracy and coverage, PSLT2 is constructed using all three modules.
Table 1 Ten-Fold Cross-Validation Test Accuracy and Coverage for Every Module Combination
Table 2 shows the positive predictive values and sensitivity for all compartments considered, calculated using a 10-fold cross-validation test. The results are shown for all proteins and for only informative proteins. The numbers in parentheses show the results of the second-best test, in which the predictor is allowed to predict the two most likely compartments and a prediction is considered a success if at least one of the two predictions is correct. The second-best test is biologically relevant since many multi-compartmental proteins are not annotated as such in current public databases. The overall prediction accuracy is 72% for all yeast proteins, 76% for the 83% of yeast proteins that are informative, and the second-best test accuracy is above 85% for all yeast proteins. In general, proteins in the nucleus, mitochondria, and secretory pathway organelles are better predicted than proteins located elsewhere in the cell. This is particularly interesting in the case of organelles of the secretory pathway, since available predictors generally either group these organelles into one multi-organelle compartment (thus providing a very unrefined prediction), or they simply achieve low prediction accuracy for these organelles. Predictions for all yeast proteins are available in Table S1.
Table 2 Ten-Fold Cross-Validation Test Results of the Full PSLT2 Localization Predictor
Proteome-Wide Multi-Compartmental Prediction
PSLT2 can be used to predict the localization of all proteins in the cell. Because it generates localization likelihoods for all compartments, it can also be used to evaluate whether a protein is present in more than one organelle. To do so, proteins for which the localization likelihood of the second highest scoring compartment is above a certain threshold are predicted to be present in both high-scoring compartments. Our previous investigations [5] of this threshold indicate optimal prediction accuracy when the likelihood of the second highest scoring compartment is greater than half of the likelihood of the highest scoring compartment.
Table 3 shows the distribution of proteins in the different compartment pairs. Proteins on the diagonal of Table 3 are predicted to be in only one compartment. Four general classes of multi-compartmental predictions can be identified.
Table 3 Number of Yeast Proteins Predicted in Every Compartment Pair
The first class contains proteins that shuttle between compartments. A large set of known examples of this class are in the nuclear–cytosolic protein group, many of which are well predicted by PSLT2, including kinases, transcription factors, and proteins that bind RNA (for example, HXK2, CAD1, RIO2, GLE1, PBP1). Another such group of proteins shuttle between the endoplasmic reticulum (ER) and the Golgi, including proteins involved in cargo transport between these two organelles (for example, SEC31, SEC24, ERV14).
The second class of multi-compartmental proteins involves proteins localized at the boundary of organelles. This class includes cytosolic proteins that are associated with the ER (for example, many proteasomal proteins, CDC48, and an ER-associated glutathione GTT1), cytosolic proteins associated with the Golgi apparatus (for example, SEC14, IMH1), as well as membrane proteins of certain organelles that can interact with components of the cytosol (DPM1 in the ER membrane) and nuclear pore complex proteins predicted to be nuclear and cytosolic (for example, NUP84, NUP157).
The third class of predicted compartment pairs involves intermediate compartments through which some proteins transit before reaching their final destination (for example, the ER-plasma membrane and ER-secreted groups including PDR12, DAL5, and some soluble cell wall (SCW) proteins.
The last class of predicted multi-compartmental proteins involves compartment pairs that are less likely to share proteins. The largest sets of such compartment pairs predicted by PSLT2 are the mitochondria–ER group and the mitochondria–plasma membrane group. It is interesting to note that the ER and mitochondria share several proteins in mammalian cells, including components of the apoptosis machinery such as Bcl2 and Bak [20]. Further studies will be necessary to determine the biological significance of the large number of proteins predicted to be localized to both the ER and mitochondrion in yeast, including a possible role in calcium signalling.
Comparison with High-Throughput Tagged Datasets
Several large-scale high-throughput protein localization experiments have been conducted recently [1–3,21]. The datasets from these screens have substantially increased the number of proteins of known localization. However, since such efforts involve tagging proteins in order to allow their visualization by microscopy, these datasets likely contain a non-negligible number of incorrect localization annotations. For example, N-terminal tagging of proteins has the potential to disrupt signal peptides, and mitochondrial targeting peptides and C-terminal protein tags can mask motifs such as the HDEL and SKL signals used, respectively, for retention and targeting to the ER and peroxisome. Internal tags, although possibly less disruptive for many localization targeting signals, could destabilize the protein, interfere with proper folding, and, in some cases, cause mislocalization.
In order to assess how well PSLT2 predictions agree with localization annotations generated by high-throughput studies and to determine where the disagreements occur, we chose to use the publicly available protein localization data from TRIPLES (http://ygac.med.yale.edu/triples/default.htm) and YeastGFP (http://yeastgfp.ucsf.edu/) due to the high coverage afforded by these datasets. The YeastGFP dataset consists of 4,156 yeast proteins whose localization was established by visualization of chromosomally tagged C-terminal GFP fusion proteins [2]. The TRIPLES dataset was derived by immunolocalization of 2,744 randomly or C-terminally tagged yeast proteins [1].
The results of the comparison are shown in Figure 2. Not all compartments were used in the experimental high-throughput localization studies, so we grouped the nine compartments predicted by PSLT2 into five mega-compartments: secretory pathway (abbreviated SecPath, it encompasses ER and Golgi proteins), cytosol (Cyt), nucleus (Nuc), mitochondria/peroxisome (Mit), and plasma membrane and periphery (PM & Periphery, it contains vacuolar proteins, plasma membrane proteins, and secreted proteins). It should be noted that the TRIPLES and YeastGFP datasets use the “cytoplasmic” annotation (which is defined by the Gene Ontology as “all of the contents of a cell excluding the plasma membrane and nucleus, but including other subcellular structures”), whereas PSLT2 uses “cytosolic”, which provides much greater specificity as no organellar proteins are part of this group. As a consequence, proteins in the ER, Golgi, vacuole, mitochondrion, and peroxisome can be classified as cytoplasmic (and some are, especially in the case of some organelles), resulting in confusing and very vague annotation. This is a widespread problem that affects large databases as well as most previous localization predictors whose training sets use cytoplasmic annotations to mean cytosolic localization.
Figure 2 Comparison between PSLT2, TRIPLES, and YeastGFP Datasets
Panels A through C represent an illustration of the Pearson correlation for the probability of localization between all compartment pairs for each pair of datasets (see Materials and Methods for details). SecPath, secretory pathway (ER and Golgi); Cyt, cytosol; Nuc, nucleus; Mit, mitochondrial or peroxisomal; PM & P, plasma membrane and periphery (including secreted and vacuolar proteins).
To quantify the degree of similarity between the three datasets, we calculated the Pearson correlation coefficients for localization probability for all pairs of compartments and datasets, over all proteins considered. If the datasets strongly agree, we would see very high Pearson correlation between same compartment pairs, and low Pearson correlation between different compartment pairs (and thus very dark diagonal and very light off-diagonal squares in Figure 2). As shown in Figure 2, there is a much higher Pearson correlation, and thus general better agreement, between all datasets for all same compartment pairs than for different compartment pairs, except in the case of the plasma membrane and periphery group. In general, there is low Pearson correlation between the mitochondria–cytosol and mitochondria–nucleus groups between all datasets, indicating that disagreements do not occur between these compartment pairs. However, in general, much higher Pearson correlation coefficient values are observed involving the secretory pathway, and plasma membrane and periphery groups with other compartments. This indicates that disagreements between datasets involve mostly proteins annotated as being in the secretory pathway, or the plasma membrane and periphery mega-compartments in at least one of the datasets.
We examined general classes of proteins that contain many members known to transit through the secretory pathway or localize in its compartments as well as other groups of proteins whose localization is not agreed upon by the three methods. Table 4 summarizes this analysis and helps explain the discrepancies. In general, in all three datasets, many proteins containing signal peptides as predicted by SignalP [9] are present in the secretory pathway. The situation is similar in the case of the plasma membrane and periphery group for the YeastGFP annotations and PSLT2 predictions, but the TRIPLES dataset annotates more signal-peptide containing proteins as being cytoplasmic rather than in the plasma membrane and periphery group. The high numbers of proteins containing a signal peptide and annotated as being in the mitochondria group is probably in part due to failure of SignalP to discriminate between signal peptides and mitochondrial targeting peptides.
Table 4 Comparison of Localization Annotations between PSLT2, YeastGFP, and TRIPLES Datasets for Different Classes of Proteinsa
Only five proteins containing the C-terminal ER retention motif HDEL are present in the three datasets. PSLT2 and the TRIPLES dataset annotated four of them as being in the secretory pathway group. The YeastGFP dataset annotates four of them as being in the plasma membrane and periphery group. This could be due to the C-terminal GFP tag that might be masking the HDEL signal.
Nine of the 56 tail-anchored proteins known to exist in yeast cells [22] are present in the three datasets. All nine of these proteins are predicted by PSLT2 to be in the secretory pathway. The YeastGFP and TRIPLES datasets annotate four of the known tail-anchored proteins as being in the secretory pathway.
Most of the multi-spanning membrane proteins are annotated by the YeastGFP and PSLT2 datasets as being either in the secretory pathway or in the periphery of the cell membrane, whereas a large number of these proteins are annotated by the TRIPLES dataset as being cytoplasmic.
We also looked at the distribution of C-terminal SKL motif–containing proteins, which are expected to localize to the peroxisome. There are four such proteins in the dataset. Two are predicted as being peroxisomal by PSLT2 (one of which is also predicted to be mitochondrial). None are annotated as peroxisomal in the YeastGFP and TRIPLES datasets.
Nine proteins in the dataset contain a C-terminal CaaX-box motif (which is a targeting signal that specifies prenylation) as defined in PROSITE (motif ID: PS00294). These motifs are believed to be used to anchor proteins in diverse cellular membranes including the plasma membrane [23]. Six of these proteins are predicted to be in the plasma membrane and periphery group by PSLT2. The YeastGFP and TRIPLES datasets annotate these proteins as being elsewhere in the cell.
Comparison with Previous Methods
To accurately and fairly compare predictors, one should ideally use an independent set that does not contain any sequences used to train the predictors. Unfortunately, such a dataset does not exist for yeast proteins, as most predictors use all the UniProt/SwissProt [24] annotations available. Annotations resulting from high-throughput datasets cannot presently be considered as gold standards as they suffer from some biases caused by the experimental procedure, as discussed in the previous section. Related fungi species could be used in an independent test on PSLT2, but many previous methods use these proteins in their training sets. Furthermore, interaction datasets for fungal species other than Saccharomyces cerevisiae contain very few interactions. As a consequence, in Table 5, we compare PSLT2 features and accuracy to those reported for previous methods, instead of performing an independent test. We considered publicly available predictors for which the prediction accuracy is reported for all compartments. These predictors include SubLoc [8] and PLOC [25], both of which are based on support vector machines and consider amino acid composition (SubLoc and PLOC), amino acid pair, and gapped amino acid pair compositions (PLOC) in eukaryotic proteins. We also consider the non-plant eukaryotic version of LOCtree [26] and the Proteome Analyst predictor trained on fungal sequences [12]. LOCtree is a hierarchical system that combines support vector machines and other prediction methods based on homology and keywords. The Proteome Analyst predictor uses homology information and SwissProt annotations to predict localization. As shown in Table 5, not all predictors consider transmembrane proteins, and PSLT2 is the only predictor to attempt to predict multi-compartmental proteins in a systematic way. The predictors vary in the number of compartments considered and the accuracy achieved for each compartment. PSLT2 is particularly successful for proteins localized to secretory pathway organelles and mitochondria, which are notoriously difficult to predict. In general, these five predictors achieve a similar overall prediction accuracy and coverage. However, care must be taken when choosing a predictor, as not all predictors consider all compartments and most do not predict all proteins (note that PSLT2 can predict localization for all proteins but achieves a slightly lower accuracy in this case; see Table 2). In general, predictors that consider homology achieve high prediction accuracy for proteins with close homologues but much lower accuracy for proteins that do not have close homologues. Methods such as PSLT2 are complementary to the homology-based methods and allow accurate predictions in the absence of well-annotated homologues.
Table 5 Comparison with Previous Methods
Sub-Compartmental Prediction
In searching through several of the large protein databases, we have observed that many soluble proteins annotated as being inside the ER are actually peripherally associated with the ER on the cytosolic side [27]. This is a widespread problem that also affects other organelles. Such misleading annotations likely result from the difficulty of distinguishing experimentally between these precise localizations. Methods to address this problem are urgently needed because of the large number of proteins it affects and the numerous research groups that rely on these annotations.
As illustrated in a previous section, the multi-compartmental prediction capacity of PSLT2 allows us to identify some groups of proteins that localize to the boundary of organelles. This is an interesting feature that can be exploited to refine protein subcellular localization predictions and, in particular, to distinguish between lumenal organellar proteins and cytosolic proteins peripherally associated with the organelle. We decided to use the cell-wide PSLT2 predictions in an attempt to annotate with greater detail the yeast proteome. We classify yeast proteins into 18 sub-compartments using PSLT2 localization predictions and targeting motif information (which is necessary because not all PSLT2 predictions allow us to distinguish between different sub-compartments; see Materials and Methods). The classification scheme used (shown in Figure 3) is a decision tree that was generated using the C4.5 software [28] and a manually curated training set consisting of 1,167 yeast proteins. We chose to use decision trees to perform such an analysis because of the nature of the data (the data are complete, and the input variables likely highly influence the output variables). In such situations, decision trees can encode the relationships between the variables with few parameters, which can result in more accurate classification [15]. Furthermore, decision trees provide a classification scheme from which it is easy to identify the rules used for the prediction. Decision trees have already been used to predict subcellular localization [29].
Figure 3 Sub-Compartmental Prediction Scheme
Proteins are predicted to be localized in one specific sub-compartment by first considering the most likely PSLT2 compartment (blue boxes). Further decisions depend on the PSLT2 second most likely compartment prediction (orange boxes) and targeting information (green boxes). Once all information has been analyzed, the protein is predicted to be in one of 18 sub-compartments (pink boxes). When the second most likely sub-compartments are considered, the default prediction is shown with a star (this branch of the tree is used in particular when proteins have no second most likely compartment as predicted by PSLT2). Pero, peroxisome; Vac, vacuole; Cyt, cytosolic; memb, membrane; TMD, number of transmembrane domains in protein; ER, endoplasmic reticulum; GPI, presence of GPI anchor; Nuc, nuclear; PM, plasma membrane; Mito, mitochondria; S, signal peptide; A, signal anchor; Q, neither signal peptide nor signal anchor.
The training set was constructed using high-quality annotations in UniProt [24] and information in the literature. Full manual curation was necessary because few proteins are annotated with such detailed information. As shown in Figure 3, to predict sub-compartmental localization, the learnt decision tree first considers the most likely compartment as predicted by PSLT2. When this information is not sufficient to predict sub-compartmental localization, the decision tree next considers either the second most likely compartment as predicted by PSLT2 (if such a prediction exists), or other protein features, such as the number of transmembrane domains or the presence of GPI anchors. In some rare cases, further information is required, such as the presence of signal peptides. We originally also used mitochondrial targeting peptide predictions as an attribute considered by the decision tree software, but this resulted in a larger tree with no gain in accuracy. We chose to model the extracellular group of yeast proteins as consisting of both secreted and cell surface proteins. In the original PSLT2 training set, several proteins annotated as secreted are in fact soluble proteins anchored on the cell surface (yeast has few truly secreted proteins that do not remain in the periphery of the plasma membrane).
The accuracy of the predictor is measured to be 83% using a 10-fold cross-validation test. We used this decision tree predictor to annotate all yeast proteins with sub-compartmental information. The sub-compartmental training set and predictions are available in Table S2. Our new predictions have substantially increased the number of yeast proteins annotated in most of these sub-compartments (as shown in Table S3). For example, five proteins were previously annotated as being in the ER periphery. Our decision tree predictor brings this number up to 152. Such sub-compartmental predictions should be useful to the biology research community, especially in the case of proteins with no previous sub-compartmental annotations.
We further assessed the quality of our sub-compartmental predictions using a specific example that involves many proteins known to localize to sub-compartments: the UPR (unfolded protein response) and ERAD (ER-associated degradation) pathways related to quality control in the ER. Using a literature search, we determined the sub-compartmental localization of 30 proteins involved in these pathways. Our sub-compartmental predictions agree remarkably well with the experimentally determined localizations of these proteins (see Table S4).
Localizome–Interactome Maps of the Secretory Pathway
As mentioned previously, PSLT2 achieves a good prediction accuracy for proteins in the secretory pathway, which is often not the case for most available localization predictors. In general, proteins in organelles of the secretory pathway and the plasma membrane are less well represented in large datasets generated by high-throughput strategies, probably because they are not amenable to some of the high-throughput experimental methods such as yeast-two hybrid and mass spectrometry. Many are membrane or membrane associated or require chemical environments that are different from those in which most proteins function—for example, the lumen of the ER. Our sub-compartment localization predictions provide a very useful new tool to investigate these proteins.
In Figure 4, we present the first (to our knowledge) such model of interaction–localization networks focused on the secretory pathway. To generate these networks, we use only the CORE protein–protein interactions available from the Database of Interacting Proteins [18], since these interactions are considered highly validated. The localization–interaction maps are generated by selecting all protein pairs annotated as being either in a secretory pathway sub-compartment or cytosolic and that are involved in protein–protein interactions annotated in the CORE dataset of the DIP. Figure 4B shows the localizome–interactome of the full secretory pathway. Visual inspection of this map suggests that large groups of co-localized proteins interact together and that protein sub-compartment annotations and protein–protein interaction data are generally consistent. For a large group of ER periphery proteins (lower left group, Figure 4B), interactions are almost exclusively between members of this group or with cytosolic proteins. Many peripheral Golgi and membrane proteins interact amongst themselves and with some cytosolic, peripheral ER, ER membrane, and vacuolar proteins (large middle group). And several cytosolic proteins interact with plasma membrane periphery proteins, plasma membrane integral membrane proteins, and vacuolar proteins (lower middle group). In general, interactions involve proteins that are localized in the same or adjacent sub-compartments, and there is extensive cross-talk between cytosolic proteins and proteins at the periphery of organelles.
Figure 4 Localizome–Interactomes
The protein–protein interaction maps for all proteins in the secretory pathway (B) or all proteins in the ER (C). Proteins are depicted as circles coloured according to their predicted sub-compartmental localization, as specified in the legend in (A). Interactions are shown as lines between proteins.
Figure 4C shows the localizome–interactome of all proteins predicted to be in one of the three ER sub-compartments and their interactors (according to DIP). Once again, a large group of ER periphery proteins interact amongst themselves, and with cytosolic proteins. Several interactions involve Golgi periphery and ER periphery or membrane proteins. Literature searches indicate that many of these Golgi periphery proteins are known to localize to COPII vesicles that traffic between the ER and the Golgi, and could thus interact with proteins in the periphery and the membrane of the ER. The lower left-most area in Figure 4C also shows some ER membrane proteins interacting with ER lumenal and cytosolic proteins. These proteins are involved in protein translocation across the ER membrane, vesicle formation between the ER and Golgi, N-linked glycosylation, phospholipid metabolism, and oxidative folding, all known to be important functions of the ER. Dispersed throughout the network, a few ER membrane and lumenal proteins interact with cargo proteins that likely transit through the ER, such as plasma membrane, vacuole membrane, and extracellular proteins.
Discussion
We introduce here PSLT2, a subcellular localization predictor for yeast proteins that achieves high prediction accuracies for most compartments considered, including secretory pathway organelles. PSLT2 can identify proteins that are potentially present in more than one compartment and has been used to predict the localization of all yeast proteins. PSLT2 combines a methodology initiated for PSLT [5] (mostly in the motif module) with additional information that is widely available for yeast. This is mainly interaction data and the presence of various targeting signals. Integration results in a more complete model to predict localization. As large-scale protein–protein interaction maps are being generated for a growing number of organisms [30–33], protein interaction data will certainly become an important type of evidence for the prediction of localization. We show here that the integration of these different types of data in an appropriate framework greatly increases the prediction accuracy and coverage.
A comparison between our predictions and annotations from two large high-throughput datasets shows that disagreements between these methods generally occur in proteins that transit through or localize in organelles of the secretory pathway. While it is probable that some of these proteins are mislocalized in the procedures for high-throughput tagging, possibly due to masking or disruption of targeting signals, it is certainly also the case that some PSLT2 predictions are incorrect. Examples of proteins that are poorly predicted by PSLT2 include enzyme families of which two isozymes are known to localize to separate compartments. We are currently investigating other types of additional information that could be used by our predictor to increase its accuracy in such cases. In general, high-throughput experimental methods and bioinformatic predictors should be used in parallel, and disagreements between these two types of methods can be viewed as indicators of cases where further experimental validation using independent methods is required.
Sub-Compartmental Localization
Using PSLT2 predictions and the presence of specific targeting signals, we have investigated how localization annotations can be refined to distinguish between lumenal organellar and organellar periphery annotations. Such refined predictions are of particular interest because they offer a new picture of the protein composition of the surface of organelles, which in turn can help understand mechanisms of communication between different cellular compartments. In addition, these predictions will likely also be of particular use in the validation of high-throughput protein–protein interaction datasets, which are currently often validated using conventional protein localization annotations [34] that do not distinguish between lumenal and peripheral organellar proteins and use the “cytoplasmic” annotation rather the “cytosolic” annotation, which is more specific.
We have used our sub-compartmental predictions to initiate the investigation of the localizome–interactome of the secretory pathway. The models that can be derived from this analysis provide a novel global view of interactions and complexes in these organelles, as well as of cross-talk between these organelles at a level that has never, to our knowledge, been investigated previously.
The large-scale study of protein–protein interactions in the secretory pathway has lagged behind such efforts in other parts of the cell. This will soon change as several methods are currently being developed that can specifically overcome the challenges offered by these organelles [35,36]. Our sub-compartmental localization predictions can be useful in the choice and technical manipulation of protein targets to study using these novel protein–protein interaction discovery platforms. The data generated by these novel strategies, in combination with methods such as ours, will provide a better global understanding of these organelles.
Materials and Methods
Components of the PSLT2 Bayesian network.
The PSLT2 predictor is composed of three independent modules that can predict localization on their own or in combination, as shown in Figure 1.
The motif module predicts localization based on the co-occurrence of InterPro motifs [16] in proteins as previously described [5]. Briefly, the likelihood of localization to all compartments considered is calculated and stored in an XML file for all combinations of motifs present in all proteins in the training set, using a dynamic programming algorithm. During this training phase of the algorithm, several thousands of motif combinations are considered. At the completion of this phase, given an unknown protein and the InterPro motifs it contains, the likelihood of localization to all compartments can be found by looking up this information in the motif-likelihood XML file. It should be noted that the motif module is a Bayesian Network structural learning problem that we solve by looking at all combinations of motifs present in proteins in the training set using dynamic programming. This approach is feasible with the number of proteins and motifs considered here but might become infeasible as more motifs are added to InterPro and more proteins are considered. In such a case, other approaches, including structural expectation-maximization (EM) could be used [15].
The targeting module is a full Bayesian network that considers all combinations of the presence of signal peptides, mitochondrial targeting peptides, GPI anchors, and the number of transmembrane domains in proteins to predict localization. The presence of these signal/domains is predicted respectively by SignalP [9], TargetP [10], DGPI (http://129.194.185.165/dgpi/index_en.html) and TMHMM [17]. The localization likelihood to all compartments given all these targeting signals is evaluated by using Bayes rule:
where S indicates the presence of signal peptides; M, the presence of mitochondrial targeting motifs; T, the number of transmembrane domains; and G, the presence of GPI anchors. The compartment prior Pr[C], the targeting signal prior Pr[S,T,M,G], and the targeting signal posterior Pr[S,T,M,G|C] are each evaluated by counting proteins in the training set. The information used by the targeting module is available for all proteins regardless of whether they have been previously characterized.
The interaction module predicts localization likelihoods to all compartments considered, given the localization of known interacting partners. This module is trained using the CORE DIP dataset [18], which represents the most reliable interactions within the full DIP dataset. For each protein X, we create a vector representing the three compartments that contain the most protein interacting partners for protein X according to the CORE DIP dataset. Then the localization likelihood to all compartments considered can be evaluated for all vectors present in the training set using Bayes rule.
All three modules output the likelihood of localization to all compartments considered and as such can predict localization on their own. When used in combination, their localization likelihoods are combined in a naïve Bayesian network according to:
where C is the cellular compartment, M and T represent the motifs and targeting signals present in the protein, respectively, and I represents the localization vector of the interaction partners of the protein whose localization is being predicted.
To avoid a situation where at least one of the three modules predicts a localization likelihood of zero for a given compartment while the other module(s) predict a non-zero localization likelihood for the compartment (which would result in the full predictor outputting a likelihood of zero for that compartment), we add pseudo-counts to the localization likelihoods of all three modules. These pseudo-counts are very small and reflect the underlying probability of presence in a compartment (here, we use one-tenth of the compartment prior P[C] for the pseudo-counts).
As mentioned previously, predictions for all of the targeting motifs used by the targeting modules are available for all yeast proteins. Such is not the case for the motif and interaction modules. Indeed, not all proteins contain InterPro motifs or interaction partners. We say a protein is uninformative if it lacks both motif and interaction information. Otherwise, we say it is informative. It should be noted that the use of pseudo-counts for all three modules is equivalent to eliminating the module (either motif or interaction) for which no informative information is available.
The compartment priors P[C] are estimated as described previously [5].
PSLT2 datasets.
The PSLT2 Bayesian network is trained using information in UniProt [24] and curated from the literature. UniProt is a curated database of protein sequences that provides a high level of annotation, including subcellular localization information. Localization annotations were kept if they clearly indicated which compartment(s) the protein is localized to. No annotations described as “possible,” “potential,” “probable,” or “by similarity” were kept. The PSLT2 training set contains localization information for 1521 S. cerevisae proteins localized to nine different compartments (several are localized to more than one compartment).
Two high-throughput localization datasets were also used in this study: the YeastGFP dataset [2] and the TRIPLES dataset [1]. The YeastGFP dataset consists of 4,156 yeast proteins whose localization was established by visualization of chromosomally tagged GFP fusion proteins. The TRIPLES dataset was derived by immunolocalization of 2,744 tagged yeast proteins.
Comparison between datasets.
The annotations of three datasets (PSLT2, YeastGFP, and TRIPLES) were compared by calculating the Pearson correlation for the probability of localization between all compartment pairs for each pair of datasets. More specifically, the vector of probabilities of localization to all compartments for all proteins in a dataset is compared to an equivalent vector for all proteins in a second dataset.
Sub-compartment prediction.
We predict sub-compartmental localization using the C4.5 software for the creation of decision trees [28]. The sub-compartmental annotation mainly involves distinguishing between lumenal, membrane, and peripherally associated proteins for most organelles. The attributes considered by the predictor are the PSLT2 predictions, the number of transmembrane domains, and some targeting signals. This new predictor is trained using a fully manually curated dataset that incorporates high-quality annotations from UniProt [24] and information available in the literature. This dataset contains subcompartmental annotation for 1,167 S. cerevisiae proteins (these annotations are available in Table S2, column “Sub-compartmental localization annotation from UniProt or literature”). The manual construction of such a dataset is necessary because few proteins are annotated with such precise localization information. Table S2 also contains all the training information used during the prediction (columns “Signal type,” “Topology,” “Targeting type,” and “GPI anchor”). Proteins that are predicted to contain only one transmembrane domain by TMHMM [17] that overlaps with the signal peptide predicted by SignalP [9] are considered to be soluble.
The predictor is created using the iterative mode of C4.5 during which the software sequentially creates more accurate decision trees by using more training examples until no further improvement is obtained. One hundred different trees were created (almost identical to one another in this case) and the most accurate tree is chosen as the decision tree that provides the classification scheme for our sub-compartmental predictions.
Localizome–interactome maps.
The localizome-interactome maps are generated using Cytoscape [37]. The datasets are created by selecting all protein pairs annotated as being in a secretory pathway sub-compartment or cytosolic and that are involved in protein–protein interactions annotated in the CORE dataset of the Database of Interacting Proteins [18]. The nodes are coloured according to their sub-compartmental predictions.
Supporting Information
Table S1 PSLT2 Predictions
(511 KB XLS)
Click here for additional data file.
Table S2 Sub-Compartmental Localization Dataset
(876 KB XLS)
Click here for additional data file.
Table S3 Number of Proteins Previously Annotated and Newly Predicted in Each Sub-Compartment Considered
(26 KB DOC)
Click here for additional data file.
Table S4 Sub-Compartmental Analysis of the UPR and ERAD Pathway Components
(54 KB PDF)
Click here for additional data file.
We wish to thank François Pepin for logistical support. This work was supported by grants to DYT and MTH from Genome Quebec/Genome Canada, as well as to DYT from the Canadian Institutes of Health Research (CIHR). MSS is a recipient of a Canada Graduate Scholarship (CGS) Doctoral Award from CIHR.
Competing interests. The authors have declared that no competing interests exist.
Author contributions. MSS and MTH conceived and designed the experiments. MSS performed the experiments. MSS, DYT, and MTH analyzed the data. SJC and MTH contributed reagents/materials/analysis tools. MSS wrote the paper.
Abbreviations
ERendoplasmic reticulum
GPIglycosyl-phosphatidylinositol
==== Refs
References
Kumar A Agarwal S Heyman JA Matson S Heidtman M 2002 Subcellular localization of the yeast proteome Genes Dev 16 707 719 11914276
Huh WK Falvo JV Gerke LC Carroll AS Howson RW 2003 Global analysis of protein localization in budding yeast Nature 425 686 691 14562095
Wiemann S Arlt D Huber W Wellenreuther R Schleeger S 2004 From ORFeome to biology: A functional genomics pipeline Genome Res 14 2136 2144 15489336
Davis TN 2004 Protein localization in proteomics Curr Opin Chem Biol 8 49 53 15036156
Scott MS Thomas DY Hallett MT 2004 Predicting subcellular localization via protein motif co-occurrence Genome Res 14 1957 1966 15466294
Dönnes P Höglund A 2004 Predicting protein subcellular localization: Past, present, and future Genomics Proteomics Bioinformatics 2 209 215 15901249
Reinhardt A Hubbard T 1998 Using neural networks for prediction of the subcellular location of proteins Nucleic Acids Res 26 2230 2236 9547285
Hua S Sun Z 2001 Support vector machine approach for protein subcellular localization prediction Bioinformatics 17 721 728 11524373
Bendtsen JD Nielsen H von Heijne G Brunak S 2004 Improved prediction of signal peptides: SignalP 3.0 J Mol Biol 340 783 795 15223320
Emanuelsson O Nielsen H Brunak S von Heijne G 2000 Predicting subcellular localization of proteins based on their N-terminal amino acid sequence J Mol Biol 300 1005 1016 10891285
Marcotte EM Xenarios I van Der Bliek AM Eisenberg D 2000 Localizing proteins in the cell from their phylogenetic profiles Proc Natl Acad Sci U S A 97 12115 12120 11035803
Lu Z Szafron D Greiner R Lu P Wishart DS 2004 Predicting subcellular localization of proteins using machine-learned classifiers Bioinformatics 20 547 556 14990451
Nakai K Kanehisa M 1992 A knowledge base for predicting protein localization sites in eukaryotic cells Genomics 14 897 911 1478671
Drawid A Gerstein M 2000 A Bayesian system integrating expression data with sequence patterns for localizing proteins: Comprehensive application to the yeast genome J Mol Biol 301 1059 1075 10966805
Heckerman D 1998 A tutorial on learning with Bayesian networks Jordan M Learning in graphical models Cambridge (Massachusetts) MIT Press 302 303
Mulder NJ Apweiler R Attwood TK Bairoch A Barrell D 2003 The InterPro Database, 2003 brings increased coverage and new features Nucleic Acids Res 31 315 318 12520011
Krogh A Larsson B von Heijne G Sonnhammer EL 2001 Predicting transmembrane protein topology with a hidden Markov model: Application to complete genomes J Mol Biol 305 567 580 11152613
Salwinski L Miller CS Smith AJ Pettit FK Bowie JU 2004 The Database of Interacting Proteins: 2004 update Nucleic Acids Res 32 D449 D451 14681454
von Mering C Krause R Snel B Cornell M Oliver SG 2002 Comparative assessment of large-scale data sets of protein-protein interactions Nature 417 399 403 12000970
Breckenridge DG Germain M Mathai JP Nguyen M Shore GC 2003 Regulation of apoptosis by endoplasmic reticulum pathways Oncogene 22 8608 8618 14634622
Sutherland HG Mumford GK Newton K Ford LV Farrall R 2001 Large-scale identification of mammalian proteins localized to nuclear sub-compartments Hum Mol Genet 10 1995 2011 11555636
Burri L Lithgow T 2004 A complete set of SNAREs in yeast Traffic 5 45 52 14675424
Hulo N Sigrist CJ Le Saux V Langendijk-Genevaux PS Bordoli L 2004 Recent improvements to the PROSITE database Nucleic Acids Res 32 D134 D137 14681377
Bairoch A Apweiler R Wu CH Barker WC Boeckmann B 2005 The Universal Protein Resource (UniProt) Nucleic Acids Res 33 D154 D159 15608167
Park KJ Kanehisa M 2003 Prediction of protein subcellular locations by support vector machines using compositions of amino acids and amino acid pairs Bioinformatics 19 1656 1663 12967962
Nair R Rost B 2005 Mimicking cellular sorting improves prediction of subcellular localization J Mol Biol 348 85 100 15808855
Scott M Lu G Hallett M Thomas DY 2004 The Hera database and its use in the characterization of endoplasmic reticulum proteins Bioinformatics 20 937 944 14751973
Quinlan J 1993 C4.5: Programs for machine learning San Francisco Morgan Kaufmann 302 p.
Horton P Nakai K 1997 Better prediction of protein cellular localization sites with the k nearest neighbors classifier Proc Int Conf Intell Syst Mol Biol 5 147 152 9322029
Butland G Peregrin-Alvarez JM Li J Yang W Yang X 2005 Interaction network containing conserved and essential protein complexes in Escherichia coli
Nature 433 531 537 15690043
Ito T Chiba T Ozawa R Yoshida M Hattori M 2001 A comprehensive two-hybrid analysis to explore the yeast protein interactome Proc Natl Acad Sci U S A 98 4569 4574 11283351
Uetz P Giot L Cagney G Mansfield TA Judson RS 2000 A comprehensive analysis of protein-protein interactions in Saccharomyces cerevisiae
Nature 403 623 627 10688190
Giot L Bader JS Brouwer C Chaudhuri A Kuang B 2003 A protein interaction map of Drosophila melanogaster
Science 302 1727 1736 14605208
Jansen R Gerstein M 2004 Analyzing protein function on a genomic scale: The importance of gold-standard positives and negatives for network prediction Curr Opin Microbiol 7 535 545 15451510
Thaminy S Miller J Stagljar I 2004 The split-ubiquitin membrane-based yeast two-hybrid system Methods Mol Biol 261 297 312 15064465
Pollock S Kozlov G Pelletier MF Trempe JF Jansen G 2004 Specific interaction of ERp57 and calnexin determined by NMR spectroscopy and an ER two-hybrid system Embo J 23 1020 1029 14988724
Shannon P Markiel A Ozier O Baliga NS Wang JT 2003 Cytoscape: A software environment for integrated models of biomolecular interaction networks Genome Res 13 2498 2504 14597658
|
16322766
|
PMC1289393
|
CC BY
|
2021-01-05 09:18:23
|
no
|
PLoS Comput Biol. 2005 Nov 25; 1(6):e66
|
utf-8
|
PLoS Comput Biol
| 2,005 |
10.1371/journal.pcbi.0010066
|
oa_comm
|
==== Front
PLoS PathogPLoS PathogppatplpaplospathPLoS Pathogens1553-73661553-7374Public Library of Science San Francisco, USA 10.1371/journal.ppat.001002905-PLPA-RA-0125R1plpa-01-03-08Research ArticleBiochemistryMicrobiologyParasitologyPlasmodiumMolecular Identification of a Malaria Merozoite Surface Sheddase PfSUB2 Is a Merozoite Surface SheddaseHarris Philippa K 1Yeoh Sharon 1Dluzewski Anton R 23O'Donnell Rebecca A 1Withers-Martinez Chrislaine 1Hackett Fiona 1Bannister Lawrence H 3Mitchell Graham H 2Blackman Michael J 1*
1 Division of Parasitology, National Institute for Medical Research, London, United Kingdom
2 Department of Immunobiology, Guy's, King's and St. Thomas' Hospitals School of Medicine, London, United Kingdom
3 Wolfson Centre, Guy's, King's and St. Thomas' Hospitals School of Biomedical and Life Sciences, London, United Kingdom
Schneider David Samuel EditorStanford University, United States of America* To whom correspondence should be addressed. E-mail: [email protected] 2005 25 11 2005 1 3 e2910 8 2005 12 10 2005 Copyright: © 2005 Harris et al.2005This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are properly credited.Proteolytic shedding of surface proteins during invasion by apicomplexan parasites is a widespread phenomenon, thought to represent a mechanism by which the parasites disengage adhesin-receptor complexes in order to gain entry into their host cell. Erythrocyte invasion by merozoites of the malaria parasite Plasmodium falciparum requires the shedding of ectodomain components of two essential surface proteins, called MSP1 and AMA1. Both are released by the same merozoite surface “sheddase,” but the molecular identity and mode of action of this protease is unknown. Here we identify it as PfSUB2, an integral membrane subtilisin-like protease (subtilase). We show that PfSUB2 is stored in apical secretory organelles called micronemes. Upon merozoite release it is secreted onto the parasite surface and translocates to its posterior pole in an actin-dependent manner, a trafficking pattern predicted of the sheddase. Subtilase propeptides are usually selective inhibitors of their cognate protease, and the PfSUB2 propeptide is no exception; we show that recombinant PfSUB2 propeptide binds specifically to mature parasite-derived PfSUB2 and is a potent, selective inhibitor of MSP1 and AMA1 shedding, directly establishing PfSUB2 as the sheddase. PfSUB2 is a new potential target for drugs designed to prevent erythrocyte invasion by the malaria parasite.
Synopsis
Malaria causes immense suffering and loss of life across the globe. In the face of growing resistance to available drugs and no licensed vaccine, new approaches are urgently required to tackle its control. Fundamental to these is an improved understanding of the basic biology of the malaria parasite. The parasite invades and replicates within red blood cells. During invasion a number of important proteins need to be shed from the parasite surface, probably in order to disengage the adhesive interactions that enable initial binding. Shedding of these surface proteins is achieved by a parasite enzyme called a protease, and compounds or antibodies that block the action of this protease prevent invasion, killing the parasite. Here the authors identify this protease as PfSUB2, a large, membrane-bound member of the subtilisin-like protease superfamily. They find that PfSUB2 is secreted from apical organelles called micronemes at the point of invasion to migrate rearwards over the surface of the parasite, and that a protein designed to be a specific inhibitor of PfSUB2 potently prevents shedding of parasite surface proteins. This work sets the scene for the development of inhibitors of PfSUB2 as a new generation of antimalarial drugs.
Citation:Harris PK, Yeoh S, Dluzewski AR, O'Donnell RA, Withers-Martinez C, et al. (2005) Molecular identification of a malaria merozoite surface sheddase. PLoS Pathog 1(3): e29.
==== Body
Introduction
Malaria is a devastating global health problem, responsible for up to 3 million deaths annually [1]. The disease results from cyclical replication within erythrocytes of protozoan parasites of the genus Plasmodium. The parasite divides asexually within its host cell to produce a number of progeny merozoites. Upon eventual rupture of the schizont, these are released to rapidly invade fresh red cells and perpetuate the cycle. Like all apicomplexan parasites, the Plasmodium merozoite enters its host cell by an active invasion process that is mediated by adhesive receptor–ligand interactions and driven by an actinomyosin motor [2]. Light and electron microscopic studies have shown that initial attachment to the host erythrocyte is followed by reorientation of the merozoite such that its apical end contacts the cell surface. This results in the formation of an irreversible zone of contact, or tight junction, between the apical prominence and the host cell surface. The host cell membrane then invaginates, forming a parasitophorous vacuole (PV) into which the parasite is propelled; in the process, the junction sweeps around the periphery of the parasite with concomitant “shaving” of bristle-like structures from the parasite surface [3,4], eventually sealing behind the intracellular parasite. The initial low-affinity binding appears to be mediated by a large, glycosylphosphatidyl inositol (GPI)-anchored protein complex which is uniformly distributed around the parasite surface and is composed of fragments of merozoite surface protein-1 (MSP1) plus associated partner proteins [5–7]. Many subsequent interactions in the invasion pathway are mediated by proteins released from micronemes, secretory vesicles at the apical end of the merozoite [8]. One of these proteins, apical membrane antigen-1 (AMA1), is a type I integral membrane protein that is secreted onto the merozoite surface just prior to interaction with the host cell and may play a role in reorientation, junction formation, or government of the release of a second set of apical organelles called rhoptries [9–11]. Both MSP1 and AMA1 play essential roles in the blood-stage cycle of the malaria parasite [12,13].
During invasion both AMA1 and the MSP1 complex are quantitatively shed from the parasite surface, in each case as a result of a single proteolytic cleavage at a juxtamembrane site. Shedding of MSP1 results from cleavage just distal to a tandem EGF (epidermal growth factor)-like domain called MSP119 at its C-terminus [14]. MSP119 remains bound to the parasite surface via its GPI anchor and is the only part of the MSP1 complex to be carried into the host cell. AMA1 is cleaved precisely 29 residues away from the transmembrane domain (TMD), releasing the bulk of the ectodomain and resulting in just the juxtamembrane “stub” being carried into the host cell with its cognate TMD and cytoplasmic domain [15–17]. Shedding of these proteins is required for productive invasion [7,18,19], and may be important to release adhesive interactions between the parasite and host cell surface in order to allow unimpeded passage into the nascent PV [20]. The MSP1 complex is an abundant merozoite component, and together with AMA1 likely corresponds to the surface structures shed adjacent to the moving junction. Importantly, shedding of both AMA1 and the MSP1 complex can occur even in the absence of invasion, and a simple assay based on the use of isolated merozoites has shown that both proteins are shed by the same parasite-derived, membrane-bound, calcium-dependent serine protease, called merozoite surface sheddase, or MESH [16,20]. MESH is active against its physiological substrates only when in the same membrane [21]. Its molecular identification has proved elusive, but the accumulated microscopic and biochemical evidence suggests that at invasion it must distribute across the surface of the parasite, probably localising at the moving junction [7,20]. We and others have previously identified a large, membrane-bound subtilisin-like serine protease called SUB2 that is expressed in the late stages of intraerythrocytic development and accumulates within the apical domain of the merozoite [22,23]. The enzyme has not been expressed in a recombinant, enzymatically active form, but SUB2 is conserved throughout Plasmodium, and attempts to disrupt both the Plasmodium falciparum gene (pfsub2) and that of the rodent malaria P. berghei have been unsuccessful [24,25], indicating an essential function within the blood-stage cycle. Here we present the first experimental evidence that PfSUB2 is MESH, and show that it has the remarkable capacity to translocate across the merozoite surface exactly as predicted.
Results
Epitope Tagging of PfSUB2 by Targeted Homologous Recombination
PfSUB2 is a poorly abundant parasite protein, hindering our earlier attempts at precise sub-cellular localization of the mature protein. To overcome this obstacle, we set out to modify the endogenous pfsub2 gene by fusing it to a sequence encoding three consecutive haemagglutinin (HA) epitope tags using targeted homologous recombination. Parasites transfected with construct pHH1-T996HA3 and the control plasmid pHH1-T996w (Figure 1A) were cultured in the presence of the antifolate WR99210 followed by rounds of drug cycling to remove parasites harbouring free episomes (which are rapidly lost in the absence of drug pressure). The resulting parasite lines, called PfSUB2HA and PfSUB2w, were cloned by limiting dilution and maintained in the presence of WR99210. Southern blot analysis (Figure 1B) showed that all transgenic clones contained more than one copy of the targeting plasmid integrated into the pfsub2 locus, as is common in single-crossover homologous recombination in P. falciparum as a result of the plasmids being maintained as concatamers (e.g., [13]). Note that despite this, only one copy of the modified gene is expected to be functional, because all downstream copies contain either the targeting fragment only or a promoterless, truncated remnant of the endogenous locus lacking the epitope tag. In over 6 mo of continuous culture, none of the clones exhibited any growth defect compared to the parental D10 clone (not shown), indicating that neither the addition of the HA tags to the C-terminus of PfSUB2 nor replacement of the pfsub2 3′ UTR with the heterologous P. berghei sequence was detrimental to parasite replication.
Figure 1 Epitope Tagging of PfSUB2
(A) Schematic depiction of integration plasmids pHH1-T996HA3 and pHH1-T996w, and single-crossover homologous recombination events (for clarity shown only for pHH1-T996HA3). Integration reconstitutes a full-length pfsub2 gene either with or without a triple HA tag at the 3′ end. Correct transcription termination and polyadenylation of this gene is regulated by the presence of the P. berghei dihydrofolate reductase 3′ UTR (PbDT 3′ UTR). hdhfr, human dihydrofolate reductase, conferring resistance to the antifolate WR99210. S, indicates the ScaI sites. P indicates the position and size of the probe used for Southern analysis.
(B) Southern blots of ScaI-digested input plasmid pHH1-T996HA3 or genomic DNA from PfSUB2w clones F7 and D8, PfSUB2HA clones 10E and 2D, and wild-type D10 parasites. The position of the genomic 3.4-kb fragment derived from the wild-type (wt) pfsub2 locus is arrowed. Its disappearance from all the transgenic clones indicates disruption of this locus, whilst the appearance of the additional species at 5.8 kb and 4.4 kb indicates integration as predicted. The fragments seen at the position of full-length input plasmid derive from integration of more than one copy of the entire plasmid, as described in the text.
(C) Western blot of transgenic clones using anti-HA mAb 3F10 (identical results were obtained using anti-HA mAb 12CA5; not shown) or anti-RAP2 mAb H5, used as a loading control. Identities of the authentic PfSUB2 species are indicated, and the 55-kDa PfSUB2 degradation product (the abundance of which varied widely between experiments, being most prominent in non-ionic detergent extracts incubated for prolonged periods) is marked with an asterisk. Note that the predicted molecular mass of an HA-tagged minimal PfSUB2 catalytic domain (extending from the catalytic Asp755 to the end of the C-terminal tag) is approximately 70 kDa. Molecular masses are in kDa.
Analysis of the transgenic clones by Western blot using two different HA-specific monoclonal antibodies (mAbs), 12CA5 and 3F10, confirmed the in-frame fusion of the epitope tag to PfSUB2. Pulse-chase studies have shown that PfSUB2 maturation involves two post-translational processing steps in which the approximately 150-kDa primary translation product (called PfSUB2P) is converted first to a 75-kDa intermediate form (PfSUB2I) and finally to a 72-kDa terminal intracellular form (PfSUB2T) [23,25]. All were detected in schizont extracts probed with the anti-HA mAbs (Figure 1C; the closely spaced PfSUB2I/PfSUB2T doublet is not easily discernible here, but see below), consistent with our earlier evidence that maturation of PfSUB2 involves truncation from the N-terminus of the protein. An additional anti-HA reactive species of 55 kDa, also often evident on Western blots, has previously been observed as a minor component in pulse-chase experiments, and is likely a non-specific degradation product of PfSUB2.
PfSUB2 Is a Microneme Protein
To establish the sub-cellular location of PfSUB2, parasites were then analysed by indirect immunofluorescence (IFA) using both of the anti-HA mAbs. Neither mAb showed any reactivity with any stages of the PfSUB2w line or clones thereof, or with the parental D10 clone (not shown). In contrast, both mAbs produced a strong punctate pattern in mature schizonts of both the uncloned PfSUB2HA line (not shown) and its clones. To examine this localisation in detail, dual labelling was performed using mAbs specific for a plasma membrane marker (MSP1), two different rhoptry markers (RhopH2 and RAP2; data not shown for RAP2), and a microneme protein (AMA1). The results were clear (Figure 2A); the anti-HA signal was quite different from that of MSP1, it was adjacent to but distinct from that of the rhoptry-specific signals, but it co-localised completely with that of AMA1, indicating that in mature schizonts, PfSUB2 accumulates in micronemes.
Figure 2 PfSUB2 Is a Microneme Protein
(A) IFA images of schizonts of PfSUB2HA clone 2D dual-labelled with mAbs X509 (anti-MSP1), 61.3 (anti-RhopH2), or 4G2 (anti-AMA1), plus in each case mAb 3F10 (anti-HA) The anti-HA signal co-localised only with the anti-AMA1 signal. Identical results were obtained with the uncloned transgenic PfSUB2HA line, and/or when anti-RAP2 mAb H5 was used as the rhoptry marker instead of mAb 61.3 (not shown). Parasite nuclei are stained throughout with DAPI (blue). Scale bar represents 2 μm.
(B and C) Electron micrographs showing immuno-gold labelling of micronemes within late-stage schizonts of PfSUB2HA clone 2D using: (B) anti-HA mAb 3F10, detecting epitope-tagged PfSUB2; and (C) a polyclonal antibody specific for PfAMA1. The inset in (B) shows another example of micronemal staining with mAb 3F10 from another schizont. Rh, rhoptry.
(D and E) Posterior labelling of with anti-HA mAb 3F10 in a free merozoite of PfSUB2HA clone 2D. Arrows indicate immuno-gold labelling. Rh, rhoptry.
These findings were confirmed by immuno-electron microscopy (IEM) using the anti-HA mAb 3F10, which showed light but consistent labelling of typical elongated micronemes within late-stage schizonts (Figure 2B), similar to that of AMA1 (Figure 2C). No labelling with mAb 3F10 was observed in less-mature schizont stages lacking micronemes, or apically in free merozoites, and labelling was also absent from wild-type schizont controls (not shown). There was no evidence of any dense granule labelling with the anti-HA antibody.
PfSUB2 Is Secreted to Translocate across the Surface of Free Merozoites
To follow the fate of PfSUB2 upon merozoite release, naturally released merozoites were next examined by IFA as described above. As with intracellular parasites in mature schizonts, all merozoites exhibited an intense, punctate anti-HA signal. Unexpectedly, however, in the great majority of cases (≥70% of free merozoites), it comprised a single tight focus of fluorescence that clearly localised to the opposite pole of the merozoite to that recognised by the anti-rhoptry mAbs. This posterior localisation was also found in the free merozoite by IEM, labelling being present just interior to the cell surface, consistent with antibodies binding to the epitope tag at the extreme C-terminus of the PfSUB2 cytoplasmic domain (Figure 2D and 2E; Figure 3A, top row). There are no known organelles situated at the posterior of the merozoite, so this suggested that, just after schizont rupture, PfSUB2 is secreted from the micronemes onto the merozoite surface and redistributed rearwards to accumulate at its posterior end. In a portion of cases (~10%), the anti-HA IFA signal presented as two separate foci that lay lateral to the anterior–posterior axis of the merozoite, suggesting the presence of a mid-bodied ring of fluorescence encircling the parasite at right angles to this axis and perhaps representing PfSUB2 en route to the posterior (Figure 3A, middle row). Only in less than 20% of cases did the anti-HA IFA signal remain adjacent to the rhoptries as reproducibly observed in intact schizonts. In contrast, the anti-AMA1 IFA signal in free merozoites took the form of a uniform circumferential pattern (Figure 3A, bottom row) as observed previously (e.g., [16]). In no case did the anti-AMA1 IFA signal concentrate at the posterior pole of the parasite as seen for the anti-HA signal. Several apicomplexan microneme proteins translocate onto the parasite surface upon release from the host cell. The best-studied example, the Toxoplasma adhesin TgMIC2/M2AP, is secreted spontaneously at low basal levels and capped to the posterior pole of the free tachyzoite. As the parasite invades a host cell, secretion is markedly up-regulated and driven rapidly to completion [26]. Our observations suggest that PfSUB2 behaves in a somewhat similar manner, although complete translocation of PfSUB2 can apparently occur even in the absence of host cell invasion. These findings document the first demonstration of capping of a Plasmodium merozoite surface protein. Importantly, the trafficking pattern observed indicates that PfSUB2 has the capacity to track along the merozoite surface, a characteristic predicted of a membrane-bound protease involved in shedding of merozoite surface proteins.
Figure 3 PfSUB2 Redistributes to the Free Merozoite Posterior in an Actin-Dependent Manner
(A) IFA images of free merozoites of PfSUB2HA clone 2D dual-labelled with mAb 61.3 (anti-RhopH2) and either mAb 3F10 (anti-HA) or 4G2 (anti-AMA1). The anti-HA signal formed either a single punctate signal at the extreme posterior of the merozoite (top), or twin foci lateral to its anterior–posterior axis (middle). Identical results were obtained using PfSUB2HA clone 10E, or using anti-RAP2 mAb H5 as the rhoptry marker (not shown).
(B) Latrunculin inhibits rearward translocation of PfSUB2. Merozoites of PfSUB2HA clone 2D released into medium containing 1% (v/v) DMSO only (solvent control), or 5 μM latrunculin A (Latrunc), or 4 μM cytochalasin D (Cyt D), probed as above with mAb 61.3 (anti-RhopH2) and mAb 3F10 (anti-HA). Note that neither compound had any effect on the efficiency of schizont rupture and merozoite release, nor on AMA1 relocalisation, but both blocked erythrocyte invasion by more than 95% at the concentrations used (not shown).
(C) PfSUB2 remains at the plasma membrane of the newly invaded parasite. Ring stages of PfSUB2HA clone 2D (≤2 h post-invasion) probed with anti-MSP1 mAb 1E1 and anti-HA mAb 3F10. Parasite nuclei are stained throughout with DAPI (blue). Scale bar represents 2 μm.
Translocation of PfSUB2 Is Actin Dependent
The unique form of substrate-dependent gliding motility exhibited by many Apicomplexa is mediated by interactions between the cytoplasmic domains of transmembrane adhesins, such as TgMIC2/M2AP or its Plasmodium sporozoite homologue, TRAP, and a subplasmalemmal actinomyosin motor [2]. Compounds that interfere with this motor block motility and invasion, and prevent capping of several micronemal adhesins, though not their secretion per se (e.g., [27,28]). To test the possibility that this or a similar actin-dependent system might drive PfSUB2 translocation, we investigated the effects of latrunculin A and cytochalasin D, two inhibitors of actin polymerisation with distinct mechanisms of action. As shown in Figure 3B, 5 μM latrunculin A clearly inhibited posterior accumulation of PfSUB2 on released merozoites, with approximately 90% of free parasites exhibiting instead a diffuse distribution of the anti-HA signal at or around the apical region of the merozoite. This result indicates that rearward translocation of PfSUB2 is actin dependent. In contrast, cytochalasin D had no effect on PfSUB2 capping, even at concentrations as high as 4 μM (Figure 3B). This was surprising given the known inhibitory effect of cytochalasins on invasion by the malaria merozoite [29], but identical results were obtained in three independent experiments using two different batches of the drug.
PfSUB2 Remains on the Plasma Membrane of Invading Parasites
Probing newly invaded ring-stage parasites from highly synchronised cultures of the PfSUB2HA clones with mAb 3F10 produced a “ring-shaped” signal that co-localised with that of a mAb specific for MSP119, the fragment of MSP1 that remains on the parasite surface after invasion (Figure 3C). By 4–6 h post invasion, the anti-HA IFA signal was no longer detectable, only to appear again in a punctate form in the late stages of schizont maturation (not shown). PfSUB2 expression peaks in the latter stages of intraerythrocytic maturation [22,30,31], so the signal observed in early rings must derive from PfSUB2 carried into the cell on the merozoite surface. Collectively, the developmental profile of PfSUB2 expression and sub-cellular trafficking, its ability to mobilise onto and across the merozoite surface, and its presence at the plasma membrane of newly invaded ring stages are entirely consistent with an involvement in MESH activity.
Recombinant PfSUB2 Propeptide Is Folded and Binds Specifically to Mature Parasite-Derived PfSUB2
Subtilases are synthesised as zymogens that comprise minimally a secretory signal peptide, a propeptide, and a catalytic domain. The propeptide acts as an intramolecular chaperone, being essential for folding of the catalytic domain. Subtilase propeptides are also highly potent, selective inhibitors of their cognate proteases (e.g., [32–34]), a property that is probably important in regulating subtilase activation during secretion. Prompted by the finding that PfSUB2 is actively distributed across the free merozoite surface, we investigated whether its propeptide could interfere with MESH activity.
By analogy with processing of PfSUB1 [33] and other subtilases, conversion of PfSUB2P to the 75-kDa PfSUB2I form likely represents autocatalytic cleavage of the propeptide. From the mass of this species we calculated that propeptide cleavage occurs between residues Tyr680 and Lys720 of the PfSUB2 sequence. We therefore expressed in Escherichia coli a series of propeptide constructs of varying size, in each case extending from Asn22 (the predicted N-terminus after signal peptide removal) and terminating at different residues within the above range. Constructs were fused to a short N-terminal peptide containing a His6 tag and an S-tag (Figure 4A), an approach previously used successfully to express and characterise the PfSUB1 propeptide [33]. The protein comprising Asn22-Leu687 (called PfSUB2PD) was expressed at the highest levels, and so further work focused exclusively on this. Preliminary purification trials showed PfSUB2PD to be susceptible to degradation, so to optimise yields, much of the purification was routinely performed in the presence of 8 M urea (Figure 4B). Purified PfSUB2PD was refolded into aqueous buffer and its identity confirmed by Western blot and mass spectrometry (not shown).
Figure 4 Recombinant PfSUB2 Propeptide Is Soluble and Structured
(A) Schematic of the PfSUB2PD construct. Full-length PfSUB2 is represented with the secretory signal peptide and TMD indicated in black and the catalytic and cytoplasmic domains in shades of grey. PfSUB2PD possesses a cleavable 46-residue N-terminal fusion peptide containing a His6 and an S-tag, shown in dark grey.
(B) Coomassie-stained SDS PAGE gel showing stages in the purification of PfSUB2PD. Lane 1, Ni-NTA agarose eluate enriched in PfSUB2PD; lane 2, peak fraction from Superdex 200 gel filtration of previous lane in 8 M urea; lane 3, peak fraction from the final RP-HPLC step; lane 4, purified PfSUB1 propeptide (PfSUB1PD). Lane M contains molecular mass markers, the sizes of which are indicated.
(C) Far-ultraviolet CD spectrum of purified refolded PfSUB2PD. Deconvolution of the spectrum obtained calculated secondary structure values of 24% α-helix, 23% β-sheet, 21% reverse turn, and 32% random coil.
Circular dichroism (CD) analysis revealed that PfSUB2PD possesses substantial secondary structure (Figure 4C). Subtilisin propeptides broadly divide into two groups according to their folding capacity; whereas some are unfolded when expressed on their own (e.g., [35]), others—including the PfSUB1 propeptide—adopt significant secondary structure in isolation [33,34,36]. The CD data indicate that the propeptide of PfSUB2 falls into this second class. Subtilase propeptides vary considerably in size and exhibit little sequence homology across evolution, but structural determination of folded propeptides from bacteria [37–39] and mammals [36] have shown that they share a core structure with a simple αβ fold. The observed αβ content of PfSUB2PD is consistent with such a structure.
To seek direct evidence that PfSUB2PD was capable of binding to its cognate protease, purified PfSUB2PD was incubated with detergent extracts of PfSUB2HA clone 2D then recovered using S-protein agarose beads. Purified PfSUB1 propeptide (PfSUB1PD) was used as a control in these experiments. Figure 5 shows that of the four PfSUB2-derived species present in the detergent extracts, only PfSUB2I and PfSUB2T—those forms predicted to lack the propeptide—were significantly bound by PfSUB2PD. Only traces of all four proteins were bound by PfSUB1PD. Reprobing these Western blots with high-titre polyclonal antibodies specific for AMA1, MSP1, or RAP2, all abundant schizont proteins, revealed no non-specific binding of any of these proteins to PfSUB2PD (not shown). These results demonstrate that PfSUB2PD can form a specific molecular complex with the mature forms of PfSUB2.
Figure 5 PfSUB2PD Binds to Mature Parasite PfSUB2
Sodium deoxycholate extracts of PfSUB2HA clone 2D were incubated with either purified PfSUB1PD or PfSUB2PD, then the propeptides were recovered with S-protein agarose and subjected to SDS PAGE and Western blot using anti-HA mAb 3F10. The PfSUB2I and PfSUB2T species bound by PfSUB2PD are indicated. Molecular mass markers are in kDa.
The PfSUB2 Propeptide Is a Potent and Selective Inhibitor of MESH Activity
Purified PfSUB2PD was then tested for its effect on MESH activity in isolated intact merozoites, taking advantage of an assay previously used extensively to characterise the inhibitor sensitivity of the sheddase. Figure 6A shows that PfSUB2PD potently inhibited shedding of both AMA1 and MSP1 in a dose-dependent manner, with a calculated IC50 (inhibitory concentration 50%) of approximately 300 nM, whilst PfSUB1PD had no detectable effect. Identical levels of inhibition were obtained with slightly less-pure preparations of PfSUB2PD, similar to those in lane 2 in Figure 4B, that had been purified in the absence of urea and had not been subjected to reversed phase–high performance liquid chromatography (RP-HPLC) (not shown).
Figure 6 PfSUB2PD Inhibits MESH in a Selective, Dose-Dependent Manner
(A) Shedding of AMA1 and MSP1 from isolated merozoites was assayed in the presence of the indicated additives. Two different batches (1 and 2) of purified PfSUB2PD were tested, in each case alongside a similar dilution of the membrane flow-through (Conc FT) from the final step of concentration by ultrafiltration. Control additives used were bovine serum albumin (BSA), purified PfSUB1PD, and EDTA. No effect on processing was seen if PfSUB2PD was added to buffer control reactions at the end of the incubation period (extreme right-hand lane). The lower panel shows a titration of inhibition of MSP1 shedding in the presence of 2-fold serially decreasing concentrations of PfSUB2PD (from 1.2 μM to 0.075 μM).
(B) Effect of propeptides on PfSUB1 activity. Progress curves showing cleavage of substrate pepF1-6R by recombinant PfSUB1 before and following addition of 600 nM purified PfSUB2PD or PfSUB1PD, or BSA. Note that this concentration is in at least 300-fold molar excess over the amount of recombinant PfSUB1 present in the assay. The time point of addition of test proteins is arrowed.
To assess the selectivity of the inhibition mediated by PfSUB2PD, we examined its effect on the activity of recombinant PfSUB1 (Figure 6B). PfSUB2PD had no effect at the highest concentrations tested (1.2 μM), whereas PfSUB1PD rapidly abolished protease activity as reported previously [33]. To further assess whether PfSUB2PD had any capacity to act as a general serine protease inhibitor, we examined its activity against a range of other serine proteases using standard colorimetric or fluorescence-based assays. At concentrations equivalent to between 12- and 300-fold molar excess over protease, PfSUB2PD had no effect on the activity of chymotrypsin, trypsin, and elastase, though it exhibited just-detectable inhibitory activity against subtilisin Carlsberg and measurable activity against subtilisin BPN′ with a calculated Ki value of 1.2 μM (data not shown). Similar inhibition of heterologous subtilases by subtilase propeptides has been noted previously (e.g., [32,34]). Taken together with the binding data above, these results provide compelling evidence that the inhibition of AMA1 and MSP1 shedding mediated by recombinant PfSUB2PD is a direct result of it binding to and inhibiting the activity of mature PfSUB2 on the merozoite surface.
Discussion
Proteolytic shedding of surface proteins during invasion by apicomplexan parasites is thought to be a strategy by which adhesion–receptor complexes are removed as the parasite enters its host cell. Shedding of MSP1 by cleavage just upstream of its C-terminal EGF-like region occurs in all species of Plasmodium examined, and new interest in the protease responsible was sparked by the revelation that the same protease also mediates release of another essential surface protein, AMA1. Here we have identified this protease as PfSUB2, an integral membrane merozoite subtilase.
PfSUB2 is a very minor parasite component, and in an earlier study [23], we were insufficiently confident in our antibody localisation data to assign it to a specific sub-cellular location. By epitope tagging the endogenous pfsub2 gene and using a range of mAbs to study its expression profile compared to other well-defined markers, we have now definitively localised PfSUB2 to micronemes. This differs from the conclusions of Barale et al. (1999) [22] who used polyclonal antibodies to obtain evidence suggesting that PfSUB2 localizes to a different set of organelles called dense granules. However, cross-reactivity of polyclonal antibodies combined with the small dimensions of the malaria merozoite can render localization of sub-cellular components particularly challenging, and merozoite proteins have been previously mis-localised (e.g., [40]). A possible alternative explanation for the discrepancy between these results is that epitope tagging of PfSUB2 resulted in its mistargeting. This phenomenon has previously been observed in Apicomplexa, but seems most unlikely in this case; given the evidence that PfSUB2 is an essential gene product, a defect in sorting would likely have phenotypic consequences. Our transgenic lines and clones displayed normal growth and morphology. Importantly, moreover, our conclusions are lent credence by the finding that PfSUB2 is secreted onto the parasite surface prior to or at invasion, an activity typical of microneme proteins. Dense granule proteins, in contrast, are discharged predominantly following invasion [41].
Collectively, our findings support the following model. PfSUB2 undergoes maturation by post-translational proteolytic processing that is initiated by propeptide cleavage and results in the formation of a membrane-bound protease. By analogy with PfSUB1 maturation [25], both PfSUB2I and PfSUB2T are likely to be enzymatically active, but it is probably the PfSUB2T form that accumulates in micronemes where it is stored until schizont rupture. At merozoite release PfSUB2 is secreted onto the parasite surface where it engages with an actin-based motor. This could occur through its cytoplasmic domain or through ectoplasmic interactions with other micronemal proteins that are themselves connected to the motor. PfSUB2 is then shuttled towards the posterior pole of the merozoite, in the process encountering and cleaving the MSP1 complex and AMA1 from the surface of the parasite. Our present results show that this capping process can occur efficiently in the absence of invasion, satisfactorily explaining the limited shedding of MSP1 and AMA1 seen in preparations of free merozoites. However, our observations may not accurately reflect the dynamics of PfSUB2 translocation at invasion. The invasive half-life of P. falciparum merozoites is extremely short, and shedding of MSP1 is known to go to completion during the brief (~30 s) time course of erythrocyte invasion [7]. We therefore propose that—similar to TgMIC2/M2AP—capping of PfSUB2 on the invading merozoite likely occurs predominantly concomitant with invasion, the protease perhaps concentrating at the moving junction. IEM studies of merozoites “captured” in the process of invasion will be required to test this postulate. Unfortunately this is a highly ambitious goal in P. falciparum in which, as a result of the short invasive half-life of merozoites, no surface protein has ever been localised by this technique on the invading parasite. The issue would be best addressed in the simian malaria model P. knowlesi, free merozoites of which are relatively long lived. This will be a major focus of future work.
An additional prediction of our model is that PfSUB2 may not be substantially exposed at the surface of the invasive merozoite prior to the point of junction formation. In support of this, we have found that addition of purified PfSUB2PD to parasite cultures has no effect on invasion (P. Harris, unpublished data). In an analogous example, antibodies against sporozoite TRAP or TgMIC2 have no effect on invasion, because these adhesins are not displayed at high levels on the parasite surface until the point of invasion, and only those molecules inhabiting the moving junction are functional during invasion [42] (V. Carruthers, personal communication). Nonetheless, we expect that it should be possible to target PfSUB2 with appropriately selective low-molecular weight, membrane-permeable compounds.
Capping of PfSUB2 to the merozoite posterior could be prevented by latrunculin A but not by cytochalasin D, a puzzling finding in view of the latter's known potent effects on invasion by the malaria merozoite and the fact that both compounds prevent invasion at the concentrations used here. Both drugs induce rapid disruption of the actin cytoskeleton in mammalian cells, but through distinct mechanisms; whereas cytochalasin D binds to filamentous actin, inhibiting polymerization at both barbed and pointed ends [43], latrunculin binds to and sequesters actin monomers, preventing their de novo polymerization [44,45]. One consequence of this is that cytochalasin D does not induce net depolymerization of actin, whereas latrunculin does [46,47], and there are now numerous examples of cellular actin-based processes that exhibit differential sensitivity to these drugs (e.g., [47,48]). Only a single actin gene is expressed in P. falciparum merozoites, and a recent study has shown that filaments formed by this actin are unusually short [49]. One explanation for the discrepancy between the effects observed with the two actin antagonists used in our experiments is that the degree of actin polymerisation required to provide the traction for invasion may be much greater that than required to translocate the bulk of secreted PfSUB2 the short distance (~1.2 μm) from the apical region of the merozoite to its posterior. This could explain why invasion is more sensitive than capping of PfSUB2 to perturbation by cytochalasin D.
The motility of apicomplexan zoites is stringently regulated, primarily by control of actin polymerization [50], and in highly motile apicomplexan zoites such as the Toxoplasma tachyzoite, motility ceases abruptly after invasion, implying rapid disassembly of the actinomyosin motor. The malaria merozoite is one of the smallest known eukaryotic cells, and so removal of the forces powering and maintaining retrograde flow of PfSUB2 after invasion would allow subsequent redistribution of the protease by simple diffusion in the parasite membrane. Our observation that PfSUB2 uniformly decorates the plasma membrane of the intracellular ring is consistent with this.
The estimated IC50 of approximately 300 nM for inhibition of MESH by PfSUB2PD is somewhat higher than Ki values measured for other subtilase propeptides against their cognate enzymes, which range from around 1–6 nM [32,33] to around 200 nM [51]. However, the combination of a relatively large molecular mass and the fact that it has to interact with membrane-bound protease may preclude ready access of PfSUB2PD to merozoite surface PfSUB2 in our assay. Moreover, PfSUB2PD may not represent the optimal propeptide construct; the exact internal sites at which processing takes place during PfSUB2 maturation are unknown, and it is well documented that even small truncations of propeptide constructs can compromise their inhibitory capacity (e.g., [32,33]). Despite this, PfSUB2PD exhibited clear selectivity, showing no effect on a number of serine proteases and only weak inhibition of a phylogenetically related bacterial subtilisin. A total of only three subtilases are encoded in the Plasmodium genome [25]. Previously, we have conclusively excluded a role for PfSUB1 in MESH activity, but one criticism of our present conclusions is that the inhibition of MESH mediated by PfSUB2PD could be the result of interactions with PfSUB3, a subtilase that is poorly characterised but known to be transcribed in parasite blood-stages [25]. However, we have recently successfully disrupted the pfsub3 gene in blood-stages of P. falciparum; the resulting parasites are viable, and processing of both MSP1 and AMA1 occurs normally (R. O'Donnell and M. Blackman, unpublished data), demonstrating that PfSUB3 is not essential for MESH activity.
In Toxoplasma gondii, several integral membrane microneme adhesins that cap to the parasite posterior through interactions with the actinomyosin motor are shed in the final seconds of invasion via cleavage within their TMD by a class of intramembrane proteases called rhomboids [52,53]. A malaria merozoite rhomboid-like activity has also recently been identified [17], the role of which may be to shed merozoite adhesins that are functional homologues of these transmembrane adhesins [20]. The merozoite rhomboid(s) cannot be involved in juxtamembrane shedding or shedding of GPI-anchored proteins such as MSP1, because rhomboids can only cleave within a TMD [54]. A Toxoplasma rhomboid, TgROM5, was recently localised to the posterior surface of the tachyzoite [55], suggesting that temporal and spatial regulation of release of capped microneme proteins is achieved through the simple expedient of maintaining the protease at the rear of the parasite; substrates are effectively recruited to the protease by the capping process. Our present findings show that PfSUB2 has evolved to function in a converse fashion; because its substrates (MSP1 and AMA1) are uniformly distributed about the parasite circumference, the sheddase must be tracked across the parasite surface in order to efficiently engage with them. PfSUB2 probably acts in concert with the merozoite rhomboid(s) during invasion.
Do functional homologues of PfSUB2 exist in other Apicomplexa? Aside from the rhomboid studies alluded to above, little is known of the molecular details of shedding of surface and microneme proteins in other apicomplexan genera. Two subtilisin-like serine proteases, TgSUB1 and TgSUB2, have been identified in T. gondii tachyzoites, but neither of these likely performs a role analogous to that of PfSUB2; TgSUB1 is a nonessential GPI-anchored microneme protease, whereas TgSUB2 is located in rhoptries in which it appears to be involved in maturation of other rhoptry proteins [56]. It remains to be seen whether Apicomplexa other than Plasmodium use a subtilisin-like sheddase at invasion.
Erythrocyte invasion is an obvious but thus far underexploited target for drugs designed to block the malarial lifecycle. PfSUB2 may represent an excellent target for new protease inhibitor-based drugs, and to this end, recombinant expression of the protease in an enzymatically active form is now a priority.
Materials and Methods
Parasite culture and transfection.
P. falciparum clones D10, 3D7, and T9/96 were maintained and free merozoites produced as described [57]. For some experiments, medium was supplemented with cytochalasin D or latrunculin A (Sigma-Aldrich, Poole, United Kingdom) added from stock solutions in DMSO. To obtain highly synchronised preparations of newly invaded ring-stage parasites, mature schizonts isolated by centrifugation on cushions of 63% isotonic Percoll were cultured with fresh erythrocytes for 2 h to allow rupture and reinvasion before removal of residual schizonts as described previously [14]. For transfection, ring-stage parasites were electroporated with plasmid DNA using standard procedures [58]. Parasites containing integrated plasmid were selected by drug cycling in the presence of 10 nM WR99210 (Jacobus Pharmaceuticals, Princeton, New Jersey, United States), then cloned by limiting dilution.
Transfection constructs.
The 3HA sequence was amplified from pREP(HA3)42 [59] using primers HA-1 5′-AGA CAC TGC TCG AGT ACC CTT ACG ATG TT-3′ and HA-2 5′-GGA CAC TGG TCG AC
T CAA GCG TAA TCT GG −3′ (XhoI and SalI sites are underlined, and the incorporated translational stop codon is in bold). The product was ligated into the unique XhoI site just upstream of the PbDT 3′ UTR in pHH1-ΔSERA4, kindly provided by Dr. Brendan Crabb, the Walter and Eliza Hall Institute, Melbourne, Australia, [60] to produce construct pHH1-HA3. The 3′ region of the pfsub2 gene, which contains an internal BglII site, was amplified from T9/96 P. falciparum genomic DNA using primers T996–1 (5′-GCC CCA GGT CAT CAC ATA TAT TCT ACT ATT CC-3′), and T996–2 (5′-GGA CAC TGC TCG AGT TTC ATA AAC ATA TCA TC-3′; XhoI site underlined); primers were designed to remove the stop codon from the end of the pfsub2 coding sequence and form a continuous reading frame into the 3HA region. The sera4 gene fragment was excised from pHH1-HA3 using BglII and XhoI and replaced with the BglII/XhoI-digested PCR fragment to create pHH1-T996HA3. Control plasmid pHH1-T996w was created in a similar manner but using a product amplified from T9/96 genomic DNA with primers T9961 (5′-GCC CCA GGT CAT CAC ATA TAT TCT ACT ATT C-3′) and T996w (5′-CTC GAG TCA TTT CAT AAA CAT ATC ATC AAG TTG ATT CAT TGC-3′; XhoI site underlined); this amplifies a similar fragment of the 3′ end of the pfsub2 gene but including the stop codon. The sera4 gene fragment was excised from pHH1-ΔSERA4 using BglII and XhoI and replaced with the BglII/XhoI-digested pfsub2 fragment. Nucleotide sequences of all cloned products were confirmed by sequencing on both strands.
Southern blot.
Parasite genomic DNA was prepared using the DNeasy Tissue Kit (Qiagen, Valencia, California, United States) and digested using ScaI (Roche). Digested products were electrophoresed on a 0.7% agarose gel and transferred to Hybond N+ Nylon membrane (Amersham Biosciences, Buckinghamshire, United Kingdom). The blot was probed with a [32P]-labelled, gel-purified, 569-base pair PCR product extending from nucleotides 3,047 to 3,615 of the T9/96 pfsub2 gene.
Indirect immunofluorescence assay and fluorescence microscopy.
Thin films of P. falciparum cultures were acetone-fixed and incubated with a 1:500 dilution of the anti-HA mAb 12CA5 (mouse) or 3F10 (rat; Roche, Basel, Switzerland) for 30 min, then with a biotinylated goat anti-mouse or anti-rat IgG (Chemicon, Temecula, California, United States) diluted 1:500, followed by incubation with FITC streptavidin (Vector Laboratories, Burlingame, California, United States) diluted 1:500. For dual labelling, samples were additionally probed with mAb 4G2 (anti-PfAMA1; [10]), mAb 61.3 (anti-RhopH2; [61]), mAb H5 (anti-RAP2; I. Ling, unpublished data), or mAb 1E1 (anti-MSP119; [62]), in each case conjugated directly to Alexa Fluor 594 (Molecular Probes, Invitrogen, Carlsbad, California, United States), or mAb X509 (anti-MSP1; [63]) followed by incubation with Alexa Fluor 594–conjugated anti-human IgG (Molecular Probes) diluted 1:500. Slides were stained with DAPI (4,6-diamidino-2-phenylindole) and mounted in Citifluor (Citifluor, Canterbury, United Kingdom). Images were collected using AxioVision 3.1 software on an Axioplan 2 Imaging system (Zeiss, Oberkochen, Germany) using a Plan-APOCHROMAT 100×/1.4 oil immersion objective, and annotated using Adobe PhotoShop.
IEM.
Samples of late schizonts were fixed in 0.075% (v/v) double-distilled glutaraldehyde and 4% (w/v) paraformaldehyde made up in culture medium (RPMI, pH 7.2), for 20 min on ice, then washed four times in ice-cold RPMI and dehydrated through a progressively low-temperature ethanol series, infiltrated with LR White resin (EMSCOPE, London, United Kingdom) and polymerized by ultraviolet light at room temperature for 48 h. For PfSUB2HA detection, sections were immunostained with the anti-HA mAb 3F10 (diluted 1:50 or 1:100) followed by biotinylated goat anti-rat antibody (diluted 1:100) then streptavidin conjugated to 10 nm gold particles (British Biocell International, Cardiff, United Kingdom) diluted 1:100. Detection of PfAMA1 was carried out as described in a previous study [64] using a rabbit polyclonal antiserum (#680) raised against recombinant PfAMA1 (kindly donated by C. Kocken and A. Thomas, Biomedical Primate Research Centre, Rijswijk, the Netherlands). For control purposes, parallel samples were treated with irrelevant primary antibodies. Sections were stained for 4 min with 2% (w/v) aqueous uranyl acetate. Images were viewed and captured digitally with a Hitachi 7600 electron microscope.
Expression and purification of the PfSUB2 propeptide.
A wholly synthetic gene called pfsub2
synth, encoding the T9/96 PfSUB2 but lacking the intron present in the authentic pfsub2 gene and with codon usage optimised for expression in Trichoplusia ni insect cells, was synthesised as described previously [65], using a total of 168 40-mer oligonucleotides. Sequence encoding Asn22-Leu687 was amplified from pfsub2
synth using forward primer 5\′-GGT ATT GAG GGT CGC AAC AAC AAC AAC AAC TAC TAC TTG-3' plus reverse primer 5\′-AGA GGA GAG TTA GAG CCC TAC AGG AAG CTG TAT TTG TTG-3' (PRO2) and cloned into the ligation-independent vector pET-30Xa/LIC (Novagen, Madison, Wisconsin, United States). The encoded propeptide construct was expressed in E. coli BL21-Gold (DE3) (Stratagene, La Jolla, California, United States). Cells were disrupted by sonication in 20 mM Tris-HCl, 300 mM NaCl, 20 mM imidazole (pH 8.2) supplemented with protease inhibitors, and the PfSUB2PD bound to Ni-NTA agarose (Qiagen) then eluted in the same buffer containing 250 mM imidazole. Usually, solid urea was then added to 8 M and the protein chromatographed on a Superdex 200 prep grade 26/60 column equilibrated in 20 mM Tris-HCl, 150 mM NaCl, 8 M urea, 1 mM DTT (pH 8.2). Polishing was performed on a Vydac 4.6 × 150 mm C4 RP-HPLC column eluted at 1 ml min−1 with a 45%−63% (v/v) gradient of acetonitrile in 0.1% (v/v) TFA over 20 min, monitoring elution at 215 nm. Eluted protein was taken up into gel filtration buffer and refolded by rapid 400-fold dilution into 20 mM Tris-HCl, 150 mM NaCl (pH 8.2). In some cases the gel filtration step was performed in the absence of urea, and the RP-HPLC step omitted to obtain lower yields of less-pure protein that had not been exposed to denaturing conditions. Protein was concentrated by ultrafiltration and the membrane flow-through retained as a control buffer for subsequent experiments. The identity of purified PfSUB2PD was confirmed by Western blot using an anti-polyhistidine antibody (mAb HIS-1; Sigma) and mass spectrometric peptide fingerprinting of tryptic digests [15] Its concentration was calculated from absorbance measurements at 280 nm, based on the tyrosine and tryptophan content that predicts a molar extinction coefficient of 53,305 M−1 cm−1 [66]. Purification of PfSUB1 propeptide was as described previously [33].
CD.
CD spectra of PfSUB2PD were recorded from 250 to 190 nm at 20 °C in a fused silica cuvette of 1-mm path length using a Jasco J-715 spectropolarimeter. Protein concentrations were approximately 0.1 mg ml−1 in 20 mM Tris-HCl, 150 mM NaCl (pH 8.2). Spectra presented are averages of multiple scans recorded using a scan speed of 200 nm min−1 and a response time of 0.25 s. Buffer blanks were subtracted from all spectra. Spectra were subjected to multicomponent secondary structure analysis using the SELCON3 algorithm [67].
Pull-down experiments and Western blot.
Percoll-enriched schizonts of PfSUB2HA clone 2D were extracted into 10 volumes of 20 mM Tris-HCl, 150 mM NaCl (pH 7.6) containing 1.5 mM sodium deoxycholate. Clarified extract (200 μl, corresponding to approximately 2 × 108 schizonts) was mixed with up to 0.5-μg purified PfSUB2PD or control proteins, then further supplemented with 100-μl S-protein agarose beads (Novagen) to bind the propeptide plus any associated proteins. Beads were washed in the above buffer then bound proteins analysed by SDS PAGE and Western blot as described previously [33], using mAb 3F10 to probe the blots.
Processing assays.
The effect of PfSUB2PD on shedding of AMA1 and MSP1 from isolated merozoites was assayed as described [16,57]. Briefly, 3D7 merozoites purified from cultures containing 10 mM EGTA were washed, suspended in 50 mM Tris-HCl, 5 mM CaCl2 (pH 7.6) and divided into equal aliquots of approximately 4 × 108 merozoites, on ice. Aliquots were supplemented with purified PfSUB2PD, purified PfSUB1 propeptide, BSA, 10 mM EDTA or control buffers and transferred to 37 °C for 2 h. Merozoites were pelleted, and shed AMA1 or MSP1 detected in merozoite supernatants by Western blot using mAb 4G2 or mAb X509 respectively.
Protease assays.
Production of recombinant PfSUB1 and an assay of activity based on cleavage of the fluorogenic substrate pepF1-6R has been described previously [68]. Cleavage was monitored at room temperature with a Cary Eclipse fluorescence spectrophotometer (Varian, Palo Alto, California, United States) equipped with a microplate reader accessory, using excitation and emission wavelengths of 552 nm and 580 nm respectively. Starting concentrations of substrate and PfSUB1 in these assays were 0.2 μM (well below the KM for this substrate) and approximately 2 nM respectively. For other protease assays, enzymes and substrates were obtained from Sigma. Subtilisins BPN′ and Carlsberg were assayed using N-succinyl-L-Ala-L-Ala-L-Pro-L-Phe-p-nitroanilide as described previously [33]. The KM value for hydrolysis of this substrate by subtilisin BPN′ under these conditions was 242 ± 8 μM. Ki determinations were performed as described [33]. Assays of bovine trypsin, α-chymotrypsin and porcine pancreatic elastase were performed using published procedures with substrates Nα-benzoyl-DL-Arg-7-amido-4-methylcoumarin [69], N-succinyl-L-Ala-L-Ala-L-Pro-L-Phe-p-nitroanilide [70], and N-succinyl-Ala-Ala-Ala-p-nitroanilide [71] respectively.
Supporting Information
Accession Numbers
The GenBank (http://www.ncbi.nlm.nih.gov/Genbank/) accession numbers for the genes discussed in this paper are pfsub2 (AJ132422) and pfsub2
synth (AY998616).
We thank Lee Graham and Muni Grainger for technical assistance and provision of purified merozoites, and John Hopkins for his valuable help in IEM specimen preparation and imaging. We are indebted to Steve Martin for CD analysis and Steve Howell for mass spectrometry. PKH is in receipt of a Medical Research Council (MRC) Graduate Studentship grant, and SY was funded by a pilot grant from MRC Technology's Development Gap Fund. ARD, LHB, and JMH acknowledge with thanks funding from the Wellcome Trust (Grant 069515), and the technical assistance of the Centre for Ultrastructural Imaging, Guy's Hospital. This work was supported by the Medical Research Council, UK.
Competing interests. The authors have declared that no competing interests exist.
Author contributions. PKH, SY, and MJB conceived and designed the experiments. PKH, SY, ARD, CWM, FH, LHB, GHM, and MJB performed the experiments. PKH, SY, ARD, LHB, GHM, and MJB analyzed the data. PKH, SY, RAOD, FH, LHB, and MJB contributed reagents/materials/analysis tools. PKH, SY, and MJB wrote the paper.
Abbreviations
AMA1apical membrane antigen-1
CDcircular dichroism
GPIglycosylphosphatidyl inositol
HAhaemagglutinin
IEMimmuno-electron microscopy
IFAimmunofluorescence
mAbmonoclonal antibody
MESHmerozoite surface sheddase
MSP1merozoite surface protein-1
PVparasitophorous vacuole
RP-HPLCreversed phase–high performance liquid chromatography
TMDtransmembrane domain
==== Refs
References
Hay SI Guerra CA Tatem AJ Noor AM Snow RW 2004 The global distribution and population at risk of malaria: Past, present, and future Lancet Infect Dis 4 327 336 15172341
Soldati D Foth BJ Cowman AF 2004 Molecular and functional aspects of parasite invasion Trends Parasitol 20 567 574 15522666
Bannister LH Butcher GA Dennis ED Mitchell GH 1975 Structure and invasive behaviour of Plasmodium knowlesi merozoites in vitro Parasitology 71 483 491 1202413
Aikawa M Miller LH Johnson J Rabbege J 1978 Erythrocyte entry by malarial parasites. A moving junction between erythrocyte and parasite J Cell Biol 77 72 82 96121
Goel VK Li X Chen H Liu SC Chishti AH 2003 Band 3 is a host receptor binding merozoite surface protein 1 during the Plasmodium falciparum invasion of erythrocytes Proc Natl Acad Sci U S A 100 5164 5169 12692305
Li X Chen H Oo TH Daly TM Bergman LW 2004 A co-ligand complex anchors Plasmodium falciparum merozoites to the erythrocyte invasion receptor band 3 J Biol Chem 279 5765 5771 14630931
Blackman MJ 2000 Proteases involved in erythrocyte invasion by the malaria parasite: Function and potential as chemotherapeutic targets Curr Drug Targets 1 59 83 11475536
Tomley FM Soldati DS 2001 Mix and match modules: Structure and function of microneme proteins in apicomplexan parasites Trends Parasitol 17 81 88 11228014
Mitchell GH Thomas AW Margos G Dluzewski AR Bannister LH 2004 Apical membrane antigen 1, a major malaria vaccine candidate, mediates the close attachment of invasive merozoites to host red blood cells Infect Immun 72 154 158 14688092
Pizarro JC Normand BV Chesne-Seck ML Collins CR Withers-Martinez C 2005 Crystal structure of the malaria vaccine candidate apical membrane antigen 1 Science 308 408 411 15731407
Mital J Meissner M Soldati D Ward GE 2005 Conditional expression of Toxoplasma gondii apical membrane antigen-1 (TgAMA1) demonstrates that TgAMA1 plays a critical role in host cell invasion Mol Biol Cell 16 4341 4349 16000372
Triglia T Healer J Caruana SR Hodder AN Anders RF 2000 Apical membrane antigen 1 plays a central role in erythrocyte invasion by Plasmodium species Mol Microbiol 38 706 718 11115107
O'Donnell RA, Saul A, Cowman AF, Crabb BS 2000 Functional conservation of the malaria vaccine antigen MSP-119 across distantly related Plasmodium species Nat Med 6 91 95 10613831
Blackman MJ Heidrich HG Donachie S McBride JS Holder AA 1990 A single fragment of a malaria merozoite surface protein remains on the parasite during red cell invasion and is the target of invasion-inhibiting antibodies J Exp Med 172 379 382 1694225
Howell S Withers-Martinez C Kocken C Thomas A Blackman M 2001 Proteolytic processing and primary structure of Plasmodium falciparum apical membrane antigen-1 J Biol Chem 276 31311 31320 11399764
Howell SA Well I Fleck SL Kettleborough C Collins CR 2003 A single malaria merozoite serine protease mediates shedding of multiple surface proteins by juxtamembrane cleavage J Biol Chem 278 23890 23898 12686561
Howell S Hackett F Jongco A Withers-Martinez C Kim K 2005 Distinct mechanisms govern proteolytic shedding of a key invasion protein in apicomplexan pathogens Mol Microbiol 57 1342 1356 16102004
Fleck SL Birdsall B Babon J Dluzewski AR Martin SR 2003 Suramin and suramin analogues inhibit merozoite surface protein 1 secondary processing and erythrocyte invasion by the malaria parasite, Plasmodium falciparum
J Biol Chem 278 47670 47677 13679371
Dutta S Haynes JD Barbosa A Ware LA Snavely JD 2005 Mode of action of invasion-inhibitory antibodies directed against apical membrane antigen 1 of Plasmodium falciparum
Infect Immun 73 2116 2122 15784553
Carruthers VB Blackman MJ 2005 A new release on life: Emerging concepts in proteolysis and parasite invasion Mol Microbiol 55 1617 1630 15752188
Blackman MJ Chappel JA Shai S Holder AA 1993 A conserved parasite serine protease processes the Plasmodium falciparum merozoite surface protein-1 Mol Biochem Parasitol 62 103 114 8114810
Barale JC Blisnick T Fujioka H Alzari PM Aikawa M 1999
Plasmodium falciparum subtilisin-like protease 2, a merozoite candidate for the merozoite surface protein 1–42 maturase Proc Natl Acad Sci U S A 96 6445 6450 10339607
Hackett F Sajid M Withers-Martinez C Grainger M Blackman MJ 1999 PfSUB-2: A second subtilisin-like protein in Plasmodium falciparum merozoites Mol Biochem Parasitol 103 183 195 10551362
Uzureau P Barale JC Janse C Waters A Breton C 2004 Gene targeting demonstrates that the Plasmodium berghei subtilisin PbSUB2 is essential for red cell invasion and reveals spontaneous genetic recombination events Cell Microbiol 6 65 78 14678331
Withers-Martinez C Jean L Blackman M 2004 Subtilisin-like proteases of the malaria parasite Mol Microbiol 53 55 63 15225303
Carruthers VB Giddings OK Sibley LD 1999 Secretion of micronemal proteins is associated with toxoplasma invasion of host cells Cell Microbiol 1 225 235 11207555
Bumstead J Tomley F 2000 Induction of secretion and surface capping of microneme proteins in Eimeria tenella
Mol Biochem Parasitol 110 311 321 11071285
Carruthers VB Sibley LD 1999 Mobilization of intracellular calcium stimulates microneme discharge in Toxoplasma gondii
Mol Microbiol 31 421 428 10027960
Miller LH Aikawa M Johnson JG Shiroishi T 1979 Interaction between cytochalasin B-treated malarial parasites and erythrocytes. Attachment and junction formation J Exp Med 149 172 184 105074
Bozdech Z Zhu J Joachimiak MP Cohen FE Pulliam B 2003 Expression profiling of the schizont and trophozoite stages of Plasmodium falciparum with a long-oligonucleotide microarray Genome Biol 4 R9 12620119
Le Roch KG Zhou Y Blair PL Grainger M Moch JK 2003 Discovery of gene function by expression profiling of the malaria parasite life cycle Science 301 1503 1508 12893887
Boudreault A Gauthier D Lazure C 1998 Proprotein convertase PC1/3-related peptides are potent slow tight-binding inhibitors of murine PC1/3 and Hfurin J Biol Chem 273 31574 31580 9813073
Jean L Hackett F Martin S Blackman M 2003 Functional characterization of the propeptide of Plasmodium falciparum subtilisin-like protease-1 J Biol Chem 278 28572 28579 12764150
Fugere M Limperis PC Beaulieu-Audy V Gagnon F Lavigne P 2002 Inhibitory potency and specificity of subtilase-like pro-protein convertase (SPC) prodomains J Biol Chem 277 7648 7656 11723118
Li Y Hu Z Jordan F Inouye M 1995 Functional analysis of the propeptide of subtilisin E as an intramolecular chaperone for protein folding. Refolding and inhibitory abilities of propeptide mutants J Biol Chem 270 25127 25132 7559646
Tangrea MA Bryan PN Sari N Orban J 2002 Solution structure of the pro-hormone convertase 1 pro-domain from Mus musculus
J Mol Biol 320 801 812 12095256
Jain SC Shinde U Li Y Inouye M Berman HM 1998 The crystal structure of an autoprocessed Ser221Cys-subtilisin E-propeptide complex at 2.0 A resolution J Mol Biol 284 137 144 9811547
Comellas-Bigler M Maskos K Huber R Oyama H Oda K 2004 1.2 A crystal structure of the serine carboxyl proteinase pro-kumamolisin; structure of an intact pro-subtilase Structure (Camb) 12 1313 1323 15242607
Kim JS Kluskens LD de Vos WM Huber R van der Oost J 2004 Crystal structure of fervidolysin from Fervidobacterium pennivorans, a keratinolytic enzyme related to subtilisin J Mol Biol 335 787 797 14687574
Healer J Crawford S Ralph S McFadden G Cowman AF 2002 Independent translocation of two micronemal proteins in developing Plasmodium falciparum merozoites Infect Immun 70 5751 5758 12228305
Mercier C Adjogble KD Daubener W Delauw MF 2005 Dense granules: Are they key organelles to help understand the parasitophorous vacuole of all apicomplexa parasites? Int J Parasitol 35 829 849 15978597
Gantt S Persson C Rose K Birkett AJ Abagyan R 2000 Antibodies against thrombospondin-related anonymous protein do not inhibit Plasmodium sporozoite infectivity in vivo Infect Immun 68 3667 3673 10816526
Sampath P Pollard TD 1991 Effects of cytochalasin, phalloidin, and pH on the elongation of actin filaments Biochemistry 30 1973 1980 1899622
Coue M Brenner SL Spector I Korn ED 1987 Inhibition of actin polymerization by latrunculin A FEBS Lett 213 316 318 3556584
Morton WM Ayscough KR McLaughlin PJ 2000 Latrunculin alters the actin-monomer subunit interface to prevent polymerization Nat Cell Biol 2 376 378 10854330
Morris A Tannenbaum J 1980 Cytochalasin D does not produce net depolymerization of actin filaments in HEp-2 cells Nature 287 637 639 6893622
Valderrama F Duran JM Babia T Barth H Renau-Piqueras J 2001 Actin microfilaments facilitate the retrograde transport from the Golgi complex to the endoplasmic reticulum in mammalian cells Traffic 2 717 726 11576448
Forest T Barnard S Baines JD 2005 Active intranuclear movement of herpesvirus capsids Nat Cell Biol 7 429 431 15803134
Schmitz S Grainger M Howell S Calder LJ Gaeb M 2005 Malaria parasite actin filaments are very short J Mol Biol 349 113 125 15876372
Wetzel DM Hakansson S Hu K Roos D Sibley LD 2003 Actin filament polymerization regulates gliding motility by apicomplexan parasites Mol Biol Cell 14 396 406 12589042
Toure BB Munzer JS Basak A Benjannet S Rochemont J 2000 Biosynthesis and enzymatic characterization of human SKI-1/S1P and the processing of its inhibitory prosegment J Biol Chem 275 2349 2358 10644685
Opitz C Di Cristina M Reiss M Ruppert T Crisanti A 2002 Intramembrane cleavage of microneme proteins at the surface of the apicomplexan parasite Toxoplasma gondii
EMBO J 21 1577 1585 11927542
Zhou XW Blackman MJ Howell SA Carruthers VB 2004 Proteomic analysis of cleavage events reveals a dynamic two-step mechanism for proteolysis of a key parasite adhesive complex Mol Cell Proteomics 3 565 576 14982962
Freeman M 2004 Proteolysis within the membrane: Rhomboids revealed Nat Rev Mol Cell Biol 5 188 197 14990999
Brossier F Jewett TJ Sibley LD Urban S 2005 A spatially localized rhomboid protease cleaves cell surface adhesins essential for invasion by Toxoplasma
Proc Natl Acad Sci U S A 102 4146 4151 15753289
Kim K 2004 Role of proteases in host cell invasion by Toxoplasma gondii and other Apicomplexa Acta Trop 91 69 81 15158690
Blackman M 1994 Purification of Plasmodium falciparum merozoites for analysis of the processing of merozoite surface protein-1 Methods Cell Biol 45 213 220 7707987
Fidock DA Wellems TE 1997 Transformation with human dihydrofolate reductase renders malaria parasites insensitive to WR99210 but does not affect the intrinsic activity of proguanil Proc Natl Acad Sci U S A 94 10931 10936 9380737
Craven RA Griffiths DJ Sheldrick KS Randall RE Hagan IM 1998 Vectors for the expression of tagged proteins in Schizosaccharomyces pombe
Gene 221 59 68 9852950
Miller SK Good RT Drew DR Delorenzi M Sanders PR 2002 A subset of Plasmodium falciparum SERA genes are expressed and appear to play an important role in the erythrocytic cycle J Biol Chem 277 47524 47532 12228245
Ling IT Kaneko O Narum DL Tsuboi T Howell S 2003 Characterisation of the rhoph2 gene of Plasmodium falciparum and Plasmodium yoelii
Mol Biochem Parasitol 127 47 57 12615335
Burghaus PA Holder AA 1994 Expression of the 19-kilodalton carboxy-terminal fragment of the Plasmodium falciparum merozoite surface protein-1 in Escherichia coli as a correctly folded protein Mol Biochem Parasitol 64 165 169 8078519
Blackman M Whittle H Holder A 1991 Processing of the Plasmodium falciparum major merozoite surface protein-1: Identification of a 33-kilodalton secondary processing product which is shed prior to erythrocyte invasion Mol Biochem Parasitol 49 35 44 1723148
Bannister LH Hopkins JM Dluzewski AR Margos G Williams IT 2003
Plasmodium falciparum apical membrane antigen 1 (PfAMA-1) is translocated within micronemes along subpellicular microtubules during merozoite development J Cell Sci 116 3825 3834 12902400
Withers-Martinez C Carpenter E Hackett F Ely B Sajid M 1999 PCR-based gene synthesis as an efficient approach for expression of the A+T-rich malaria genome Protein Eng 12 1113 1120 10611405
Pace C Vajdos F Fee L Grimsley G Gray T 1995 How to measure and predict the molar absorption coefficient of a protein Protein Sci 4 2411 2423 8563639
Sreerama N Woody RW 2000 Estimation of protein secondary structure from circular dichroism spectra: Comparison of CONTIN, SELCON, and CDSSTR methods with an expanded reference set Anal Biochem 287 252 260 11112271
Withers-Martinez C Saldanha J Ely B Hackett F O'Connor T 2002 Expression of recombinant Plasmodium falciparum subtilisin-like protease-1 in insect cells J Biol Chem 277 29698 29709 12052828
Kanaoka Y Takahashi T Nakayama H Takada K Kimura T 1977 Organic fluorescence reagent. IV. Synthesis of a key fluorogenic amide, L-arginine-4-methylcoumaryl-7-amide (L-Arg-MCA) and its derivatives. Fluorescence assays for trypsin and papain Chem Pharm Bull 25 3126 3128
DelMar EG Largman C Brodrick JW Geokas MC 1979 A sensitive new substrate for chymotrypsin Anal Biochem 99 316 320 574722
Bieth J Spiess B Wermuth CG 1974 The synthesis and analytical use of a highly sensitive and convenient substrate of elastase Biochem Med 11 350 357 4429553
|
16322767
|
PMC1291349
|
CC BY
|
2021-01-05 12:11:28
|
no
|
PLoS Pathog. 2005 Nov 25; 1(3):e29
|
utf-8
|
PLoS Pathog
| 2,005 |
10.1371/journal.ppat.0010029
|
oa_comm
|
==== Front
PLoS PathogPLoS PathogppatplpaplospathPLoS Pathogens1553-73661553-7374Public Library of Science San Francisco, USA 10.1371/journal.ppat.001003105-PLPA-RA-0004R2plpa-01-03-11Research ArticleImmunologyMicrobiologyRespiratory MedicineGenetics/Gene DiscoveryGenetics/Disease ModelsEubacteriaAnimalsMus (Mouse)Homo (Human)Comparative Signature-Tagged Mutagenesis Identifies Pseudomonas Factors Conferring Resistance to the Pulmonary Collectin SP-A Pseudomonas Factors Resistant to SP-A
Zhang Shiping 1Chen Yi 1Potvin Eric 2Sanschagrin Francois 2Levesque Roger C 2McCormack Francis X 1Lau Gee W 1*
1 Division of Pulmonary and Critical Care Medicine, University of Cincinnati College of Medicine, Cincinnati, Ohio, United States of America
2 Centre de Recherche sur la Fonction Structure et Ingenierie des Proteines, Universite Laval, Ste-Foy, Quebec, Canada
Ausubel Fred EditorHarvard Medical School, United States of America* To whom correspondence should be addressed. E-mail: [email protected] 2005 25 11 2005 1 3 e3114 3 2005 13 10 2005 Copyright: © 2005 Zhang et al.2005This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are properly credited.The pulmonary collectin, surfactant protein A (SP-A), is a broad spectrum opsonin with microbicidal membrane permeabilization properties that plays a role in the innate immune response of the lung. However, the factors that govern SP-A's microbial specificity and the mechanisms by which it mediates membrane permeabilization and opsonization are not fully understood. In an effort to identify bacterial factors that confer susceptibility or resistance to SP-A, we used comparative signature-tagged mutagenesis to screen a library of 1,680 Pseudomonas aeruginosa mutants for evidence of differential pulmonary clearance in SP-A-sufficient (SP-A+/+) and SP-A-deficient (SP-A−/−) mice. Two SP-A-sensitive P. aeruginosa mutants harboring transposon insertions in genes required for salicylate biosynthesis (pch) and phosphoenolpyruvate-protein-phosphotransferase (ptsP) were recovered. The mutants were indistinguishable from the parental wild-type PA01 with regard to opsonization by SP-A, but they exhibited increased susceptibility to SP-A-mediated membrane permeabilization. These results suggest that bacterial gene functions that are required to maintain membrane integrity play crucial roles in resistance of P. aeruginosa to the permeabilizing effects of SP-A.
Synopsis
Everyday, normal breathing deposits numerous microorganisms on the surfactant membrane that lines the air-exchanging surfaces of the lung. Surfactant protein SP-A, a component of the surfactant membrane, helps to maintain the lung in a germ-free state by aggregating inhaled microorganisms and facilitating their ingestion by immune cells, and by increasing the permeability of their cell membranes. However, the bacterial pathogen Pseudomonas aeruginosa is resistant to SP-A-mediated membrane disruption.
Using a genetic tool called comparative signature-tagged mutagenesis, the authors have identified two P. aeruginosa genes, pch and ptsP, that are required to resist SP-A-mediated membrane permeabilization. Molecular analyses indicate that the pch gene encodes an enzyme that synthesizes salicylate, a compound utilized by bacteria to acquire essential metal ions. The ptsP gene encodes an enzyme called phosphoenolpyruvate-protein-phosphotransferase. The loss of salicylate and phosphoenolpyruvate-protein-phosphotransferase weakens the P. aeruginosa cell membrane, which allows SP-A to poke holes on the membrane and kill the bacteria.
This is the first known report of the roles played by salicylate and phosphoenolpyruvate-protein-phosphotransferase in maintenance of bacterial membrane, and consequently, protecting bacteria from killing by SP-A, through disruption of membrane integrity.
Citation:Zhang S, Chen Y, Potvin E, Sanschagrin F, Levesque RC, et al. (2005) Comparative signature-tagged mutagenesis identifies Pseudomonas factors conferring resistance to the pulmonary collectin SP-A. PLoS Pathog 1(3): e31.
==== Body
Introduction
Pulmonary surfactant is a protein-phospholipid complex that lines the air-liquid interface of alveoli, forming the first point of contact with inhaled microbial pathogens [1,2]. A multitude of in vitro studies have demonstrated that the pulmonary collectins, surfactant protein A (SP-A) and surfactant protein D (SP-D), bind and aggregate bacterial, fungal, viral, and mycobacterial organisms, directly activate macrophages, and enhance the in vitro phagocytosis and intracellular killing of a variety of pulmonary pathogens [3,4]. Recognition of diverse microbial species by the pulmonary collectins is mediated by the C-type lectin domain, which selectively binds to carbohydrates that are prevalent on the surface of microbes, but not to the predominant terminal sugars on surface molecules of mammalian cells [5]. More recently, SP-A and SP-D have been reported to possess potent antimicrobial properties [6–8]. They directly inhibit the proliferation of bacteria and fungi in a macrophage- and aggregation-independent manner, by increasing the permeability of the microbial cell membrane. In particular, rough strains of Gram-negative bacteria, containing truncated lipopolysaccharide (LPS) domains are uniquely vulnerable to permeabilization by the collectins [6], as has been well known for other antimicrobial peptides and proteins.
Although the collectins inhibit microbial growth, the factors that govern the specificity and the mechanisms of membrane disruption by the collectins are not understood. In this study, we employed comparative signature-tagged mutagenesis (STM) as a tool to identify bacterial factors required to resist SP-A-mediated clearance, in vivo. STM relies on the ability of the pathogen in question to replicate in vivo as a mixed population and allows for the identification of the mutants whose phenotypes cannot be trans-complemented by other virulent strains present in the same inoculum [9]. Two Pseudomonas factors that are specifically required to protect P. aeruginosa from killing by SP-A were identified and characterized.
Results
SP-A−/− Mice Are Susceptible to Infection by P. aeruginosa
Levine et al [10] have reported that outbred SP-A−/− mice exhibit delayed clearance of a clinical P. aeruginosa strain from the lung, suggesting that SP-A plays a role in pulmonary innate immunity against this organism. We also found that the inbred C3H/HeN SP-A−/− mice were more susceptible to P. aeruginosa infection, in this case by strain PA01. SP-A+/+ mice cleared 63-fold (1.8 log) more of intranasally inoculated PA01 by 16 h (Figure 1A). In contrast, the bacterial load increased by 22 fold (1.34 log) in SP-A−/− mice (Figure 1A). In the absence of infection, SP-A−/− mice are healthy, and the histopathological analyses of their non-infected lungs revealed no significant differences when compare to the lungs of SP-A+/+ mice (unpublished data). When infected with PA01 however, SP-A+/+ mice developed broncho-pneumonia with mild pulmonary infiltrates (Figure 1B). In contrast, SP-A−/− mice developed more extensive consolidation with areas of lobar pneumonia.
Figure 1 SP-A Deficient Mice Are More Susceptible to Infection by P. aeruginosa
(A) SP-A+/+ and SP-A−/− mice (group of six) were infected intranasally with 1 × 107
P. aeruginosa PA01. Bacteria were recovered from homogenized lung tissue 16 h after inoculation. p < 0.01.
(B) Representative lung histology sections (stained with hematoxylin and eosin) from SP-A+/+ and SP-A−/− mice 16 h post intranasal instillation of PA01. Original magnification: 10 ×.
Identification of P. aeruginosa Genes Conferring Resistance to SP-A by Comparative STM Screening
We exploited the PCR-based STM technique and the P. aeruginosa PA01 STM mutant library [11] to identify putative microbial targets for SP-A. SP-A+/+ and SP-A−/− mice, in groups of three, were intranasally challenged with pools of 72 oligo-tagged STM mutants. Lungs were harvested at 16 h, homogenized, and plated. At least 10,000 bacterial colonies grown on Luria broth (LB) agar plates were harvested from each lung for bacterial genomic DNA extractions and oligo-tag amplification. We screened for differential clearance of individual STM mutants using PCR. If the PCR product of a mutated gene was absent on the output gel in SP-A+/+ mice, but present in SP-A−/− mice (Figure 2A, see arrows), we postulated that the bacterial protein encoded by the mutated gene was required to overcome the host defense activities of SP-A.
Figure 2 PCR-Based Signature-Tagged Mutagenesis
(A) Schematic drawing of PCR-based STM. Pools of 72 uniquely tagged mutants were intranasally inoculated into SP-A+/+ and SP-A−/− mice; 16 h later, lungs were harvested, homogenized, and plated. Approximately 10,000 colonies were harvested from the plates for genomic DNA extraction. PCR-amplification of tags was performed on the genomic DNA to screen for the presence or absence of DNA tags of each of the 72 mutants. Mutants whose DNA tags were present in the input pool and in the SP-A−/− pool, but absent in the SP-A+/+ pool (see white arrows) were further screened for susceptibility to SP-A.
(B) Agarose gel of PCR-based STM, identifying the pch mutant (left panel). Attenuation in the first SP-A+/+ mouse (right panel, m1) was confirmed in two additional SP-A+/+ mice (right panel, m2 and m3).
(C) Genetic loci of P. aeruginosa, when mutated, conferred increased sensitivity to in vivo killing by SP-A. DNA regions flanking pUTmini-Tn5 transposon insertions were cloned and sequenced. Similarity BLAST searches were performed against P. aeruginosa PA01 genomic sequence on NCBI and on http://www.pseudomonas.com. Black arrows indicate the approximate insertion site within the mutated ORFS.
(D) TLC analyses indicate that the pch STM mutant is unable to synthesize pyochelin (see arrows). Wild-type PA01 grown in LB and the pch mutant grown in LB supplemented with 1 mM salicylate produced green color pyochelin (see arrows). In contrast, no observable pyochelin was produced by the pch and PA06331 strains. Pure pyochelin was used as control.
(E) Restriction fragment length polymorphism analysis between parental wild-type PA01 and STM mutants confirmed that mutations in pch and ptsP were caused by a single insertion.
We comparatively screened 1,680 STM mutant strains in SP-A+/+ and SP-A−/− mice. Two mutants with increased sensitivity to SP-A (i.e., failed to survive in SP-A+/+ mice) were recovered, for a 0.12% recovery rate. We have cloned and sequenced both STM mutants by plasmid rescue [11]. DNA sequences were analyzed by comparison with online databases (http://www.pseudomonas.com). The output from Pool 4 (mutant pch, see below) is shown in Figure 2B (left panel). In the presence of SP-A, the pch mutant, which is disrupted in a gene required for salicylate biosynthesis, was unable to survive in the SP-A+/+ lung (missing band in the box, Figure 2B). In absence of SP-A, the pch mutant was able to grow in the lung, as indicated by the presence of an oligo-tag amplified by PCR (Figure 2B, left panel, SP-A−/− lane; right panel, lane m1). The results were confirmed by PCR analysis of the output pool of two additional mice (Figure 2B, right panel, lanes m2 and m3).
The pch STM mutant carries a transposon pUTminiTn5Km2 insertion within an intergenic region in between the 10th and 11th nucleotide after the stop codon of the pchA (PA4231) gene (Figure 2C). The pchA gene encodes an isochorismate synthase that participates in the biosynthesis of the siderophores, salicylate, and pyochelin [12–14]. Because transposon inserted within the intergenic region immediately downstream of the pchA gene, it might have affected the mRNA stability of the pchDCBA transcript. Thus, we named the STM mutant as pch. We examined whether the pch mutant had lost the ability to synthesize pyochelin by using thin layer chromatography (TLC) assays. TLC analysis of pyochelin extracted from bacteria cultured in M63 minimal medium confirmed that the pch mutant is unable to synthesize the green-colored pyochelin (Figure 2D, pch lane). In contrast, the parental wild-type PA01 produced pyochelin (Figure 2D, PA01 lane). When the pch STM mutant was grown in the presence of 1 mM salicylate, however, the ability to synthesize pyochelin was restored (Figure 2D, pch + 1mM salicylate lane). The expression of the genes within the downstream operon (PA4232 and PA4233), which is transcribed in opposite orientation to the pchADCBA operon, was found to be unaffected by transposon insertion as assessed by mRNA expression analyses (unpublished data). Collectively, these data suggest that susceptibility to SP-A-mediated clearance is caused by the loss of pch gene function, most probably by affecting the mRNA stability of the pchDCBA operon.
The second SP-A-susceptible STM mutant had the transposon pUTminiTn5Km2 inserted in between the 2,213th and 2,214th nucleotides of the ptsP (PA0337) gene (Figure 2C). The ptsP gene encodes a phosphoenolpyruvate-protein-phosphotransferase, a homolog of the E. coli Enzyme Initrogen (EINtr) [15]. The transposon pUTminiTn5Km2 insertion into ptsP is predicted to cause polarity and interrupt the transcription of PA0338.
Both STM mutants were judged to harbor a single transposon pUTminiTn5Km2 insertion after hybridization with the transposon vector (unpublished data). DNA fragments that flanked both ends of the transposon integrations in pchA and ptsP genes were cloned from the wild-type PA01. They were used to confirm gene interruption by the integrating transposon, by comparing restriction fragment length polymorphisms (RFLP) between wild-type PA01 and the STM mutants. As shown in Figure 2E, transposon-flanking DNA probes hybridized to DNA fragments with higher molecular weights in the chromosomes of both the pch and ptsP mutants than their parental strain PA01, indicating transposon insertion.
SP-A-Sensitive STM Mutants Are as Virulent as Parental Wild-Type in SP-A−/− Mice
Apart from confronting pulmonary host defenses, our STM screens require each mutant to compete with 71 other STM mutants within individual library pools. Competitive mixed infection assays have been widely used to assess the fitness of individual mutants versus their parental strains during in vivo infection [16,17]. As shown in Figure 3A, when co-administered with the parental wild-type PA01 by intranasal inoculation into mice, both STM mutants were less competitive than PA01 in SP-A+/+ mice. Specifically, mutant strains pch and ptsP were only 53% and 46% as competitive as PA01, respectively. In contrast, both mutants performed equally well as, if not better than, the wild-type strain in SP-A−/− mice (Figure 3A).
Figure 3 SP-A-Sensitive P. aeruginosa STM Mutants Compete Effectively with Parental Wild-Type PA01 in SP-A−/− , but Not in SP-A+/+ Mice
(A) In vivo competition assays between the wild-type and individual STM mutants after intranasal pulmonary inoculation of 6-wk-old SP-A+/+ and SP-A−/− mice. The mean CFU recovered from four mice in each group is shown. The CI is defined as the output ratio of mutant to wild-type bacteria divided by the input ratio of mutant to wild-type bacteria. *p < 0.013, #p < 0.011.
(B) Single respiratory tract infection of pch and ptsP mutants in SP-A+/+ and SP-A−/− mice was performed as described for Figure 1. Attenuation is defined as the log10 difference in CFU between wild-type and mutant bacteria recovered from lung tissue 16 h after inoculation. The mean ± standard error of six mice is shown. *p < 0.019, #p < 0.045.
We further tested the mutants, pch and ptsP, in single infection studies. When infected singly into wild-type mice, the viable bacterial counts of pch and ptsP mutant bacteria were 32-fold (1.5 log) and 25-fold (1.4 log) higher, respectively, in SP-A−/− mice than in SP-A+/+ mice (Figure 3B). Our results indicate that SP-A plays an important role in clearance of PA01 mutants with defective pch or ptsP gene function.
Preferential Clearance of the STM Mutants from SP-A+/+ Lung Is Independent of Macrophages
Previous studies have suggested that SP-A opsonizes P. aeruginosa and increases uptake by macrophages [10,18]. To determine if SP-A-mediated opsonization contributes to the early clearance of STM mutants, we compared the phagocytosis of pch and ptsP to PA01 in the presence and absence of SP-A. When live, green fluorescence protein (GFP)-expressing bacteria were exposed to mouse alveolar macrophages in the presence of SP-A, the uptake of PA01, its pch, or ptsP mutant was increased by 1.8, 1.3, and 1.6 fold, respectively, over that which occurred in the absence of SP-A (Figure 4A). However, the SP-A mediated increase in opsonization was only statistically significant in PA01 (p<0.01). The magnitude of the increase in the uptake of PA01 in the presence of SP-A was similar to previously published reports [10,18]. Although the addition of SP-A increased the macrophage uptake of STM mutants, enhancement of phagocytosis was slightly lower than that of wild-type PA01. Therefore, it does not appear that SP-A-mediated opsonization is responsible for the preferential clearance of pch and ptsP mutants from SP-A+/+ mice.
Figure 4 Bacterial Opsonization by SP-A Does Not Result in Differential In Vivo Phagocytosis and Killing of STM Mutants
(A) In vitro phagocytosis assays were performed using live, GFP-expressing PA01 and isogenic pch and ptsP bacteria, with alveolar macrophages isolated from SP-A+/+ mice. Internalized bacteria were counted using a phase contrast fluorescence microscope. The means of three experiments are shown. Two hundred macrophages were counted for each mouse. p = 0.0058 (PA01 vs PA01 + SP-A), p = 0.075 (pch vs pch + SP-A), and p = 0.109 (ptsP vs ptsP + SP-A).
(B) In vitro phagocytosis of PA01, pch, and ptsP by human neutrophils isolated from healthy volunteers. Phagocytosis experiments were performed as described for (A). p = 0.037 (PA01 vs PA01 + SP-A), p = 0.577 (pch vs pch + SP-A), and p = 0.919 (ptsP vs ptsP + SP-A).
(C–E) SP-A did not affect uptake or killing of STM mutants. PA01 (C), but not the STM mutants pch (D) and ptsP (E) were killed in significant amounts by neutrophils, independent of SP-A. The mean of three separate experiments is shown. p < 0.001 when comparing PA01 versus PA01 + neutrophils, and when comparing PA01-SP-A versus PA01-SP-A + neutrophils.
Preferential Clearance of the STM Mutants in SP-A+/+ Lung Is Independent of Neutrophils
Neutrophils are known to play a major role in protecting the lung against bacterial infection. We examined whether bacterial phagocytosis and killing by neutrophils might have led to preferential clearance of pch and ptsP. As shown in Figure 4B, the uptake of PA01 and the pch and ptsP mutants was not increased in the presence of SP-A. In fact, SP-A-opsonized PA01 bacteria were phagocytosed less efficiently by neutrophils than untreated bacteria. Interestingly, only wild-type PA01 bacteria were susceptible to killing by neutrophils, with approximately 73.7% killed in the absence of SP-A, and 57.2% killed in the presence of SP-A (Figure 4C). In contrast, both pch and ptsP mutants were not susceptible to neutrophil killing (Figure 4D and 4E). These results indicate that neutrophils kill P. aeruginosa by using a mechanism independent of SP-A opsonization, and lend support to the notion that neutrophils play a less critical role in the preferential clearance of pch and ptsP mutants from the SP-A+/+ mice than other anti-pseudomonal mechanisms.
The Loss of pch and ptsP Gene Functions Render P. aeruginosa Susceptible to SP-A-Mediated Membrane Permeabilization
Recent studies have indicated that SP-A is capable of directly killing bacteria and fungi by membrane permeabilization and inhibition of macro-molecular synthesis, independent of macrophage-mediated phagocytosis [6,7]. We examined whether direct membrane permeabilization contributed to the clearance of the SP-A-sensitive mutants. Due to the previously described importance of LPS in resistance to SP-A-mediated membrane permeabilization [6,19], we used a LPS mutant wbpL, which is deficient in the production of “initial” glycosyltransferase essential for the biosynthesis of both the A-band and B-band O-antigen of LPS [20], as positive control for membrane permeabilization assays. Permeabilization of the wbpL mutant was approximately 2.7-fold greater than the wild-type strain PA01 at 60 min (Figure 5A). In contrast, STMG2A7, an irrelevant mutant that is virulent in both SP-A+/+ and SP-A−/− mice, was found to be as resistant as parental wild-type PA01 to SP-A-mediated membrane permeability (Figure 5B).
Figure 5 The pch and ptsP Mutants Are Preferentially Permeabilized and Killed by SP-A
(A) Fluorescence of a cleavage-activated phosphatase substrate ELF97 was examined in the presence or absence of 50 μg hSP-A for a period of 60–90 min. Human SP-A (hSP-A) preferentially permeabilized LPS mutant wbpL, but not the parental wild-type PA01. The wbpL strain was used as positive control for membrane permeability. The difference in membrane permeabilization between wbpL mutant and PA01 was significant from the 28th min onward. *p < 0.05.
(B) A mutant strain STMG2A7 that is virulent in both SP-A+/+ and SP-A−/− mice was as resistant to SP-A-mediated membrane permeabilization as the parental wild-type PA01.
(C–D) Membrane permeability of the STM mutant pch and PA06331, a P. aeruginosa strain which is deleted in all of the structural genes required for pyochelin biosynthesis (Table 1.). The difference in membrane permeabilization between the mutants against wild-type PA01 was significant from the 35th and 18th min onward for mutant strains pch and PA06331, respectively. *p < 0.05.
(E–F) Membrane permeability of the STM mutant ptsP and the ΔptsP, a P. aeruginosa strain with a nonpolar, inframe deletion of the ptsP gene. The difference in membrane permeabilization between the mutants against wild-type PA01 was significant from the 29th and 28th min onward, for mutant strains ptsP and ΔptsP, respectively. *p < 0.05.
Table 1 Bacterial Strains and Plasmids Used in this Work
The STM mutant pch, and a mutant strain, PA06331, in which the pyochelin biosynthetic genes encompassing pchDCBA, pchR, and pchEFGHI operons had been removed [14], were permeabilized by SP-A to an extent that was 3.4- and 2.55-fold greater than the wild-type strain PA01, respectively (Figure 5C and 5D) after 60–90 min exposure to SP-A. Similarly, STM mutant ptsP, and an in-frame, nonpolar deletion mutant, ΔptsP, were permeabilized by SP-A to an extent that was 3.0- and 3.8-fold greater than the wild-type strain PA01, respectively (Figure 5E and 5F), at 60–90 min post-exposure to SP-A. These data are comparable to the approximately 3-fold increase in membrane permeability that SP-A exposure induces in E. coli K12 [6].
SP-A-Mediated Membrane Permeabilization Directly Kills the pch and ptsP Mutants
We next determined if SP-A mediated permeability kills STM mutants. Live/dead staining, based on exclusion of propidium iodide from the live cells, was performed on bacterial cells following membrane permeability assays. Viable cells are stained green while dead cells are stained red. As shown in Figure 6A, SP-A did not kill wild-type PA01 (green stained). In contrast, mixtures of green- and red-stained cells of pch and ptsP were visible following treatment with SP-A. Approximately 11% and 9 % of pch and ptsP cells, respectively, were killed (Figure 6B) within 60 min, in comparison to about 40% of E. coli K12 (unpublished data). In addition, in contrast to the robust SP-A-induced aggregation of E. coli K12 (Figure 6A), SP-A did not aggregate P. aeruginosa PA01, pch, or ptsP. These results suggest that the loss of intact LPS (i.e. rough LPS), or the inability to synthesize salicylate and phosphoenolpyruvate-protein-phosphotransferase, renders the cells vulnerable to membrane permeabilization and killing by SP-A. Furthermore, bacterial aggregation does not play a role in preferential clearance of the pch and ptsP mutants.
Figure 6 SP-A-Mediated Membrane Permeabilization Directly Kills the pch and ptsP Mutants
(A) Live/Dead staining was performed on E. coli K12, PA01, pch, and ptsP cells, following 1 h membrane permeabilization with 100 μg/ml SP-A. Green-stained cells are alive whereas red-stained cells are dead.
(B) Enumeration of live or dead bacteria following a 1 h exposure to hSP-A. At least 600–1,000 bacterial cells were counted under a fluorescence microscope. The mean of two experiments is shown.
Complementation of the pch and ptsP Mutants Restores Their Resistance to SP-A-Mediated Membrane Permeabilization
The susceptibility of STM mutants to SP-A mediated membrane permeabilization can be exploited for complementation analyses. As the pch mutant is disrupted in a gene required for salicylate biosynthesis, we performed complementation analysis by growing the pch mutant bacteria in the presence of 1 mM sodium salicylate. As shown in Figure 7A, the pch mutant bacteria grown in LB were susceptible to SP-A-mediated membrane permeabilization. In contrast, the pch mutant bacteria grown in LB supplemented with 1 mM sodium salicylate were fully resistant to membrane permeabilization by SP-A. We further examined whether the increased susceptibility to SP-A-mediated membrane permeability was due to the inability pch bacteria to synthesize the siderophore pyochelin. The pch bacteria cultured in pyochelin-supplemented LB or iron-deficient minimal medium were still membrane permeabilized by SP-A (unpublished data). Collectively, these results suggest that the increased susceptibility to SP-A-mediated membrane permeability in the pch bacteria is due to their inability to synthesize salicylate, or both salicylate and pyochelin. However, we cannot completely rule out the possibility that pyochelin may be involved in resistance to SP-A-mediated membrane permeability until we have exhausted all the growth conditions and optimal pyochelin concentrations for the assays.
Figure 7 Complementation Studies of SP-A-Sensitive Mutants pch and ptsP
(A) Culturing pch bacteria in LB supplemented with sodium salicylate restored its resistance to SP-A mediated membrane permeabilization.
(B) In control experiments, we demonstrated that the enzymatic activity of bacterial phosphatases in ptsP bacterial cells was not inhibited by 1 mM sodium salicylate.
(C) Provision of the wild-type ptsP gene on the pUCP19 in trans (pUCP-ptsP) restores the ability of STM mutant ptsP to resist SP-A mediated membrane permeabilization. In contrast, ptsP mutant expressing GFP gene on the pUCP19 (pstP-gfp) is permeabilized to similar levels as ptsP.
To rule out the possibility that 1 mM sodium salicylate may inactivate the enzymatic activity of the periplasmic phosphatases that cleaves the impermeant substrate, ELF97, in our permeability assay, we also tested ptsP mutant bacteria cultured in the presence or absence of 1 mM sodium salicylate. The ptsP mutant bacteria were permeabilized to the same extent, regardless of whether they were grown in the presence or absence of 1 mM sodium salicylate (Figure 7B), suggesting that salicylate did not inactivate phosphatase. These results suggest that the loss of pch gene function results in susceptibility of P. aeruginosa to clearance by SP-A by rendering the organism susceptible to membrane permeabilization.
We next tested if providing the wild-type ptsP gene in trans to the ptsP STM mutant bacteria could restore the resistance to SP-A-mediated membrane permeabilization. As shown in Figure 7C, the resulting strain, pUCP-ptsP was as resistant to SP-A-mediated membrane permeabilization as the wild-type PA01. In contrast, providing an irrelevant gfp gene in the ptsP bacteria (ptsP-gfp) did not restore the resistance of ptsP STM mutant to SP-A, nor did it alter the resistance or susceptibility of the parental wild-type PA01 (PA01-gfp) to permeabilization by SP-A (Figure 7C). These results suggest that ptsP gene function protects P. aeruginosa from membrane permeabilization by SP-A.
LPS Profiles of STM Mutants
Previous studies in E. coli [6] and Bordetella pertussis [19] have demonstrated that the structures of LPS on the surface of bacterial cells determine resistance to SP-A-mediated aggregation and membrane permeabilization. P. aeruginosa LPS mutants are also sensitive to SP-A-mediated membrane permeability, but are not aggregated by the collectin (Figure 5A; unpublished data). To examine if the pch and ptsp mutants have LPS defects, we compared the LPS profiles after separation by SDS-PAGE and silver staining. The LPS from wild-type PA01 and the pch mutant were indistinguishable (unpublished data). However, the two major LPS core bands of the ptsP mutant showed a reversal in relative abundance compared with the PA01 (unpublished data). Thus, there may be changes in LPS, at least in terms of the proportion of two different cores made. More detailed molecular analyses will be required to determine if LPS modification occurs within pch and ptsP mutants.
Discussion
The genome-wide mutagenesis scheme called STM was originally developed to identify novel vaccine and antibiotic targets [9]. STM relies on the ability of the pathogen in question to replicate in vivo as a mixed population, and allows the identification of the virulence genes whose mutant phenotypes cannot be trans-complemented by other virulent strains present in the same inoculum. Hundreds of virulence factors and putative drug targets have been identified by STM in at least 20 different bacterial and fungal pathogens [21]. In all of these screens, a pathogen of interest is mutated by transposon or suicide vector-based insertional strategies. The vectors are each individually tagged with a DNA oligonucleotide to allow mutant identification after selection within the host. Attenuated insertion mutants that are killed by host defense mechanisms such as exposure to SP-A, will fail to amplify from lung homogenates harvested 16 h after intratracheal inoculation (Figure 2A). Comparative STM screening using wild-type and genetically altered mice, as presented herein, represents a novel application of the STM technique. This approach allows the identification of bacterial determinants targeted by specific components of host defense, rather than the entire immune system as a whole. Thus, in contrast to conventional STM screens in wild-type mice, which yield an attenuated mutant recovery rate of 2%–10% [21], our comparative STM screen to identify factors that protect bacteria against a specific host innate immunity protein, SP-A, had a yield of only 0.12%. As STM screens, by definition, only recover mutations in non-essential genes, another reason for detecting few mutants in this screen is that these genes involved in resistance to SP-A may be essential.
In this study, we have employed STM technology to identify bacterial factors that are required for P. aeruginosa strain PA01 to confer resistance to SP-A. Both mutants were susceptible to SP-A-mediated membrane permeabilization, confirming that disruption of membrane integrity is one of the mechanisms by which SP-A mediates bacterial clearance. The degree of bacterial membrane permeabilization of the STM mutants by SP-A was comparable to that of E. coli K12, which has rough LPS [6]. Based on the results of in vitro assays, macrophages and neutrophils do not seem to play important roles in preferential clearance of the pch and ptsP mutants. However, we cannot exclude the possibility that phagocytosis contributes to the differential clearance of these two mutants in vivo.
Thus far, we have not recovered STM mutants that are susceptible to SP-A-mediated opsonization, or mutants that are disrupted in the genes required for LPS biosynthesis, which play critical roles in resistance to SP-A-mediated membrane permeability. The lack of recovery of opsonization-sensitive and LPS mutants are not particularly surprising, considering that we screened only approximately 30% of the 5,570 mutants required for 1 × genome coverage (http://www.pseudomonas.com). Additional screenings are being carried out to catalog other P. aeruginosa factors conferring resistance to SP-A.
The mechanisms by which mutations in pch and ptsP genes result in susceptibility to SP-A-mediated clearance are briefly discussed below. The pchDCBA gene encodes enzymes involved in the synthesis of iron-chelating siderophores salicylate and pyochelin [12–14]. In another closely related pathogen, Burkholderia cepacia, salicylate is a major siderophore [22]. Salicylate has also been reported to chelate magnesium (Mg2+) and calcium [23]. Thus, it is possible that P. aeruginosa requires salicylate and its downstream product pyochelin to acquire cations. Divalent cations, especially Mg2+, are required to structurally stabilize LPS on the outer membrane. The recently resolved structure of SP-A protein indicates that there are two metal binding domains within SP-A, which are most likely occupied by calcium [24]. Thus, SP-A may sequester cations from alveolar lining fluid and result in a growth disadvantage for the salicylate-deficient pch mutant. Furthermore, inability to acquire cations may potentially alter the expression, stability, and modification of various transmembrane proteins and glycolipids by enzymes that require cations as cofactors. Thus, we speculate that the pch mutation may have contributed to its increased sensitivity to SP-A-mediated membrane permeabilization through direct or indirect effects on LPS integrity and membrane stability.
The ptsP gene encodes phosphoenolpyruvate-protein phosphotransferase EINtr [15], a transcriptional regulator of RpoN-dependent operons [25]. RpoN is known to regulate the expression of variety of virulence factors in pathogenic bacteria, including the upregulation of the rfaH gene, which is required for increased production of the O-antigen of LPS in Salmonella enterica serovar Typhi during the stationary phase of growth [25]. The additional O-antigen production during the stationary phase of bacterial growth may be necessary for resistance to SP-A-mediated membrane permeabilization. PtsP also coordinates carbon and nitrogen metabolism and is required for virulence in P. aeruginosa, by an unknown mechanism [26]. Mutational inactivation of the Azotobacter vinelandii ptsP ortholog affects lipidic poly-hydroxybutyrate accumulation within its cytoplasm as a carbon source for long-term starvation survival [27]. Interestingly, poly-hydroxybutyrate is found in the plasma membranes of E. coli complexed to calcium polyphosphate, and forms divalent cation-selective channels [28]. Thus, increased sensitivity to SP-A-mediated membrane permeabilization in ptsP mutants may be caused by its reduced ability to transport divalent cations necessary to maintain LPS and outer membrane integrity.
Our results suggest that the killing of P. aeruginosa PA01 and STM mutants is more robust in vivo than in vitro. One potential explanation for this discrepancy is that our in vitro membrane permeability conditions do not accurately model the environment in the lung. In addition, it is possible that multiple factors contribute to the killing of P. aeruginosa including SP-A, SP-D, lactoferrin, β-defensins, and other antimicrobial peptides and proteins [29,30]. We are currently examining whether these peptides act synergistically or cooperatively with SP-A in clearance of P. aeruginosa.
Decreased SP-A levels have been found in several respiratory diseases including bacterial pneumonia, adult respiratory distress syndrome [31,32], and cystic fibrosis [33–35]. There is some in vitro evidence to suggest that decreased SP-A and SP-D levels in some of these disease states result from degradation by neutrophil proteases of the host [36,37], or by proteases of bacterial origin [38]. These proteases include bacterial elastases, and neutrophil-derived cathepsin G, elastase, and proteinase-3. The reduced levels of SP-A and SP-D in these lung diseases may contribute to increased susceptibility to infections by variety of microbial pathogens. Degradation of SP-A by P. aeruginosa proteases may also explain our failure to develop informative in vitro STM screens (unpublished data).
Antibiotic-resistant P. aeruginosa is an emerging clinical problem that can lead to denial for lung transplantation and death [39]. Thus, there is an added urgency to explore the use of novel, non-antibiotic-based, anti-Pseudomonal peptides to combat life-threatening infections with this organism. The discovery of factors governing resistance or susceptibility to SP-A may ultimately have therapeutic value.
Materials and Methods
Bacterial strains, media, and growth conditions.
Shaking cultures of E. coli DH5-α [40], SM17λpir [41], and P. aeruginosa strain PA01 [42] were grown in LB broth at 37 °C. When needed, LB was supplemented with 1.5% of bacto agar. Antibiotic selections were performed for E. coli or P. aeruginosa using the following concentrations: carbenicillin (50 or 300 μg/ml), kanamycin (50 or 250 μg/ml), and tetracycline (5 or 15–30 μg/ml).
Animal husbandry.
The C3H/HeN SP-A−/− mice were developed as previously described [6,7]. C3H/HeN control (SP-A+/+) mice were purchased from Charles River Laboratory (Boston, Massachusetts, United States). All infections were performed intranasally with age-matched, 6-wk-old mice. Animals were housed in positively ventilated microisolator cages with automatic recirculating water, located in a room with laminar, high-efficiency, particle accumulation-filtered air. The animals received autoclaved food, water, and bedding. Mice used in experimental procedures were handled in accordance with protocols approved by the Institutional Animal Care and Use Committee at University of Cincinnati College of Medicine.
Comparative screening of the P. aeruginosa STM library in SP-A+/+ and SP-A−/− mice.
The P. aeruginosa PA01 STM library and screening methods were previously described [11]. An adapted acute mouse lung pneumonia model of infection was employed, where adult mice rather than the infant mice were used [16]. In earlier experiments, we detected no significant difference in SP-A expression by ELISA assays at 8, 16, and 24 h in SP-A+/+ mouse lungs infected with bacteria (1 × 107 cells) from a representative pool of STM library (unpublished data). We chose the 16 h time point to screen the STM library because it gave us the most reproducible PCR amplification of DNA tags from mutants, with ≥ 10,000 colony forming units (CFU) still recoverable from the SP-A+/+ mice (unpublished data). In contrast, prolonged infection (≥ 36 h) with 1 × 107 bacteria from STM pool frequently killed SP-A−/− mice (unpublished data). Briefly, SP-A+/+ and SP-A−/− mice [6,7] were anaesthetized using isofluorane, and were intranasally inoculated with 50 μl (1 × 107 cells) of each individual STM pool. After 16 h, lungs were removed from sacrificed mice, and homogenized tissues were plated on LB agar plates. At least 10,000 bacterial colonies recovered after in vivo selection were used for multiplex PCR as described previously [11]. Twenty sets of 72 mutants (total = 1,440) were pooled and screened. An additional 240 mutants previously shown to be attenuated in a chronic model of rat lung infection [11] were also screened. The mutants that were differentially cleared by SP-A during the in vivo passaging were subsequently retested by single PCR in bacterial cells recovered from three separate mice.
In vivo competitive and single infection assays.
For competition assays, mouse lungs were inoculated intranasally with bacterial cells (1 × 107 in 50 μl) composed of a 1:1 ratio of wild-type PA01 and its isogenic STM mutants. Infected lungs were recovered 16 h after infection for bacterial load determination. The competitive index (CI) is defined as the output ratio of mutant to wild-type bacteria divided by the input ratio of mutant to wild-type bacteria [16,17]. Thus, if a mutant strain is less competitive than the parental strain from which it was derived, a CI value of < 1 will be achieved, indicating that the mutant is attenuated. Single organism inoculations with individual bacterial strains were performed by the intranasal route in SP-A+/+ and SP-A−/− mice (group of six). Attenuation was defined as the log10 difference in CFU between wild-type and mutant bacteria recovered from lung tissue 16 h after inoculation.
Mutant cloning and bioinformatics analysis.
Cloning of putative STM mutants were performed as previously described [11]. The DNA sequences were analyzed with MacVector or DNA Star software. Sequence data was compared with the database available for P. aeruginosa at http://www.pseudomonas.com.
Non-polar deletion and complementation of the ptsP mutant.
Non-polar, in-frame deletion of ptsP was generated by PCR: 0.802 and 1.155-kb 5′ and 3′ segments were amplified from target PA01 genomic DNA, and each amplicon, which included an engineered restriction site, was ligated into pEX18Ap [43] to produce replacement plasmids. Inframe deletion mutants were generated in PA01 via homologous recombination by sucrose resistance selection, and confirmed by hybridization and PCR analysis (unpublished data). To complement the ptsP mutation, a 4.5-kb DNA fragment containing both the promoter and the ygdP-ptsP-PA0338 operon (Figure 2C) was PCR-amplified from the wild-type parental strain PA01 using the Expand Long Template PCR System (Roche Diagnostic, Basel, Switzerland), and cloned into E. coli-P. aeruginosa shuttle plasmid pUCP19 [44] to obtain pUCP19-ptsP. The cloned fragment was sequenced to rule out any mutation that was introduced during the amplification. The pUCP19-ptsP was introduced into P. aeruginosa STM mutant strain ptsP by heat shock. Carbenicillin-resistant transformants were selected, and verified by Southern hybridization and PCR methods (unpublished data). Complementation was assessed by resistance to SP-A-mediated membrane permeabilization.
Complementation of the pch STM mutant.
Complementation of the pch mutant was performed by culturing the bacteria into stationary phase in LB supplemented with 1 mM sodium salicylate.
Pyochelin extraction and detection.
Pyochelin production in P. aeruginosa was analyzed using the method previously described [45]. M63 minimal medium (5 ml) was inoculated with fresh colonies of individual P. aeruginosa strains and incubated at 37 °C for 24 h with shaking. A fresh 30 ml of M63 supplemented with 0.5% w/v casamino acids was then inoculated with 0.3 ml of starter culture, and incubated for 36 h at 37 °C with shaking. The culture was then centrifuged at 13,000 rpm for 10 min in 1.5 ml aliquots in a microcentrifuge. The supernatants were removed into a 50 ml polypropylene screw-capped centrifuge tube, pooled, adjusted to (pH 2.0) with 1M HCl, using universal indicator paper to estimate pH, and filter sterilized through a 0.22 μM membrane. Pyochelin was extracted by the addition of 0.4 volumes of ethyl acetate, followed by vigorous vortexing. The two phases were separated by centrifugation at 2,000 rpm for 5 min in a bench top centrifuge, and the upper, organic phase was removed to a separate tube. Pyochelin extract was concentrated by rotary evaporation at 40 °C in 100 ml round bottomed flasks, and the residue was dissolved in 300 μl of methanol. Pyochelin extracts (10 μl) was separated by TLC (silica gel 60, plastic backed, from Sigma, St. Louis, Missouri, United States), using acetone:methanol:0.2M acetic acid (5:2:1) as a development solvent. The chromatography tank was left to stand for several minutes with its lid on to allow the air inside to become saturated with solvent vapor. Pyochelin, which is fluorescent, naturally occurs as two stereoisomers. When visualized under a UV transilluminator, it is revealed as two green fluorescent bands. Pyochelin standard was generously provided by Dr. J. M. Meyer (Université Louis Pasteur).
Purification of human SP-A.
Human SP-A (hSP-A) was purified from lung washings of patients with the lung disease alveolar proteinosis, as previously described [6]. SP-A preparations were re-suspended in 5 mM Tris, 150 mM NaCl. SP-A preparations were determined to contain 140–190 pg of LPS/μg of SP-A by the Limulus Amebocyte Lysate (BioWhittaker, Rockland, Maryland, United States).
Assays of P. aeruginosa permeability.
The effect of the SP-A on P. aeruginosa cell wall integrity was assessed by determining permeability to a phosphatase activity substrate, Enzyme-Labeled Fluorescence 97 (ELF-97, Molecular Probes, Eugene, Oregon, United States) as described [7]. SP-A was incubated with 1 × 108 bacterial cells/ml in 100 μl of 5 mM Tris, 150 mM NaCl for 15 min at 37 °C, and 100 μM ELF97 phosphatase substrate was added. Fluorescence was measured at excitation and emission wavelengths of 355 and 535 nm, respectively, for a period of 60–90 min.
In vitro macrophage and neutrophil phagocytosis assays.
Freshly lavaged mouse macrophages (~ 5 × 105 from 3 SPA+/+ mice) were cultured on chamber plastic culture slides, coated with 5% poly-D-lysine with 200 μl RPMI (Dulbecco's, containing 2.5 mg/L gentamycin and 0.1% BSA) and used immediately. Macrophages were allowed to adhere for 2 h at 37 °C in 5% CO2. Live GFP-expressing PA01 or isogenic STM mutants in RPMI were opsonized with or without 25 μg/ml human SP-A for 1 h at 37 °C with rotation. The medium was removed from each well and replaced with 250 μl of opsonized PA01 or STM mutants, and incubated for 1 h at 37 °C in 5% CO2 at a ratio of 100 bacteria to 1 macrophage. Chamber slides were washed three times with PBS containing 1 mM CaCl2. Extracellular bacteria were quenched with crystal violet (0.8 mg/ml). Following two additional washes, cells were fixed with 1% paraformaldehyde in PBS plus 1 mM CaCl2 for 10 min, and stained with Evans blue for 2 min. The percentage of macrophages with engulfed bacteria was quantified under a phase contrast fluorescence microscope. At least 200 macrophages were counted. Human neutrophils were purified and cultured as previously described [46]. Neutrophil phagocytosis was performed as described for macrophages.
LIVE/DEAD bacterial staining.
The viability of SP-A exposed bacteria was determined by LIVE/DEAD BaclightTM Bacterial Viability Kit (Molecular Probes). Briefly, stationary phase E. coli or P. aeruginosa bacteria were washed three times with 5 mM Tris, 150 mM NaCl, and adjusted to an of OD600 = 1.0. A 100-μl aliquot of the cells was added to culture tubes containing 100 μl of SP-A (100 μg/ml) or 100 μl of 5 mM Tris, 150 mM NaCl, as a control. The mixtures were incubated at 37 °C, 300 rpm for 1 h, and stained according to the instructions provided by the supplier. The viability of the cells was checked under a fluorescence microscope. At least 1, 000 cells were counted.
LPS preparations.
LPS was isolated from P. aeruginosa by acetone/phenol extraction [47]. LPS samples were size-fractionated on 4%–12% Bis-Tris PAGE gels (NuPAGE Novex Gels; Invitrogen, San Diego, California, United States) and stained with silver.
Statistical analysis.
Statistical analysis was performed using the Student's t-test and one-way analyses of variance (ANOVA). A significant difference was considered to be p < 0.05.
This work was supported by grants from the NIH (1R03AI057915), an American Lung Association research grant (RG-131-N), and a University of Cincinnati College of Medicine Dean Discovery Fund to GWL; and the NIH (HL-61612) and the Department of Veterans Affairs (Merit Award) to FXM. Work on STM in R. C. Levesque's laboratory is funded by the Canadian Institute for Health Research as part of the CIHR Genomics Program. R. C. Levesque is also funded by a Research Scholar of Exceptional Merit from Le Fonds de la Recherche en Santé du Québec. E. Potvin obtained a studentship from CIHR. We thank Drs. Jeff Whitsett and Tom Korfhagen for the gift of Swiss Black SP-A−/−mice. We thank Dr. C. Reimmann for the gifts of the PA06331 and pyochelin purification methods, Dr. J.M. Meyer (Laboratoire de Microbiologie et Génétique, Université Louis Pasteur) for her generous supply of purified pyochelin, and Dr. Herbert Budzikiewicz (Institut fuer Organische Chemie, Koeln) for his generous advice on pyochelin TLC and purification.
Competing interests. The authors have declared that no competing interests exist.
Author contributions. SZ, YC, FXM, and GWL conceived and designed the experiments. SZ, YC, FS, and GWL performed the experiments. SZ, YC, FS, FXM, and GWL analyzed the data. EP, FS, RCL, FXM, and GWL contributed reagents/materials/analysis tools. GWL wrote the paper.
Abbreviations
CFUcolony forming unit
CIcompetitive index
GFPgreen fluorescence protein
LBLuria broth
LPSlipopolysaccharide
RFLPrestriction fragment length polymorphism
SP-Asurfactant protein A
SP-Dsurfactant protein D
STMsignature-tagged mutagenesis
TLCthin layer chromatography
==== Refs
References
McCormack FX Whitsett JA 2002 The pulmonary collectins, SP-A and SP-D, orchestrate innate immunity in the lung J Clin Invest 109 707 712 11901176
Wright JR 2003 Pulmonary surfactant: A front line of lung host defense J Clin Invest 111 1453 1455 12750392
Crouch E Wright JR 2001 Surfactant proteins A and D and pulmonary host defense Annu Rev Physiol 63 521 554 11181966
LeVine AM Whitsett JA 2001 Pulmonary collectins and innate host defense of the lung Microbes Infect 3 161 166 11251302
Kuhlman M Joiner K Ezekowitz RA 1989 The human mannose-binding protein functions as an opsonin J Exp Med 169 1733 1745 2469767
Wu H Kuzmenko A Wan S Schaffer L Weiss A 2003 Surfactant proteins A and D inhibit the growth of Gram-negative bacteria by increasing membrane permeability J Clin Invest 111 1589 1602 12750409
McCormack FX Gibbons R Ward SR Kuzmenko A Wu H 2003 Macrophage-independent fungicidal action of the pulmonary collectins J Biol Chem 278 36250 36256 12857753
Piboonpocanun S Chiba H Mitsuzawa H Martin W Murphy RC 2005 Surfactant protein A binds Mycoplasma pneumoniae with high affinity and attenuates its growth by recognition of disaturated phosphatidylglycerols J Biol Chem 280 9 17 15498759
Hensel M Shea JE Gleeson C Jones MD Dalton E 1995 Simultaneous identification of bacterial virulence genes by negative selection Science 269 400 403 7618105
LeVine AM Kurak KE Bruno MD Stark JM Whitsett JA 1998 Surfactant protein-A-deficient mice are susceptible to Pseudomonas aeruginosa infection Am J Respir Cell Mol Biol 19 700 708 9761768
Potvin E Lehoux D Kukavica-Ibrujl I Richard K Sanschagrin F 2003 In vivo functional genomics of Pseudomonas aeruginosa for high-throughput screening of new virulence factors and antibacterial targets Environ Microbiol 5 1294 1308 14641575
Gaille C Reimann C Haas D 2003 Isochorismate synthase (PchA), the first and rate-limiting enzyme in salicylate biosynthesis of Pseudomonas aeruginosa
J Biol Chem 278 16893 16898 12624097
Cox CD Rinehart KL Jr Moore ML Cook JC Jr 1981 Pyochelin: Novel structure of an iron-chelating growth promoter for Pseudomonas aeruginosa
Proc Natl Acad Sci USA 78 4256 4260 6794030
Reimmann C Patel HM Serino L Barone M Walsh CT 2001 Essential PchG-dependent reduction in pyochelin biosynthesis of Pseudomonas aeruginosa
J Bacteriol 183 813 820 11208777
Reizer J Reizer A Merrick MJ Plunkett G 3rd Rose DJ 1996 Novel phosphotransferase-encoding genes revealed by analysis of the Escherichia coli genome: A chimeric gene encoding an Enzyme I homologue that possesses a putative sensory transduction domain Gene 181 103 108 8973315
Lau GW Ran H Kong F Hassett DJ Mavrodi D 2004
Pseudomonas aeruginosa pyocyanin is critical for lung infection in mice Infect Immun 72 4275 4278 15213173
Chiang SL Mekalanos JJ 1998 Use of signature-tagged transposon mutagenesis to identify Vibrio cholerae genes critical for colonization Mol Microbiol 27 797 805 9515705
Mariencheck WI Savov J Dong Q Tino MJ Wright JR 1999 Surfactant protein A enhances alveolar macrophage phagocytosis of a live, mucoid strain of P. aeruginosa
Am J Physiol 277 777 786
Schaeffer LM McCormack FX Wu H Weiss AA 2004
Bordetella pertussis lipopolysaccharide resists the bactericidal effects of pulmonary surfactant protein A J Immunol 173 1959 1965 15265930
Rocchetta HL Burrows LL Lam JS 1999 Genetics of O-antigen biosynthesis in Pseudomonas aeruginosa
Microbiol Mol Biol Rev 63 523 553 10477307
Mecsas J 2002 Use of signature-tagged mutagenesis in pathogenesis studies Curr Opin Microbiol 5 33 37 11834366
Nair BM Cheung KJ Jr Griffith A Burns JL 2004 Salicylate induces an antibiotic efflux pump in Burkholderia cepacia complex genomovar III (B. cenocepacia)
J Clin Invest 113 464 473 14755343
Domenico P Schwartz S Cunha BA 1989 Reduction of capsular polysaccharide production in Klebsiella pneumoniae by sodium salicylate Infect Immun 57 3778 3782 2680983
Head JF Mealy TR McCormack FX Seaton BA 2003 Crystal structure of trimeric carbohydrate recognition and neck domains of surfactant protein A J Biol Chem 278 43254 43260 12913002
Bittner M Saldias S Altamirano F Valvano MA Contreras I 2004 RpoS and RpoN are involved in the growth-dependent regulation of rfaH transcription and O antigen expression in Salmonella enterica serovar Typhi
Microb Pathog 36 19 24 14643636
Tan MW Mahajan-Miklos S Ausubel FM 1999 Killing of Caenorhabditis elegans by Pseudomonas aeruginosa used to model mammalian bacterial pathogenesis Proc Natl Acad Sci USA 96 715 720 9892699
Segura D Espin G 1998 Mutational inactivation of a gene homologous to Escherichia coli ptsP affects poly-hydroxybutyrate accumulation and nitrogen fixation in Azotobacter vinelandii
J Bacteriol 180 4790 4798 9733679
Reusch RN 1999 Polyphosphate/poly-(R)-3-hydroxybutyrate ion channels in cell membranes Prog Mol Subcell Biol 23 151 182 10448676
Singh PK Tack BF McCray PB Jr Welsh MJ 2000 Synergistic and additive killing by antimicrobial factors found in human airway surface liquid Am J Physiol Lung Cell Mol Physiol 279 799 805
Yan H Hancock RE 2001 Synergistic interactions between mammalian antimicrobial defense peptides Antimicrob Agents Chemother 45 1558 1560 11302828
Baughman RP Sternberg RI Hull W Buchsbaum JA Whitsett J 1993 Decreased surfactant protein A in patients with bacterial pneumonia Am Rev Respir Dis 147 653 657 8442601
LeVine AM Lotze A Stanley S Stroud C O'Donnell R 1996 Surfactant content in children with inflammatory lung disease Crit Care Med 24 1062 1067 8681574
Griese M Birrer P Demirsoy A 1997 Pulmonary surfactant in cystic fibrosis Eur Respir J 10 1983 1988 9311489
Postle AD Mander A Reid KB Wang JY Wright SM 1999 Deficient hydrophilic lung surfactant proteins A and D with normal surfactant phospholipid molecular species in cystic fibrosis Am J Respir Cell Mol Biol 20 90 98 9870921
Noah TL Murphy PC Alink JJ Leigh MW Hull WM 2003 Bronchoalveolar lavage fluid surfactant protein-A and surfactant protein-D are inversely related to inflammation in early cystic fibrosis Am J Respir Crit Care Med 168 685 691 12829455
von Bredow C Wiesener A Griese M 2003 Proteolysis of surfactant protein D by cystic fibrosis relevant proteases Lung 181 79 88 12953146
Rubio F Cooley J Accurso FJ Remold-O'Donnell E 2004 Linkage of neutrophil serine proteases and decreased surfactant protein-A (SP-A) levels in inflammatory lung disease Thorax 59 318 323 15047952
Mariencheck WI Alcorn JF Palmer SM Wright JR 2003
Pseudomonas aeruginosa elastase degrades surfactant proteins A and D Am J Respir Cell Mol Biol 28 528 537 12654643
Conway SP Brownlee KG Denton M Peckham DG 2003 Antibiotic treatment of multidrug-resistant organisms in cystic fibrosis Am J Respir Med 2 321 332 14719998
Miller VL Mekalanos JJ 1988 A novel suicide vector and its use in construction of insertion mutations: Osmoregulation of outer membrane proteins and virulence determinants in Vibrio cholerae requires toxR
J Bacteriol 170 2575 2583 2836362
Taylor RK Manoil C Mekalanos JJ 1989 Broad-host range vectors for delivery of TnphoA: Use in genetic analysis of secreted virulence determinants of Vibrio cholerae
J Bacteriol 171 1870 1878 2539354
Holloway BW Krishnapillai V Morgan AF 1979 Chromosomal genetics of Pseudomonas
Microbiol Rev 43 73 102 111024
Hoang TT Karkhoff-Schweizer RR Kutchma AJ Schweizer HP 1998 A broad-host-range Flp-FRT recombination system for site-specific excision of chromosomally-located DNA sequences: Application for isolation of unmarked Pseudomonas aeruginosa mutants Gene 212 77 86 9661666
Schweizer HP 1991
Escherichia-Pseudomonas shuttle vectors derived from pUC18/19 Gene 97 109 121 1899844
Serino L Reimmann C Baur H Beyeler M Visca P 1995 Structural genes for salicylate biosynthesis from chorismate in Pseudomonas aeruginosa
Mol Gen Genet 249 217 228 7500944
Newman SL Gootee L Gabay JE 1993 Human neutrophil-mediated fungistasis against Histoplasma capsulatum . Localization of fungistatic activity to the azurophil granules J Clin Invest 92 624 631 8349801
Westphal O Jann K 1965 Methods in carbohydrate chemistry. New York Academic Press Volume 5, 83 91
|
16322768
|
PMC1291351
|
CC BY
|
2021-01-05 12:11:25
|
no
|
PLoS Pathog. 2005 Nov 25; 1(3):e31
|
utf-8
|
PLoS Pathog
| 2,005 |
10.1371/journal.ppat.0010031
|
oa_comm
|
==== Front
PLoS PathogPLoS PathogppatplpaplospathPLoS Pathogens1553-73661553-7374Public Library of Science San Francisco, USA 10.1371/journal.ppat.001003305-PLPA-RA-0010R3plpa-01-03-09Research ArticleCell BiologyImmunologyInfectious DiseasesMicrobiologyRespiratory MedicineSystems BiologyGenetics/Gene FunctionGenetics/Functional GenomicsEubacteriaMycobacterial Mutants with Defective Control of Phagosomal Acidification Mycobacterial Control of Phagosome pHStewart Graham R 12*Patel Janisha 1Robertson Brian D 1Rae Aaron 1Young Douglas B 1
1 Department of Infectious Diseases and Microbiology, Centre for Molecular Microbiology and Infection, Imperial College London, London, United Kingdom
2 School of Biomedical and Molecular Sciences, University of Surrey, Surrey, United Kingdom
Ramakrishnan Lalita EditorUniversity of Washington, United States of America*To whom correspondence should be addressed. E-mail: [email protected] 2005 25 11 2005 1 3 e333 3 2005 18 10 2005 Copyright: © 2005 Stewart et al.2005This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are properly credited.The pathogenesis of mycobacterial infection is associated with an ability to interfere with maturation of the phagosomal compartment after ingestion by macrophages. Identification of the mycobacterial components that contribute to this phenomenon will allow rational design of novel approaches to the treatment and prevention of tuberculosis. Microarray-based screening of a transposon library was used to identify mutations that influence the fate of Mycobacterium bovis bacille Calmette-Guérin (BCG) following uptake by macrophages. A screen based on bacterial survival during a 3-d infection highlighted genes previously implicated in growth of Mycobacterium tuberculosis in macrophages and in mice, together with a number of other virulence genes including a locus encoding virulence-associated membrane proteins and a series of transporter molecules. A second screen based on separation of acidified and non-acidified phagosomes by flow cytometry identified genes involved in mycobacterial control of early acidification. This included the KefB potassium/proton antiport. Mutants unable to control early acidification were significantly attenuated for growth during 6-d infections of macrophages. Early acidification of the phagosome is associated with reduced survival of BCG in macrophages. A strong correlation exists between genes required for intracellular survival of BCG and those required for growth of M. tuberculosis in mice. In contrast, very little correlation exists between genes required for intracellular survival of BCG and those that are up-regulated during intracellular adaptation of M. tuberculosis. This study has identified targets for interventions to promote immune clearance of tuberculosis infection. The screening technologies demonstrated in this study will be useful to the study of pathogenesis in many other intracellular microorganisms.
Synopsis
The pathogenesis of Mycobacterium tuberculosis relies on an ability to survive inside host macrophages. Macrophages kill most other bacteria by engulfment into an intracellular compartment called a phagosome, which quickly matures to an acidic, hydrolytic organelle, resulting in bacterial death. M. tuberculosis and the related vaccine strain M. bovis bacille Calmette-Guérin (BCG) possess the ability to stop phagosome maturation and thus avoid its microbicidal properties. In this study, the researchers screened a library of mutant BCG bacteria to identify the bacterial genes responsible for preventing phagosome acidification. The predicted products of these genes span many different functional groups, but tend to be associated with the outside of the cell or secreted to the extracellular milieu. The researchers also demonstrated that mutant mycobacteria whose phagosomes acidify are unable to replicate in macrophages. This study identifies targets for new vaccines against tuberculosis.
Citation:Stewart GR, Patel J, Robertson BD, Rae A, Young DB (2005) Mycobacterial mutants with defective control of phagosomal acidification. PLoS Pathog 1(3): e33.
==== Body
Introduction
Mycobacterium tuberculosis presents a major challenge to global health, claiming around two million lives every year. The success of this pernicious pathogen centres around an ability to avoid destruction by the immune response throughout the course of an infection that may last the lifetime of the host. To achieve this, the bacilli have adapted to survive and replicate inside the cells of the host, principally within the very immune cell that is designed to destroy invading microorganisms, the macrophage. Multiple factors are involved in intracellular survival [1]. A degree of inherent resistance to killing is conferred by the mycobacterial cell wall, which presents a relatively impervious physical barrier to the hydrolytic enzymes encountered within macrophages, and by mycobacterial enzymes that detoxify reactive radicals. To avoid activation of macrophages by T cells, M. tuberculosis interferes with dendritic cell function [2,3], antigen presentation [4,5], and cytokine signalling by infected cells [6]. But perhaps the most striking feature of pathogenic mycobacteria is an ability to arrest the normal process of phagosome maturation, allowing survival in a non-acidified intracellular compartment [1].
Normal phagosomes undergo a process of maturation that is driven by continual fusion and fission events with other endosomal compartments [7], resulting in acquisition of late endosome/lysosome-associated proteins via a biosynthetic trafficking pathway from the trans-Golgi network (TGN) [1,8]. Early phagosomes share many properties with early endosomes and are characterised by markers such as the small GTPase Rab5 [9] and its effector, early endosomal antigen 1 [10]. Within minutes of formation, however, phagosomes begin to accumulate other markers corresponding to those of late endosomes. For example, while early endosomal antigen 1 and Rab5 are lost, there is a gradual accumulation of another GTPase, Rab7 [11], which is involved in fusion of late endosomes and lysosomes [12]. Concurrent with these changes in membrane markers and fusogenicity of the phagosome are changes in the physiology of the phagosomal lumen, including its acidification from a pH of 5.5 in late phagosomes to 4.5 in phagolysosomes. This occurs by the accumulation of the vacuolar proton ATPase (V-ATPase). The V-ATPase may be sourced from the TGN [13] and from fusion with other compartments, particularly V-ATPase-rich lysosomes.
Slow-growing mycobacteria—such as M. tuberculosis, M. avium, and the attenuated bacille Calmette-Guérin (BCG) vaccine strain of M. bovis—interfere with this process, occupying an atypical phagosome arrested at the early phagosome stage [14–16]. The mycobacterial phagosome retains Rab5 [17–19] and maintains fusogenicity with early endosomes in the rapid recycling pathway, as evidenced by its access to transferrin [20,21]. However, it does not fuse with late endosomes/lysosomes or accumulate associated markers. In addition, it does not accumulate V-ATPase, and it fails to acidify, remaining at pH 6.3 [22]. Phagosomal arrest is specific to the live organism; phagosomes occupied by killed mycobacteria rapidly acidify and fuse with lysosomes [1].
Elucidation of the molecular mechanisms by which mycobacteria arrest phagosome maturation may uncover novel intervention targets for disease control. It is clear that the surface properties of the organism have an important influence on this process, and several mycobacterial lipids and glycolipids have been implicated in altered phagosome biogenesis [13,23–25]. The involvement of mycobacterial viability in efficient phagosomal arrest suggests that such structural effector mechanisms are complemented by active interference in macrophage cell biology. Recent reports have described a serine-threonine kinase, PknG, of pathogenic mycobacteria that could modify host proteins involved in control of intracellular trafficking pathways, for example [26], and exclusion of phosphatidylinositol-3-phosphate from phagosomes containing live mycobacteria [27]. In a study based on screening of mycobacterial mutants, Pethe and colleagues identified multiple genes required for phagosomal arrest [28], including those involved in biosynthesis of surface lipids, transport of possible effector molecules, and production of isoprenoid compounds. Taken together, these studies suggest that phagosomal arrest represents a complex multigenic phenotype.
The combination of whole-genome transposon mutagenesis with microarray-based, high-throughput analysis presents an attractive approach to study such multicomponent systems. In two seminal papers, Sassetti and colleagues have used this approach to provide an overview of mycobacterial genes required for optimal growth in laboratory culture and in a mouse infection model [29,30]. The aim of the present study was to use an analogous approach to investigate phagosomal arrest. We analysed a transposon library in M. bovis BCG using biological screens that were designed to identify mutants that were defective (a) in intracellular survival, and (b) in prevention of early phagosomal acidification. We anticipated that this would allow us to identify novel individual genes involved in each of these processes and, by comparing the two datasets, to determine whether the two phenotypes are functionally linked.
Results
Transposon Screen by Microarray
A derivative of the Epicentre EZ::TN transposon carrying a hygromycin-resistance determinant (EZ::TNhyg ) was used to generate a transposon library in M. bovis BCG. Sequencing of insertion sites within specific genes revealed 9-bp direct repeats flanking the transposon, thus indicating genuine Tn5 transposition [31]. Detailed analysis of 30 insertions revealed some site preference, with position 1 of the direct repeat biased to G and position 9 to C in greater than 60% of events (Table S1). In GC-rich genomes such as BCG and M. tuberculosis, this bias may represent an advantage that makes EZ:TN a useful addition to Himar1 [29,30] and IS1086-based transposons [32–34]. Pooled mutants were analysed by hybridisation of fluorescently labelled transposon insertion sites to whole-genome microarrays. We evaluated the reproducibility of the transposon screen by microarray (TSM) labelling and hybridisation process by comparing fluorescent intensities of hybridising spots derived from two independent labelling experiments performed on the same 2,500-clone library (Figure S1A). A highly significant correlation between the different experiments (r
2 = 0.9289) demonstrated the reproducibility of the method. The microarray readout of insertion site locations within the transposon pool provides an ideal method for examining the genome-wide distribution of transposition, and showed that the EZ::TNhyg transposon was generally well distributed throughout the entire genome (Figure S1B). We have not observed any occurrences of multiple transposon insertions within a single cell. We next examined the effect on the pool of exposure to biological selections based on intracellular survival in macrophages and on phagosomal acidification (Figure 1).
Figure 1 High-Throughput Screening for Mycobacterial Mutants with Reduced Survival in Macrophage Culture, or Defective Ability to Arrest Phagosome Acidification
For the survival screen, the BCG transposon library was selected through three rounds of macrophage culture (72 h each) and survival of mutants assessed by fluorescent labelling of transposon insertion sites in input and output mutant pools and competitive hybridisation to a microarray. To screen for mutants unable to inhibit phagosome acidification, the transposon mutants were FITC-labelled and used to infect macrophages in the presence of LysoTracker for 1 h. Organelles were prepared and acidic phagosomes sorted by flow cytometry. Mutants in the acidic phagosomes were grown in broth and used for a second round of selection. The identity and abundance of mutants recovered from acidic and non-acidic phagosomes were compared by fluorescent labelling of transposon insertion sites and hybridisation to a microarray.
Transposon Screen to Quantify Intracellular Fitness
TSM was used to identify insertions associated with decreased intracellular fitness by comparing transposon pools before and after three rounds of 72-h culture in macrophages. Results of four replicate screens from each of two independent experiments with the same input pool were combined to generate a mean fitness ratio (FR) for more than 1,500 genes, representing approximately 60% of approximately 2,600 non-essential genes in the BCG genome; this information is provided together with an ordered list of the top 100 attenuating insertions in Tables S2 and S3. The 20 most strongly attenuating insertions are shown in Table 1. As the BCG genome has yet to be annotated and most existing datasets relate to nomenclature based on the M. tuberculosis H37Rv sequence, we have used “Rv” numbers to identify genes throughout this study. We screened this list initially for the presence of genes previously shown to be required for survival of mycobacteria in macrophages. Insertions with a high fitness cost include those in a known virulence locus involved in synthesis and transport of phthiocerol dimycocerosate (DIM), a complex lipid component of the mycobacterial cell wall [32,33], and in the mycobacterial cell entry 1 (mce1) region (Rv0169–Rv0178), which has been implicated in mycobacterial entry to host cells [35–37] and is required for virulence in a mouse model (Figure 2) [30]. Reduced fitness (FR < 0.5; false discovery rate [FDR] = 1%) was also associated with a series of insertions in the pathway for conversion of isocitrate to glucose (Table S3)—isocitrate lyase (icl, Rv0467), PEP carboxykinase (pcka, Rv0211) and fructose-1,6-bisphosphatase (Rv1099c)—consistent with previous proposals that mycobacteria rely predominantly on lipid substrates for intracellular metabolism [38]. The prominence of previously described attenuating mutations provides a validation of the effectiveness of the intracellular fitness screen.
Table 1 Mutants That Incurred the Highest Fitness Cost in Macrophage Culture
Figure 2 Gene Regions with Low Intracellular Fitness Ratios
The identification of known virulence regions such as the mce1 operon and the DIM synthesis and transport region, validate the TSM intracellular fitness assay. The strategy also identified virulence regions such as VAMP. Region coordinates derived from M. tuberculosis H37Rv are shown. FRs (calculated by dividing output signal intensity by input signal intensity) are shown inside horizontal grey arrows representing each open reading frame (open reading frames without ratios did not generate hybridisation signals). CHP, conserved hypothetical protein; HP, hypothetical protein; MP, membrane protein; TMP, transmembrane protein.
We next searched the data for other operons or regions that contained multiple transposon hits with low intracellular fitness scores. The four spots representing Rv0199–Rv0202c had significantly reduced fitness ratios (FR = 0.19–0.55; FDR < 2%), with Rv0200 and Rv0201c amongst the most highly attenuated mutants (Tables 1, S2, and
S3). Insertional mutation of several genes in the region from Rv0199 through Rv0208c has been reported to confer a significant growth defect in mice [30], and Rv0204c was picked out as required for optimal in vivo growth in a previous signature-tagged mutagenesis screen [32]. Due to the abundance of genes encoding membrane-associated proteins in this region, we have termed it the virulence-associated membrane protein region, or VAMP (Figure 2). The region is one example of several that highlight a central requirement for export of macromolecules during mycobacterial growth in macrophages. Rv0199 and Rv0200 encode transmembrane proteins related to Rv1972/1973 and Rv0177/0178 present in the mce3 and mce1 loci, respectively. Rv0202c and Rv0206c encode members of the RND superfamily, referred to as MmpL11 and MmpL3 and predicted to act as lipid transporters. Rv0204c and Rv0205 encode conserved transmembrane proteins, predicted in the latter case to function as a permease. Rv0203 encodes an exported protein, the secretion signal of which contains a characteristic motif with adjacent twin arginine residues, that is associated with proteins that are exported by the Sec-independent Tat pathway [39]. Tat-exported proteins have been implicated in the pathogenesis of several bacteria; homologues of the tatA and tatC genes are present in the genome of M. tuberculosis, but the function of this secretion pathway has yet to be investigated in mycobacteria. Further examples of attenuating lesions associated with predicted secreted proteins include insertions in adjacent genes encoding members of the ESAT-6 family (Rv3444c and Rv3445c) [40,41]. The insertion with the second highest fitness costs mapped to Rv3812, a gene encoding a member of the PE PGRS family of predicted cell surface proteins that has been linked to chronic infection by M. marinum [42].
Small molecule transporters were also prominent amongst the most attenuating mutations. These include a dipeptide transporter (dppC, Rv3664c) that also displays a marked phenotype in the mouse virulence model [30], two genes involved in uptake of asparagine (ans, Rv2127 and aroP2, Rv0346c), and transporters for sugars (sugA, Rv1236), ribonucleotides (mkl, Rv0655), and sialic acid (nanT, Rv3294). Other transporters associated with virulence impairments are predicted to mediate uptake or export of cations (ctpG, Rv1992c and mgtE, Rv0362) and drugs (Rv1272c and Rv1410c).
Transposon Screen for Mycobacterial Genes Involved in Control of Phagosome Acidification
To screen for mutations affecting phagosomal acidification, we developed flow cytometry techniques for analysis and sorting of these organelles. Macrophages infected for 1 h with fluorescence-tagged BCG were lysed and, after disruption of the cytoskeleton and removal of nuclei, organelles were analysed by flow cytometry. The fluorescent tag, combined with forward- and side-scatter profiles, provided a marker to identify the mycobacteria-containing compartments (Figure 3A). To verify that these fluorescent events represented intact mycobacterial phagosomes, we sorted the gated particle population into fixative and confirmed by electron microscopy that it contained large numbers of mycobacteria enclosed by unbroken membranes (Figure 3B). Phagosomes were further sorted into acidified and non-acidified compartments using the fluorescent dye LysoTracker DND99, which accumulates in acidic compartments and has previously been successfully exploited to monitor acidification of the mycobacterial phagosome by confocal microscopy [18]. We used flow cytometry to compare the proportion of LysoTracker -positive phagosomes obtained after infection with live and dead BCG. Approximately 26% (standard deviation [SD] = 8.6%) of live BCG phagosomes were LysoTracker-positive compared to 62% (SD = 6.7%) of phagosomes containing dead bacteria (Figure 3C). This is consistent with previous reports that live mycobacteria inhibit acidification of the phagosome and validated the use of flow cytometric detection of LysoTracker accumulation in phagosomes as a method to differentiate bacteria by their ability to arrest acidification.
Figure 3 Fluorescence-Activated Cell Sorting of Mycobacterial Phagosomes
(A) Forward and side scatter plot of organelle preparation from BCG-infected J774 macrophages. Gated area denotes the population which includes mycobacterial phagosomes, based on the signal from FITC-labelled mycobacteria.
(B) Using fluorescently labelled mycobacteria, phagosomes were sorted into fixative for confirmation of identity and integrity. Fluorescent events in the gated population predominantly consisted of mycobacterial phagosomes with bacteria surrounded by intact membranes.
(C) Flow cytometric analysis of acidification of mycobacterial phagosomes using the acidotropic dye LysoTracker DND99. Significantly more UV-killed BCG localised to acid compartments at 1 h post-infection (p < 0.01).
We infected macrophages with a pool of 2,500 BCG EZ::TNhyg mutants and used flow cytometry to select the LysoTracker-positive phagosome population that we anticipated would be enriched for mutants unable to arrest acidification. This was confirmed by the observation that the proportion of acidified phagosomes increased from 35% during the first flow-cytometric selection to 51% in a second, subsequent round of enrichment. A final pool of mutants from the two rounds of enrichment was compared to the pool of mutants present in the LysoTracker-negative phagosomes by TSM (see Figure 1). Fold enrichment in the “acidified” pool was calculated as the ratio of signal intensity in the LysoTracker-positive pool over signal intensity in the LysoTracker-negative pool. Spots with a high fold enrichment indicated genes in or near which a transposon insertion conferred the greatest reduction in ability to control phagosome acidification.
Results from four replicate screens from two independent experiments were combined to generate the mean fold enrichment. Data were considered statistically significant when the FDR was calculated to be less than 5%. Insertion in 43 genetic regions had a significant effect on phagosome acidification (Table S4); the top ten insertions (FDR < 2%) are shown in Table 2. Sequence analysis showed that eight of the ten insertions had occurred within the gene indicated by the microarray, one was located just within the 5′ end of an adjacent gene, and one could not be amplified and sequenced (Table 2). We thus consider that the hybridisation pattern revealed by the TSM technique relates well to the actual mutagenised genes.
Table 2 The Ten Mutants Most Significantly Enriched in Acidified Phagosomes
To confirm by an alternative technique that the screen had correctly identified mutations with a reduced ability to inhibit phagosome acidification, we isolated four transposon mutants from the acidification-enriched pool (Tn::Rv3707c, Tn::PPE10, Tn::cut2, and Tn::glyA1). Macrophages infected with the individual mutants were examined by confocal microscopy. After a 1-h infection, all four mutant strains co-localised with LysoTracker at a significantly increased frequency compared to wild-type (Table 3), thus further confirming the validity of the flow cytometric selection procedure.
Table 3 Confocal Microscopy Analysis of FACS-Selected Mutants Confirms Increased Localisation in Acidic Phagosomes
The top ten genes (see Table 2) span a range of predicted functional classes, but it is notable that eight share the property of encoding proteins associated either with the outside of the cell or with the extracellular milieu. These include three predicted surface proteins of unknown function: a member of the PPE family (PPE10), a lipoprotein (LppN), and a conserved hypothetical protein (Rv3707c). The list includes two secreted enzymes: cut2, encoding a member of a family of serine esterases with significant similarity to fungal cutinases, and glyA1, a serine hydroxymethyltransferase that is found in the culture filtrate of M. tuberculosis [43]. Insertion in kefB provides the most obvious connection between gene function and altered phagosomal pH. KefB is a potassium efflux system that functions via potassium/proton antiport [44,45] and in Escherichia coli acts to acidify the bacterial cytoplasm to protect the cell from electrophile toxicity [46,47]. For a bacterium in the phagosome, the release of potassium from the cytoplasm via KefB would be compensated by uptake of protons from the phagosomal lumen, resulting in an increase in lumenal pH. A mycobacterial mutant lacking kefB would thus have a reduced ability to actively alter cation proton balance in its own cytoplasm and consequently also in the phagosomal lumen.
Intracellular Fitness of Mutants Whose Phagosomes Acidify
While each of the biological screens confirmed and expanded our understanding of the biology of mycobacteria-macrophage interactions, we were surprised to observe little overlap between the two datasets. Only two (Rv2506 and Rv3057) of the 43 insertions identified on the basis of their failure to control phagosome acidification (Table S4) were amongst the top 100 insertions that compromised intracellular growth over the 72-h infection (see Table S3). Indeed, the overall distribution of intracellular fitness values for the acidification-enriched mutants was not significantly different from that of the whole transposon library (p < 0.38, t-test) (Figure S2). In light of this result, we examined the intracellular fitness of the individual Tn::Rv3707c, Tn::PPE10, Tn::cut2, and Tn::glyA1 “acid phagosome” mutants during 6-d (144-h) macrophage infections. All four mutants were significantly attenuated (p < 0.05) for survival, but this was most pronounced after 4 d and 6 d of infection (Figure 4).
Figure 4 Reduced Intracellular Fitness of Mutants That Failed to Arrest Phagosome Acidification
Mutants that failed to inhibit the acidification of their phagosomes were assessed for survival over 6 d in J774 macrophages. BCGPasteur (solid squares), Tn::Rv3707c (open squares), Tn::PPE10 (closed circles), Tn::cut2 (solid triangles), Tn::glyA1 (open circles). Error bars indicate ± 1 SD.
Discussion
The EZ::TN system provides a tool that is complementary to IS1096-based transposons and Himar1 [29,30,32–34] for generation of transposon libraries in slow-growing mycobacteria. Application of a simple PCR amplification allows it to be used in combination with high-thoughput microarray screening, bypassing the need for construction of a customised transcription-based readout required for related systems [29,48]. Approaches based on TSM are attractive in providing a global overview of genes involved in a complex biological system, particularly in the search for mutations that are negatively selected due to loss of biological function or viability. Drawbacks of such approaches can include a lack of precision—depending on the precise location of an insertion, it may generate a signal from adjacent genes, for example—and aberrant behaviour displayed by individual clones when present as a member of a pool of organisms. It is therefore often necessary to confirm results from such screens by targeted analysis of individual mutants. The two screens described in the present study have been successful in identifying candidates for such follow-up studies, including a locus (VAMP) encoding membrane and secreted proteins that are required for intracellular survival of mycobacteria in macrophages, and KefB, a potassium transporter involved in control of early phagosomal pH and survival in the chronic phase of infection in mice [30].
A central aim of the present study was to test whether comparison of independent TSM datasets could provide insight into links between two complex biological phenotypes. Based on our understanding of mycobacterial interactions with macrophages, we had anticipated that we would observe an overlap between TSM datasets from screens involving mycobacterial survival and control of acidification. Unexpectedly, our results appeared to demonstrate that inability to inhibit phagosome acidification at 60 min post-infection did not confer attenuated survival and/or growth over the first 72 h of parasitism. However, a longer (6 d) infection with four individual mutants did demonstrate these strains to be defective for intracellular survival, but the phenotype was only just evident at 72 h and not overtly manifest until at least 96 h post-infection. Thus, while the TSM survival assay was successful in detecting those mutations that confer the most pronounced attenuation of survival, either it was not sensitive enough or it encompassed an insufficient infection time to detect the more subtle levels of attenuation observed in mutants that fail to control the pH of their phagosomes. It is also possible that mixed-infection experiments involving complex pools of mutants, such as the TSM assay, will fail to detect attenuation in certain mutants due to trans-complementation between strains. This may be particularly pertinent, given that many of the acid phagosome mutants have insertions in genes encoding secreted products that may be easily complemented by secretion from neighbouring bacteria carrying the wild-type gene. Our observations are in general agreement with the findings of Pethe et al. [28] who, using a selection procedure targeted at a later stage of phagosome maturation (interaction with dextran-containing lysosomes), isolated a set of mycobacterial mutants that were unable to block acidification and were also impaired in intracellular survival over a 9-d macrophage infection. However, none of the genes identified by Pethe et al. were found in our TSM dataset. One explanation for this lack of overlap may be that the different selection procedures used in the two studies have selected for different subsets of “phagosome maturation” mutants.
A broader comparison of the current data with published reports demonstrates a clear overlap between mutations that impair survival of BCG in macrophages and those that influence pathogenesis of M. tuberculosis in mice—for example, from Table 1, nine of 17 genes on which data are available in the study by Sassetti and Rubin [30] are also attenuated in the early stages of murine infection (using a cut-off of p < 0.05). This is consistent with previous reports of a concordance between macrophage survival and mouse pathogenesis, particularly during the initial stage of acute infection, when mycobacterial growth is controlled by innate immune mechanisms [32]. In contrast, very little overlap was observed between genes required for macrophage survival and those that are up-regulated during the early stages of macrophage infection [49]. Only two of the top 20 genes listed in Table 1 (the peptide transporter dppC and Rv1357c) were significantly up-regulated during macrophage infection, while expression of others, such as the mce1 locus, was significantly depressed. Amongst the proteins encoded by genes implicated in control of acidification, PPE10 and two further PPE family members were significantly up-regulated during infection, as was the gene encoding the LppN lipoprotein. It is unlikely that there will be a simple linear relationship between essential function and changes in expression level, and linking these two datasets will require a deeper understanding of interactions between genes and their protein products at different time points during the course of infection.
Materials and Methods
Bacterial strains and growth conditions.
DNA plasmid construction was performed in E. coli DH5α using standard procedures. Mycobacterium bovis BCG (Pasteur) was cultured at 37 °C in Middlebrook 7H9 medium supplemented with 10% albumin/dextrose/catalase (ADC) and 0.05% Tween 80, or on Middlebrook 7H11 solid medium containing 10% oleic acid/albumin/dextrose/catalase (OADC) supplement. Hygromycin was included as appropriate at 50 μg/ml for BCG and 150 μg/ml for E. coli.
Transposon mutagenesis.
The hygromycin-B phosphotransferase gene from Streptomyces hygroscopicus was cloned into the multiple cloning site of the EZ::TN pMOD transposon construction vector (Epicentre Biotechnologies, Madison, Wisconsin, United States) between the transposon Mosaic End sequences. The EZ::TN transposon containing the hygromycin resistance gene, EZ::TNhyg, was excised from the plasmid by PvuII restriction digest and gel-purified using QIAEX II (Qiagen, Valencia, California, United States). The transposome was generated by mixing EZ::TNhyg DNA with EZ::TN transposase (Epicentre) and glycerol according to manufacturer's instructions and incubated at room temperature for 30 min. Transposome mix (1 μl) was electroporated into 200 μl of competent M. bovis BCG, and transposition events were selected by plating on Middlebrook 7H11 containing hygromycin. Approximately 800 insertion clones were generated per electroporation, and 2,500 mutants were pooled by scraping into 7H9 broth with hygromycin and subcultured four times before macrophage infection. This pool of mutants did not provide complete coverage of the BCG genome, but we were reluctant to feed larger libraries into flow cytometry selection for fear of a bottleneck effect that would introduce stochastic variation into the selection output.
Transposon screen by microarray.
We screened transposon libraries by microarray to reveal insertion sites and relative abundances of mutants using a modification of the techniques described by Badarinarayana [48]. Genomic DNA was prepared from pools of transposon mutants by standard procedures and restricted with BssHII to generate an average fragment size of approximately 900 bp. Oligonucleotides TA1 (5′-ACTACGCACGCGACGAGACGTAGCGTC-3′) and TA2 (5′-CGCGGACGCTACGTCCGTGTTGTCGGTCCTG-3′) were annealed to make a Y-shaped linker [48] with a BssHII-compatible 5′ overhang. Approximately 100 ng of digested genomic DNA was ligated to 100 pmol of Y-linker, and then 10 ng of ligated DNA was used as template for PCR amplification of transposon-flanking regions using the transposon-specific primer TA3 (5′-GCCTTCACCTTCCTGCACGACTTCGAGGT-3′) and the Y-linker-specific primer TA4 (5′-ACGCACGCGACGAGACGTAGC-3′) in the presence of 8% DMSO. Following an initial denaturation step for 2 min at 95 °C, the reaction was hot-started and cycled between 94.5 °C (30 s) and 72 °C (90 s) for 22 cycles. Amplification products less than 500 bp were gel purified and used in a second round of PCR amplification between the Y-linker-specific primer TA4 and a nested transposon-specific primer TA6 (3′-GTGTTCGAGGAGACCCCGCTGGATCTCTC-5′) to further enrich for transposon-flanking products and to incorporate Cy3 or Cy5-dCTP (Amersham, Little Chalfont, United Kingdom). The reaction mix included 20 μM Cy-dCTP, 180 μM dCTP, and 200 μM dGTP/dATP/dTTP and the cycling conditions were as before, but for 12 cycles. The labelled PCR products were cleaned up using a Qiagen MinElute kit, eluting in water.
The fluorescently labelled transposon insertion sites were hybridised to whole-genome microarrays, prepared by spotting the M. tuberculosis 70-mer oligonucleotide set (Qiagen and Operon [Huntsville, Alabama, United States]) onto Corning GAPS Coated Slides. Signal intensities of hybridisation were collected using Genepix Pro 3.0 and an Axon 4000B microarray scanner (Molecular Devices, Sunnyvale, California, United States).
Macrophage infections.
J774 murine macrophage-like cells were grown in DMEM (Invitrogen, Carlsbad, California, United States) containing 10% heat-inactivated foetal bovine serum and 5 mM glutamine. For TSM screening of the BCG transposon mutants during infection, culture dishes containing 3 × 107 macrophages were infected with 1.5 × 108 bacteria from the BCG mutant library. Infections were synchronised by allowing the bacteria to adhere to host cells for 30 min at 4 °C before replacement of the inoculum with fresh complete DMEM and incubation at 37 °C and 5% CO2 for 1 h. The monolayer was washed three times with PBS to remove free bacteria and then either harvested for phagosome analysis or cultured in DMEM at 37 °C and 5% CO2 for analysis of intracellular survival and growth (Figure 1). We observed that approximately 35% of macrophages took up at least one bacterium, resulting in an expected internalisation of more than 4,200 bacteria of each mutant clone.
Intracellular survival of transposon mutants.
To select for mutants with intracellular growth and/or survival defects, infected macrophages were cultured for 72 h and then the bacteria were harvested by washing the monolayer in PBS and lysing the macrophages by the addition of 7H9 medium containing Tween 80 and hygromycin. The resultant pool of transposon mutants was allowed to grow to late logarithmic phase and then used for a second round of macrophage infection and subsequently for a third selection through macrophages. The final output pool was grown to late logarithmic phase and processed for extraction of genomic DNA. The transposon-flanking regions were amplified, Cy-labelled, and co-hybridised with Cy-labelled genomic DNA from the input pool. The FR was calculated as output signal intensity divided by input signal intensity. Four replicates were performed and the entire experiment repeated.
Individual mutants isolated from the library were compared to wild-type BCG for survival and/or growth in J774 macrophages. Macrophages were seeded into 24-well plates at 2 × 105 cells per well and infected with mycobacteria at a multiplicity of 1. The monolayers were washed and bacterial survival and growth assessed by plating of lysed monolayers on 7H11 and counting of colony-forming units.
Flow cytometric analysis of mycobacterial phagosomes.
To detect mycobacterial phagosomes by flow cytometry, macrophages were infected with fluorescently labelled BCG. The bacteria were surface labelled by incubation with 1 μg/ml FITC in carbonate buffer (pH 9.2). During preliminary experiments, BCG were also labelled by heterologous expression of EGFP from plasmid pEGFPLux. No difference was observed in the localisation of FITC- and EGFP-labelled bacteria with respect to LysoTracker (unpublished data). After a 1-h infection, the cells were placed on ice, washed with ice-cold PBS, and scraped into homogenisation buffer (250 mM sucrose, 20 mM Hepes, 0.5 mM EGTA, 0.1% gelatine [pH 7.0]). The cells were lysed by multiple passage through a 22-gauge needle, and 20 mM ATP was added to break up the cytoskeleton. To remove nuclei, the homogenate was centrifuged at 900 rpm for 10 min and the supernatant retained and subjected to two further spins. The final post-nuclei supernatant was analysed on Becton Dickinson (Palo Alto, California, United States) FACSCalibur or FACSVantage flow cytometry systems. The frequency of EGFP- or FITC-positive events in different organelle populations, based on forward and side scatter, gave an indication of the position of the mycobacterial phagosomes. This was verified by sorting of different organelle populations and fixation for electron microscopy. Organelles were pelleted and fixed in 4% paraformaldehyde and 2.5% glutaraldehyde in PBS, followed by treating with 1% osmium tetroxide in sodium cacodylate buffer, rinsing with 1% tannic acid to highlight the phagosome membrane, dehydrating in an ethanol series, and embedding in Epon resin. Ultrathin 60-nm sections were further contrasted with uranyl acetate and lead citrate and viewed in a Philips CM100 transmission electron microscope.
To demonstrate the applicability of flow cytometry to differentiate phagosomes on the basis of the degree of acidification, we made a comparison of phagosomes isolated from macrophages infected with live and UV-killed BCG. BCG were irradiated in a Stratalinker UV crosslinker (Stratagene, La Jolla, California, United States) for 10 min. Preliminary experiments showed that this treatment rendered the bacteria unable to produce growth on 7H11 medium. As a marker for acidification of the phagosome we used the fluorescent acidotropic dye LysoTracker DND-99 (Invitrogen), which accumulates in acidified organelles and was included in media at 50 nM during 1-h infections of macrophages (see Figure 1). The threshold for determining LysoTracker positivity was determined by analysis of uninfected macrophages. The percentage of phagosomes that were positive for LysoTracker was determined by collecting data on at least 3,000 phagosomes.
Flow cytometry selection of transposon mutants in acidified phagosomes.
Three culture plates of 3 × 107 macrophages were infected with the FITC-labelled BCG transposon mutant library. After a 1-h infection in the presence of LysoTracker, the macrophages were harvested and organelles were prepared as above. The post-nuclei supernatant was analysed using the FACSVantage flow cytometer, and 2 × 106 LysoTracker-positive mycobacterial phagosomes were sorted into 7H9 medium containing hygromycin. The sorted phagosomes were enriched with transposon mutants that were unable to arrest phagosome acidification. To further enrich for these mutants, the sorted bacteria were allowed to grow in 7H9 broth and were used for a second infection and flow cytometry selection, from which LysoTracker-positive and LysoTracker-negative phagosomes were collected. These pools of bacteria were grown in broth for extraction of genomic DNA.
The transposon insertion sites from the two genomic DNA samples were fluorescently labelled with different Cy-dyes as described above and co-hybridised to an M. tuberculosis microarray. Signal intensities of the transposon-directed labels were collected for the genomic DNA pools. The transposon labelling and hybridisations were repeated four times and included dye-swaps. The entire experiment was repeated.
Confocal microscopy of phagosome acidification.
Macrophage infections with individual FITC-labelled transposon mutant strains and wild-type BCG were examined by fluorescence microscopy to assess acidification of phagosomes. Individual transposon mutant strains were identified and isolated from the library using insertion-specific PCR. Macrophages were seeded into chamber well slides overnight and then infected with BCG strains for 1 h in the presence of LysoTracker. The monolayers were washed three times and fixed in 2% paraformaldehyde in PBS before visualization using a Zeiss LSM510 confocal laser scanning microscope (Oberkochen, Germany).
Data analysis.
Absent and poor-quality microarray spots were flagged using GenePix Pro 3.0 and median signal intensities calculated. For the intracellular survival experiment, data for input and output transposon pools were normalised to median signal intensities. Microarray data were analysed by significance analysis of microarrays [50]. This generates a q value, or FDR, for each data point. Survival of individual mutants in macrophages, and frequencies of co-localisation with LysoTracker obtained by flow cytometry and fluorescence microscopy, were analysed by Student's t-test. Predictions of prokaryote protein subcellular localization were made using PSORTb v.2.0 software (http://www.psort.org/psortb/).
Supporting Information
Figure S1 Reproducibility of the TSM Labelling and Genome-Wide Distribution of Transposon Insertions
(A) Reproducibility of the TSM labelling and hybridisation. Hybridisation signal intensities were compared from two independent labelling and hybridisation operations on a 2,500-clone EZ::TNhyg insertion library. A strong correlation was observed (r2 = 0.9289).
(B) Genome-wide distribution of transposon insertions in a pool of 2,500 M. bovis BCG EZ::TNhyg mutants revealed by PCR-based TSM using an M. tuberculosis H37Rv microarray. The horizontal bar represents the genome of H37Rv with mutagenised genes represented as vertical black lines. Three of the BCG deletion regions are indicated by arrows along with a region of defective oligonucleotide probes that do not efficiently hybridise target DNAs.
(848 KB PDF)
Click here for additional data file.
Figure S2 Mutants with Defects in the Inhibition of Phagosome pH Are no more Likely to Have Reduced Intracellular Fitness than Those in a Random Mutant Library
(A) Distribution of intracellular fitness values (FR) in the acid phagosome mutants. Mean, 0.146; 95% confidence interval = 0.223.
(B) Distribution of intracellular fitness values in the whole transposon library. Mean, 0.024; 95% confidence interval = 0.042.
(394 KB PDF)
Click here for additional data file.
Table S1 Ez::TNhyg Insertion Sites
(21 KB XLS)
Click here for additional data file.
Table S2 Fitness Ratios of All Mutants During Growth in Macrophage Culture
(276 KB XLS)
Click here for additional data file.
Table S3 The 100 Mutants Most Attenuated During Growth in Macrophage Culture
(28 KB XLS)
Click here for additional data file.
Table S4 Mutants Enriched in Acidic Phagosomes
(22 KB XLS)
Click here for additional data file.
This work was supported by the Wellcome Trust. We thank the Bacterial Microarray Group at St. Georges Hospital Medical School and Dave Goulding for electron microscopy.
Competing interests. The authors have declared that no competing interests exist.
Author contributions. GRS, JP, BDR, and DBY conceived and designed the experiments. GRS, JP, and AR performed the experiments. GRS and AR analyzed the data. BDR and DBY contributed reagents/materials/analysis tools. GRS, BDR, and DBY wrote the paper.
Abbreviations
BCGbacille Calmette-Guérin
DIMphthiocerol dimycocerosate
FDRfalse discovery rate
FRfitness ratio
SDstandard deviation
TGNtrans-Golgi network
TSMtransposon screen by microarray
VAMPvirulence-associated membrane protein
V-ATPasevacuolar proton ATPase
==== Refs
References
Russell DG 2001
Mycobacterium tuberculosis: Here today, and here tomorrow Nat Rev Mol Cell Biol 2 569 577 11483990
Tailleux L Schwartz O Herrmann JL Pivert E Jackson M 2003 DC-SIGN is the major Mycobacterium tuberculosis receptor on human dendritic cells J Exp Med 197 121 127 12515819
Geijtenbeek TB Van Vliet SJ Koppel EA Sanchez-Hernandez M Vandenbroucke-Grauls CM 2003 Mycobacteria target DC-SIGN to suppress dendritic cell function J Exp Med 197 7 17 12515809
Hmama Z Gabathuler R Jefferies WA de Jong G Reiner NE 1998 Attenuation of HLA-DR expression by mononuclear phagocytes infected with Mycobacterium tuberculosis is related to intracellular sequestration of immature class II heterodimers J Immunol 161 4882 4893 9794422
Noss EH Pai RK Sellati TJ Radolf JD Belisle J 2001 Toll-like receptor 2-dependent inhibition of macrophage class II MHC expression and antigen processing by 19-kDa lipoprotein of Mycobacterium tuberculosis
J Immunol 167 910 918 11441098
Ting LM Kim AC Cattamanchi A Ernst JD 1999
Mycobacterium tuberculosis inhibits IFN-gamma transcriptional responses without inhibiting activation of STAT1 J Immunol 163 3898 3906 10490990
Duclos S Diez R Garin J Papadopoulou B Descoteaux A 2000 Rab5 regulates the kiss and run fusion between phagosomes and endosomes and the acquisition of phagosome leishmanicidal properties in RAW 264.7 macrophages J Cell Sci 113 3531 3541 10984443
Chua J Vergne I Master S Deretic V 2004 A tale of two lipids: Mycobacterium tuberculosis phagosome maturation arrest Curr Opin Microbiol 7 71 77 15036144
Desjardins M Huber LA Parton RG Griffiths G 1994 Biogenesis of phagolysosomes proceeds through a sequential series of interactions with the endocytic apparatus J Cell Biol 124 677 688 8120091
Simonsen A Lippe R Christoforidis S Gaullier JM Brech A 1998 EEA1 links PI(3)K function to Rab5 regulation of endosome fusion Nature 394 494 498 9697774
Desjardins M 1995 Biogenesis of phagolysosomes: The “kiss and run” hypothesis Trends Cell Biol 5 183 186 14731444
Harrison RE Bucci C Vieira OV Schroer TA Grinstein S 2003 Phagosomes fuse with late endosomes and/or lysosomes by extension of membrane protrusions along microtubules: Role of Rab7 and RILP Mol Cell Biol 23 6494 6506 12944476
Fratti RA Chua J Vergne I Deretic V 2003
Mycobacterium tuberculosis glycosylated phosphatidylinositol causes phagosome maturation arrest Proc Natl Acad Sci U S A 100 5437 5442 12702770
Brown CA Draper P Hart PD 1969 Mycobacteria and lysosomes: A paradox Nature 221 658 660 4975468
Frehel C de Chastellier C Lang T Rastogi N 1986 Evidence for inhibition of fusion of lysosomal and prelysosomal compartments with phagosomes in macrophages infected with pathogenic Mycobacterium avium
Infect Immun 52 252 262 2870027
Hasan Z Schlax C Kuhn L Lefkovits I Young D 1997 Isolation and characterization of the mycobacterial phagosome: Segregation from the endosomal/lysosomal pathway Mol Microbiol 24 545 553 9179848
Kelley VA Schorey JS 2003 Mycobacterium's arrest of phagosome maturation in macrophages requires Rab5 activity and accessibility to iron Mol Biol Cell 14 3366 3377 12925769
Via LE Fratti RA McFalone M Pagan-Ramos E Deretic D 1998 Effects of cytokines on mycobacterial phagosome maturation J Cell Sci 111 897 905 9490634
Clemens DL Lee BY Horwitz MA 2000 Deviant expression of Rab5 on phagosomes containing the intracellular pathogens Mycobacterium tuberculosis and Legionella pneumophila is associated with altered phagosomal fate Infect Immun 68 2671 2684 10768959
Sturgill-Koszycki S Schaible UE Russell DG 1996 Mycobacterium-containing phagosomes are accessible to early endosomes and reflect a transitional state in normal phagosome biogenesis EMBO J 15 6960 6968 9003772
Clemens DL Horwitz MA 1996 The Mycobacterium tuberculosis phagosome interacts with early endosomes and is accessible to exogenously administered transferrin J Exp Med 184 1349 1355 8879207
Sturgill-Koszycki S Schlesinger PH Chakraborty P Haddix PL Collins HL 1994 Lack of acidification in Mycobacterium phagosomes produced by exclusion of the vesicular proton-ATPase Science 263 678 681 8303277
Fratti RA Backer JM Gruenberg J Corvera S Deretic V 2001 Role of phosphatidylinositol 3-kinase and Rab5 effectors in phagosomal biogenesis and mycobacterial phagosome maturation arrest J Cell Biol 154 631 644 11489920
Indrigo J Hunter RL Jr. Actor JK 2003 Cord factor trehalose 6,6′-dimycolate (TDM) mediates trafficking events during mycobacterial infection of murine macrophages Microbiology 149 2049 2059 12904545
Anes E Kuhnel MP Bos E Moniz-Pereira J Habermann A 2003 Selected lipids activate phagosome actin assembly and maturation resulting in killing of pathogenic mycobacteria Nat Cell Biol 5 793 802 12942085
Walburger A Koul A Ferrari G Nguyen L Prescianotto-Baschong C 2004 Protein kinase G from pathogenic mycobacteria promotes survival within macrophages Science 304 1800 1804 15155913
Vergne I Chua J Lee HH Lucas M Belisle J 2005 Mechanism of phagolysosome biogenesis block by viable Mycobacterium tuberculosis
Proc Natl Acad Sci U S A 102 4033 4038 15753315
Pethe K Swenson DL Alonso S Anderson J Wang C 2004 Isolation of Mycobacterium tuberculosis mutants defective in the arrest of phagosome maturation Proc Natl Acad Sci U S A 101 13642 13647 15340136
Sassetti CM Boyd DH Rubin EJ 2001 Comprehensive identification of conditionally essential genes in mycobacteria Proc Natl Acad Sci U S A 98 12712 12717 11606763
Sassetti CM Rubin EJ 2003 Genetic requirements for mycobacterial survival during infection Proc Natl Acad Sci U S A 100 12989 12994 14569030
Berg DE Schmandt MA Lowe JB 1983 Specificity of transposon Tn5 insertion Genetics 105 813 828 6315531
Camacho LR Ensergueix D Perez E Gicquel B Guilhot C 1999 Identification of a virulence gene cluster of Mycobacterium tuberculosis by signature-tagged transposon mutagenesis Mol Microbiol 34 257 267 10564470
Cox JS Chen B McNeil M Jacobs WR Jr. 1999 Complex lipid determines tissue-specific replication of Mycobacterium tuberculosis in mice Nature 402 79 83 10573420
McAdam RA Quan S Smith DA Bardarov S Betts JC 2002 Characterization of a Mycobacterium tuberculosis H37Rv transposon library reveals insertions in 351 ORFs and mutants with altered virulence Microbiology 148 2975 2986 12368431
Flesselles B Anand NN Remani J Loosmore SM Klein MH 1999 Disruption of the mycobacterial cell entry gene of Mycobacterium bovis BCG results in a mutant that exhibits a reduced invasiveness for epithelial cells FEMS Microbiol Lett 177 237 242 10474190
Arruda S Bomfim G Knights R Huima-Byron T Riley LW 1993 Cloning of an M. tuberculosis DNA fragment associated with entry and survival inside cells Science 261 1454 1457 8367727
Chitale S Ehrt S Kawamura I Fujimura T Shimono N 2001 Recombinant Mycobacterium tuberculosis protein associated with mammalian cell entry Cell Microbiol 3 247 254 11298648
McKinney JD Honer zu Bentrup K Munoz-Elias EJ Miczak A Chen B 2000 Persistence of Mycobacterium tuberculosis in macrophages and mice requires the glyoxylate shunt enzyme isocitrate lyase Nature 406 735 738 10963599
Berks BC Palmer T Sargent F 2003 The Tat protein translocation pathway and its role in microbial physiology Adv Microb Physiol 47 187 254 14560665
Cole ST Brosch R Parkhill J Garnier T Churcher C 1998 Deciphering the biology of Mycobacterium tuberculosis from the complete genome sequence Nature 393 537 544 9634230
Gey Van Pittius NC Gamieldien J Hide W Brown GD Siezen RJ 2001 The ESAT-6 gene cluster of Mycobacterium tuberculosis and other high G+C gram-positive bacteria Genome Biol 2 44
Ramakrishnan L Federspiel NA Falkow S 2000 Granuloma-specific expression of Mycobacterium virulence proteins from the glycine-rich PE-PGRS family Science 288 1436 1439 10827956
Chaturvedi S Bhakuni V 2003 Unusual structural, functional, and stability properties of serine hydroxymethyltransferase from Mycobacterium tuberculosis
J Biol Chem 278 40793 40805 12913008
Booth IR Epstein W Giffard PM Rowland GC 1985 Roles of the trkB and trkC gene products of Escherichia coli in K+ transport Biochimie 67 83 89 3888294
Bakker EP Booth IR Dinnbier U Epstein W Gajewska A 1987 Evidence for multiple K+ export systems in Escherichia coli
J Bacteriol 169 3743 3749 3301813
Ferguson GP McLaggan D Booth IR 1995 Potassium channel activation by glutathione-S-conjugates in Escherichia coli : Protection against methylglyoxal is mediated by cytoplasmic acidification Mol Microbiol 17 1025 1033 8594323
Ferguson GP Nikolaev Y McLaggan D Maclean M Booth IR 1997 Survival during exposure to the electrophilic reagent N-ethylmaleimide in Escherichia coli: Role of KefB and KefC potassium channels J Bacteriol 179 1007 1012 9023177
Badarinarayana V Estep PW 3rd Shendure J Edwards J Tavazoie S 2001 Selection analyses of insertional mutants using subgenic-resolution arrays Nat Biotechnol 19 1060 1065 11689852
Schnappinger D Ehrt S Voskuil MI Liu Y Mangan JA 2003 Transcriptional adaptation of Mycobacterium tuberculosis within macrophages: Insights into the phagosomal environment J Exp Med 198 693 704 12953091
Tusher VG Tibshirani R Chu G 2001 Significance analysis of microarrays applied to the ionizing radiation response Proc Natl Acad Sci U S A 98 5116 5121 11309499
|
16322769
|
PMC1291353
|
CC BY
|
2021-01-05 12:11:27
|
no
|
PLoS Pathog. 2005 Nov 25; 1(3):e33
|
utf-8
|
PLoS Pathog
| 2,005 |
10.1371/journal.ppat.0010033
|
oa_comm
|
==== Front
PLoS PathogPLoS PathogppatplpaplospathPLoS Pathogens1553-73661553-7374Public Library of Science San Francisco, USA 10.1371/journal.ppat.001003405-PLPA-RA-0079R2plpa-01-03-10Research ArticleImmunologyInfectious DiseasesMicrobiologyEubacteriaHomo (human)In VitroNOD2 and Toll-Like Receptors Are Nonredundant Recognition Systems of Mycobacterium tuberculosis
Recognition of
M. tuberculosis by NOD2
Ferwerda Gerben 12Girardin Stephen E 3Kullberg Bart-Jan 12Le Bourhis Lionel 3de Jong Dirk J. 4Langenberg Dennis M. L 5van Crevel Reinout 12Adema Gosse J 6Ottenhoff Tom H. M 5Van der Meer Jos W. M. 12Netea Mihai G 12*
1 Department of Internal Medicine, Radboud University Medical Center, Nijmegen, The Netherlands
2 Nijmegen University Center for Infectious Diseases, Nijmegen, The Netherlands
3 Unité de Pathogénie Microbienne Moléculaire, INSERM U389, Institut Pasteur, Paris Cedex, France
4 Department of Gastroenterology, Radboud University Medical Center, Nijmegen, The Netherlands
5 Department of Immunohematology and Blood Transfusion, Leiden University Medical Center, Leiden, The Netherlands
6 Department of Tumor Immunology, Radboud University Medical Center, Nijmegen, The Netherlands
Filler Scott EditorUCLA Research and Education Institute, United States of America*To whom correspondence should be addressed. E-mail: [email protected] 2005 25 11 2005 1 3 e3421 6 2005 20 10 2005 Copyright: © 2005 Ferwerda et al.2005This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are properly credited.Infection with Mycobacterium tuberculosis is one of the leading causes of death worldwide. Recognition of M. tuberculosis by pattern recognition receptors is crucial for activation of both innate and adaptive immune responses. In the present study, we demonstrate that nucleotide-binding oligomerization domain 2 (NOD2) and Toll-like receptors (TLRs) are two nonredundant recognition mechanisms of M. tuberculosis. CHO cell lines transfected with human TLR2 or TLR4 were responsive to M. tuberculosis. TLR2 knock-out mice displayed more than 50% defective cytokine production after stimulation with mycobacteria, whereas TLR4-defective mice also released 30% less cytokines compared to controls. Similarly, HEK293T cells transfected with NOD2 responded to stimulation with M. tuberculosis. The important role of NOD2 for the recognition of M. tuberculosis was demonstrated in mononuclear cells of individuals homozygous for the 3020insC NOD2 mutation, who showed an 80% defective cytokine response after stimulation with M. tuberculosis. Finally, the mycobacterial TLR2 ligand 19-kDa lipoprotein and the NOD2 ligand muramyl dipeptide synergized for the induction of cytokines, and this synergism was lost in cells defective in either TLR2 or NOD2. Together, these results demonstrate that NOD2 and TLR pathways are nonredundant recognition mechanisms of M. tuberculosis that synergize for the induction of proinflammatory cytokines.
Synopsis
Tuberculosis is one of the most prevalent infections worldwide, with 2 billion people believed to be infected, and 2 million deaths each year. In addition to representing a major health care problem in developing countries, concern is also growing about the increased incidence of tuberculosis in developed countries, especially in immunocompromised patients such as those with AIDS, transplantation, and immunosuppressive therapy. The present study describes the pathways that enable leukocytes to recognize M. tuberculosis, and demonstrates for the first time that NOD2, member of a new class of intracellular receptors, is an independent recognition mechanism for mycobacteria. NOD2 acts together with the earlier-described Toll-like receptors for the activation of host defenses during the encounter of leukocytes with M. tuberculosis. Understanding the mechanisms through which the cells of the immune system recognize M. tuberculosis can be an important step in designing new therapeutic approaches, as well as improving the limited success of current vaccination strategies.
Citation:Ferwerda G, Girardin SE, Kullberg B, Le Bourhis L, De Jong DJ, et al. (2005) NOD2 and Toll-like receptors are nonredundant recognition systems of Mycobacterium tuberculosis. PLoS Pathog 1(3): e34.
==== Body
Introduction
Worldwide, 2 billion people are currently believed to be infected with Mycobacterium tuberculosis, with an estimated death toll of 2 million patients each year [1]. M. tuberculosis is an intracellular pathogen capable of infecting and surviving within the host's mononuclear cells (MNCs), and a coordinated response of the innate and adaptive immune systems is required for an effective host defense. This involves sequestration of the microorganism in macrophages within organized granulomas, and elimination of the pathogen through a combination of killing mechanisms and apoptosis of host macrophages [2]. These responses are coordinated by T helper 1-type proinflammatory cytokines, which are synthesized by phagocytes upon recognition of pathogen-associated molecular patterns of mycobacteria by pattern recognition receptors (PRRs).
Toll-like receptors (TLRs) are believed to be an important pattern recognition system of M. tuberculosis. A soluble, heat-stable mycobacterial fraction was initially reported to signal through TLR2, whereas heat-labile components associated with the cell wall were reported to signal through TLR4 [3]. Later, several components of mycobacteria were identified as being responsible for TLR2-dependent activation: the 19-kDa lipoprotein [4], lipomannan [5], phosphatidyl-myo-inositol mannoside [6], but not mannosyl-caped lipoarabinomannan from virulent M. tuberculosis, which mainly has anti-inflammatory effects through its interaction with the mannose receptor and dendritic cell-specific intercellular adhesion molecule 3-grabbing nonintegrin, also termed DC-SIGN [7,8]. Despite an abundance of in vitro data regarding the recognition of mycobacterial structures by TLR2 and TLR4, knock-out mice deficient for these receptors display remarkably little enhanced susceptibility to infection with M. tuberculosis. TLR4
−/− mice showed variable responses to the challenge with M. tuberculosis, with either normal resistance to infection [9,10] or chronic pneumonia and increased mortality [11,12]. In one study, TLR2
−/− mice had a decreased clearance of the bacteria and developed chronic pneumonia when infected with a low dose of microorganisms [13], whereas in other studies only minor effects have been found [9,14]. The role of TLRs in the recognition of mycobacteria was further highlighted in mice deficient for MyD88 (an adaptor molecule shared by almost all the receptors of the TLR family), which are highly susceptible to M. tuberculosis infection [15].
Nucleotide-binding oligomerization domain 2 (NOD2), initially described as a susceptibility gene for Crohn's disease [16,17], is an intracellular protein containing leucine-rich repeats (LRRs) similar to those found in TLRs. NOD2 is the PRR responsible for recognition of bacterial peptidoglycans from both gram-positive and gram-negative bacteria, through its interaction with muramyl dipeptide (MDP) [18]. In contrast, NOD1 recognizes peptidoglycans of gram-negative bacteria only [19]. The intracellular localization of both NOD2 and M. tuberculosis, the cell wall of which contains peptidoglycans, makes NOD2 a highly suitable candidate for the recognition of mycobacteria. Our results support the hypothesis that TLRs and NOD2 represent two nonredundant recognition systems of M. tuberculosis, and that an efficient activation of innate immunity requires both classes of receptors.
Results
The Role of TLR2 and TLR4 in the Recognition of M. tuberculosis
TLR2 and TLR4 have been suggested to recognize bacterial structures of M. tuberculosis [20]. Indeed, a sonicate of M. tuberculosis strongly activated a Chinese hamster ovary fibroblast (CHO) cell line cotransfected with human TLR2 and CD14, whereas cells transfected with CD14 alone or a combination of CD14 and TLR4 displayed no signaling upon activation with the sonicated mycobacterial preparation (Figure 1A). However, when cells were stimulated with a preparation of whole mycobacteria, both TLR2 and TLR4-transfected cells were activated, although TLR2 activation was stronger (Figure 1A).
Figure 1 TLR2 and TLR4 Are Partially Responsible for Induction of Cytokine Production by M. tuberculosis
(A) CHO cells cotransfected with CD14 and TLR2 (CD14/TLR2, hatched bars) induced potent expression of CD25 on the cell membrane as measured by FACS analysis, after stimulation with both a sonicated M. tuberculosis (MTB-son, 1 μg/ml) or whole mycobacteria (MTB-whole, 1 × 107/ml). In contrast, cells transfected with CD14 and TLR4 (CD14/TLR4, black bars) were only moderately activated by whole M. tuberculosis, but not by the sonicated material. LPS (1 μg/ml) and Pam3Cys (5 μg/ml) served as control stimuli for TLR4 and TLR2, respectively. As controls, cells transfected only with CD14 were used (white bars).
(B and C) Stimulation of TNF production by M. tuberculosis in peritoneal macrophages of mice deficient for TLR2 (TLR2-KO) (B) or TLR4 (TLR4-def) (C). Groups of five mice were stimulated with the indicated reagents, and the experiment was repeated twice. Medium-stimulated cells resulted in cytokine concentrations below detection limit.
Data are presented as mean ± SD (n = 5; *p < 0.05).
In line with these data, macrophages isolated from TLR2
−/− mice displayed a 50%–75% reduction in tumor necrosis factor (TNF) production after stimulation with both M. tuberculosis preparations (Figure 1B), whereas TLR4-deficient macrophages showed a 30%–40% reduction of TNF release only when stimulated with the whole mycobacteria (Figure 1C). To confirm the role of TLR4 in the stimulation of cytokines by M. tuberculosis, we stimulated human MNCs with the whole mycobacterial preparation in the absence or presence of 10 μg/ml of a blocking anti-TLR4 antibody. TLR4 blockade completely inhibited lipopolysaccharide (LPS)-induced TNF secretion, and reduced M. tuberculosis-induced TNF secretion from 0.9 ± 0.2 to 0.5 ± 0.2 ng/ml (p < 0.05). These data confirm the role played by TLR2 and TLR4 in the recognition of M. tuberculosis; however, the significant remaining production of cytokines induced by M. tuberculosis in TLR2−/− or TLR4-defective mice points to the presence of additional signaling pathway(s) for cytokine induction.
The Role of Intracellular Recognition Systems in the Recognition of M. tuberculosis
The role of internalization in cytokine induction by M. tuberculosis was assessed by blocking it with cytochalasin B, an inhibitor of actin polymerization. Blocking internalization of M. tuberculosis partially inhibited M. tuberculosis-induced cytokine release in freshly isolated human MNCs (Figure 2). In contrast, cytochalasin B increased the zymosan-induced cytokine production in these MNCs, which likely results from prolonged stimulation of receptors at the cell surface by zymosan. The difference in the effect of cytochalasin B on M. tuberculosis- or zymosan-induced cytokines strongly suggests that internalization of M. tuberculosis is important for the induction of cytokine production through intracellular receptors.
Figure 2 Blockade of Internalization of M. tuberculosis Impairs Recognition and Cytokine Production
Blockade of M. tuberculosis internalization by cytochalasin B (20 μg/ml) impairs TNF (A) and IL-10 (B) stimulation in human MNCs stimulated with sonicated M. tuberculosis (1 × 106 microorganisms/ml), but not zymosan (1 μg/ml). Data are presented as mean ± SD (n = 5; *p < 0.05).
The Role of NOD1 and NOD2 in the Recognition of M. tuberculosis
Next, human embryonic kidney 293T cells (HEK293Ts) were transfected with either NOD1 or NOD2 expression vectors, and the ability of M. tuberculosis sonicates to activate these pathogen recognition receptors was followed by monitoring the level of NOD-dependent activation of a nuclear factor κB (NF-κB)-driven luciferease reporter gene. Strikingly, M. tuberculosis sonicates efficiently stimulated NOD2, while NOD1-transfected cells responded only modestly to the bacteria (Figure 3).
Figure 3 Stimulation of HEK Cells Transfected with NOD1 or NOD2
HEKs transfected with NOD2 were strongly activated by whole M. tuberculosis, in contrast to cells transfected with NOD1, which were only weakly activated by M. tuberculosis. MDP (100 nM) and TriDAP (100 nM) served as control stimuli for NOD2 and NOD1, respectively. Data are presented as fold increase over unstimulated cells (mean ± SD).
To investigate further the role of NOD1 and NOD2 in the recognition of M. tuberculosis, we stimulated peritoneal macrophages from NOD1
−/− or NOD2
−/− mice. We observed that peritoneal macrophages from NOD2
−/− mice produced significantly less TNF than did control cells, supporting a role of NOD2 in the recognition of M. tuberculosis (Figure 4). In contrast, macrophages from NOD1
−/− mice responded normally to M. tuberculosis (Figure 4), although as a control, their response to the NOD1 ligand FK156 was abrogated (unpublished data). Together, these results strongly support the notion that NOD2 represents an intracellular recognition system for M. tuberculosis.
Figure 4 Stimulation of Cells from Mice Deficient in NOD1 or NOD2
Stimulation of cytokine production by sonicated M. tuberculosis in peritoneal macrophages of mice deficient for either NOD1 (A) or NOD2 (B). Groups of four mice were stimulated; medium-stimulated cells resulted in cytokine concentrations below detection limit.
Data are presented as mean ± SD (*p < 0.05).
Stimulation of Cells from Patients with Defective NOD2 Recognition
In line with the hypothesis that NOD2 is involved in the recognition of mycobacteria, MNCs isolated from patients homozygous for the 3020insC mutation synthesized 65%–80% less cytokines after stimulation with M. tuberculosis than did Crohn's disease patients heterozygous for the mutation or patients and volunteers homozygous for the wild-type variant (Figure 5A). This demonstrates that NOD2 is critical for the recognition of M. tuberculosis cell wall. In control experiments, cells from patients homozygous for the 3020insC mutation were defective in their response to MDP, but not LPS (Figure 5B), as previously described [21].
Figure 5 Human NOD2 Is a Receptor for M. tuberculosis
(A) MNCs isolated from four patients with Crohn's disease homozygous for the 3020insC NOD2 mutation (NOD2fs), five patients heterozygous for NOD2 mutations (NOD2het), five patients with the wild-type NOD2 allele (NOD2WT), and five healthy volunteers with wild-type NOD2 (control) were stimulated with 10 μg/ml sonicated M. tuberculosis. TNF (white bars) and IL-10 (black bars) were measured after 24 h of stimulation at 37 °C by specific RIA and ELISA, respectively.
(B) MNCs isolated from four patients with Crohn's disease homozygous for the 3020insC NOD2 mutation (NOD2fs, black bars), five patients with the wild-type NOD2 allele (NOD2WT, gray bars), and five healthy volunteers with wild-type NOD2 (control), were stimulated with 1 μg/ml MDP or 10 ng/ml LPS. TNF were measured after 24 h of stimulation at 37 °C by specific RIA. Medium-stimulated cells resulted in cytokine concentrations below detection limit.
Data are presented as mean ± SD (*p < 0.05).
TLR2 and NOD2 Synergize for the Induction of Cytokine Release
In order to study the cross talk between TLR2 and NOD2 in the recognition of M. tuberculosis, we investigated whether mycobacterial TLR2 and NOD2 ligands synergize for the induction of cytokine release. Indeed, the NOD2 ligand MDP, the minimal component of peptidoglycan responsible for NOD2 activation, had strong synergistic effect on TNF production induced by the 19-kDa lipoprotein of M. tuberculosis, a specific mycobacterial TLR2 ligand (Figure 6A). Similar synergistic effects between the 19-kDa lipoprotein and MDP were observed when IL-6 (4.5-fold synergism), and IL-1β (7.2-fold synergism) were measured. This synergism was lost in individuals homozygous for the NOD2 3020insC mutation (Figure 6B) or macrophages harvested from TLR2−/− mice (5.7-fold synergism between lipoprotein 19-kDa and FK-156 in control macrophages, and 1.9-fold synergism in TLR2−/− cells). No such synergy was observed between MDP and mannosyl-caped lipoarabinomannan, a component of M. tuberculosis that does not interact with TLR2 (unpublished data).
Figure 6 NOD2 and TLR2 Have Synergistic Effects on Cytokine Production
(A) MNCs isolated from five healthy volunteers with wild-type NOD2 were costimulated with 1 μg/ml MDP and increasing concentrations of 19-kDa lipoprotein (indicated on the x-axis). TNF was measured after 24 h of stimulation at 37 °C by specific RIA.
(B) The synergistic effects observed in the five control volunteers, as well as in five Crohn's disease patients heterozygous for NOD2 mutations (NOD2-het), and in five patients with the wild-type NOD2 allele (NOD2-WT), was lost in patients homozygous for the NOD2 3020insC mutation. Medium-stimulated cells resulted in cytokine concentrations below detection limit. Data are presented as mean ± SD (*p < 0.05).
Discussion
In the present study, we investigated the role of TLRs and NODs, the two most important classes of PRRs in the recognition by macrophages of M. tuberculosis. Although we confirmed the role of TLR2 and TLR4 for in mycobacterial recognition, strong residual activity was detectable in cells lacking TLR2, which suggests the existence of TLR-independent recognition mechanisms; this idea is supported by a study demonstrating MyD88-independent pathways of macrophage stimulation by M. tuberculosis [22]. Using cell lines transfected with NOD1 and NOD2, as well as primary MNCs defective in these receptors, we demonstrated that, in addition to TLRs, NOD2 represents a nonredundant recognition system of M. tuberculosis. We also demonstrated that mycobacterial TLR2 and NOD2 ligands synergize for the production of proinflammatory cytokines, and that this synergism is lost in cells lacking either of these receptors.
Several studies have demonstrated the role of TLR2 and TLR4 in the recognition of M. tuberculosis. The 19-kDa lipoprotein [4], lipomannan [5], and phosphatidyl-myo-inositol mannoside [6], all components of mycobacteria, have been identified as being responsible for TLR2-dependent activation, whereas heat-labile components associated with the cell wall were found to signal via TLR4 [3]. A role for TLRs in antimycobacterial defense was also suggested by the enhanced susceptibility to M. tuberculosis infection in mice deficient for MyD88, an adapter molecule shared by almost all TLR family members [15]. Similarly, TLR2−/− mice had a decreased clearance of the bacteria and developed chronic pneumonia when infected with low doses of microorganisms [9,13,14], whereas TLR4−/− mice showed variable responses to challenge with M. tuberculosis [9,10].
Our data confirm the important role played by TLRs, and especially TLR2, in the recognition of M. tuberculosis, but at the same time demonstrate strong TLR-independent induction of cytokines by this microorganism. The contribution of TLR4 was found to be less crucial in our study, and could be observed only when cells were stimulated with intact microorganisms.
Because M. tuberculosis is an intracellular pathogen, we hypothesized that intracellular recognition systems could contribute to the sensing of mycobacteria and stimulation of innate immunity. To test this hypothesis, we studied the effect of blocking the internalization of M. tuberculosis with cytochalasin B. Blocking internalization of M. tuberculosis partially inhibited M. tuberculosis-induced cytokine release in MNCs, whereas cytochalasin B potentiated the cytokine induction by zymosan, likely due to prolonged stimulation of receptors at the cell surface by zymosan. The differential effects of cytochalasin B on M. tuberculosis- or zymosan-induced cytokine secretion demonstrate that, in addition to the interaction with cell-membrane bound TLRs, M. tuberculosis is recognized by and induces cytokine production through intracellular receptors.
NOD2 and NOD1 are members of the expanding CATERPILLER family of proteins, which share an LRR domain similar to that found in TLRs [23]. NOD2 has been linked genetically to increased risk for Crohn's disease [16,17], and is a sensor of bacterial peptidoglycans [18]. When HEKs transfected with either NOD1 or NOD2 were stimulated with M. tuberculosis cell wall preparations, both of them—but most markedly those transfected with NOD2—showed a dose-dependent response. These data are consistent with the conclusion that NOD2 is a general sensor of bacteria through the detection of MDP, a peptidoglycan substructure present in bacterial cell walls [18]. However, the fact that NOD1 was found to be a poor sensor of M. tuberculosis not only in NOD1-transfected HEK cells, but also through the lack of a defective response in NOD1−/− macrophages, is somewhat puzzling. Indeed, NOD1 detects diaminopimelic acid (DAP)-type peptidoglycans, and earlier reports had suggested that M. tuberculosis peptidoglycan is of this category. Further investigation is therefore required to discover why NOD1 detects M. tuberculosis poorly.
To test whether the data in transfected cell lines could be reproduced in primary cells, we stimulated MNCs isolated from Crohn's disease patients homozygous for the 3020insC null-mutant allele with M. tuberculosis. This mutation leads to the deletion of the last 32 amino acids of the LRR region responsible for the detection of peptidoglycan, and we have recently shown that cells isolated from these patients are completely unable to recognize MDP or gram-positive peptidoglycan [21]. In line with the hypothesis that NOD2 is involved in the recognition of mycobacteria, both peritoneal macrophages from NOD2-deficient mice and MNCs isolated from patients homozygous for the 3020insC mutation of NOD2 were found to synthesize significantly less cytokines after stimulation with M. tuberculosis. Interestingly, the very strong defect in the response to M. tuberculosis of the cells of patients with the NOD2 3020insC mutation suggests that NOD1 is not able to compensate for the defective NOD2 recognition. This is in line with the weak stimulation of NOD1-transfected HEKs by M. tuberculosis and our recent finding that NOD2 is needed for normal signaling by NOD1 ligands such as Mur-Tri-DAP [24]. The reverse is not true in the case of NOD1: Macrophages harvested from NOD1−/− mice responded normally to M. tuberculosis, demonstrating that the absence of NOD1 can be compensated by other recognition systems, most likely NOD2.
The finding of strongly reduced cytokine production after stimulation with M. tuberculosis in cells of patients with a defective NOD2 and of NOD2 knock-out mice demonstrates that NOD2 is a key sensor of M. tuberculosis in mammalian cells. Interestingly, inhibition of both NOD2 and TLR2 systems blocked stimulation of cytokines by M. tuberculosis by substantially more than 50%. This suggests that the signaling pathways induced by these receptors interact and potentiate each other, and, conversely, that defects in one pathway lead to a loss of synergy. We have recently shown that NOD2 signals strongly synergize with specific TLR pathways such as TLR2, TLR4, and TLR3 [25]. We therefore investigated whether NOD2 activation leads to similar synergistic cytokine stimulation by TLR2 ligands specifically derived from M. tuberculosis. Indeed, MDP had strong synergistic effect on the cytokine production induced by the 19-kDa lipoprotein of M. tuberculosis, and this synergy was lost both in individuals homozygous for the NOD2 3020insC mutation and in macrophages harvested from TLR2−/− mice.
An issue yet to be resolved is represented by the mechanism through which mycobacterial peptidoglycans come in contact with NOD2. NOD2 is an intracytoplasmic molecule, while M. tuberculosis remains located mainly in phagosomes. Although shedding of cell wall components from the microorganism is likely responsible for the release of peptidoglycans that are ultimately recognized by NOD2, the precise mechanism through which peptidoglycans translocate from the phagosome into the cytoplasm remains to be identified.
The data presented in this study demonstrate that NOD2 and TLRs are two nonredundant recognition mechanisms of M. tuberculosis. Both are essential for effective activation of the proinflammatory cytokine production by M. tuberculosis, and they strengthen each other's activity through synergistic effects. This demonstrates that host cells sense the presence of M. tuberculosis using multiple recognition systems in which different classes of receptors, in this case NOD2 and TLRs, interact with each other.
The involvement of NOD2 in the recognition of M. tuberculosis has several implications. First, it is possible that NOD2 is involved in recognition of other gram-positive bacteria with cell walls rich in peptidoglycans. The recent report of NOD2 serving as a receptor for Streptococcus pneumoniae supports this idea [26]. Second, our data suggest that NOD2 is involved in the recognition of other Mycobacteria species. From this point of view, M. paratuberculosis is of particular interest, due to its possible involvement in the pathogenesis of Crohn's disease [27]. A defective host defense against M. paratuberculosis in individuals bearing loss-of-function NOD2 mutations may be responsible for the invasion of the intestine and may promote chronic inflammation ultimately leading to Crohn's disease. This hypothesis is supported by the recently described presence of M. paratuberculosis in the circulation of patients with Crohn's disease [28]. Regarding the role of NOD2 in tuberculosis, one must, however, acknowledge that the present study provides information only on the in vitro recognition of M. tuberculosis, and studies investigating the role of NOD2 in infection models are warranted.
In conclusion, NOD2 and TLRs appear to serve as independent, nonredundant PRRs of M. tuberculosis. The intracellular pathways induced by NOD2 and TLR2 during recognition of mycobacterial components synergize, and the stimulation of cytokine production by M. tuberculosis is greatly impaired in individuals with NOD2 mutations.
Materials and Methods
Reagents and microorganisms.
Synthetic Pam3Cys and 19-kDa lipoprotein were purchased from EMC Microcollections (Tübingen, Germany). MDP and LPS (E. coli serotype 055:B5) were purchased from Calbiochem (San Diego, California, United States) and Sigma (St. Louis, Missouri, United States), respectively. The synthetic MDP was tested for contamination with lipoproteins or LPS in mice deficient in TLR2 or TLR4, respectively. No defect in cytokine production was apparent in these mice after stimulation with MDP, demonstrating the absence of contamination.
Cultures of M. tuberculosis H37Rv were grown to mid-log phase in Middlebrook 7H9 liquid medium supplemented with oleic acid/albumin/dextrose/catalase (Difco, Becton-Dickinson, Palo Alto, California, United States), washed three times in sterile saline, and resuspended in RPMI 1640 medium at the various concentrations. Separate culture suspensions were sonicated for 10 min on ice, in order to obtain cell lysates.
Genotyping of NOD2 variants.
Blood was collected from 74 patients with Crohn's disease and ten healthy volunteers. PCR amplification of NOD2 gene fragments containing the polymorphic site 3020insC was performed in 50-μl reaction volumes containing 100–200 ng of genomic DNA as previously described [21]. The 3020insC polymorphism was analyzed by Genescan analysis on an ABI-Prism 3100 Genetic Analyzer according to the protocol of the manufacturer (Applied Biosystems, Nieuwerkerk a/d IJssel, The Netherlands).
Four patients with Crohn's disease were found to be homozygous for the 3020insC mutation, and they were further investigated in the cytokine studies. As control groups, five patients with Crohn's disease who were heterozygous for the 3020insC NOD2 mutation, five patients with Crohn's disease bearing the wild-type allele, and five healthy volunteers homozygous for the wild-type NOD2 allele were included. None of the patients with Crohn's disease used immunosuppressive medication at the time of the study.
Isolation of MNCs and stimulation of cytokine production.
After informed consent, venous blood was drawn from the cubital vein of patients and healthy volunteers into three 10-ml EDTA tubes (Monoject, s-Hertogenbosch, The Netherlands). Isolation of MNCs was performed as described elsewhere [29], with minor modifications. The MNC fraction was obtained by density centrifugation of blood diluted 1:1 in pyrogen-free saline over Ficoll-Paque (Pharmacia Biotech AB, Uppsala, Sweden). Cells were washed twice in saline and suspended in culture medium (RPMI 1640 DM) supplemented with 10 μg/ml gentamicin, 10 mM L-glutamine, and 10 mM pyruvate. The cells were counted in a Coulter counter (Coulter Electronics, Mijdrecht, The Netherlands), and the number was adjusted to 5 × 106 cells/ml.
MNCs (5 × 105) were added in a 100-μl volume to round-bottom 96-well plates (Greiner, Alphen a/d Rijn, The Netherlands) and incubated with either 100 μl of culture medium (negative control), or the various stimuli: M. tuberculosis bacteria (1 × 105 to 1 × 107 microorganisms/ml), M. tuberculosis sonicate (10 μg/ml), MDP (1 μg/ml), LPS (10 ng/ml), 19-kDa lipoprotein (concentrations as described below), or combinations of MDP and 19-kDa lipoprotein. The influence of internalization of M. tuberculosis on cytokine production was investigated by adding 20 μg/ml cytochalasin B during stimulation with the microorganism. Positive control stimulation for the effects of cytochalasin B was provided by stimulation of cells with zymosan (1 μg/ml; Sigma). All stimuli were checked for the contamination with LPS in the Limulus amoebocyte lysate assay and found to be negative. Cytochalasin B did not influence cell viability (unpublished data). To evaluate the role of TLR4 in the induction of cytokines, cells were preincubated with 10 μg/ml of a blocking monoclonal anti-TLR4 antibody (eBioscience, San Diego, California, United States).
Cytokine production by murine peritoneal macrophages.
Resident peritoneal macrophages from either ScCr (TLR4-defective) or C57Bl/10J (TLR4 control) mice, TLR2−/− or control TLR2+/+ mice (kindly provided by S. Akira, Osaka, Japan) [30], NOD1−/− and NOD1+/+ littermates (kindly provided by E. Abraham, Denver, Colorado, United States), or NOD2+/+ and NOD2−/− (from M. Giovannini, CEPH, Paris), backcrossed to the seventh generation into the C57BL6/J background, were harvested by injection of 4 ml of sterile PBS containing 0.38% sodium citrate [31]. After centrifugation and washing, the cells were resuspended in RPMI 1640 containing 1 mM pyruvate, 2 mM L-glutamine, 100 μg/ml gentamicin, and 2% fresh mouse plasma. Cells were cultured in 96-well microtiter plates (Greiner) at 1 × 105 cells/well, in a volume of 100 μl. The cells were stimulated with purified LPS (1 μg/ml), Pam3Cys (1 μg/ml), FK-156 ligand of murine NOD1 (1 μg/ml) as a control stimulation, or a sonicate of M. tuberculosis (1 μg/ml). After 24 h incubation at 37 °C, the supernatants were collected and stored at −70 °C until cytokine assays were performed.
Cytokine measurements.
Human and murine TNFα concentrations were determined by specific RIAs as described [32,33]. IL-10 and IL-6 were measured by a commercial ELISA kits (Pelikine Compact, CLB, Amsterdam, The Netherlands), according to the instructions of the manufacturer.
Signaling through human TLR2 and TLR4 in transfected cell lines.
CHOs stably transfected with human CD14 (3E10-CD14), a combination of CD14 and TLR2 (3E10-TLR2), or TLR4 (3E10-TLR4), were a kind gift from R. Ingalls [34]. All cell lines express inducible membrane CD25 under control of a region from the human E-selectin (ELAM-1) promoter containing NF-κB binding sites. Cells were maintained at 37 °C and 5% CO2 in HAM's F12 medium (Gibco, Invitrogen, Breda, The Netherlands) supplemented with 10% FCS, 0.01% L-glutamine, 50 μg/ml gentamicin, and either 400 U/ml hygromycin and 0.5 mg/ml of G418 (for 3E10-TLR2) or 0.05 mg/ml of puromycin (for 3E10-TLR4) as additional selection antibiotics. TLR2 and TLR4 expression was confirmed by flow cytometry (Coulter Epics XL-MCL, Beckman Coulter, Mijdrecht, the Netherlands) using PE-labeled anti-TLR2 (clone TL2.1) or anti-TLR4 (clone HTA125) (Immunosource, Halle-Zoersel, Belgium).
For stimulation experiments, 500 μl of cells in culture medium at a density of 1 × 105/ml were plated in 24-well culture plates. After an overnight incubation, cells were incubated with control medium, M. tuberculosis sonicate (10 μg/ml), Pam3Cys (10 μg/ml), or LPS (1 μg/ml) for 20 h, and thereafter cells were harvested using trypsin/EDTA (Cambrex, East Rutherford, New York, United States) and prepared for flow cytometry (Coulter FACS-scan). CD25 expression of the CHOs was measured using FITC-labeled anti-CD25 (DAKO, Glostrup, Denmark), and expressed as folds-over-mean increase.
Stimulation of NOD-transfected cell lines and NF-κB translocation.
Studies examining the activation of NF-κB by M. tuberculosis in cells overexpressing NOD1 or NOD2 were carried out as previously described [19]. Briefly, 1 × 106/ml HEK293T cells were transfected overnight with 1 ng of either NOD1 or NOD2 plus 75 ng luciferase reporter plasmid. At the same time, heat-killed whole M. tuberculosis preparations (ratio of microorganisms and effector cells 1:10, 1:1, and 10:1) were added to cell culture medium, and the NF-κB–dependent luciferase activation was then measured following 24 h of incubation. NF-κB–dependent luciferase assays were performed in duplicate, and data represent three independent experiments [19].
Statistical analysis.
The human experiments were performed in triplicate with blood obtained from patients and volunteers. The mouse experiments were performed twice in 10 mice per group, and the data are presented as cumulative results of all experiments performed. The differences between groups were analyzed by unpaired Student t-test, and where appropriate by paired t-test. The level of significance between groups was set at p < 0.05. The data are given as means ± standard deviation (SD).
We thank Dr. Shizuo Akira, Osaka University, Japan for providing us with TLR2
−/− mice, Dr. Edward Abraham, University of Colorado, Denver, Colorado, United States for the NOD1
−/− mice, and Dr. Marco Giovannini and Dr. Jean-Pierre Hugot, Robert Debre Hospital, Paris, France, for the initial generation of the NOD2
−/− mice. MGN was supported by a VIDI-grant of the Netherlands Organization for Scientific Research. We thank Trees Jansen and Liesbeth Jacobs for the help with the cytokine measurements.
Competing Interests. The authors have declared that no competing interests exist.
Author Contributions. GF, SEG, BK, THMO, JWMV, and MGN conceived and designed the experiments. SEG, LL, DJD, and DMLL performed the experiments. GF, SEG, RV, GJA, and MGN analyzed the data. LL, DJD, DMLL, RV, and THMO contributed reagents/materials/analysis tools. GF, SEG, BK, GJA, THMO, JWMV, and MGN wrote the paper.
Abbreviations
CHOChinese hamster ovary fibroblast
HEK293Thuman embryonic kidney 293T cell
LPSlipopolysaccharide
LRRleucine-rich repeat
MDPmuramyl dipeptide
MNCmononuclear cell
NODnucleotide-binding oligomerization domain
PRRpattern recognition receptor
SDstandard deviation
TLRToll-like receptor
TNFtumor necrosis factor
==== Refs
References
Frieden TR Sterling TR Munsiff SS Watt CJ Dye C 2003 Tuberculosis Lancet 362 887 899 13678977
Van Crevel R Ottenhoff TH van der Meer JW 2002 Innate immunity to Mycobacterium tuberculosis
Clin Microbiol Rev 15 294 309 11932234
Means TK Wang S Lien E Yoshimura A Golenbock DT 1999 Human Toll-like receptors mediate cellular activation by Mycobacterium tuberculosis
J Immunol 163 3920 3927 10490993
Aliprantis AO Yang R-B Mark MR Sugget S Devaux B 1999 Cell activation and apoptosis by bacterial lipoproteins through Toll-like receptor-2 Science 285 736 739 10426996
Vignal C Guerardel Y Kremer L Masson M Legrand D 2003 Lipomannans, but not lipoarabinomannans, purified from Mycobacterium chelonae and Mycobacterium kansasii induce TNF-alpha and IL-8 secretion by a CD14-Toll-like receptor-2 -dependent mechanism J Immunol 171 2014 2023 12902506
Gilleron M Quesniaux VF Puzo G 2003 Acylation state of the phosphatidylinositol hexamannosides from Mycobacterium bovis bacillus Calmette Guérin and Mycobacterium tuberculosis H37Rv and its implication in Toll-like receptor response J Biol Chem 278 29880 29889 12775723
Nigou J Zelle-Rieser C Gilleron M Thurnher M Puzo G 2001 Mannosylated lipoarabinomannans inhibit IL-12 production by human dendritic cells: Evidence for a negative signal delivered through the mannose receptor J Immunol 166 7477 7485 11390501
Geijtenbeek TB Van Vliet SJ Koppel EA Sanchez-Hernandez M Vanderbroucke-Grauls CM 2003 Mycobacteria target DC-SIGN to suppress dendritic cell function J Exp Med 197 7 17 12515809
Reiling N Holscher C Fehrenbach A Svenja K Kirschning CJ 2002 Toll-like receptor (TLR)2- and TLR4-mediated pathogen recognition in resistance to airborne infection with Mycobacterium tuberculosis
J Immunol 169 3480 3484 12244136
Kamath AB Alt J Debbabi H Behar SM 2003 Toll-like receptor-4 defective C3H/HeJ mice are not more susceptible than other C3H substrains to infection with Mycobacterium tuberculosis
Infect Immun 71 4112 4118 12819102
Abel B Thieblemont N Quesniaux VF Brown N Mpagi J 2002 Toll-like receptor 4 expression is required to control chronic Mycobacterium tuberculosis infection in mice J Immunol 169 3155 3162 12218133
Branger J Leemans JC Florquin S Weijer S Speelman P 2004 Toll-like receptor 4 plays a protective role in pulmonary tuberculosis in mice Int Immunol 16 509 516 14978024
Drennan MB Nicolle D Quesniaux VF Jacobs M Allie N 2004 Toll-like receptor 2 deficient mice succumb to Mycobacterium tuberculosis infection Am J Pathol 164 49 57 14695318
Sugawara I Yamada H Li C Mizuno S Takeuchi O 2003 Mycobacterial infection in TLR2 and TLR6 knockout mice Microbiol Immunol 47 327 336 12825894
Fremond C Yeremeev V Nicolle DM Jacobs M Quesniaux VF 2004 Fatal Mycobacterium tuberculosis infection despite adaptive immune response in the absence of MyD88 J Clin Invest 114 1790 1799 15599404
Hugot J-P Chamaillard M Zouali H Lesage S Cezard J-P 2001 Association of NOD2 leucine-rich repeat variants with susceptibility to Crohn's disease Nature 411 599 603 11385576
Ogura Y Bonen DK Inohara N Nicolae L Chen FF 2001 A frameshift mutation in NOD2 associated with susceptibility to Crohn's disease Nature 411 603 606 11385577
Girardin SE Boneca IG Viala J Chamaillard M Labigne A 2003 Nod2 is a general sensor of peptidoglycan through muramyl dipeptide (MDP) detection J Biol Chem 278 8869 8872 12527755
Girardin SE Boneca IG Carneiro LAM Antignac A Jehanno M 2003 Nod1 detects a unique muropeptide from gram-negative bacterial peptidoglycan Science 300 1584 1587 12791997
Quesniaux VF Fremond C Jacobs M Parida S Nicolle D 2004 Toll-like receptor pathways in the immune responses to mycobacteria Microbes Infect 6 946 959 15310472
Netea MG Kullberg BJ de Jong D Francke B Sprong T 2004 NOD2 mediates induction of the antiinflammatory signals induced by TLR2-ligands: Implications for Crohn's disease Eur J Immunol 34 2052 2059 15214053
Shi S Nathan C Schnappinger D Drenkow J Fuortes M 2003 MyD88 primes macrophages for full-scale activation by interferon-gamma yet mediates few responses to Mycobacterium tuberculosis
J Exp Med 198 987 997 14517275
Ting JP-Y Davis BK 2005 CATERPILLER: A novel gene family important in immunity, cell death, and disease Annu Rev Immunol 23 387 414 15771576
Netea MG Ferwerda G De Jong DJ Werts C Boneca IG 2005 The frameshift mutation in Nod2 results in unresponsiveness not only to Nod2-, but also Nod1-activating peptidoglycan agonists J Biol Chem In press.
Netea MG Ferwerda G De Jong DJ Jansen T Jacobs L 2005 NOD2 modulates specific Toll-like receptor pathways for the induction of cytokine release J Immunol 174 6518 6523 15879155
Opitz B Puschel A Schmeck B Hocke AC Rosseau S 2004 Nucleotide-binding oligomerization domain proteins are innate immune receptors for internalized Streptococcus pneumoniae
J Biol Chem 35 36426 36432
Greenstein RJ 2003 Is Crohn's disease caused by a mycobacterium? Comparisons with leprosy, tuberculosis, and Johne's disease Lancet Infect Dis 3 507 514 12901893
Naser SA Ghobrial G Romero C Valentine JF 2004 Culture of Mycobacterium avium subspecies paratuberculosis from the blood of patients with Crohn's disease Lancet 364 1039 1044 15380962
Endres S Ghorbani R Lonnemann G Van der Meer JWM Dinarello CA 1988 Measurement of immunoreactive interleukin-1 beta from human mononuclear cells: Optimization of recovery, intrasubject consistency, and comparison with interleukin-1 alpha and tumor necrosis factor Clin Immunol Immunopathol 49 424 438 2461270
Takeuchi O Hoshino K Kawai T Sanjo H Takada H 1999 Differential roles of TLR2 and TLR4 in recognition of gram-negative and gram-positive bacterial cell wall components Immunity 11 443 451 10549626
Kullberg BJ Van't Wout JW Hoogstraten C Van Furth R 1993 Recombinant interferon-γ enhances resistance to acute disseminated Candida albicans infection in mice J Infect Dis 168 436 443 8335982
Drenth JPH Van Uum SHM Van Deuren M Pesman GJ Van der Ven-Jongekrijg J 1995 Endurance run increases circulating IL-6 and IL-1ra but downregulates ex vivo TNF-α and IL-1β production J Appl Physiol 79 1497 1503 8594005
Netea MG Demacker PNM Kullberg BJ Boerman OC Verschueren I 1996 Low-density-lipoprotein receptor deficient mice are protected against lethal endotoxinemia and severe gram-negative infections J Clin Invest 97 1366 1372 8617867
Lien E Means TK Heine H Yoshimura A Kusumoto S 2000 Toll-like receptor 4 imparts ligand-specific recognition of bacterial lipopolysaccharide J Clin Invest 105 497 504 10683379
|
16322770
|
PMC1291354
|
CC BY
|
2021-01-05 12:11:25
|
no
|
PLoS Pathog. 2005 Nov 25; 1(3):e34
|
utf-8
|
PLoS Pathog
| 2,005 |
10.1371/journal.ppat.0010034
|
oa_comm
|
==== Front
PLoS PathogPLoS PathogppatplpaplospathPLoS Pathogens1553-73661553-7374Public Library of Science San Francisco, USA 10.1371/journal.ppat.001003805-PLPA-OP-0177plpa-01-03-12OpinionsCan Specialized Pathogens Colonize Distantly Related Hosts? Schistosome Evolution as a Case Study OpinionsBrant Sara V *Loker Eric S **To whom correspondence should be addressed. E-mail: [email protected] (SVB); [email protected] (ESL)Sara V. Brant and Eric S. Loker are at the Department of Biology, University of New Mexico, Albuquerque, New Mexico, United States of America.
11 2005 25 11 2005 1 3 e38Copyright: © 2005 Brant and Loker.2005This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are properly credited. Citation:Brant SV, Loker ES (2005) Can specialized pathogens colonize distantly related hosts? Schistosome evolution as a case study. PLoS Pathog 1(3): e38.
==== Body
Parasites that live in intimate contact with the immune system of their hosts require specialized adaptations to survive in such exposed environments. Once adapted to these demanding environments, it seems such parasites could not easily switch to distantly related hosts [1], and, thus, would be good candidates to diversify congruently with their hosts, i.e., cospeciation. One of the best-known parasite groups is the schistosomes, digenetic (having alternating sexually and asexually reproducing generations in their life cycle) trematodes that live in the vascular system of their vertebrate hosts. Schistosomes achieve notoriety because six of the roughly 100 described species [2,3] cause schistosomiasis, a disease that afflicts 200 million people, mostly in tropical Africa.
Schistosomiasis is usually chronic and debilitating in its course, with most of the pathogenesis caused not directly by the adult worms but by the eggs they produce. Eggs become lodged in the viscera and incite tissue responses, often causing pronounced enlargements in the liver and spleen, and abnormalities in blood flow through these organs. If worms colonize the urinary system, hematuria (blood in the urine) and kidney and bladder damage often result. Schistosomiasis negatively affects growth and productivity, and has largely underappreciated, insidious effects on the people with this disease. Adult worms can be killed by drugs, but the limited availability and high cost of these drugs and the potential for emergence of drug resistance are important concerns. Immunity is slow to develop, though hope for an effective vaccine remains high.
Schistosomes infect birds or mammals, but one species, Griphobilharzia amoena, often considered the missing link in schistosome evolution, is known to infect freshwater crocodiles [4]. Schistosomes share the habit of living in a vascular habitat with other trematodes, including the Spirorchiidae of turtles and the Sanguinicolidae of fishes. Worms in these three families have two-host life cycles—a snail host and a vertebrate host—and also share a distinctive tegument, or body covering, consisting of two lipid bilayers instead of the typical single bilayer. The double bilayer is believed to be an adaptation for survival in the immunologically hostile environment of the blood [5]. Schistosomes differ from the other two families of blood flukes, though, by being dioecious (having separate male and female worms) and dimorphic (with the two sexes different in morphology), and by having specialized habitat requirements. The remarkable biology of schistosomes has precipitated considerable discussion regarding their origins and their evolution of dioecy (the change from the typical state in trematodes of being hermaphrodites to a state with separate males and females) [2,4,6–10].
The discovery of G. amoena [4], the only species of schistosome known in an ectotherm, gave rise to a hypothesis that schistosomes arose in early ectothermic archosaurs, for example, ancestors of modern crocodiles, and radiated into endothermic archosaurs (birds). This view was supported by a phylogenetic analysis of adult morphology, which placed G. amoena as basal, or ancestral, among schistosomes [11], and challenged the role of endothermy as the pivotal factor driving schistosome diversification [10–12]. Molecular phylogenetic studies to date [7,13,14] have been hampered by an incomplete sampling of the 13 widely recognized schistosome genera, including the provocative and putatively basal G. amoena. The molecular phylogeny in Figure 1 includes representatives of all the commonly recognized genera of schistosomes, and spirorchiids from both freshwater and marine turtles [7]. Included in this molecular phylogeny is G. amoena, specimens of which were recovered from the Australian freshwater crocodile, Crocodylus johnstoni, obtained near Darwin, Australia.
Figure 1 Maximum Likelihood Estimated Tree from the Combined Data Partitions of 6,335 Bases Derived from 18S and 28S Ribosomal DNA and Partial Cytochrome Oxidase I Mitochondrial DNA Genes
Nodal support values are indicated on the branch as bootstrap values for maximum parsimony/minimum evolution/Bayesian posterior probabilities [27–28]. To the right of the tree are listed the vertebrate host groups: birds are light green, mammals are dark green, marine turtles, freshwater turtles, freshwater crocodiles, (GenBank accession numbers 28S: AY899914; 18S: AY899915; cytochrome oxidase I: AY899916), and fish are black, and the families of snail hosts used by the various schistosome species are blue indicating families belonging to the Pulmonata, orange to the Caenogastropoda, and red to the Opisthobranchia. At our present state of knowledge, schistosomes that infect birds are not united in a single lineage, nor are those that infect mammals, which is suggestive of at least two separate colonizations of at least one of these lineages. With respect to snail hosts, in at least two instances, even congeneric schistosomes depend on markedly divergent gastropod lineages (pulmonates versus opisthobranchs or caenogastropods), suggesting that, historically, host switching has also been extensive within molluscan hosts [7,13,14,20].
Our analysis shows that G. amoena is, in fact, not a basal schistosome, but is more closely related to spirorchiids from freshwater turtles. This expands the host range of spirorchiids to include reptiles other than turtles, and suggests that schistosomes parasitize only endotherms. Our analysis confirms that the sister group to the schistosomes are the spirorchiids from marine turtles [7], and that the basal schistosomes are parasites of marine birds and snails (Figure 1). This pattern supports the idea that a long-range host switch from turtles to avian hosts occurred in marine habitats, and that schistosomes subsequently colonized birds, mammals, and freshwater snails. This argues against a hypothesis of a long-term association between schistosomes and archosaurs (crocodilians), and suggests that exploitation of endotherms has been the key factor leading to the emergence of schistosomes and their dioecious condition [2,8]. We speculate that the transferal of a spirorchiid protoschistosome to an endothermic host was favored by the partial endothermy [15,16] of their ancestral marine turtle hosts.
Endotherms have metabolic rates that are roughly one order of magnitude higher than those of ectotherms of comparable size, and they consequently ingest more food [17]. Nutrients are then conveyed to the liver via the hepatic portal system. Most species of schistosomes live in the hepatic portal system. We argue that schistosomes colonized this specific habitat in endotherms to take advantage of its high nutrient content. In contrast, spirorchiids prefer the heart and arterial system [10], but can be found in widely scattered locations in the vascular system [18,19]. One of the unique demands imposed by the hepatic portal system is the need for the adult worms to move against the flow of blood to reach the small blood vessels surrounding the intestinal wall where oviposition would best lead to expulsion of the eggs in the feces. This requires a difficult combination of strength, to allow movement against the flow of blood, and delicacy, to allow extension of the body into the smallest vessels for oviposition.
One solution is to forego the normal digenean condition of hermaphroditism to relegate each of these functions to separate sexes [2]. The muscular schistosome males are specialized to move females to preferred oviposition sites, whereas the delicate, thread-like females [10] are adapted to insinuate into tiny vessels for oviposition. Precision oviposition is crucial for schistosomes because eggs swept into the liver are not only a dead end from the parasite's point of view, but they also impose significant pathology on their hosts [10].
Schistosomes remain paradoxical because they are specialized in morphology and habitat, and given the intimacy of contact with their host, it seems these worms would become restricted to particular hosts and consequently exhibit cophylogenetic patterns with those hosts. Our results (Figure 1), however, suggest that, historically, schistosomes have undertaken numerous long-range host switches among both vertebrates and snail hosts [7,13,14,20]. Observations that avian schistosomes can persist for a surprisingly long period in mammalian hosts, and likewise for some mammalian schistosomes in avian hosts, provide further evidence for this propensity [21–23].
We propose that schistosomes retain a pleisiomorphic (ancestral trait), nonspecific immune evasion strategy, such that long-distance host switches occasionally occur. The double tegumental membrane and the ability of the tegument to acquire macromolecules from many different host species [5] may contribute to this switching ability. With the possible exception of some species in derived taxa, such as duck schistosomes of the genus Trichobilharzia (see Figure 1) [24], the phylogeny of schistosomes does not mirror that of its hosts. Free-swimming larval schistosomes, the miracidia (for snails) and cercariae (for vertebrates), no doubt incessantly attempt to colonize new potential hosts, creating opportunities for host switches. Our results suggest that there are fundamental immune evasive mechanisms dictating schistosome success in intravascular environments in humans and other vertebrate hosts that we do not yet fully appreciate. They also suggest that morphological and microhabitat specialization need not preclude successful colonization of new hosts, indicating that congruent patterns in host and parasite phylogenies may be rare in these specialized worms [25,26].
We thank Scott Snyder for his help in the field and in finding these worms, and we thank D. Blair, O. Pugh, N. Robinson, V. Simlesa, L. Small, and M. Turner for their help on various aspects of obtaining worms from crocodiles, and use of facilities and expertise in Australia. We acknowledge technical support from the University of New Mexico's Molecular Biology Facility and support from NIH grant 1P20RR18754 from the Institute Development Award Program of the National Center for Research Resources.
==== Refs
References
Adamson ML Caira JN 1995 Evolutionary factors influencing the nature of parasite specificity Parasitology 109 S85 S95
Basch PF 1991 Schistosomes: Development, reproduction and host relations Oxford Oxford University Press 248 p.
Khalil LF 2002 Schistosomatoidea Gibson DI Jones A Bray RA Keys to the Trematoda Wallingford CABI Publishing 419 432
Platt TR Blair D Purdie J Melville L 1991
Griphobilharzia amoena n. gen., n. sp. (Digenea: Schistosomatidae), a parasite of the freshwater crocodile Crocodylus johnstoni (Reptilia: Crocodylia) from Australia, with the erection of a new subfamily, Griphobilharziinae
J Parasitol 77 65 68
McLaren DJ Hockley DJ 1977 Blood flukes have a double outer membrane Nature 269 147 149 71658
Carmichael AC 1984 Phylogeny and historical biogeography of the Schistosomatidae [PhD thesis] East Lansing (Michigan) Michigan State University 1 246
Snyder SD 2004 Phylogeny and paraphyly among tetrapod blood flukes (Digenea: Schistosomatidae and Spirorchiidae) Int J Parasitol 34 1385 1392 15542099
Short RB 1983 Sex and the single schistosome J Parasitol 69 3 22 6338197
Despres L Maurice S 1995 The evolution of dimorphism and separate sexes in schistosomes Proc R Soc Lond B Biol Sci 262 175 180
Platt TR Brooks DR 1997 Evolution of the schistosomes (Digenea: Schistosomatoidea): The origin of dioecy and colonization of the venous system J Parasitol 83 1035 1044 9406775
Morand S Muller-Graf CD 2000 Muscles or testes? Comparative evidence for sexual competition among dioecious blood parasites (Schistosomatidae) of vertebrates Parasitology 120 45 56 10726265
Combes C Després L Establet D Fournier A Jourdane J 1991 Schistosomatidae (Trematoda): Some views on their origin and evolution Res Rev Parasitol 51 25 28
Snyder SD Loker ES 2000 Evolutionary relationships among the Schistosomatidae (Platyhelminthes: Digenea) and an Asian origin for Schistosoma
J Parasitol 86 283 288 10780546
Lockyer AE Olson PD Ostergaard P Rollinson D Johnston DA 2003 The phylogeny of the Schistosomatidae based on three genes with emphasis on the interrelationships of Schistosoma Weinland, 1858 Parasitology 126 203 224 12666879
Spotila JR Standora EA 1985 Environmental constraints on the thermal energetics of sea turtles Copeia 3 694 702
Paladino FV O'Connar MP Spotila JR 1990 Metabolism of leatherback turtles, gigantothermy, and thermoregulation of dinosaurs Nature 344 858 860
Pough FH Janis CM Heiser JB 2005 Vertebrate life. 7th edition Upper Saddle River (New Jersey) Prentice Hall 752 p.
Platt TR 2002 Spirorchiidae Gibson DI Jones A Bray R Keys to the Trematoda Wallingford CABI Publishing 453 467
Ulmer MJ 1959 Studies on Spirorchis haematobium (Stunkard, 1922) Price, 1934 (Trematoda: Spirorchiidae) in the definitive host Trans Am Microsc Soc 78 81 89
Blair D Davis GM Wu B 2001 Evolutionary relationships between trematodes and snails emphasizing schistosomes and paragonimids Parasitology 123 S229 S243 11769286
Oliver L 1953 Observations on the migration of avian schistosomes in mammals previously unexposed to cercariae J Parasitol 39 237 243 13053304
Horak P Dvorak J Kolarova L Trefil L 1999
Trichobilharzia regenti , a pathogen of the avian and mammalian central nervous systems Parasitology 119 577 581 10633919
Bayssade-Dufour C Vuong PN Rene M Martin-Loehr C Martins C 2002 Lesions viscerales de mammiferes et oiseaux, exposes aux agents de dermatite cercarienne humaine Bull Soc Pathol Exot 95 229 237 12596366
Blair D Islam KS 1983 The life cycle and morphology of Trichobilharzia australis n.sp. (Digenea: Schistosomatidae) from the nasal blood vessels of the black duck (Anas superciliosa ) in Australia, with a review of the genus Trichobilharzia
Syst Parasitol 5 89 117
Johnson KP Clayton DH 2003 Coevolutionary history of ecological replicates: Comparing phylogenies of wing and body lice to columbiform hosts Page RDM Tangled trees: Phylogeny, cospeciation, and coevolution Chicago University of Chicago Press 262 286
Criscione C Blouin M 2004 Life cycles shape parasite evolution: Comparative population genetics of salmon trematodes Evolution 58 198 202 15058733
Swofford DL 2000 PAUP*. Phylogenetic Analysis Using Parsimony* and other methods, Version 4 [software] New York Sinauer Associates
Huelsenbeck JP Ronquist F 2001 MrBayes: Bayesian inference of phylogenetic trees Bioinformatics 17 754 755 11524383
|
16322771
|
PMC1291355
|
CC BY
|
2021-01-05 12:11:25
|
no
|
PLoS Pathog. 2005 Nov 25; 1(3):e38
|
utf-8
|
PLoS Pathog
| 2,005 |
10.1371/journal.ppat.0010038
|
oa_comm
|
==== Front
Aust New Zealand Health PolicyAustralia and New Zealand Health Policy1743-8462BioMed Central London 1743-8462-2-261626290810.1186/1743-8462-2-26ReviewGeneral practice and the New Zealand health reforms – lessons for Australia? McAvoy Brian R [email protected] Gregor D [email protected] Department of General Practice, School of Primary Health Care, Monash University 867 Centre Road, East Bentleigh, Victoria 3165, Australia2 Department of General Practice and Primary Health Care, University of Auckland Private Bag 92019, Auckland, New Zealand2005 2 11 2005 2 26 26 8 7 2005 2 11 2005 Copyright © 2005 McAvoy and Coster; licensee BioMed Central Ltd.2005McAvoy and Coster; licensee BioMed Central Ltd.This is an Open Access article distributed under the terms of the Creative Commons Attribution License (), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
New Zealand's health sector has undergone three significant restructures within 10 years. The most recent has involved a Primary Health Care Strategy, launched in 2001. Primary Health Organisations (PHOs), administered by 21 District Health Boards, are the local structures for implementing the Primary Health Care Strategy. Ninety-three percent of the New Zealand population is now enrolled within 79 PHOs, which pose a challenge to the well-established Independent Practitioner Associations (IPAs).
Although there was initial widespread support for the philosophy underlying the Primary Health Care Strategy, there are concerns amongst general practitioners (GPs) and their professional organisations relating to its implementation. These centre around 6 main issues:
1. Loss of autonomy
2. Inadequate management funding and support
3. Inconsistency and variations in contracting processes
4. Lack of publicity and advice around enrolment issues
5. Workforce and workload issues
6. Financial risks
On the other hand, many GPs are feeling positive regarding the opportunities for PHOs, particularly for being involved in the provision of a wider range of community health services. Australia has much to learn from New Zealand's latest health sector and primary health care reforms.
The key lessons concern:
• the need for a national primary health care strategy
• active engagement of general practitioners and their professional organisations
• recognition of implementation costs
• the need for infrastructural support, including information technology and quality systems
• robust management and governance arrangements
• issues related to critical mass and population/distance trade offs in service delivery models
==== Body
Review
Preamble
The most recent New Zealand health reforms can be viewed from a wide range of perspectives, depending on whether you are a consumer, general practitioner, practice nurse, policy maker or health services manager. This review has been written from the general practice perspective by an Australian GP academic who has practised in both Australia and New Zealand and by a New Zealand GP academic. The paper is based on a review of policy and discussion documents, peer-reviewed publications, websites and discussions with New Zealand colleagues.
Context
Over the past 15 years or so, across most developed nations of the world the combination of burgeoning medical technology, ageing populations and increased patient expectations has led to dramatic escalations in the costs of providing health and social care. In 2001 Australia spent US$2,504 per capita on health, 9.1% of its Gross Domestic Product (GDP). New Zealand, with 20% of Australia's population, spent US$ 1,710 per capita, 8% of its GDP [1]. Not surprisingly, reforms of health care and social welfare have been undertaken across Europe, North America, South Africa, Australia and New Zealand. Barbara Starfield's review of health care in 11 industrialised nations concluded that a primary care orientation is associated with lower costs of care, higher satisfaction of the population with its health services, better health levels and lower medication use [2]. The upsurge of interest in primary care at such high levels of policy-making in so many different countries is related to the recognition of its potential to limit the escalating costs of secondary and tertiary care.
Background
New Zealand's health sector has been subject to continual change since the early 1990s, undergoing three significant restructures within 10 years. These recent developments in the funding and organisation of the New Zealand health sector have been reviewed by Ashton who notes that after a decade of turbulence the sector now appears to be more stable [3]. The 1991 Green and White Paper ushered in an era of market oriented reforms which assumed that a purchaser-provider split and competition between health care providers would result in more efficient delivery of health services and, implicitly, improved health outcomes [4]. The reforms were intended to:
• Increase choice and access for all New Zealanders in a health care system that was effective, fair and affordable
• Encourage efficiency, flexibility and innovation in health care delivery
• Increase accountability to purchasers
• Reduce hospital waiting times
• Enhance the working environment for health professionals.
Four Regional Health Authorities (RHAs) were established, and the hospital and community services previously provided by 14 Area Health Boards were reconfigured into 23 Crown Health Enterprises (CHEs). The CHEs were required to manage their resources in a business-like fashion with the objective of being 'as successful and efficient as comparable businesses that are not owned by the Crown' [5]. However, the anticipated benefits of this 'experiment with competition' were not delivered [6]. In 1996 a briefing to the incoming Minister of CHEs stated 'the health reforms have yet to yield the original expectations. By a range of measures the pace of performance seems, if anything, to have weakened since the advent of the reforms' [7]. The CHEs' experience raises questions about the degree to which business models such as quasi-markets can be applied to public health provision. There were reductions in general practice subsidies, erosion of practice nurse subsidies and many primary care services including maternity, well-child and sexual health services were fragmented [8]. On a more positive note there emerged Maori health providers, community health trusts and most significantly, Independent Practitioner Associations (IPAs). These are similar in many respects to Australian Divisions of general practice and UK primary care groups but are owned and controlled independently by GPs themselves.
Currently over 75% of New Zealand GPs are members of over 30 IPAs which vary in size from 7 to 340 GP members (mean 74), and there is now an IPA Council of New Zealand (IPAC). IPAs cover more than 800 community-based practices, attended by some 2,200 GPs and more than 2,000 practice nurses. Developments in contracting and alternative methods of funding and managing services were initially either resisted strongly or treated with caution by the majority of GPs. The main opposition was voiced by the GP Action Group and to a lesser extent, for a short period, by the then New Zealand General Practitioners Association and the New Zealand Medical Association (NZMA). Early successes in contracting, in budget holding for pharmaceutical and laboratory services and establishing new services, arising in part from budget holding savings, led to gradual and progressive recruitment of IPA membership [9]. In both New Zealand and Australia the 'public' provision of primary health care remains via market driven private practice. Fortunately the reforms did not have to deal with any market failures in the quasi-market arrangements.
In 1996 New Zealand's first proportionally elected National (conservative) led coalition government signalled a change of direction. A single purchaser, the Health Funding Authority, replaced the four RHAs, the CHEs became 'Hospitals and Health Services' and had their 'for profit' status removed. 'Cooperation' replaced 'competition' as the new political catch-cry [10]. During this time the IPAs consolidated, developing well-established infrastructures, including staff, information systems, clinical guidelines, peer discussion groups, and personalised feedback on clinical performance. They also began to develop expertise in budget holding for laboratory tests and pharmaceuticals, making savings to develop new and better services [11]. The IPAs made significant efforts to manage pharmaceutical and laboratory expenditure with the savings achieved providing significant funding for a variety of new service developments. The ability to use some of these savings was important for the development of the IPAs [12]. However, the acquisition of such funding from savings also led to ongoing conflict with the Health Funding Authority which became embroiled in bureaucratic controlling processes in order to endeavour to recoup some of that funding.
Further radical changes followed the election of the Labour led coalition in 1999. The main structural change was abolition of the Health Funding Authority and its replacement by 21 new District Health Boards (DHBs) commencing in 2001, comprising a majority of locally elected and a minority of ministerially appointed members, accountable to the Minister of Health [13]. This was intended to strengthen local democratic input to decisions. Funding was now to be allocated between DHBs according to a formula based on the local population weighted for relative health need. Coupled with these structural changes a series of national strategies have been developed to guide the system; these identify objectives and priorities for improving health and independence levels in the population, aim to reduce the 'health gap' between Maori and non-Maori, and specify how services should be delivered [10].
The New Zealand Health Strategy was published in 2000 [14]. This provided an overall framework for the heath sector, with the aim of directing health services at those areas that would provide the greatest benefit for the population, focusing in particular on tackling inequalities in health (see Table 1). Primary health care is one of five service delivery areas in the New Zealand Health Strategy (see Table 2), which identifies seven fundamental principles for the health sector (see Table 3), and out of a total of ten goals and 61 objectives, highlighted 13 population health objectives (see Table 4). Particular priorities included cancer, cardiovascular disease, diabetes and mental health. The New Zealand Health Strategy has set out the strategic direction for the development of health services in New Zealand, based on a model for improving health outcomes. The lesson for Australia is that Australia should have a national health strategy, including a national primary health policy. Many providers have argued this for years – to deaf Commonwealth ears.
Table 1 New Zealand Health Strategy Priority Objectives to Reduce Inequalities
• Ensure accessible and appropriate services for people from lower socio-economic groups
• Ensure accessible and appropriate services for Maori
• Ensure accessible and appropriate services for Pacific Peoples.
Table 2 New Zealand Health Strategy Service Delivery Priority Areas
• Public health
• Primary heath care
• Reducing waiting times for public hospital elective services
• Improving responsiveness of mental health services
• Accessible and appropriate services for people living in rural areas
Table 3 New Zealand Health Strategy Principles
• Acknowledging the special relationship between Maori and the Crown under the Treaty of Waitangi
• Good health and wellbeing for all New Zealanders throughout their lives
• An improvement in health status of those currently disadvantaged
• Collaborative health promotion and disease and injury prevention by all sectors
• Timely and equitable access for all New Zealanders to a comprehensive range of health and disability services, regardless of ability to pay.
• A high performing system in which people have confidence.
• Active involvement of consumers and communities at all levels.
Table 4 New Zealand Health Strategy Population Health Objectives
• Reduce smoking
• Improve nutrition
• Increase the level of physical activity
• Reduce the rates of suicide and suicide attempts
• Minimise harm caused by alcohol, illicit and other drug use to both individuals and the community
• Reduce the incidence and impact of cancer
• Reduce the incidence and impact of cardiovascular disease
• Reduce the incidence and impact of diabetes
• Improve oral health
• Reduce violence in interpersonal relationships, families, schools and communities
• Improve the health status of people with severe mental illness
• Ensure access to appropriate child health care services including well child and family health care, and immunisation.
A New Zealand Disability Strategy was also developed, with fifteen objectives [15]. A significant policy shift towards population-based approaches was signalled by the National Health Committee [16], based on a paper by Coster and Gribben who proposed new primary health organisations with a focus on population-based health outcomes [17]. The New Zealand Health Strategy and Disability Strategy both informed the Primary Health Care Strategy, published in February 2001 [18]. The latter is a key document and promised to achieve a new vision over five to ten years with the following:
• People will be part of local primary health care services that improve their health, keep them well, are easy to get to and coordinate their ongoing care
• Primary health care services will focus on better health for a population, and actively work to reduce health inequalities between different groups.
Recent health policy developments in primary health care in New Zealand are redefining general practice to align with the 1978 Alma Ata Declaration [19]. The vision involves a new direction for primary health care with a greater emphasis on population health and the role of the community, health promotion and preventive care, the need to involve a range of professionals, and the advantage of funding based on population needs rather than fee for service. This reflects a desire by the New Zealand government to reduce health inequalities between different population groups, and protect and promote the health of its population. Central to the Primary Health Care Strategy are the new arrangements for primary health care, which are administered through the DHBs, supported by the Ministry of Health, which is the national policy advice, regulatory, funding and monitoring agency (see Figure 1). Primary Health Organisations (PHOs) are the local structures for implementing the Primary Health Care Strategy, and have the following features, set out in the Minimum Requirements released by the Health Minister in November 2001 [20]:
Figure 1 The New Primary Health Care Sector. This diagram reflects the sector as envisaged under this Strategy, however, as noted previously primary health care practitioners will be free to decide whether or not they join a Primary Health Organisation
• PHOs will aim to improve and maintain the health of their populations and restore people's health when they are unwell. They will provide at least a minimum set of essential population-based and personal first-line general practice services
• PHOs will be required to work with those groups in their populations (for example, Maori, Pacific and lower income groups) that have poor health or are missing out on services to address their needs
• PHOs must demonstrate that they are working with other providers within their regions to ensure that services are coordinated around the needs of their enrolled populations
• PHOs will receive most of their funding through a population needs-based formula (capitation)
• PHOs will enrol people through primary providers using consistent standards and rules
• PHOs must demonstrate that their communities, iwi and consumers are involved in their governing processes and that the PHO is responsive to its community
• PHOs must demonstrate how all their providers and practitioners can influence the organisation's decision-making
• PHOs are to be not-for-profit bodies with full and open accountability for the use of public funds and the quality and effectiveness of services.
It is useful to clarify the relationships between IPAs and PHOs. IPAs, which were the original developments in organised general practice, have either been merged into or remain partially independent entities from PHOs. In general, and despite coming under the control of PHOs, IPAs have preserved the autonomy of general practice for both GPs and practice nurses. Whilst some loss of autonomy may be felt by rank-and-file GPs, this has been balanced to a certain extent by gains which organised general practice has made through the IPAs, and now especially through PHOs. The achievements have been significant in advancing the status and influence of general practice/primary health care. General practice, through PHOs, now has a much more important voice in its engagement within the health system, and particularly with DHBs, than was ever possible through individual general practice and general practice organisations.
The first two PHOs got underway in South Auckland in July 2002. To date 79 PHOs have been established with 3.72 million New Zealanders (93% of the population) enrolled. Cumming et al in a recent report note that there is a great variation between PHOs in terms of size, structure, age and context [21]. As a generalisation, there are two main types of PHOs (Table 5). Of the 77 PHOs established and studied as at April 2005, 38 were small with <20,000 enrolees; while these PHOs made up 49% of PHOs, they enrolled only 11% of the total enrolled population. Small PHOs tend to have difficulty in supplying management input within their organisation and meeting DHB requirements. Small PHOs are characterised by being made up of 76% access funded practices (see later); large PHOs are more commonly interim funded or mixed (72%) [21]. The issues surrounding critical mass are both interesting and vital. They also reflect the Australian problem of trading off population for distance in service organisation models – and implicit in this is how catastrophic risk can be managed across a population (e.g. a flu pandemic). If either country accepts a system design that does not provide a critical population under a population resource funding formula then we are setting up primary care to fail.
Table 5 Characteristics of PHOs (simplified)
Small (< 20,000 enrolees)
Inadequate management resources Large (>20,000 enrolees)
Well resourced, efficiently managed
Access funded
History – Previous NGO, capitated
Low Investment in IT, premises
Salaried doctors Interim funded
History – Previous IPA, fee-for-service
Established IT, premises etc
Doctors own practice
Low co-payments
Full/increasing use of nurses
Established community governance
Maori and Pacific focus Higher co-payments
Use of nurses dependent on busy-ness
Establishing community governance
General population focus
The NZ Government has committed additional base funding of $NZ284 million for 2004/05, $NZ338 million for 2005/06 and NZ$425 million for 2006/07 to implement the Primary Health Care Strategy. The PHOs are funded under two formulae – Access and Interim. PHOs serving areas with people who have high health needs, i.e. Maori, Pacific Island people and those on low incomes, receive a higher level of funding, according to what is known as the Access Formula. Patients belonging to these PHOs are able to get free or very cheap visits to their GP. For example, a child under 6 years will pay NZ$14 to the practice; other age groups will pay between NZ$20–27 per consultation (normal total fee is approximately NZ$43–50). They also pay no more than $3 for a prescription. The remaining PHOs that are not on the Access Formula are funded according to the Interim Formula – so named because the Government would eventually like to see all PHOs on the Access Formula. Most patients belonging to PHOs on the Interim Formula will have to pay much the same as they do now to go to their GP. Subsides available with Community Services Cards and High Use Health Cards will still apply, but are intended to be phased out over time. Whilst historically the percentage of government funding of general practice has been low, it is now increasing but there still remain high and widely variable levels of co-payments [22].
Care Plus is a new service that was introduced to Interim PHOs in July 2004. It is aimed at people with significant chronic illness who need to visit a GP frequently. The service covers such conditions as diabetes, heart disease, mental health, terminal care and others. Care Plus provides an additional 10% capitation funding for these patients and 8.5% of PHO enrolled patients are eligible for Care Plus [23]. The key criterion is that the person is expected to need at least two hours of clinical contact time in the coming six months. All Care Plus patients will have a care plan developed for them, including quarterly reviews to check on health status, treatment, medications etc. The government introduced Care Plus around the time of the Interim Formula to assuage the concerns of the GPs who were not on the Access Formula and who felt that their high-needs patients were being unfairly disadvantaged. Care Plus aims to improve the management of chronic conditions, reduce health inequalities between population groups, improve teamwork within PHOs, and lower the cost for high-need patients [24]. An early evaluation suggests that this is a successful programme, with moderate levels of satisfaction among patients and the primary health care team. In the experience of the pilot practices, the time involved for patients and practitioners, patient apathy towards a more active role in their own care, and staffing, were the main barriers to implementation of the programme [25].
However, not all of the primary health care services will be supplied by all PHOs, and not all of the services will be subsidised by the government. From 1 October 2003 low cost healthcare for those under 18 years of age has been administered through the PHOs, and this was extended to cover all enrolled people aged 65 years and over from July 2004. Progressive introduction of the new funding means that those aged 18–24 were covered from 1 July 2005, and for 46–64 years are covered from 1 July 2006.
Discussion
The Primary Health Care Strategy, launched in 2001, promised a new vision over five to ten years. Implementation of the Strategy is now well underway and the DHBs are in place and 79 PHOs have been established. Such radical changes will always be accompanied by teething problems, and it is still relatively early in the evolution of the changes to make absolute or definitive judgements. A formative evaluation on the development of DHBs has been published [26] and three early formative evaluations have been undertaken of the process of implementing PHO development [27-29]. The Royal New Zealand College of General Practitioners (RNZCGP) published an overview of the Primary Health Care Strategy implementation in April 2003 expressing concerns regarding the vulnerability of primary care early on in the reforms [30]. Also of relevance is the recently published report from NZMA indicating growing concerns regarding the New Zealand general practitioner workforce [31]. These issues are discussed further later. Typically, reforms and restructures happen with each parliament (in New Zealand 3 years), but a longer time frame was proposed in order to allow time for the primary care reforms to bed in properly. The recent re-election of the Labour-led Government for a third term in New Zealand should allow the reforms to be completed.
On a positive note the Primary Health Care Strategy has acknowledged the key role of general practice and primary health care in the health system. It is providing opportunities to:
• Address health inequalities
• Improve access to services
• Enhance population health, health promotion and preventative care
• Develop coordination and continuity of care
• Foster multidisciplinary care and collaboration.
There is goodwill present within the sector despite difficulties with the implementation processs, and a willingness to 'make it work' to achieve health gains for all [27]. Strong support has been expressed by PHOs for the philosophy of the Primary Health Care Strategy [21]. Although there was widespread support for the philosophy underlying the Primary Health Care Strategy both on its introduction and presently, there are concerns relating to its implementation from the RNZCGP, the NZMA and many general practitioners. These centre around six main issues:
1. Loss of autonomy
2. Inadequate management funding or support
3. Inconsistency and variations in contracting processes
4. Lack of publicity and advice around enrolment issues
5. Workforce and workload issues
6. Financial risks.
1. Loss of Autonomy
The majority of New Zealand GPs are self-employed and until recently only 30% of their income came from the public purse [30]. Despite being technically 'independent operators', running their own businesses and setting their own fees, they are significantly influenced by public policy. There is a long-standing tradition of autonomy, and strong suspicion of government moves to make GPs 'salaried servants'. The IPAs have been a great success story for general practice, bringing together GPs in an environment that embraces quality, education and accountability, coherent management and political strength [32]. IPAs have pioneered extensive development of information systems including merging and managing practice registers, analysing laboratory and pharmaceutical data, and providing personalised feedback to members. They have also formulated and monitored guidelines and pharmaceutical and laboratory services [12]. Nevertheless, about 20% of GPs have resisted the move to join IPAs. While some GPs have been persuaded for financial reasons to join, now at least 90% have joined either PHOs either directly or through IPAs, and 93% of the population is now enrolled with a PHO. There is a real risk of further fragmentation within general practice if the IPAs are not able to adapt to the transition to PHOs. Less than 25% of IPAs include members other than GPs and a few have community representation on their Boards [8], yet these are essential features of PHOs as set out in the Health Minister's minimum requirements.
GPs' fears of losing control of their destinies by becoming part of a PHO are perhaps best illustrated by considering an example. Partnership Health Canterbury Te Kei o Te Waka is the largest PHO in New Zealand with 350,000 enrolled individuals. Its governance body comprises nine individuals representing primary health care services and nine individuals representing broad consumer and community interests with an independent chair. These board members have been elected or selected by six electoral groups:
(on the provider side)
• General practice teams (GPs and practice nurses), as the contracted providers (5)
• Representatives of providers who are not general practice (4)
(on the consumer side)
• Maori (2)
• Pacific Island community (1)
• Territorial local authorities (2)
• Consumer and community representation (4)
This represents a very different balance of power compared to current IPA Boards. The IPAs are private organisations that may or may not hold government contracts. Although privately owned, the PHOs are more like instruments of government and are accountable as such to DHBs, which are crown owned entities. Consequently, there has been a considerable shift in the locus of control within primary care, one which the IPAs are presently resisting. More recently, GPs have expressed concerns regarding the potential for more controls, including restrictions on prescribing and laboratory testing. While these have not eventuated, there is significant mistrust of the Ministry of Health and Government [21]. One PHO (Middlemore) has folded because GPs and community representatives on the Board were at loggerheads on how the funding would be spent (Mike Lamont, personal communication). The other loss of autonomy comes in the form of the shifts in care, particularly to nursing, that is occurring within the GP practices. The advent of needs-based capitation means that patients can be seen and treated by any health professional including nurses where appropriate. Whilst this means that the nursing profession can feel that it's expertise can be properly recognised, GPs have felt threatened in some circumstances. Others have welcomed the change stating that it allows them to spend more time managing more complex consultations that require their skills.
2. Inadequate management funding and support
Concerns have been expressed that implementation funding is inadequate, particularly for management and health promotion [27]. Set up costs of PHOs have been high and establishment funding inadequate, especially for smaller PHOs. The report of the Referred Services Advisory Group to the Ministry of Health stated [33]:
'The group recognises that there are significant organisational and infrastructure costs involved in the functions required of PHOs. Much better information systems are required than many of those currently available, including at practice level. The group considers that the management payments currently proposed as part of the PHO funding are inadequate and should be increased. This will be particularly important as referred service savings will no longer be available to any PHO spending above its equitable level.' Management costs include those of governance, general management, planning control and coordination, performance monitoring and reporting, and referred services management.
If effective implementation of the Primary Health Care Strategy is to occur, these infrastructure needs and costs need to be much more clearly identified and addressed. Given the fragile position of many providers, it is also critical that the government funders ensure that payment mechanisms are efficient and timely. Late or omitted payments to providers do nothing to engender confidence or assist the viability of the organisations concerned.
Failure to recognise the cost of implementing reforms is also a common ongoing problem for Divisions of General Practice in Australia. The importance of proper management and governance has been illustrated in several Australian cases recently, including Divisions going under. Primary Care organisations do not have the luxury of going back to Treasury if things get tough. The need to develop robust governance arrangements is central to success.
These difficulties in New Zealand have been compounded by the failure to create an economic model of a PHO in order to understand the minimum size needed for viability, sustainability, and development. PHOs have enrolled populations varying in size from 2,000 to 350,000. In a recent review of the management costs for PHOs it was found that smaller PHOs (fewer than 20,000 enrolees) were not delivering on all required management services, primarily because they do not have the resources, including staff, to undertake all the requirements. Medium PHOs (between 21,000 and 75,000 enrolees) were better able to meet the requirements than small PHOs. So far the Ministry has not fully addressed the issue of management costs, and consequently significant numbers of PHOs are not meeting their targets for service delivery to patient populations [34]. Modelling and experience would suggest a critical mass of at least 100,000 as optimal (Jonathan Simon, personal communication). There are some lessons here for viability of some Australian Divisions with small populations in large areas of land. It is interesting to note that the UK Primary Care trusts have reached a similar conclusion – that populations of around 150,000 are needed for viability. This implies that there may be some trade offs in a pluralistic model in terms of performance and accountability expectations if small PHOs are deemed desirable.
The Ministry of Health has recently approved a Performance Management Programme by which PHOs will be assessed on their performance regarding 15 performance/quality indicators, including prescribing and laboratory use. This will provide additional funding to PHOs rewarding quality use of these services.
3. Inconsistency and variation in contracting processes
This was one of the main concerns identified in the Victoria University's Health Services Research Centre report [27].
• 'Inconsistency and variations in the contracting process was noted. Several different versions in the PHO contract have been signed. Not all PHOs had an agreed contract before going 'live'.
• 'Dissatisfaction was expressed over many of the processes involved with implementation of PHOs, especially with regard to funding and payment processes. Greater definition and clarity around the rules pertaining to qualification for funding was also felt to be required. Streamlining of payment processes, greater accuracy and appropriateness of reports, and timely payments, were believed to be necessary.'
Some of these difficulties have arisen due to a lack of clarity and consistency with regard to implementation of PHOs [35]. This is a reflection of the various degrees of enthusiasm and commitment to PHOs expressed by the 21 DHBs. The recent Victoria University Health Services Research Centre report has suggested that there are too many DHBs, leading to high transaction costs and duplication of effort [26]. There is evidence that the introduction of capitation has been supported with poor processes and business rules leading to gaming and inaccurate payments (Jonathan Simon, personal communication). More recently, DHBs are transitioning PHOs to the new PHO contract version 17 which amalgamates and standardises previous contracts.
4. Lack of publicity and advice around enrolment issues
Although the Primary Health Care Strategy promised 'a public education campaign to explain enrolment and promote its benefits for communities', there has been poor public awareness of PHOs and an associated lack of understanding by the general public as to the concept and implications of PHO enrolments. Duplicate enrolments, i.e. people enrolling in two or more PHOs, have been estimated to average 8.6% of the enrolled population [27], with a range of 1.6% to 13.6%. Tightening the rules of enrolment, particularly in regions which have more than one PHO in close proximity, has been considered important.
Further difficulties encountered have included large quarterly fluctuations in revenue as a result of mobile populations and rapid changes in numbers of enrolees in areas where new PHOs are being established [27].'Clawbacks' of funding between PHOs have created a new round of administration.
5. Workforce and workload issues
The New Zealand health reforms are unrolling against a backdrop of a substantial and growing shortage of GPs. In May 2004 the NZMA published An Analysis of the New Zealand General Practitioners Workforce [31], showing that the actual number of active GPs has decreased by 6.5% from 1997 to 2002 (and 8% over the last two of those years). Using the workforce definition 'GPs identifying general practice as their main type of work' the decrease is 13.4% from 1998 to 2002. Compared to Australia there are fewer GPs per 100,000 population. The report which collated information from many sources identified a number of key factors affecting the GP workforce:
• An ageing GP population
• Few new medical graduates choosing general practice as a career option
• The effects of increased numbers of women in the workforce
• The reliance on overseas-trained doctors
• GPs who are departing or intending to depart are not being replaced by incoming GPs
• Negative working conditions.
The workforce issues identified in this report are identical to those in Australia.
The report also concluded that 'if early action is not taken the problem will get progressively worse'. Compounding these concerns over workforce is a growing expectation that GPs will play an increasing role in chronic disease management, taking on responsibilities previously assumed by hospitals and secondary care providers. This requires additional time and skills, and is seen by some as a cost-shifting exercise – moving patients from hospital to community care. Finally, the additional administrative work (much of it non-reimbursed) associated with becoming part of a PHO is imposing increasing pressures on a shrinking, beleaguered workforce, with predictable effects on morale, especially for those GPs in small PHOs (<20,000). However, for those in large PHOs there is considerable centralised resource to provide management, administrative and clinical support.
6. Financial risks
Until recently only 30% of primary health care has been government subsidised, the balance being funded through private co-payments, health insurance and Accident Compensation payments [36]. In 2001, public sources of Vote Health in New Zealand were 76.7% (down from 82.4% in 1989) and private sources 23.3% (up from 17.6% in 1989) [37]. With the introduction of PHOs, levels and types of funding in general practice vary, particularly during this period of phased introduction of the new funding. This creates uncertainty of revenue for GPs during the establishment phase and high financial risk on an on-going basis [35]. This has recently been acknowledged by the Ministry of Health which has announced a 12-month 'funding floor' initiative for Access-funded PHOs and practices that have experienced financial hardship since becoming part of a PHO.
There is also a potentially destabilising effect associated with the establishment of low cost (i.e. more highly funded) PHOs located alongside practices that are being paid a lower level of subsidy. Although a majority of GPs have now joined PHOs, many have done so reluctantly. While joining a PHO can expose them to some financial risks, not joining is often not a viable option if other GPs in the district who are members of PHOs are able to reduce co-payments for their services [35]. However, since the new funding has been more fully introduced to the PHOs, GPs can expect a significant increase in their incomes. Many GPs have already benefited, reporting improved incomes and financial benefits from joining the PHOs. It is important to note that in most cases GPs are the unit of service provision in terms of payments from PHOs, even though practice revenues will also increase as a consequence of increased funding to the GPs who contribute to practice revenue. However, increasingly general practices (as companies, or other legal structure) are becoming the unit of service provision through a contract with the PHO. There are also indications that GPs, particularly in access funded practices, are now less worried and more supportive of the Strategy [21].
The RNZCGP has recently published a paper The Public and Private Interface in New Zealand Primary Health Care which states that 'in order to stabilise the primary care workforce, and achieve the intended outcomes of the Primary Health Strategy, it is essential that the interface between the private and public sector be more robustly and explicitly addressed. It is therefore critical that such a framework is developed as soon as possible and explicit models of engagement become an essential part of any government policy development, or contracting process with non-government providers' [36].
Conclusion
The issues arising from the implementation of the Primary Health Care Strategy have been encapsulated well in the RNZCGP's recent report The Public and Private Interface in New Zealand Primary Health Care [36]. There is a broad range of issues arising from the implementation of the Primary Care Strategy. The following are identified for their relevance to the private/public interface.
1. DHBs started to contract with PHOs with considerable variation; the resulting outcome has been a shifting of hospital/DHB risk and responsibility on to primary care and in particular, general practice.
2. Minimal funding was allocated for infrastructure development, quality, information technology and governance capacity building.
3. Compliance and administrative demands ballooned for providers, with development of enrolment registers, information technology compatibility issues.
4. Government information technology and capacity were inadequate and underdeveloped for reporting and payment requirements – compromising provider viability.
5. Graduated funding introduction increased funding for those most at need, at risk populations became low-cost access PHOs while other PHOs took the interim funding formula. This created inequalities in boundaries, with instability of provider viability, and neighbouring practices being funded better resulting in patients leaving general practices to enrol with another down the road.
6. Some general practitioners are concerned about PHO governance requirements; where others in governance make strategic decisions directly affecting their (the GP) future. (The practitioner may be carrying personal financial risk such as capital investments, mortgages on homes to finance health services which are now affected by public governance requirements).
7. The gaps in knowledge and expertise of those in governance making strategic decisions affecting practices.
It is important to note that the majority of these issues are related to the administration, contracting and reporting of service provision, rather than the delivery of the service to the patient per se.
Evaluations undertaken so far using qualitative interviews with 12 PHOs and surveys of another 22, indicate that there was strong support for the philosophy that underlies PHOs, particularly the focus on populations, the increased collaboration across professional groups and the opportunity for improving service integration [27]. Not surprisingly, consumers are also strongly supportive of reduced co-payments.
The concluding paragraph of the evaluation of PHOs noted:
'However, there is still goodwill present within the sector, despite difficulties with the implementation process (hampered by delays in documentation regarding requirements); confusion and inconsistency in application of rules; and inadequate funding streams. There is also a willingness to 'make it work' for the longer term benefits to patients and the community, and to achieve overall health gains for the wider population. This is despite some concern that unless certain implementation difficulties are addressed, there is a danger that the restructuring of the primary care sector will not be viable in the long term' [27].
Now, two years later, the most recent evaluation of the implementation of the Strategy noted:
'In the four years since the government published the Primary Health Care Strategy much has been achieved and there is wide support for the goals of the Strategy. More than 90% of the population are enrolled in one of 77 Primary Health Organisations, an uptake considerably faster than originally anticipated. PHOs report that much of the set-up work has been completed and that effort can now be redirected towards substantive changes.
For many, better access has already been achieved through lower fees and PHOs report that they are better able to identify and meet the need of a known, enrolled, population. Community representation on PHO boards appears to be increasingly effective and many valuable initiatives are underway.
General medical practitioners, freed from the incentives of a fee-for-service subsidy, have noted a greater flexibility in how they use their time. Some have found in the PHO environment a welcome opportunity to cooperate with other practitioners and one went so far to say that the changes would rejuvenate general practice' [21].
Australia has much to learn from New Zealand's latest health sector and primary health care reforms. The key lessons concern:
• the need for a national primary health care strategy
• active engagement of general practitioners and their professional organisations
• recognition of implementation costs
• the need for infrastructural support, including information technology and quality systems
• robust management and governance arrangements
• issues related to critical mass and population/distance trade offs in service delivery models
Competing interests
The author(s) declare that they have no competing interests.
Authors' contributions
Brian McAvoy drafted the manuscript and Gregor Coster revised it, and added material from the latest evaluation of the New Zealand Primary Health Care Strategy.
Acknowledgements
This review is based on a discussion document commissioned from Brian McAvoy by the Australian Medical Association. The authors would like to thank the two peer reviewers who provided thoughtful advice to improve the paper.
==== Refs
Organisation for Economic Cooperation and Development The OECD in 2003 2004 Paris, OECD
Starfield B Primary care: Concept, evaluation and policy 1992 Oxford, Oxford University Press
Ashton T Recent developments in the funding and organisation of the New Zealand health system Aust New Zealand Health Policy 2005 2 9 15877820 10.1186/1743-8462-2-9
Upton S Your Health and the Public Health 1991 Wellington, Ministry of Health
Health and Disability Services Act 1993 1993 Wellington. New Zealand Government
Hornblow A Barnett P A turbulent decade: lessons from the "health reforms" NZMJ 2000 113 133 4
Ashton T Boston J The health reforms: to market and back Redesigning the welfare state in New Zealand: problems, policies and prospects 1999 Auckland, Oxford University Press 134 53
The Okataina Group The Future Shape of Primary Health Care: A discussion document Joint submission (unpublished) 2000
Malcolm L Wright L Barnett P The development of primary care organisations in New Zealand: a review undertaken for Treasury and the Ministry of Health 1999 Lyttelton: Aotearoa Health
Devlin N Maynard A Mays N New Zealand's new health sector reforms: back to the future? BMJ 2001 322 1171 4 11348917 10.1136/bmj.322.7295.1171
Malcolm L Mays N New Zealand's independent practitioner associations: a working model of clinical governance in primary care? BMJ 1999 319 1340 2 10567141
Majeed A Malcolm L Unified budgets for primary care groups BMJ 1999 318 772 6 10082703
New Zealand Government New Zealand Public Health and Disability Act 2000 Wellington
Minister of Health New Zealand Health Strategy 2000 Wellington, Ministry of Health
Minister for Disability Issues The New Zealand Disability Strategy. Making a World of Difference Whakanui Oranga 2001 Wellington, Ministry of Health
National Advisory Committee on Health and Disability Improving Health for New Zealanders by Investing in Primary Health Care 2000 Wellington, National Advisory Committee on Health and Disability
Coster G Gribben B Primary care models for delivering population-based health outcomes 1999 Wellington, National Advisory Committee on Health and Disability
Minister of Health The New Zealand Primary Health Care Strategy 2001 Wellington, Ministry of Health
World Health Organization Primary Health Care Alma-Ata 1978. Report of the International Conference on Primary Health Care (Alma-Ata, USSR) 1978 Geneva, World Health Organization
Minister of Health Minimum Requirements for Primary Health Organisations 2001 Wellington, Ministry of Health
Cumming J Raymont A Gribben B Horsburgh M Kent B McDonald J Mays N Evaluation of the Implementation and Intermediate Outcomes of the Primary Health Care Strategy: First Report. Health Services Resarch Centre 2005 Victoria University of Wellington, Wellington
Malcolm L How general practice is funded in New Zealand Med J Aust 2004 181 106 7 15257651
CBG Health Research Ltd Care Plus Investigation: Estimating case loads 2003 16 April 2004
Ministry of Health Care Plus: an Overview 2004 16 April 2005
CBG Health Research Ltd Care Plus Process Report 2004 16 April 2005
Health Reforms 2001 Research Team Interim Report on Health Reforms 2001 Research Project 2001 Wellington, Victoria University of Wellington
Perera R McDonald J Cumming J Goodhead A Primary Health Organisations. The first year (July 2002 – June 2003) from the PHO perspective 2003 Wellington, Victoria University of Wellington
CBG Research Limited Reducing Inequalities Contingency Funded Projects 2003 Auckland, CBG Health Research Limited
Minister of Health Counties Manukau Enrolment and CBF Process Improvement Project 2003 Wellington, Ministry of Health
Austin C Primary Care in New Zealand – Viable but Vulnerable? 2003 Wellington, RNZCGP
New Zealand Medical Association An Analysis of the New Zealand General Practitioner Workforce 2004 Wellington, NZMA
Royal New Zealand College of General Practitioners Richards J General Practice in New Zealand 1997 1997 Wellington, RNZCGP
Report of the Referred Services Advisory Group to the Ministry of Health Referred Services Management: building towards equity, quality, and better health outcomes 2002 Wellington, Ministry of Health
Capital Strategy Review of Primary Health Organisation Management Services 2004 accessed 16 April 2005
Ashton T Cumming J Primary Health Organisations Establishment. Health Policy Monitor (accessed 28 June 2004)
Austin C The Public and Private Interface in New Zealand Primary Health Care 2004 Wellington, RNZCGP
Scott C School of Government, Victoria University of Wellington Health Services Conference, Australia 15 November 2003
|
16262908
|
PMC1291356
|
CC BY
|
2021-01-04 16:38:28
|
no
|
Aust New Zealand Health Policy. 2005 Nov 2; 2:26
|
utf-8
|
Aust New Zealand Health Policy
| 2,005 |
10.1186/1743-8462-2-26
|
oa_comm
|
==== Front
BMC BioinformaticsBMC Bioinformatics1471-2105BioMed Central London 1471-2105-6-2621625314210.1186/1471-2105-6-262Methodology ArticleUsing hexamers to predict cis-regulatory motifs in Drosophila Chan Bob Y [email protected] Dennis [email protected] School of Information and Computer Science, University of California, Irvine, Irvine, California, USA2005 27 10 2005 6 262 262 30 6 2005 27 10 2005 Copyright © 2005 Chan and Kibler; licensee BioMed Central Ltd.2005Chan and Kibler; licensee BioMed Central Ltd.This is an Open Access article distributed under the terms of the Creative Commons Attribution License (), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
Background
Cis-regulatory modules (CRMs) are short stretches of DNA that help regulate gene expression in higher eukaryotes. They have been found up to 1 megabase away from the genes they regulate and can be located upstream, downstream, and even within their target genes. Due to the difficulty of finding CRMs using biological and computational techniques, even well-studied regulatory systems may contain CRMs that have not yet been discovered.
Results
We present a simple, efficient method (HexDiff) based only on hexamer frequencies of known CRMs and non-CRM sequence to predict novel CRMs in regulatory systems. On a data set of 16 gap and pair-rule genes containing 52 known CRMs, predictions made by HexDiff had a higher correlation with the known CRMs than several existing CRM prediction algorithms: Ahab, Cluster Buster, MSCAN, MCAST, and LWF. After combining the results of the different algorithms, 10 putative CRMs were identified and are strong candidates for future study. The hexamers used by HexDiff to distinguish between CRMs and non-CRM sequence were also analyzed and were shown to be enriched in regulatory elements.
Conclusion
HexDiff provides an efficient and effective means for finding new CRMs based on known CRMs, rather than known binding sites.
==== Body
Background
The development of eukaryotic organisms is tightly regulated by a variety of mechanisms. The initial step of regulation is carried out by transcription factors interacting with cis-regulatory sequences, also known as transcription factor binding sites (TFBS). In eukaryotes, multiple TFBS are often clustered together into cis-regulatory modules (CRMs). The TFBS can be thought of as inputs into an information processing element, with the output being the level of expression of the gene controlled by the CRM [1].
One of the major challenges for understanding eukaryotic gene regulation is finding CRMs. There are two main types of CRMs – promoters and enhancers. Promoters are located immediately upstream of a gene's transcriptional start site and often contain a variety of sequence signals such as the TATA box, CCAAT box, and different TFBS. These characteristics have been used in approaches for finding promoters [2]. In contrast, enhancers do not share these signals and operate in a manner that is relatively independent of orientation or distance from their target gene [3]. In fact, one enhancer, Dct, has been found almost a megabase away from Sox9, the gene it regulates [4]. Because of the lack of common signals and because the search for enhancers cannot be limited to the few hundred base pairs upstream of the transcriptional start site, finding enhancers is a more difficult problem.
Methods for predicting CRMs can be classified by the type of information they use to make the predictions – known binding sites of regulatory proteins, homologous sequences, or known CRMs. Binding sites for the first type of method are generally modeled using position weight matrices (PWMs) or consensus sequences. These models are used to search for statistically significant clusters of predicted TFBS. Examples of methods based on binding sites of multiple transcription factor proteins include one developed for human skeletal muscle [5], a logistic regression analysis model for liver-specific transcription factors [6], CIS-ANALYST [7], MCAST [8], Ahab [9], Stubb [10], Cluster Buster [11], MSCAN [12], and EMCMODULE [13]. Methods based on binding sites of single transcription factors have also been developed – SCORE [14], Fly Enhancer [15], and a method of searching for homotypic clusters [16].
Methods based on homologous sequences assume that areas of the DNA involved in regulating gene transcription are under selective pressure and are therefore more likely to be conserved than non-functional DNA [17]. These methods can be categorized by whether they search for conserved DNA by aligning homologous regions from multiple species [18-20], homologous regions from two species [21-24], or homologous regions from related genes in a single species (also referred to as co-regulated genes) [25-27].
Methods based on locations of known CRMs have been rarer. Methods of this type tend to look for statistical properties of DNA sequence that distinguish CRMs from non-regulatory DNA. One group has developed a statistical test called the "fluffy-tail test" that looks for differences in nucleotide composition, particularly in lists of words of various lengths [28]. From the related field of promoter prediction comes PromFind, an algorithm for finding promoters that uses hexamer frequencies of known promoters to search for DNA with similar frequencies [29]. Because PromFind was developed for promoters, the author could assume that every sequence being tested contained one promoter, and that the strand containing the promoter was known – assumptions that are not true for enhancers. Another recent algorithm developed to predict CRMs is based on the exhaustive analysis of local word frequencies (LWF) [30]. Unlike PromFind where the algorithm is based solely on hexamer words, the LWF algorithm considers the pattern of word frequencies in a sliding window.
While the LWF algorithm was shown to perform well at the task of predicting CRMs [30], it is difficult to put a biological meaning to the results since the algorithm depends on word frequencies, not on the words themselves. In contrast, the PromFind algorithm generates lists of hexamers that are important for distinguishing promoters from non-promoter sequences. In the paper describing PromFind, the author analyzes the hexamers used by the algorithm for their CpG dinucleotide content and their similarity to various known promoter signals such as the Sp1, TATA-box, and CCAAT-box motifs [29]. This type of analysis cannot be performed on the local word frequency distributions of the LWF algorithm, due to the way it was designed.
We developed the HexDiff algorithm to solve the same problem as the LWF algorithm – predicting the location of CRMs – while being as biologically meaningful as the PromFind algorithm. The performance of HexDiff was compared to the LWF algorithm and several other CRM prediction programs using a common data set. The hexamers used by HexDiff were then examined to see if known TFBS were recovered.
Results
Data set
The early development of the Drosophila embryo is well-studied, both biologically and computationally. Data sets detailing the locations of known CRMs have been published by several groups [7,9,30,31], but we chose the one compiled by Schroeder et al. as it clearly defined the regulatory networks involved. In their analysis, Schroeder et al. examined 29 genes with gap and pair-rule patterns. However, only 16 of those genes were associated with known CRMs [31]. Therefore, we focused our study on those 16 genes, which contained a total of 52 CRMs. Sequences for the genes were obtained from the 4.1 Release of the Drosophila genome and included the 20 kb upstream as well as downstream of each gene [see Additional file 1]. Since the known CRMs were provided as FASTA-formatted DNA sequences, their positions relative to the extracted sequences were confirmed using BLAST [see Additional file 2].
The HexDiff algorithm
The HexDiff algorithm is designed to discriminate between CRMs and non-CRM sequence by using hexamer frequencies. For evaluation, we use the leave-one-out cross-validation (LOOCV) methodology. In this case, 15 of the 16 sequences in the data set are used as a training set and the 16th sequence is used as the test set. The process is repeated 16 times; leaving out one sequence each time. During training, hexamers that appear more frequently in CRMs are selected. In order to predict CRMs in the test set sequence, a window is slid across the sequence and the set of selected hexamers (Hd) is used to calculate a score that is used to predict whether each position is either CRM or non-CRM sequence.
Performance comparison
The HexDiff algorithm was compared to five other CRM prediction algorithms: Ahab, Cluster Buster, MSCAN, MCAST, and LWF (see Table 1). The results for Ahab were obtained from the supplementary files of Schroeder et al., while Cluster Buster, MSCAN, MCAST, and LWF are all available as downloadable software or public web servers. An important criterion for the inclusion of an algorithm was that it accepted user-defined sequences and PWMs as input. Ahab, Cluster Buster, MSCAN, and MCAST are algorithms based on binding sites of known transcription factors and were given the same set of nine PWMs for the transcription factors as described in Schroeder et al.: the maternal factors Bicoid (Bcd), Hunchback (Hb), Caudal (Cad), the Torso-response element (TorRE), and Stat92E (D-Stat), and the gap factors Kruppel (Kr), Knirps (Kni), Giant (Gt), and Tailless (Tll). LWF was given the same positive and negative training sets as HexDiff. The default parameters were used for Cluster Buster, MSCAN, MCAST, and LWF. The complete list of predictions can be found in the Additional materials section [see Additional file 3].
Table 1 Key aspects of HexDiff and other algorithms. The table shows the knowledge used and the parameters required by the different algorithms.
Algorithm Knowledge Used Parameters
HexDiff CRM Locations Number of hexamers in Hd
Window size
Window score threshold
Ahab PWMs Window size
Free energy cutoff
Order of background model
Cluster Buster PWMs Motif score threshold
Gap parameter
Cluster score threshold
Residue abundance range
MSCAN PWMs Motif score threshold
Window size
Minimum hits
Maximum hits
MCAST PWMs Motif score threshold
Maximum allowed distance between adjacent hits
Pseudocount weight
LWF CRM Locations String length
Number of mismatches
Detection window size
Maximum number of channels
Channels equalized
Profile cutoff
Peak width cutoff
Smoothing window
As shown in Table 2, each algorithm was assigned a cumulative score, calculated by summing the Matthews correlation coefficients for each of the sequences in the data set. The predictions made by HexDiff have the highest correlation with the known CRMs (5.71), followed by Ahab (4.48) and Cluster Buster (4.24). An interesting result is that although four of the six algorithms used the same PWMs, their performance varied widely – from 1.81 to 4.48.
Table 2 Correlation between predicted and known CRMs. The performance of six different algorithms on a common data set is compared in this table. For each sequence, the Matthews correlation coefficient is calculated by checking whether each position is a TP, TN, FP, or FN and using the equation listed in the Methods section. The sum of the correlation coefficients gives a cumulative score for each algorithm on this data set.
Gene CRMs HexDiff Ahab Cluster Buster MSCAN MCAST LWF
btd 1 0.70 0.57 0.19 0.01 0.07 0.10
ems 3 0.00 0.00 -0.03 0.12 -0.01 -0.01
eve 6 0.55 0.63 0.65 0.50 0.41 0.06
fkh 1 -0.03 -0.02 -0.02 -0.04 -0.02 -0.01
ftz 5 0.40 0.28 0.28 0.07 0.16 0.08
gt 1 0.27 0.42 0.33 0.35 0.15 0.03
h 5 0.71 0.63 0.53 0.30 0.37 0.08
hb 2 0.35 0.63 0.39 0.34 0.24 0.04
hkb 1 0.51 0.00 -0.02 -0.02 -0.08 0.09
kni 3 0.55 0.55 0.39 0.37 0.23 -0.05
kr 3 0.43 0.00 0.77 0.20 0.11 -0.03
oc 2 0.70 -0.02 0.00 0.11 0.02 0.07
prd 7 0.01 -0.07 0.16 0.07 -0.04 0.05
run 6 0.27 0.16 0.08 0.08 0.02 0.07
slp1 3 -0.07 0.15 -0.04 0.00 0.07 0.01
tll 3 0.35 0.56 0.58 0.19 0.12 -0.04
Total 52 5.71 4.48 4.24 2.64 1.81 0.52
While the Matthews correlation coefficient was the primary performance measure for the six algorithms, a closer look at two other measures offers more information about the characteristics of the individual algorithms. Table 3 shows sensitivities and positive predictive values (PPVs) for each of the algorithms. A known CRM was considered recovered if the overlap between it and a predicted CRM exceeded 50 bp.
Table 3 Sensitivities and positive predictive values (PPVs) of HexDiff and other algorithms. A known CRM was considered recovered if a predicted CRM overlapped it by at least 50 bp. The PPVs in this table are italicized because they are estimates of the true PPVs. Without complete knowledge of all CRMs that are present in the 16 sequences, it is possible that some of the predicted CRMs that are labeled as false positives are actually true positives.
Algorithm CRMs Recovered Num CRMs Sensitivity TP/(TP + FN) True Positives CRMs Predicted PPV TP/(TP + FP)
HexDiff 36 52 69.23% 35 104 33.65%
Ahab 23 52 44.23% 20 35 57.14%
Cluster Buster 31 52 59.62% 23 88 26.14%
MSCAN 34 52 65.38% 42 226 18.58%
MCAST 43 52 82.69% 53 499 10.62%
LWF 27 52 51.92% 48 433 11.09%
One caveat about the PPVs in Table 3 is that they are not true PPVs, but estimates of the true PPVs. This is due to the fact that the 16 sequences in the data set may contain more CRMs than the 52 that have been characterized so far, which would mean that some of the predicted CRMs that are labeled as false positives could actually be true positives.
Predicted CRMs
Of the 104 predictions made by HexDiff, 36 overlapped known CRMs by at least 50 bp, leaving 68 potential CRMs. Some of these may have been false positives, so to narrow down the candidates, the 68 potential CRMs were compared to predictions made by Ahab, Cluster Buster, MSCAN, MCAST, and LWF. 37 of the 68 matched predictions made by at least one other method, while 18 matched predictions made by at least two other methods, 10 matched predictions made by at least three other methods, and 6 matched predictions made by at least four other methods.
While our analysis was focused on the gap and pair-rule regulatory networks, it's possible that some of the predicted CRMs are actually known CRMs from other regulatory networks involved in the early development of Drosophila. Therefore, the list of 18 predicted CRMs was compared to a more comprehensive compilation of 124 CRMs [32]. 8 of the 18 predicted CRMs matched CRMs from the compilation, leaving 10 that do not correspond to any known CRMs. Given the specificities of the individual methods and the breadth of approaches, it seems very likely that some of the 18 predictions listed in Table 4 constitute novel CRMs.
Table 4 Potential novel CRMs predicted by HexDiff and other algorithms. All of the predicted CRMs listed in this table were predicted by HexDiff and at least two other algorithms. The column labeled "Gene" lists the gene involved in the early development of Drosophila that is closest to the predicted CRM. The columns labeled 1–5 are the different algorithms whose predictions matched the CRMs predicted by HexDiff: 1 – Ahab, 2 – Cluster Buster, 3 – MSCAN, 4 – MCAST, and 5 – LWF. The predicted CRMs were also compared to a compilation of 124 CRMs [32] – matching CRMs are listed in the last column.
Gene Arm Begin End Length 1 2 3 4 5 Matched
btd X 9534921 9535192 271 * *
eve 2R 5492385 5493575 1190 * * eve_late2_mel
fkh 3R 24421705 24422385 680 * *
ftz 3R 2683060 2683406 346 * *
gt X 2268347 2270179 1832 * * *
gt X 2290228 2290685 457 * * * * * gt_23-bcd_mel
hb 3R 4503375 4503962 587 * * *
hb 3R 4519805 4520172 367 * *
kni 3L 20628230 20628504 274 * * * * kni_+1_mel
prd 2L 12080435 12082316 1881 * * * prd_bcd_mel
prd 2L 12089627 12089847 220 * * prd_1_mel
run X 20488169 20488643 474 * * * *
run X 20524260 20524722 462 * * * *
slp1 2L 3811050 3812092 1042 * *
slp1 2L 3822581 3823049 468 * *
slp1 2L 3824891 3825039 148 * * * * * slp_A-bcd_mel
slp1 2L 3833433 3834671 1238 * * * * slp2_-3_mel
tll 3R 26680559 26683175 2616 * * * tll_bcd_mel
Meaning of differential hexamers
One advantage of the HexDiff algorithm is that the set of hexamers (Hd) used to distinguish between CRMs and non-CRM sequence can be analyzed for further insights into the mechanisms of gene regulation. Since the Hd hexamers were selected based on their overrepresentation in CRMs relative to non-CRM sequence, it would be reasonable to expect that some of the hexamers would be similar to gap and pair-rule regulatory sites. Therefore, the top 80 Hd hexamers and their reverse complements were compared to the list of binding sites used to build the 9 PWMs from Schroeder et al [see Additional file 4]. In the 16 rounds of cross-validation, an average of 59.6 hexamers was found within the binding sites.
Simulations were carried out to estimate the likelihood of this result. 100,000 random sets of 80 hexamers and their reverse complements were compared to the binding sites. The probability of 59 or more hexamers being found within the binding sites was 0.019, indicating that the hexamers selected by HexDiff were enriched in binding sites of known regulatory proteins.
A similar study was performed using regulatory sites obtained from TRANSFAC [33], but the results were not statistically significant. TRANSFAC contains binding regions obtained through a variety of biological methods. For instance, R02491 contains the following binding region obtained using DNase I footprinting: GACTTTATTGCAGCATCTTGAACAATCGTCGCAGTTTGGTAACAC. On average, TRANSFAC binding sites are much longer than the binding sites used by Schroeder et al .and are therefore probably much less specific. While the TRANSFAC binding sites do contain regulatory DNA, it seems that the signal is washed out by extraneous sequence.
Discussion
The HexDiff algorithm is designed to solve the difficult task of distinguishing CRMs from non-CRM sequence. It is first trained on sequences containing known CRMs by selecting hexamers that discriminate between the two categories. These hexamers are then applied to novel sequences to search for predicted CRMs. The requirements of the training process mean that HexDiff works best in a well-defined regulatory system where some CRMs are already known.
Using a data set of 16 sequences containing 52 CRMs obtained from Schroeder et al., we compared the HexDiff algorithm to five other algorithms: Ahab, Cluster Buster, MSCAN, MCAST, and LWF. Hexdiff's predictions correlated best with biological knowledge, and its sensitivity and specificity were comparable to the other algorithms. This result was encouraging considering that HexDiff is a type of machine learning algorithm, which tend to do better in problems where the items being classified are separated into two roughly equal groups. In this case, the CRMs made up just 9.36% of the total data set, with non-CRM sequence making up the other 90.64%. Even with a data set where the negative data outweighed the positive data by a factor of 9, HexDiff still performed well.
While the HexDiff algorithm has a fairly high specificity in isolation, it would still be prudent to compare its results to the predictions made by other algorithms, considering the time and effort required to biologically confirm a computational prediction. The 18 CRMs listed in Table 4 were predicted by at least three algorithms and did not match any of the 52 CRMs provided by Schroeder et al. A comparison with a more comprehensive list of 124 CRMs revealed that 8 of the 18 predicted CRMs corresponded to known CRMs from other regulatory networks involved in the early development of Drosophila. The 10 remaining predicted CRMs are strong candidates for future study.
A further attempt to understand the meaning of the Hd hexamers was made by comparing them to the known binding sites used to generate the 9 PWMs provided by Schroeder et al. Simulations showed that the number of Hd hexamers found within the binding sites was significantly more than would be expected of a randomly selected set of hexamers of the same size. While this result is not unexpected, it is a confirmation that we retrieved the binding sites of relevant regulatory proteins using only the locations of known CRMs.
While recovering the known binding sites is important, it is important to note that they accounted for less than half of the hexamers in the Hd sets. When 80 hexamers were selected for Hd, their reverse complements would increase the total number of hexamers to 160, barring palindromes. And yet, the average number of hexamers that were found within the known binding sites was 59.6, leaving another 100 hexamers whose identities are unknown. This result suggests that there are novel sequence features besides the known binding sites that are important for distinguishing between CRMs and non-CRM sequence.
Conclusion
One of the major questions in studying eukaryotic gene regulation is how regulatory proteins with relatively degenerate binding sequences can precisely regulate many genes. The discovery of cis-regulatory modules, short stretches of DNA that contain multiple binding sites for multiple proteins, has provided at least a partial explanation for the regulatory specificity observed in eukaryotes, and motivated a search for ways to predict CRMs computationally.
We have developed a simple and effective algorithm for predicting CRMs. In our study of the gap and pair-rule genes in Drosophila melanogaster, the results of the HexDiff algorithm correlated best with biological knowledge, and the sensitivity and specificity of the algorithm were comparable to other algorithms. Our predictions were compared to those made by other methods and resulted in a list of 10 putative CRMs with strong computational support. Analysis of the Hd hexamers revealed that not only were we rediscovering the known binding sites, but also discovering new signals that distinguished between CRMs and non-CRM sequence.
Methods
Differential hexamer frequency algorithm
The HexDiff algorithm is based on the idea of distinguishing between two types of DNA sequence: CRMs, and non-CRM sequence. In order to accomplish this task, a model is built using sequences where the CRMs are known – the training set. The training set is split into positive and negative training sets by consolidating all of the known CRMs into a positive training set, and the remainder of the training set into a negative training set. On average, the ~1 Mb training set is split into a ~50 kb positive training set and a ~950 kb negative training set. The frequency of each hexamer h on both strands is then calculated for the positive fp(h) and negative fn(h) training sets and used to calculate a ratio, R(h).
R(h)=fp(h)fn(h)
MathType@MTEF@5@5@+=feaafiart1ev1aaatCvAUfKttLearuWrP9MDH5MBPbIqV92AaeXatLxBI9gBaebbnrfifHhDYfgasaacH8akY=wiFfYdH8Gipec8Eeeu0xXdbba9frFj0=OqFfea0dXdd9vqai=hGuQ8kuc9pgc9s8qqaq=dirpe0xb9q8qiLsFr0=vr0=vr0dc8meaabaqaciaacaGaaeqabaqabeGadaaakeaacqWGsbGucqGGOaakcqWGObaAcqGGPaqkcqGH9aqpdaWcaaqaaiabdAgaMnaaBaaaleaacqWGWbaCaeqaaOGaeiikaGIaemiAaGMaeiykaKcabaGaemOzay2aaSbaaSqaaiabd6gaUbqabaGccqGGOaakcqWGObaAcqGGPaqkaaaaaa@3DF4@
The next step is to select hexamers that discriminate between CRMs and non-CRM sequence. The hexamers with the highest values for R(h) are then placed into Hd. As a result of this selection process, Hd contains hexamers that are much more common in CRMs than in non-CRM sequence.
Once the hexamers for Hd are chosen, the final step is to use them to classify each position in an unknown sequence as either CRM or non-CRM sequence. A window is slid across the unknown sequence 1 bp at a time. At each position i, a score Si is calculated for the window by increasing the score by the product of R(hd) and the number of times that hexamer appeared n(hd) for each hexamer from Hd:
Si=∑hd∈Hd[n(hd)R(hd)]
MathType@MTEF@5@5@+=feaafiart1ev1aaatCvAUfKttLearuWrP9MDH5MBPbIqV92AaeXatLxBI9gBaebbnrfifHhDYfgasaacH8akY=wiFfYdH8Gipec8Eeeu0xXdbba9frFj0=OqFfea0dXdd9vqai=hGuQ8kuc9pgc9s8qqaq=dirpe0xb9q8qiLsFr0=vr0=vr0dc8meaabaqaciaacaGaaeqabaqabeGadaaakeaacqWGtbWudaWgaaWcbaGaemyAaKgabeaakiabg2da9maaqafabaGaei4waSLaemOBa4MaeiikaGIaemiAaG2aaSbaaSqaaiabdsgaKbqabaGccqGGPaqkcqWGsbGucqGGOaakcqWGObaAdaWgaaWcbaGaemizaqgabeaakiabcMcaPiabc2faDbWcbaGaemiAaG2aaSbaaWqaaiabdsgaKbqabaWccqGHiiIZcqWGibasdaWgaaadbaGaemizaqgabeaaaSqab0GaeyyeIuoaaaa@47D3@
Any positions where the window scores exceed a threshold score are labeled as potential CRMs. As the shortest known CRM in the data set was 66 bp, we filtered out any predicted CRMs shorter than 50 bp. The entire process takes less than 10 seconds per gene on an Athlon 64 3200+.
Evaluation of predictions
Because there are only 16 sequences in the data set, LOOCV was used to assess the performance of the HexDiff algorithm. In LOOCV, the first of the 16 sequences was withheld from the data set and used as the test set. The HexDiff algorithm was trained on the 15 remaining sequences and predictions were made on the test set sequence. Those predictions were then compared to the locations of the known CRMs in the test set sequence. The same process was repeated with the second sequence, etc., until each of the 16 sequences had been used as a test set.
The accuracy of the predictions for each sequence was measured using the Matthews correlation coefficient, which has been used extensively to evaluate the performance of various prediction algorithms [34]. It combines both sensitivity and specificity into one measure and relies on four values that satisfy TP + TN + FP + FN = N (length of the sequence): TP (the number of base pairs where known CRMs overlap with predicted CRMs), TN (the number of base pairs that are not in known CRMs or predicted CRMs), FP (the number of base pairs where predicted CRMs did not overlap known CRMs), and FN (the number of base pairs where known CRMs did not overlap predicted CRMs). The Matthews correlation coefficient is calculated as follows:
CC=TP*TN−FP*FN(TP+FP)(TP+FN)(TN+FP)(TN+FN)
MathType@MTEF@5@5@+=feaafiart1ev1aaatCvAUfKttLearuWrP9MDH5MBPbIqV92AaeXatLxBI9gBaebbnrfifHhDYfgasaacH8akY=wiFfYdH8Gipec8Eeeu0xXdbba9frFj0=OqFfea0dXdd9vqai=hGuQ8kuc9pgc9s8qqaq=dirpe0xb9q8qiLsFr0=vr0=vr0dc8meaabaqaciaacaGaaeqabaqabeGadaaakeaacqWGdbWqcqWGdbWqcqGH9aqpdaWcaaqaaiabdsfaujabdcfaqjabcQcaQiabdsfaujabd6eaojabgkHiTiabdAeagjabdcfaqjabcQcaQiabdAeagjabd6eaobqaamaakaaabaGaeiikaGIaemivaqLaemiuaaLaey4kaSIaemOrayKaemiuaaLaeiykaKIaeiikaGIaemivaqLaemiuaaLaey4kaSIaemOrayKaemOta4KaeiykaKIaeiikaGIaemivaqLaemOta4Kaey4kaSIaemOrayKaemiuaaLaeiykaKIaeiikaGIaemivaqLaemOta4Kaey4kaSIaemOrayKaemOta4KaeiykaKcaleqaaaaaaaa@5868@
The Matthews correlation coefficient ranges from -1 to +1, like the better-known Pearson correlation coefficient. A value of 0 signifies that the prediction is equivalent to a completely random prediction, while +1 signifies a perfect prediction.
Choosing parameters
The HexDiff algorithm was designed so that the number of parameters needed to be set by the user was minimized. This reduced the complexity of the model and helped to avoid overfitting. Overfitting is a problem often faced by machine learning algorithms, where a model that is too complex will not only learn the signal in the training set but will also fit the noise, reducing the algorithm's performance on the test set [35]. The three parameters HexDiff does require are: the number of hexamers in Hd, the size of the window that is slid across the sequence of interest, and the threshold score that determines whether each position is predicted as a CRM or not.
These values were chosen during LOOCV using an empirical method. During each round of cross-validation, the 16 sequences were split into a 15-sequence training set and a single-sequence test set. The HexDiff algorithm was trained on the training set and its performance evaluated on the training set, using various combinations of the three parameters. The combination that gave the best performance on the training set was then used to predict CRMs in the test set.
In general, we observed that the HexDiff algorithm was relatively insensitive to the precise values for the parameters. We ran the HexDiff algorithms with 6 values for the number of hexamers (30, 40, 50, 60, 70, 80) and 11 values for the size of the sliding window (1000, 1100, 1200, 1300, 1400, 1500, 1600, 1700, 1800, 1900, 2000). For each pair of parameters, the threshold score that gave the best performance on the training set was selected. For these 66 analyses, the average cumulative Matthews correlation coefficient for the test set sequences was 5.37 with a standard deviation of 0.51. The average number of modules recovered was 33.8 with a standard deviation of 3.29.
Choices in algorithm design
Two important choices were made during the design of the HexDiff algorithm: the model used to represent the training sets, and the type of negative training set used. In order to choose the correct model, a balance had to be struck between the expressivity of the model and the amount of training data. Since the amount of positive training data was fixed at 52 known CRMs, we tried pentamers, hexamers, and heptamers with 0, 1, or 2 mismatches (data not shown). Shorter n-mers would have resulted in a model that was insufficiently expressive to capture the difference between CRMs and non-CRM sequences, while longer n-mers would have resulted in a model whose parameters would have had high variance. Allowing mismatches would have consolidated n-mers into groups and therefore would have had the effect of simplifying the model. In the end, hexamers with no mismatches turned out to have the best performance.
For the LWF algorithm, Nazina and Papatsenko tried three different negative training sets drawn from the Drosophila genome: random samples from the whole genome, random samples of coding sequence, and random samples of non-coding sequence. They found that samples from the whole genome and samples from non-coding sequence resulted in better agreement between CRM predictions from the LWF algorithm and the biologically determined locations. Since we were predicting CRMs in a well-defined set of 16 genes, we used a local negative training set – the portions of the 16 genes that were not labeled as known CRMs. We found this approach to give higher accuracy than the negative training sets used by Nazina and Papatsenko (data not shown).
Availability and requirements
Project name: HexDiffProject home page:
Operating system: Platform independent
Programming language: Perl
Other requirements: Perl 5.6 or higher
License: GNU GPL
Any restrictions to use by non-academics: licence needed
Authors' contributions
BC carried out the computational work in this study and drafted the manuscript. DK participated in the design of the study and helped revise the manuscript. Both authors read and approved the final manuscript.
Supplementary Material
Additional File 1
Sequences comprising the data set. FASTA-formatted sequences that include the 20 kb upstream and downstream of the 16 gap and pair-rule genes that were studied. Sequences were extracted from Release 4.1 of the Drosophila genome using FlyBase's "GBrowse" Genome Browser.
Click here for file
Additional File 2
Coordinates of known CRMs. Coordinates of known CRMs, relative to the sequences in Additional file 2.
Click here for file
Additional File 3
Coordinates of CRMs predicted by the different algorithms. Coordinates of predicted CRMs for each of the algorithms compared in this study, relative to the sequences in Additional file 2.
Click here for file
Additional File 4
Hd hexamers. Top 80 Hd hexamers calculated by the HexDiff algorithm for each round of cross-validation.
Click here for file
Additional File 5
ROC curves for the HexDiff algorithm. The three curves in this plot were made by taking the combination of parameters that gave the best performance on the training sets (number of nmers: 80, window size: 1700 bp, threshold: 170) and holding two parameters constant while varying the third.
Click here for file
Additional File 6
Sensitivities and specificities for HexDiff and the other algorithms. Sensitivities and specificities for HexDiff and the other algorithms were calculated by checking whether each position was a TP, FP, TN, or FN and using the appropriate formulas.
Click here for file
Acknowledgements
This investigation was supported by National Institutes of Health, National Research Service Award 5 T15 LM00744, from the National Library of Medicine.
==== Refs
Davidson EH McClay DR Hood L Regulatory gene networks and the properties of the developmental process Proc Natl Acad Sci U S A 2003 100 1475 1480 12578984 10.1073/pnas.0437746100
Qiu P Recent advances in computational promoter analysis in understanding the transcriptional regulatory network Biochem Biophys Res Commun 2003 309 495 501 12963016 10.1016/j.bbrc.2003.08.052
Laimins LA Gruss P Pozzatti R Khoury G Characterization of enhancer elements in the long terminal repeat of Moloney murine sarcoma virus J Virol 1984 49 183 189 6690710
Qin Y Kong LK Poirier C Truong C Overbeek PA Bishop CE Long-range activation of Sox9 in Odd Sex (Ods) mice Hum Mol Genet 2004 13 1213 1218 15115764 10.1093/hmg/ddh141
Wasserman WW Fickett JW Identification of regulatory regions which confer muscle-specific gene expression J Mol Biol 1998 278 167 181 9571041 10.1006/jmbi.1998.1700
Krivan W Wasserman WW A predictive model for regulatory sequences directing liver-specific transcription Genome Res 2001 11 1559 1566 11544200 10.1101/gr.180601
Berman BP Nibu Y Pfeiffer BD Tomancak P Celniker SE Levine M Rubin GM Eisen MB Exploiting transcription factor binding site clustering to identify cis-regulatory modules involved in pattern formation in the Drosophila genome Proc Natl Acad Sci U S A 2002 99 757 762 11805330 10.1073/pnas.231608898
Bailey TL Noble WS Searching for statistically significant regulatory modules Bioinformatics 2003 19 Suppl 2 II16 II25 14534166
Rajewsky N Vergassola M Gaul U Siggia ED Computational detection of genomic cis-regulatory modules applied to body patterning in the early Drosophila embryo BMC Bioinformatics 2002 3 30 12398796 10.1186/1471-2105-3-30
Sinha S van Nimwegen E Siggia ED A probabilistic method to detect regulatory modules Bioinformatics 2003 19 Suppl 1 i292 301 12855472 10.1093/bioinformatics/btg1040
Frith MC Li MC Weng Z Cluster-Buster: Finding dense clusters of motifs in DNA sequences Nucleic Acids Res 2003 31 3666 3668 12824389 10.1093/nar/gkg540
Johansson O Alkema W Wasserman WW Lagergren J Identification of functional clusters of transcription factor binding motifs in genome sequences: the MSCAN algorithm Bioinformatics 2003 19 Suppl 1 i169 76 12855453 10.1093/bioinformatics/btg1021
Gupta M Liu JS De novo cis-regulatory module elicitation for eukaryotic genomes Proc Natl Acad Sci U S A 2005 102 7079 7084 15883375 10.1073/pnas.0408743102
Rebeiz M Reeves NL Posakony JW SCORE: a computational approach to the identification of cis-regulatory modules and target genes in whole-genome sequence data. Site clustering over random expectation Proc Natl Acad Sci U S A 2002 99 9888 9893 12107285 10.1073/pnas.152320899
Markstein M Markstein P Markstein V Levine MS Genome-wide analysis of clustered Dorsal binding sites identifies putative target genes in the Drosophila embryo Proc Natl Acad Sci U S A 2002 99 763 768 11752406 10.1073/pnas.012591199
Lifanov AP Makeev VJ Nazina AG Papatsenko DA Homotypic regulatory clusters in Drosophila Genome Res 2003 13 579 588 12670999 10.1101/gr.668403
Tagle DA Koop BF Goodman M Slightom JL Hess DL Jones RT Embryonic epsilon and gamma globin genes of a prosimian primate (Galago crassicaudatus). Nucleotide and amino acid sequences, developmental regulation and phylogenetic footprints J Mol Biol 1988 203 439 455 3199442 10.1016/0022-2836(88)90011-3
Blanchette M Tompa M Discovery of regulatory elements by a computational method for phylogenetic footprinting Genome Res 2002 12 739 748 11997340 10.1101/gr.6902
Boffelli D McAuliffe J Ovcharenko D Lewis KD Ovcharenko I Pachter L Rubin EM Phylogenetic shadowing of primate sequences to find functional regions of the human genome Science 2003 299 1391 1394 12610304 10.1126/science.1081331
Sumiyama K Kim CB Ruddle FH An efficient cis-element discovery method using multiple sequence comparisons based on evolutionary relationships Genomics 2001 71 260 262 11161821 10.1006/geno.2000.6422
Grad YH Roth FP Halfon MS Church GM Prediction of similarly acting cis-regulatory modules by subsequence profiling and comparative genomics in Drosophila melanogaster and D.pseudoobscura Bioinformatics 2004 20 2738 2750 15145800 10.1093/bioinformatics/bth320
Lenhard B Sandelin A Mendoza L Engstrom P Jareborg N Wasserman WW Identification of conserved regulatory elements by comparative genome analysis J Biol 2003 2 13 12760745 10.1186/1475-4924-2-13
Loots GG Ovcharenko I Pachter L Dubchak I Rubin EM rVista for comparative sequence-based discovery of functional transcription factor binding sites Genome Res 2002 12 832 839 11997350 10.1101/gr.225502. Article published online before print in April 2002
Yuh CH Brown CT Livi CB Rowen L Clarke PJ Davidson EH Patchy interspecific sequence similarities efficiently identify positive cis-regulatory elements in the sea urchin Dev Biol 2002 246 148 161 12027440 10.1006/dbio.2002.0618
Hertz GZ Stormo GD Identifying DNA and protein patterns with statistically significant alignments of multiple sequences Bioinformatics 1999 15 563 577 10487864 10.1093/bioinformatics/15.7.563
Hughes JD Estep PW Tavazoie S Church GM Computational identification of cis-regulatory elements associated with groups of functionally related genes in Saccharomyces cerevisiae J Mol Biol 2000 296 1205 1214 10698627 10.1006/jmbi.2000.3519
van Helden J Andre B Collado-Vides J Extracting regulatory sites from the upstream region of yeast genes by computational analysis of oligonucleotide frequencies J Mol Biol 1998 281 827 842 9719638 10.1006/jmbi.1998.1947
Abnizova I te Boekhorst R Walter K Gilks WR Some statistical properties of regulatory DNA sequences, and their use in predicting regulatory regions in the Drosophila genome: the fluffy-tail test BMC Bioinformatics 2005 6 109 15857505 10.1186/1471-2105-6-109
Hutchinson GB The prediction of vertebrate promoter regions using differential hexamer frequency analysis Comput Appl Biosci 1996 12 391 398 8996787
Nazina AG Papatsenko DA Statistical extraction of Drosophila cis-regulatory modules using exhaustive assessment of local word frequency BMC Bioinformatics 2003 4 65 14690551 10.1186/1471-2105-4-65
Schroeder MD Pearce M Fak J Fan H Unnerstall U Emberly E Rajewsky N Siggia ED Gaul U Transcriptional control in the segmentation gene network of Drosophila PLoS Biol 2004 2 E271 15340490 10.1371/journal.pbio.0020271
Wingender E Chen X Fricke E Geffers R Hehl R Liebich I Krull M Matys V Michael H Ohnhauser R Pruss M Schacherer F Thiele S Urbach S The TRANSFAC system on gene expression regulation Nucleic Acids Res 2001 29 281 283 11125113 10.1093/nar/29.1.281
Baldi P Brunak S Chauvin Y Andersen CA Nielsen H Assessing the accuracy of prediction algorithms for classification: an overview Bioinformatics 2000 16 412 424 10871264 10.1093/bioinformatics/16.5.412
Mitchell TM Machine Learning McGraw-Hill series in computer science 1997 New York , McGraw-Hill xvii, 414 p.
|
16253142
|
PMC1291357
|
CC BY
|
2021-01-04 16:27:46
|
no
|
BMC Bioinformatics. 2005 Oct 27; 6:262
|
utf-8
|
BMC Bioinformatics
| 2,005 |
10.1186/1471-2105-6-262
|
oa_comm
|
==== Front
BMC BiotechnolBMC Biotechnology1472-6750BioMed Central London 1472-6750-5-301627115610.1186/1472-6750-5-30Research ArticleUse of endogenous signal sequences for transient production and efficient secretion by moss (Physcomitrella patens) cells Schaaf Andreas [email protected] Stefanie [email protected] Armin [email protected] Ralf [email protected] Gilbert [email protected] Eva L [email protected] Department of Plant Biotechnology, Faculty of Biology, University of Freiburg, Schaenzlestr. 1, 79104 Freiburg, Germany2 greenovation Biotech GmbH, Boetzinger Str. 29b, 79111 Freiburg, Germany3 Department of Plant Biochemistry and Biotechnology, University of Münster, Hindenburgplatz 55, 48143 Münster, Germany2005 7 11 2005 5 30 30 20 5 2005 7 11 2005 Copyright © 2005 Schaaf et al; licensee BioMed Central Ltd.This is an Open Access article distributed under the terms of the Creative Commons Attribution License (), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
Background
Efficient targeting to appropriate cell organelles is one of the bottlenecks for the production of recombinant proteins in plant systems. A common practice is to use the native secretory signal peptide of the heterologous protein to be produced. Though general features of secretion signals are conserved between plants and animals, the broad sequence variability among signal peptides suggests differing efficiency of signal peptide recognition.
Results
Aiming to improve secretion in moss bioreactors, we quantitatively compared the efficiency of two human signal peptides and six signals from recently isolated moss (Physcomitrella patens) proteins. We therefore used fusions of the different signals to heterologous reporter sequences for transient transfection of moss cells and measured the extra- and intracellular accumulation of the recombinant proteins rhVEGF and GST, respectively. Our data demonstrates an up to fivefold higher secretion efficiency with endogenous moss signals compared to the two utilised human signal peptides.
Conclusion
From the distribution of extra- and intracellular recombinant proteins, we suggest translational inhibition during the signal recognition particle-cycle (SRP-cycle) as the most probable of several possible explanations for the decreased extracellular accumulation with the human signals. In this work, we report on the supremacy of moss secretion signals over the utilised heterologous ones within the moss-bioreactor system. Though the molecular details of this effect remain to be elucidated, our results will contribute to the improvement of molecular farming systems.
==== Body
Background
The product range for recombinant biopharmaceuticals includes relatively simple proteins, for example insulin, and ends with candidates that require extensive secondary modification like erythropoietin or coagulation factor IX [1,2]. Commercial-scale production of proteins (molecular farming) nowadays is mainly performed with bacteria or mammalian cell-culture systems [1,3,4]. Bacterial production systems are highly efficient and well established for the production of comparably simple proteins but they come to their limits if complex posttranslational processing is required. One major problem is the absence of protein glycosylation in prokaryotes. Mammalian cell lines, on the other hand, perform human-like posttranslational modifications and offer a high product quality but hold risks of product contamination and high overall production costs (e.g. [5]).
Plants have emerged as a safe and cost-effective alternative to the traditional systems (e.g. [6,7]). They synthesize proteins with correct assemblage and the posttranslational modifications of higher eukaryotes. In contrast to microbial systems, glycosylated plant proteins carry complex-type glycans with the same core structure as human glycoproteins. The glycosylation pattern of plant proteins, however, differs from that in humans concerning two residues linked to the core structure as well as terminal sugar moieties [2,8]. The adaptation of the plant glycan pattern to its human counterpart (humanisation) poses a challenge for the implementation of plant systems with regard to the production of recombinant pharmaceutical proteins [9] as differences in the glycan structure could substantially alter protein characteristics and even act immunogenically in patients [10-13].
The moss Physcomitrella patens is the only known plant in which homologous recombination occurs in a frequency that allows its application as an engineering tool for targeted knock out and gene replacement. With this technique a humanisation of the moss glycosylation pattern was performed recently [14,15]. Together with photoautotrophic growth in bioreactors these advantages make Physcomitrella an ideal system for the production of plant-made pharmaceuticals [16].
In molecular farming approaches, recombinant proteins, which require correct posttranslational modifications, are often targeted to the secretory pathway. This, on the one hand, is due to the fact that enzymes responsible for various modifications, e. g. glycosylation and protein trimming, are located within the membranes of endoplasmic reticulum (ER) and Golgi apparatus. On the other hand, secretion of the protein of interest into the culture medium separates it from the pool of intracellular proteins and therefore drastically facilitates its purification and reduces costs for downstream processing [1].
A common way to achieve secretion of recombinant proteins is the utilisation of the native signal peptide, if any is present. The mechanism of protein secretion depends on the recognition of an N-terminal signal peptide and is conserved among all eukaryotes [17]. Though the general hydrophobic character of secretion signals is conserved between plants and animals, they are not universally interchangeable [18]. Additional features might contribute to the efficiency of signal peptide recognition [19]. In order to establish moss as a production system for recombinant proteins, we intended to enrich our toolboxes with a set of endogenous signal sequences from the moss. The availability of Physcomitrella secretion signals gives us the opportunity to secrete even proteins that are located within the cytosol in their host organism and thus lack a secretion signal.
Here, we describe the comparison of two heterologous signal peptides from human origin, the vascular endothelial growth factor (hVEGF) and the coagulation factor IX (hFIX), respectively, in the moss. Evaluation of a heterologous signal and the signal peptide of Physcomitrella aspartic proteinase PpAP1 [20] revealed better secretion of recombinant hVEGF (rhVEGF) with the moss signal.
In addition, the secretion efficiency of five putative signal peptides originating from recently isolated and identified Physcomitrella extracellular proteins xyloglucan endotransglycosylase/hydrolase (PpXTH1), pectin methylesterase (PpPME1), fasciclin-like protein (PpFLP), carbonic anhydrase-like protein (PpCALP), and lipid-transfer protein (PpLTP) (Tintelnot et al., manuscript in preparation) was tested in comparison to the PpAP1 secretion signal. These signal sequences were fused to hVEGF and amounts of secreted rhVEGF were quantified by ELISA measurements. All of the investigated moss signal peptides showed at least comparable or even higher secretion efficiency than the human signals. Besides, we underline the functionality of a transient production technique, that allows the fast production of recombinant proteins [21]. Within the field of molecular farming these techniques are essential for proof-of-concept studies prior to the establishment of production lines.
Results
Efficiency of human secretion signals in Physcomitrella patens
In order to improve transient production and secretion of pharmaceutically interesting proteins in the moss Physcomitrella patens, we first tested the efficiency of two human signal peptides. These were derived from the human blood clotting factor IX (hFIX) and the vascular endothelial growth factor (hVEGF). The secretion signals of both proteins, FSP and VSP, respectively, were tested with hVEGF as a reporter gene. Physcomitrella protoplasts were transiently transfected with the respective plasmids followed by quantification of extra- and intracellular concentrations of the recombinant proteins by ELISA (Fig. 1). Testing the hFIX signal peptide FSP, the moss cells secreted on average 18.17 ng/ml recombinant hVEGF (rhVEGF) to the culture supernatant, whereas the hVEGF signal VSP yielded a slightly higher average value (25.27 ng/ml). Comparison with intracellular concentrations of rhVEGF proved that both heterologous signals from human origin were quite effective in secreting the reporter protein from moss cells.
Transient production of recombinant glutathione S-transferase
Having proven the general functionality of heterologous secretion signals we aimed to compare the efficiency of a human and a moss signal peptide. We chose VSP that secreted a slightly higher amount of the reporter protein, and ASP, a moss signal sequence of the recently characterised aspartic proteinase PpAP1 [20]. The signal sequences were fused to the cDNA of glutathione S-transferase (GST) from Schistosoma japonicum, which served as a reporter protein. Five days after transient transfection of moss cells, reporter concentrations were quantified by ELISA (Fig. 2). With the moss signal peptide ASP, the protoplasts secreted between 14.11 and 24.33 ng/ml rGST into the culture supernatant. On the contrary, the human signal peptide yielded values in the range from 4.50 to 7.17 ng/ml. Taken the means of these values, when using the moss signal peptide, the concentration of rGST in the medium (18.10 ng/ml) was almost three times higher than with the human signal (6.10 ng/ml) (Fig. 2A).
For the quantification of intracellular rGST, the crude protein extracts were directly measured by ELISA (Fig. 2B). The relation of the values correlated to the extracellular situation. ASP-GST-transfected protoplasts contained 3.38 ng/ml rGST whereas those transfected with VSP-GST reached a mean value of no more than 0.35 ng/ml.
For visualisation of recombinant GST on Western blots (Fig. 2), GST was affinity-purified by using GSH-Sepharose, and the eluted proteins were separated by non-reducing SDS-PAGE. GST was detected with an anti-GST antibody. The blots showed the main bands at the expected size of 26 kDa for the GST protein. The weak additional higher band was caused by dimer formation. Weak bands detected in the control lanes may derive from cross-reactivity of the anti-GST antibody to moss GST or be an artefact from the protein purification procedure.
Transient production of recombinant human VEGF
To examine whether the supremacy of the moss signal is specific for GST, we subsequently analysed the secretion efficiency of ASP and VSP with the hVEGF reporter protein. When the natural signal peptide was used, the amount of extracellular rhVEGF ranged between 21.90 and 29.50 ng/ml (mean of 5 transfections: 25.48 ng/ml; Fig. 3A). Using the moss signal peptide resulted in a more than fivefold increase of rhVEGF in the culture medium (mean of 7 transfections: 111.14 ng/ml; single values ranged from 75 ng/ml to 162 ng/ml). The differences in extracellular rhVEGF accumulation of ASP-VEGF vs. VSP-VEGF-transfected cells were also reflected in the Western analysis (Fig. 3B). Additionally, the amounts of intracellular rhVEGF were determined by ELISA. ASP-VEGF-transfected protoplasts contained 1.0 ng/ml rhVEGF. Those transfected with VSP-VEGF reached an average value of 0.51 ng/ml.
Highly efficient secretion by endogenous signal peptides
Comparison of the intracellularly retained and extracellular proportions of rGST and rhVEGF, respectively, demonstrates that secretion from the moss cells was very efficient (Fig. 4). Most of rGST, i.e. 90.6%, was secreted to the culture medium. Up to 99.8% of the produced rhVEGF accumulated in the culture medium, whereas no significant amounts were found inside the cells, neither with the moss signal nor with the human one. Furthermore, the ratio of intra- to extracellular protein did not differ significantly between the human and the moss secretion signal. With respect to the biological processes, we suggest as the most probable among several possible reasons for the difference in secretion efficiency a less efficient translation process when using the human signal (see discussion).
After having proven the high efficiency of one endogenous moss signal peptide in the moss system, we aimed to study the efficiency of additional moss signal sequences and tested five other putative secretion signals from Physcomitrella in transient transfection assays. These signal peptides originated from putative extracellular proteins isolated from moss culture medium (Tintelnot et al., manuscript in preparation). The signal sequences of PpFLP (fasciclin-like protein), PpLTP (lipid-transfer protein), PpPME1 (pectin methylesterase), PpXTH1 (xyloglucan endotransglycosylase/hydrolase), PpCALP (carbonic anhydrase-like protein) as well as PpAP1 were fused to rhVEGF in order to analyse their secretion efficiency by ELISA after transient transfection of moss protoplasts. All further investigated moss signal peptides showed comparable or even higher secretion efficiencies than the PpAP1 signal ASP (Fig. 5). The rhVEGF amount in the medium achieved by the signal peptides of PpPME1 and PpXTH1 was approximately 80% higher than with the PpAP1 signal.
A comparison of the different moss signal peptides is shown in Fig. 6. According to SignalP, for the six cloned sequences the probabilities for being a signal peptide were always higher than 99% and their functionality in vivo was proven as demonstrated above. All of the signals comprise many hydrophobic amino acid residues in their central part indicating the hydrophobic region. The N-terminal part has at least one positive residue in most of the signals (except PpCALP and VSP, respectively), whereas the C-terminal part always contains a small and neutral residue at position -1 and -3 with respect to the cleavage site as it is proposed for signal sequences [22,23]. The probability for correct prediction of the cleavage site varied from 58.3% for PpCALP to 99.8% for PpFLP. High variations were also found regarding the overall length of the signal sequences, which ranges from 20 to 27 amino acids. Analysis of available Physcomitrella sequences resulted in an average length of 26 amino acids for 86 signal peptides predicted.
Taken together, the results of the transient production of recombinant proteins demonstrate that the secretion machinery of Physcomitrella works very efficiently. Furthermore it was shown that the utilisation of an endogenous signal peptide can lead to a remarkable increase in recombinant protein production in comparison with the heterologous sorting signals used in this study.
Discussion
An important issue in molecular farming is the accumulation and downstream processing of recombinant proteins. Mostly, the proteins are targeted to the secretory pathway as the complete set of enzymes responsible for posttranslational modifications, especially for protein glycosylation, is located within the secretory compartments, i.e. ER and Golgi apparatus. In case of using suspension cultures as a production system secretion of proteins provides the additional advantage of easier protein purification.
Several systems involving secretion in protein production have been established. Among these are both, cell suspension cultures and whole plant systems, e.g. rhizosecretion [24,25] or production in the water plant Lemna [26]. Seed storage of the recombinant product also involves secretory targeting, as deposition in protein storage vacuoles is achieved by traffic via the secretory pathway [27].
A common practice for the production and secretion of human proteins in plants is to use their native signal peptide. This is possible because of the conservation of the signal recognition- and ER-insertion mechanisms throughout eukaryotes [17,28,29] and circumvents laborious steps of signal engineering. However, the amino-acid composition of signal peptides is extremely variable and the exact mechanism of their recognition by SRP and the translocon is not yet completely understood. Their determining feature is the hydrophobicity of a ten to fifteen residue core sequence, which surprisingly does not show any sequence conservation [30]. The functionality of the signals is only marginally impaired by amino-acid exchanges, as long as the hydrophobic core is not affected [22,31]. Exchanges within the core can in contrast totally impede signal recognition, even when leading to a higher hydrophobicity [32]. A closer look at the nature of the six moss signal peptides analysed in this study revealed a high percentage of the hydrophobic amino acids L, F, I, M, V and W [22] in the central part of the signal peptides, even though the hydrophobic stretches vary in length and position. N-terminal of the hydrophobic centre positively charged residues have been proposed [33]. Except for PpCALP, all of the moss signal sequence N-termini contain at least one arginine residue. For the signal peptide of PpCALP, which has no positively charged residue, the initiation methionine could perhaps compensate the lack of arginine [34]. All of the investigated signal sequences follow the (-3, -1)rule, which means that the two amino acids at position one and three previous to the cleavage site have to be small and neutral [23]. The HMM cleavage site prediction from SignalP showed a rather low probability for the signal sequences of PpPME1 (0.684) and of PpCALP (0.583). Additionally, in case of PpCALP, the SignalP-neural network predicted a signal peptide length of 32 amino acids instead of 27 amino acids which was predicted with SignalP-HMM and used in our study. However, experimental data showed that the sequences for both PpCALP and PpPME1 were not only functional, but resulted in a very efficient secretion of the rhVEGF protein. In addition, cleavage site prediction of ASP and VSP was verified experimentally. The analysis revealed that secretion dropped drastically when the last amino acid of a correctly predicted signal was deleted (data not shown).
The average length of eukaryotic signal peptides was calculated to be 22 amino acids [22,35], whereas the average length of 86 analysed Physcomitrella signal peptides is 26 residues. Although some correlations between the analysed Physcomitrella signal peptides and other eukaryotic secretion signals can be found, the variations within the amino acid composition remain high [34]. A functional role of additional, most likely structural determinants of signal peptides is assumed [19].
Previous studies evaluating endogenous vs. heterologous signal sequences in several species of the plant and animal kingdoms revealed for the majority a higher efficiency of kingdom-specific signals [18,36,37]. As secretion is one of the bottlenecks when establishing a production system, we aimed to define highly efficient signals to transport the recombinant proteins to the culture medium of the moss bioreactor system [16]. Thus we decided to compare six endogenous moss signal peptides plus two heterologous sequences from human origin. All of the moss signals turned out to drive secretion more efficiently than the human signal peptides. Two signals, the moss-derived ASP and the human VSP, were analysed in more detail with two different reporters, rGST (glutathione S-transferase) and rhVEGF (vascular endothelial growth factor), respectively. By secretion of the recombinant reporter protein GST we could show a threefold increase in extracellular accumulation with the ASP sequence compared to the human VSP. With the moss secretion signal, extracellular rhVEGF amount was up to fivefold more. Surprisingly, we did not find an increased intracellular accumulation in case of the human signal peptide. Amongst several possible alternatives (see below), our preferred explanation is that recognition by the SRP and maybe ER-insertion of the human signal peptide is suboptimal thus decreasing the translation rate of the recombinant protein. The differences in secretion efficiency clearly prove that besides hydrophobicity, some additional features, that may have a species-specific character, play a crucial role in signal recognition. As it is known that SRP occupies the EF-binding-site of the ribosome and thereby temporarily arrests translation [38], the cause for a reduced production rate with the human signal could be found in a hindered dissociation of SRP after docking of the translation complex to the translocon.
The second important interaction of the nascent signal peptide is performed with ER-membrane lipids as well as the translocon itself [39,40]. This interaction could demand unknown structural features of a signal peptide, which are not completely cross-host compatible, as mammalian and plant membranes differ significantly in their lipid composition [41].
However, besides a translational inhibition caused by hindered SRP-dissociation, alternative explanations for the worse performance of the human signals do exist: The protein in part may fail to be properly folded (perhaps because SP cleavage might not be efficient) and is subsequently degraded. Alternatively, transcription of that particular construct may be inefficient. The only possible determinant for such events can be found in the relatively short signal peptide sequences themselves, as all used vectors do not differ in any other detail. Apart from the above-mentioned translational inhibition, additional influence of the signal peptide on recombinant protein stability could only take place during, or shortly after ER-insertion. When the translation of the protein is completed, the signal peptides are cleaved. Thereafter, recombinant products do not differ in any residue and can therefore not exhibit different stabilities. Regulatory, post-targeting functions of cleaved signals are proposed by other authors and can therefore not be completely excluded in our case [42,43].
Conclusion
Even if the structural background remains to be elucidated, our results demonstrate that the six moss sequences are more efficient than the two human ones used in our study in promoting secretion of recombinant human VEGF. In the interplay with optimised gene expression [44] and product stabilisation [45], the utilisation of moss-endogenous secretion signals could greatly enhance the productivity of the moss-bioreactor system.
Methods
Plant material and growth conditions
Physcomitrella patens (Hedw.) B.S.G. was grown axenically under standard conditions (agitated liquid modified Knop medium, 250 mg/l KH2PO4, 250 mg/l MgSO4 × 7H2O, 250 mg/l KCl, 1000 mg/l Ca(NO3)2 × 4H2O, 12.5 mg/l FeSO4 × 7H2O, pH 5.8) in a growth chamber or in photobioreactors (25 ± 1°C; light provided from outside by fluorescent tubes, Philips TL-D 36W/25; light flux of 55 μmol s-1 m-2, light-dark regime of 16:8 h) as described [46].
Prediction of signal sequences
Signal sequences of the putative extracellular Physcomitrella proteins Fasciclin-like protein (PpFLP [EMBL:AJ843251]), Lipid-transfer protein (PpLTP [EMBL:AJ843246]), Pectin methylesterase1 (PpPME1 [EMBL:AJ843245]), Xyloglucan endotransglycosylase/hydrolase1 (PpXTH1 [EMBLAJ843248]) and Carbonic anhydrase-like protein (PpCALP [EMBL:AJ843247]) (Tintelnot et al., manuscript in preparation) as well as for PpAP1 [20] were predicted by SignalP 3.0 Server [23,33,35].
For analysis of the average length of moss signal peptides, all 486 publicly available full-length Physcomitrella coding sequences as of May, 2005 were retrieved from Genbank and 86 putative signal peptides predicted using the SignalP 3.0 neural network. In addition, a set of 408 Arabidopsis thaliana coding sequences were retrieved using the query keywords "mature" and "peptide" and 39 putative signal peptides predicted.
Construction of targeting vectors
For the construction of pASP-GST and pVSP-GST the coding sequence (cds) of GST was PCR-amplified with the primers GSTf (5'-AGATCTATGTCCCCTATACTAGGT-3') and GSTr (5'-GAGCTCTCACGGAACCAGATCCGATTT-3') with pGEX (Amersham Biosciences, Germany) serving as template. The PCR product was directly cloned into the pCR4-TOPO vector (Invitrogen, Germany). Introduced BglII and SalI restriction sites enabled the replacement of the GFP-cds of mAV4 [47] by the newly generated cds for GST. The resulting plasmid was named pGST. The ASP-signal was amplified with the primers ASPf (5'-AGATCTTGCCTCAGCTAAGGCTGC-3') and ASPr (5'-AGATCTTGCCTCAGCTAAGGC-3') using pSP-GFP [20] as template. The PCR product was directly cloned into pCR4-TOPO vector. For subcloning in pGST, the introduced SalI and BglII sites were used. The resulting plasmid was called pASP-GST.
For the construction of pVSP-GST, the plasmid pVSP-GFP served as a starting point. It is a derivative of mAV4, with the cds of the rhVEGF-signal sequence (VSP) fused upstream to GFP. The GST-cds was introduced by using the SalI and BglII restriction sites for replacing the GFP-cds.
The plasmids pVEGF and pASP-VEGF were constructed using the plasmids pRT101_C3_VEGF [48] containing the cds for human VEGF as starting point. The human VEGF signal was amplified from pRT101_P21 [48] with MOB_323 (5'-ATACTCGAGGAAGATGAACTTTCTGCTGTCTTGG-3') and MOB_349 (5'-CTGCCATGGGTGCAGCCTGGGACCAC-3') and cut at XhoI/NcoI restriction sites for ligation into pRT101_C3_VEGF. The ASP-signal from Physcomitrella patens (PpAP1) located on the plasmid pSP-GFP was amplified using MOB_642 (5'-GCCCTCGAGGAAGATGGGGGCATCGAGGAGT-3') and MOB_765 (5'-CTGCCATGGGTGCTGCCTCAGCTAAGGC-3'). By use of the restriction sites XhoI and NcoI the was cloned into pRT101_C3_VEGF.
Construction of the plasmid pUC-SP-VEGF was carried out by replacing the luciferase-cds and terminator in pluc-Actin [44] with the cds of rhVEGF and 35S-terminator from pAP1-SP-VEGF-His [15] via the enzymes Hind III and Nco I. The different secretion signals – plus four basepairs (GCAC) missing from the rhVEGF cds after Nco I digest – were cloned into the pUC-SP-VEGF via the restriction enzymes Nco I and Xho I after PCR amplification and direct cloning into the pCR4-TOPO vector (Invitrogen). The following primers were used for amplification: PME-SPf (5'-CGAGCGCAATGGGGAGCATGTC-3') and PME-SPr (5'-TCCATGGGTGCTGCTGACGCGGGCTTC-3'), XTH-SPf (5'-GCTGCAGAGAAATGGGGTTCAATAGAGG-3') and XTH-SPr (5'-TCCATGGGTGCAGCGTGACTGCCGAC-3'), CALP-SPf (5'-GCTCGAGGGCGATGGCGAGCCAACTTG-3') and CALP-SPr (5'-TCCATGGGTGCTGCCCACGCGCCAAC-3'), FLP-SPf (5'-GCTCGAGAACAATGGCGCTTTCGTCGG-3') and FLP-SPr (5'-TCCATGGGTGCGGCATACGCTTGAGGC-3'), GLP-SPf (5'-GCTCGAGTAACATGGCTGCCCGTTTCG-3') and GLP-SPr (5'-TCCATGGGTGCAGCATACACCATGGCC-3'), LTP-SPf (5'-GCTCGAGGAGGATGGCACAACGCATTTG-3') and LTP-SPr (5'-TCCATGGGTGCAGCAGACACTCCAGAG-3'), and AP-SPf (5'-TCTCGAGGACGATGGGGGCATCGAGGAG-3') and AP-SPr (5'-TCCATGGGTGCTGCCTCAGCTAAGGCTG-3') for the endogenous signal peptides as well as VEGF-SPfor (5'-TGCTCGAGCGAGATGAACTTTCTGCTGTC-3') and rev (5'-TTCCATGGGTGCAGCCTGGGACCACTTG-3') for the heterologous secretion signal.
Transfection of Physcomitrella protoplasts
The DNA for transfection was prepared with Qiagen's Plasmid Maxi kit. A total number of 3 × 105 protoplasts was transiently transfected with 30–50 μg of circular plasmid DNA. Protoplasts were isolated from bioreactor-grown material and transfected as described [46]. After transfection, the protoplasts were suspended in 100 μl 3 M-medium supplemented with 0.01% BSA and subsequently transferred to 96 well plates (Nunclon™ surface; Nunc, Denmark). After 24 hrs the culture medium was replaced with fresh medium. Five days after transfection the supernatant was harvested. Protoplasts were resuspended in 100 μl PBS buffer supplemented with 1% Triton X100 and 1% plant protease inhibitor cocktail (Sigma, Germany). These samples were immediately processed with protein extraction.
Extraction of intracellular protein
For the extraction of intracellular protein, glass beads (1 mm diameter) were added to the tubes. Cell disruption was carried out in a beadmill (Tissuelyzer, Qiagen) for 1 min at 30 Hz. Cell debris was sedimented (14,000 rpm, 10 min, 4°C) and the supernatant was transferred into a fresh tube for further assaying.
Measurement of protein concentration by ELISA
Extra- and intracellular concentrations of recombinant protein were determined by ELISA. In order to adjust pH and ionic conditions, the harvested culture medium was supplemented with an equal volume of 2 × PBS and kept on ice until ELISA measurement or stored at -80°C. Cell extracts were directly applied to the ELISA. All samples were diluted appropriately and measured in duplicates. Separate standard dilutions were prepared for medium- and cell extract-conditions respectively, because of the samples' different buffer conditions. For determination of recombinant GST concentrations, the 96 well detection module (Amersham Biosciences) was used following the manufacturer's instructions. Recombinant hVEGF samples were analysed by sandwich ELISA (anti-hVEGF, capture: #AF-293-NA, anti-hVEGF, detection: #BAF-293, rhVEGF121, standard: #298-VS, R&D, Germany; Nunc-Immunosorb plates). Dilutions were made in 1 × PBS/0.1% BSA (Serva, Germany).
Purification of GST
Recombinant GST was affinity-purified using GSH-Sepharose (Gluthathione Sepharose 4B, Amersham Biosciences). Therefore a 50%-slurry of the GSH-Sepharose was prepared according to the manufacturer's instructions and 1/50 volume of the slurry was added to the protein solution. In case of extracellular GST, 1/10 volume of 10 × PBS buffer was added to adjust pH of the culture medium. Binding was carried out for 2 hours at room temperature with gentle end-over-end rotation. GSH-Sepharose was sedimented at 500 × g for five minutes and the supernatant was discarded. The sepharose was washed once with ten bed volumes of PBS buffer. After adding 50 μl of SDS sample buffer, the samples were heated at 95°C for five minutes, and centrifuged for one minute at 13,000 × g. The resulting supernatant was analysed by SDS-PAGE followed by Western blot.
SDS-PAGE and western blot
For the visualisation by Western blot, the affinity-purified GST was first separated on a 12% SDS gel following standard techniques. The separated proteins were then transferred to a PVDF membrane (Immobilon-P, Millipore, Germany) with the semi-dry method. Transfer was carried out according to the instructions given by the membranes manufacturer. For signal detection, the ECL advance western blotting detection kit (Amersham Biosciences) was used. All steps were carried out according to the manufacturer's protocol. Anti-GST antibody (developed in rabbit, Sigma) was used at a 1:5,000 dilution. Anti-rabbit antibody (peroxidase-coupled, developed in donkey, Amersham Biosciences) was used for detection in a 1:500,000 dilution. For rhVEGF, anti-human VEGF (#AF-293-NA, 1:500), anti-goat IgG (#A5420, Sigma, 1:8000) and rhVEGF as standard was used.
Authors' contributions
AS carried out cloning, ELISA and western blot analyses of the GST-fusion constructs and drafted the main part of the manuscript. ST performed cloning and ELISA studies as well as computational analysis of the Physcomitrella signal peptides and was involved in preparing the manuscript. AB carried out cloning, ELISA and western blot analyses of the VEGF-fusion constructs. RR leads the chair plant biotechnology at Freiburg University. GG is scientific director of greenovation Biotech GmbH and guided the work on ASP-VEGF and VSP-VEGF. ELD was responsible for the design and development of these studies and writing of the manuscript; consequently she is the corresponding author. All authors read and approved the final manuscript.
Acknowledgements
We would like to thank Agnes Kinal, Axel Hoffmann and Ingrid Heger for their help with cloning, transient transfections, and ELISA. Many thanks to Stefan Rensing for SignalP analysis of Physcomitrella sequences. This work was funded by the German Federal Ministry of Education and Research (BMB+F; grant No. 0312624) and the Deutsche Forschungsgesellschaft (DFG, RE 837/6).
Figures and Tables
Figure 1 Comparison of heterologous signals from human origin. The secretion signals of human blood clotting factor IX (FSP) as well as human vascular endothelial growth factor (VSP) were fused to the coding sequence of vascular endothelial growth factor (hVEGF). Moss protoplasts were transiently transfected with the expression constructs and concentration of extracellular as well as intracellular rhVEGF was determined by ELISA. Mean values were taken from three transfections. Error bars indicate the absolute average deviation (AAD).
Figure 2 Concentrations of secreted recombinant GST. Physcomitrella protoplasts were transiently transfected with pVSP-GST and pASP-GST in eight independent transfections. After 5 days the concentrations of secreted and intracellularly-retained recombinant protein were measured by ELISA. Additionally, GST was affinity-purified from both medium and supernatant and detected by western blot. A: Culture medium of two transfections was pooled. Mean values of eight transfections are given. B: Protoplasts of eight transfections were pooled and GST was affinity-purified from the crude extracts. Values reflect the means of the two measurements. Control: mock-transfected protoplasts.
Figure 3 Intra- and extracellular concentrations of recombinant rhVEGF. Physcomitrella protoplasts were transiently transfected with pVSP-VEGF and pASP-VEGF in four (pVSP-VEGF and control) and seven (pASP-VEGF) independent transfections. After 5 days the concentrations of secreted and intracellularly retained recombinant protein were measured by ELISA. A: mean values of concentrations of extra- and intracellular rhVEGF measured by ELISA. B: Western blot with culture medium of transiently rhVEGF-producing cell. Control: untransfected moss protoplasts.
Figure 4 Intra- and extracellular distribution of recombinant protein. The relations between intra- and extracellular amounts of recombinant rhVEGF (A) and GST (B) are given. Values reflect the mean amounts of recombinant protein per transfection. ASP: plasmids with moss signal peptide; VSP: plasmids with human signal peptide; extra: amount of recombinant protein in culture medium; intra: intracellular amounts of recombinant protein.
Figure 5 Secretion efficiency of different moss signal peptides. Signal sequences of PpAP1, PpFLP, PpLTP, PpPME1, PpXTH1, PpCALP were fused to rhVEGF. Shown are the mean values of three independent experiments with three transfections each time (n = 9). Prior to calculation of averages, outliers were removed if the z-score was >1. RhVEGF concentration of the PpAP1-construct was set to 100%. Error bars indicate the absolute average deviation (AAD) of the three experiments. Control: untransfected Physcomitrella protoplasts.
Figure 6 Physcomitrella patens signal peptides. Amino acid sequences of signal peptides from PpAP1 (ASP), VEGF (VSP) and the extracellular proteins PpFLP, PpLTP, PpPME1, PpXTH1, PpCALP, and VEGF are shown as predicted by SignalP. Letters with yellow background represent hydrophobic amino acid residues (V, L, I, W, F, M), orange letters positive (K, R) and green letters small and neutral residues (A, C, G, N, P, S, T, V). The predicted cleavage sites are indicated as well as the positions adjacent to this site. In the right part of the figure probabilities for each signal peptide and the cleavage site given by the SignalP-HMM prediction as well as the length of the signal peptides are shown.
==== Refs
Schmidt FR Recombinant expression systems in the pharmaceutical industry Appl Microbiol Biotechnol 2004 65 363 372 15480623 10.1007/s00253-004-1656-9
Gomord V Faye L Posttranslational modification of therapeutic proteins in plants Curr Opin Plant Biol 2004 7 171 181 15003218 10.1016/j.pbi.2004.01.015
Andersen DC Krummen L Recombinant protein expression for therapeutic applications Curr Opin Biotechnol 2002 13 117 123 11950561 10.1016/S0958-1669(02)00300-2
Thiel KA Biomanufacturing, from bust to boom...to bubble? Nature Biotechnol 2004 22 1365 1372 15529157 10.1038/nbt1104-1365
Ma JK Drake PM Christou P The production of recombinant pharmaceutical proteins in plants Nature Rev Genet 2003 4 794 805 14526375 10.1038/nrg1177
Fischer R Stoger E Schillberg S Christou P Twyman RM Plant-based production of biopharmaceuticals Curr Opin Plant Biol 2004 7 152 158 15003215 10.1016/j.pbi.2004.01.007
Hellwig S Drossard J Twyman RM Fischer R Plant cell cultures for the production of recombinant proteins Nature Biotechnol 2004 22 1415 1422 15529167 10.1038/nbt1027
Lerouge P Cabanes-Macheteau M Rayon C Fischette-Laine AC Gomord V Faye L N-glycoprotein biosynthesis in plants: recent developments and future trends Plant Mol Biol 1998 38 31 48 9738959 10.1023/A:1006012005654
Larrick JW Thomas DW Producing proteins in transgenic plants and animals Curr Opin Biotechnol 2001 12 411 418 11551472 10.1016/S0958-1669(00)00236-6
Garcia-Casado G Sanchez-Monge R Chrispeels MJ Armentia A Salcedo G Gomez L Role of complex asparagine-linked glycans in the allergenicity of plant glycoproteins Glycobiol 1996 6 471 477
van Ree R Cabanes-Macheteau M Akkerdaas J Milazzo JP Loutelier-Bourhis C Rayon C Villalba M Koppelman S Aalberse R Rodriguez R Faye L Lerouge P Beta(1,2)-xylose and alpha(1,3)-fucose residues have a strong contribution in IgE binding to plant glycoallergens J Biol Chem 2000 275 11451 11458 10753962 10.1074/jbc.275.15.11451
Wilson IB Zeleny R Kolarich D Staudacher E Stroop CJ Kamerling JP Altmann F Analysis of Asn-linked glycans from vegetable foodstuffs: widespread occurrence of Lewis a, core alpha1,3-linked fucose and xylose substitutions Glycobiol 2001 11 261 274 10.1093/glycob/11.4.261
Bardor M Faveeuw C Fitchette AC Gilbert D Galas L Trottein F Faye L Lerouge P Immunoreactivity in mammals of two typical plant glyco-epitopes, core alpha(1,3)-fucose and core xylose Glycobiol 2003 13 427 434 10.1093/glycob/cwg024
Huether CM Lienhart O Baur A Stemmer C Gorr G Reski R Decker EL Glyco-engineering of moss lacking plant-specific sugar residues Plant Biol 2005 7 292 299 15912449 10.1055/s-2005-837653
Koprivova A Stemmer C Altmann F Hoffmann A Kopriva S Gorr G Reski R Decker EL Targeted knockouts of Physcomitrella lacking plant-specific immunogenic N-glycans Plant Biotechnol J 2004 2 517 523 10.1111/j.1467-7652.2004.00100.x
Decker EL Reski R The moss bioreactor Curr Opin Plant Biol 2004 7 166 170 15003217 10.1016/j.pbi.2004.01.002
Blobel G Walter P Chang CN Goldman BM Erickson AH Lingappa VR Translocation of proteins across membranes: the signal hypothesis and beyond Symp Soc Exp Biol 1979 33 9 36 524275
Tessier DC Thomas DY Khouri HE Laliberte F Vernet T Enhanced secretion from insect cells of a foreign protein fused to the honeybee melittin signal peptide Gene 1991 98 177 183 2016060 10.1016/0378-1119(91)90171-7
Matoba S Ogrydziak DM Another factor besides hydrophobicity can affect signal peptide interaction with signal recognition particle J Biol Chem 1998 273 18841 18847 9668059 10.1074/jbc.273.30.18841
Schaaf A Reski R Decker EL A novel aspartic proteinase is targeted to the secretory pathway and to the vacuole in a moss, Physcomitrella patens Eur J Cell Biol 2004 83 145 152 15260436 10.1078/0171-9335-00371
Baur A Kaufmann F Rolli H Weise A Luethje R Berg B Braun M Baeumer W Kietzmann M Reski R Gorr G A fast and flexible PEG-mediated transient expression system in plants for high level expression of secreted recombinant proteins J Biotechnol 2005 119 332 342 15993971 10.1016/j.jbiotec.2005.04.018
von Heijne G Signal sequences. The limits of variation J Mol Biol 1985 184 99 105 4032478 10.1016/0022-2836(85)90046-4
Nielsen H Engelbrecht J Brunak S von Heijne G Identification of prokaryotic and eukaryotic signal peptides and prediction of their cleavage sites Protein Eng 1997 10 1 6 9051728 10.1093/protein/10.1.1
Borisjuk NV Borisjuk LG Logendra S Petersen F Gleba Y Raskin I Production of recombinant proteins in plant root exudates Nature Biotechnol 1999 17 466 469 10331806 10.1038/8643
Gaume A Komarnytsky S Borisjuk N Raskin I Rhizosecretion of recombinant proteins from plant hairy roots Plant Cell Rep 2003 21 1188 1193 12819926 10.1007/s00299-003-0660-3
Gasdaska JR Spencer D Dickey L Advantages of therapeutic protein production in the aquatic plant Lemna Bioprocess J 2003 2 49 56
Jurgens G Membrane trafficking in plants Annu Rev Cell Dev Biol 2004 20 481 504 15473849 10.1146/annurev.cellbio.20.082503.103057
Walter P Blobel G Translocation of proteins across the endoplasmic reticulum III. Signal recognition protein (SRP) causes signal sequence-dependent and site-specific arrest of chain elongation that is released by microsomal membranes J Cell Biol 1981 91 557 561 7309797 10.1083/jcb.91.2.557
Rapoport TA Jungnickel B Kutay U Protein transport across the eukaryotic endoplasmic reticulum and bacterial inner membranes Annu Rev Biochem 1996 65 271 303 8811181 10.1146/annurev.bi.65.070196.001415
Stroud RM Walter P Signal sequence recognition and protein targeting Curr Opin Struct Biol 1999 9 754 759 10607673 10.1016/S0959-440X(99)00040-8
Zheng N Gierasch LM Signal sequences: the same yet different Cell 1996 86 849 852 8808619 10.1016/S0092-8674(00)80159-2
Rothe C Lehle L Sorting of invertase signal peptide mutants in yeast dependent and independent on the signal-recognition particle Eur J Biochem 1998 252 16 24 9523707 10.1046/j.1432-1327.1998.2520016.x
von Heijne G Signal sequences. The limits of variation J Mol Biol 1985 184 99 105 4032478 10.1016/0022-2836(85)90046-4
von Heijne G The signal peptide J Membr Biol 1990 115 195 201 2197415 10.1007/BF01868635
Bendtsen JD Nielsen H von Heijne G Brunak S Improved prediction of signal peptides: SignalP 3.0 J Mol Biol 2004 340 783 795 15223320 10.1016/j.jmb.2004.05.028
Sijmons PC Dekker BM Schrammeijer B Verwoerd TC van den Elzen PJ Hoekema A Production of correctly processed human serum albumin in transgenic plants Biotechnology (NY) 1990 8 217 221 10.1038/nbt0390-217
Ma S Huang Y Davis A Yin Z Mi Q Menassa R Bandle JE Jevnikar AM Production of biologically active human interleukin-4 in transgenic tobacco and potato Plant Biotechnol J 2005 3 309 318 10.1111/j.1467-7652.2005.00125.x
Halic M Becker T Pool MR Spahn CM Grassucci RA Frank J Beckmann R Structure of the signal recognition particle interacting with the elongation-arrested ribosome Nature 2004 427 808 814 14985753 10.1038/nature02342
Hamman BD Hendershot LM Johnson AE BiP maintains the permeability barrier of the ER membrane by sealing the lumenal end of the translocon pore before and early in translocation Cell 1998 92 747 758 9529251 10.1016/S0092-8674(00)81403-8
Van den Berg B Clemons WM JrCollinson I Modis Y Hartmann E Harrison SC Rapoport TA X-ray structure of a protein-conducting channel Nature 2004 427 36 44 14661030 10.1038/nature02218
Opekarova M Tanner W Specific lipid requirements of membrane proteins – a putative bottleneck in heterologous expression Biochim Biophys Acta 2003 1610 11 22 12586375
Martoglio B Intramembrane proteolysis and post-targeting functions of signal peptides Biochem Soc Trans 2003 31 1243 1247 14641034
von Heijne G Protein transport: Life and death of a signal peptide Nature 1998 396 111 113 9823886 10.1038/24036
Horstmann V Huether CM Jost W Reski R Decker EL Quantitative promoter analysis in Physcomitrella patens: a set of plant vectors activating gene expression within three orders of magnitude BMC Biotechnol 2004 4 13 15239842 10.1186/1472-6750-4-13
Baur A Reski R Gorr G Enhanced recovery of a secreted recombinant human growth factor using stabilizing additives and by co-expression of human serum albumin in the moss Physcomitrella patens Plant Biotech J 2005 3 331 340 10.1111/j.1467-7652.2005.00127.x
Frank W Decker EL Reski R Molecular tools to study Physcomitrella patens Plant Biol 2005 7 220 227 15912441 10.1055/s-2005-865645
Kircher S Wellmer F Nick P Rugner A Schafer E Harter K Nuclear import of the parsley bZIP transcription factor CPRF2 is regulated by phytochrome photoreceptors J Cell Biol 1999 144 201 211 9922448 10.1083/jcb.144.2.201
Gorr G Biotechnologische Nutzung von Physcomitrella patens (Hedw.) B.S.G PhD thesis 1999 University of Hamburg
|
16271156
|
PMC1291358
|
CC BY
|
2021-01-04 16:02:58
|
no
|
BMC Biotechnol. 2005 Nov 7; 5:30
|
utf-8
|
BMC Biotechnol
| 2,005 |
10.1186/1472-6750-5-30
|
oa_comm
|
==== Front
BMC CancerBMC Cancer1471-2407BioMed Central London 1471-2407-5-1431625963710.1186/1471-2407-5-143Research ArticleAssociation of paternal age at birth and the risk of breast cancer in offspring: a case control study Choi Ji-Yeob [email protected] Kyoung-Mu [email protected] Sue Kyung [email protected] Dong-Young [email protected] Sei-Hyun [email protected] Keun-Young [email protected] Daehee [email protected] Department of Preventive Medicine, Seoul National University College of Medicine, 28 Yongon-Dong Chongno-Gu, Seoul 110-799 Korea2 Department of General Surgery, Seoul National University College of Medicine, 28 Yongon-Dong Chongno-Gu, Seoul 110-799 Korea3 Department of General Surgery, Ulsan University College of Medicine, 388-1 Pungnap-2dong Songpa-gu, Seoul 138-736, Korea4 Cancer Research Institute, Seoul National University College of Medicine, 28 Yongon-Dong Chongno-Gu, Seoul 110-799 Korea2005 31 10 2005 5 143 143 8 7 2005 31 10 2005 Copyright © 2005 Choi et al; licensee BioMed Central Ltd.This is an Open Access article distributed under the terms of the Creative Commons Attribution License (), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
Background
Older paternal age may increase the germ cell mutation rate in the offspring. Maternal age may also mediate in utero exposure to pregnancy hormones in the offspring. To evaluate the association between paternal and maternal age at birth with the risk of breast cancer in female offspring, a case-control study was conducted in Korea.
Methods
Histologically confirmed breast cancer cases (n = 1,011) and controls (n = 1,011) with no present or previous history of cancer, matched on year of birth and menopausal status, were selected from several teaching hospitals and community in Seoul during 1995–2003. Information on paternal and maternal ages and other factors was collected by interviewed questionnaire. Odds ratio (OR) and 95% confidence interval (95% CI) were estimated by unconditional logistic regression model adjusting for family history of breast cancer in 1st or 2nd degree relatives, and lifetime estrogen exposure duration.
Results
The risk of breast cancer significantly increased as the paternal age increased (p for trend = 0.025). The association was stronger after controlling for maternal age; women whose fathers were aged ≥40 years at their birth had 1.6-fold increased risk of breast cancer compared with fathers aged <30 years. This association was profound in breast cancer cases in premenopausal women (OR = 1.9, 95% CI = 1.12–3.26, for paternal aged ≥40 vs. <30) (p for trend = 0.031). Although the risk of breast cancer increased as maternal age increased up to the intermediate, and then reduced; the risks in women whose mother were aged 25–29, 30–34, and ≥35 yrs at birth compared to women whose mothers were aged <25 years, were 1.2, 1.4, and 0.8, respectively, the trend was not significant (p for trend = 0.998).
Conclusion
These findings suggest that older paternal age increases the risk of breast cancer in their female offspring.
==== Body
Background
There have been growing evidences that prenatal factors may play an important role in determining breast cancer risk in adult life. These factors are hypothesized to affect breast cancer risk by altering the hormonal environment of the developing fetus [1], or by affecting the cumulative frequency of germ cell mutations [2,3].
A numbers of epidemiologic research have suggested that older parental ages may influence risk for subsequent development of breast cancer later in their lives. Several studies found a slightly increased risk of breast cancer for women whose mothers were older [4-11], but not all [12-25]. Similarly, there have been conflicting results with regard to breast cancer risk according to paternal age [4,5,8,9,13,16,18,20,24,25]. Recently, Innes et al [7] suggested that there was a positive trend in risk of breast cancer with increasing paternal age in young women.
These inconsistent results from previous studies might be due to several factors; small numbers of breast cancer cases, subject selection (e.g., age at diagnosis, different ethnicity), improper adjustment of covariates and paternal and maternal ages mutually, other study design issue (no information on known risk factors).
There is a trend toward higher paternal and maternal ages predominantly in Korea as well as in developed countries [26]. To our knowledge no studies published to date have specifically addressed the association between paternal and maternal ages and breast cancer risk in Asian women. We evaluate the independent effect of both paternal and maternal ages at birth on the risk of breast cancer of their daughters in a large case-control study in Korean.
Methods
The cases consisted of a consecutive series of breast cancer patients admitted to three teaching hospitals located in Seoul, Korea (SNUH, Borame, and Asan) between 1995 and 2003. The control subjects consisted of non-cancer patients admitted to the same hospitals as the cases in the same period and of healthy women who participated in the community health screening program provided by a teaching hospital located in Seoul (EWUMC) in 2003. The study design was approved by the Institutional Review Board of Seoul National University Hospital, and the subjects provided their informed consents prior to participation in the study.
Of 1,999 histologically confirmed incident breast cancer patients and 1,548 cancer-free controls, 1,709 breast cancer cases and 1,412 cancer-free controls were eligible after exclusion of subjects with previous history of cancer or previous history of hysterectomy and/or oophorectomy due to cervical, ovarian cancer or its precursors. The control group consisted of 577 healthy women and of 835 hospital controls with non-cancerous diseases including infection or stone of gall bladder/bile duct (26%), benign breast disease (e.g. fibroadenomas, fibrocystic disease, mastitis, etc.) (17%), acute appendicitis (14%), hemorrhoid (8%), hernia/perforation (7%), lipoma (2%), and the others (26% included liver injury, cellulitis, chronic bowl disease, benign vascular disease, ulcer, etc.). The proportion of atypical hyperplasia was estimated less than 0.2% among benign breast disease in Korean [27].
Information on demographic characteristics, current and previous residence, education, marital status, family history of breast cancer in the 1st and 2nd degree relatives, reproductive and menstrual factors, life-style habits including cigarette smoking, alcohol consumption, oral contraceptive use, and hormone replacement therapy was collected by trained interviewers using a structured questionnaire.
After exclusion of subjects with missing value of either maternal or paternal age (n = 185 in cases; n = 320 in controls), cases were frequency matched to controls by 10-year of birth group (before 1930, 1930–1939, 1940–1949, 1950–1959, 1960–1969, after 1970) and menopausal status. The final study population consisted of 1,011 cases and 1,011 controls. The distribution of matched controls was similar to that of unmatched controls (missing either paternal or maternal age) with regard to age, family history of breast cancer in 1st or 2nd degree relatives, and lifetime estrogen exposure duration. Risk factors profiles were not different between hospital and healthy community controls [28]; although community controls were older than hospital controls, the distributions of most known risk factors (e.g., education, family history of breast cancer in 1st and 2nd degree relatives) were similar. The means of paternal and maternal ages at birth were not significantly different between hospital and community controls (32.7 vs. 32.1 in paternal age, p = 0.222; 28.3 vs. 28.4 in maternal age, p = 0.807 examined by t-test). The means of paternal and maternal age at birth of patients with benign breast disease among controls were also different from neither hospital controls nor overall controls (data not shown). Although eighty-seven percent of hospital controls and 97% of community controls came from the same catchment areas (Seoul and suburbs), other demographical characteristics (e.g., age, paternal age, maternal age) were similar between hospital and community controls. Thus, the final statistical analyses were done by adjusting for all significant covariates identified from the initial analysis.
The associations between factors of interest and breast cancer risk were estimated as odds ratios (ORs) and 95% confidence intervals (CIs) by unconditional logistic regression model adjusting for family history of breast cancer in 1st or 2nd degree relatives (yes/no), and lifetime estrogen exposure duration (yrs) (presenting the number of years of exposure to menstrual cycles, which is calculated according to the age at menarche and age at interview for premenopausal women and age at menarche and age at menopause for postmenopausal women), which were identified as the significant covariates (p value < 0.05) in the initial analysis. The variables included in the logistic model were selected among age (yrs), education (under or at middle school, at high school, at or over college), family history of breast cancer in 1st or 2nd degree relatives (yes/no), lifetime estrogen exposure duration (yrs), age at full-term pregnancy or nulliparous (<25 yrs, 25–29 yrs, ≥30 yrs or nulliparity), cigarette smoking (smoked at least 400 cigarettes/lifetime, yes/no), frequency of alcohol consumption (<1/month, 1–3/month, ≥1/week) and body mass index (BMI) (<25.0 kg/m2, 25.0–30.0 kg/m2, ≥30.0 kg/m2). Only the adjusted estimates were reported in the results because the change in the β coefficient for any level of the paternal and maternal ages relative to the referent was less than 20% between unadjusted estimates and those adjusted for family history of breast cancer in 1st or 2nd degree relatives (yes/no), and lifetime estrogen exposure duration (yrs). Tests for trend in risk were conducted by treating categorical values as a continuous variable.
Paternal and maternal ages of subjects at birth were first compared using t-test and the estimates of odds ratios (ORs) with adjustment for other covariates were obtained using unconditional logistic regression analysis. We classified subjects into four groups according to paternal age at birth (<30, 30–34, 35–39, ≥40) and maternal age at birth (<25, 25–29, 30–34, ≥35). Separate analysis was conducted with and without the other paternal or maternal age mutually adjusting. To evaluate the independent effect of paternal and maternal ages, the ORs were calculated after stratified by the each paternal and maternal age group. Finally, we assessed the association of paternal and maternal ages at birth with breast cancer after stratified by menopausal status. All statistical analyses were performed using STATA version 8.0 (Stata corporation, College Station, TX).
Results
Family history of breast cancer in 1st and 2nd degree relatives (OR = 2.2, 95% CI = 1.47–3.38), lifetime estrogen exposure duration (per 10 years) (OR = 1.2, 95% CI = 1.05–1.41), ≥30 age at first full-term pregnancy or nulliparity (OR = 1.3, 95% CI = 1.04–1.76) and BMI ≥30 kg/m2 (OR = 2.0, 95% CI = 1.07–3.78) increased the risk of breast cancer significantly after adjusting for family history of breast cancer in 1st and 2nd degree relatives, and lifetime estrogen exposure duration (Table 1).
The mean of paternal age at birth was significantly different between cases and controls (33.1 yrs vs. 32.5 yrs; OR = 1.1, 95% CI = 1.00–1.28 per 10 yrs), however, the mean of maternal age was not (28.7 yrs vs. 28.3 yrs; OR = 1.1, 95% CI = 0.97–1.28 per 10 yrs) (Table 2). The risk of breast cancer showed a significantly increased as paternal age at birth increased (p for trend = 0.025). The association of paternal age with the risk of breast cancer was pronounced after controlling for maternal age; women whose fathers were aged 30–34, 35–39, and ≥40 yrs at their births, had 1.0, 1.0, and 1.6-fold increased risk of breast cancer compared with women whose fathers were aged <30 years, respectively. The risk of breast cancer increased as maternal age increased up to the intermediate, and then reduced; the risks in women whose mother were aged 25–29, 30–34, and ≥35 yrs at birth compared to women whose mothers were aged <25 years, were 1.2, 1.4, and 0.8, respectively (Table 2). The most remarkable risk of breast cancer was observed for women with higher paternal age (≥40 yrs) and intermediate maternal age (30–34 yrs) compared to women with lowest paternal (<30 yrs) and maternal ages (<25 yrs) (OR = 2.8, 95% CI = 1.74–4.62) (Table 3).
When the association was evaluated after stratified by menopausal status, the association of paternal age at birth in breast cancer was stronger in premenopausal women. Women whose fathers were aged 30–34, 35–39 and ≥40 years at their birth had 1.1, 1.2 and 1.9-fold increased risk of breast cancer compared with women whose fathers were aged <30 years, respectively (p for trend = 0.031). In contrast with paternal age, there was no significant trend between maternal age at birth and risk of breast cancer in premenopausal women (p for trend = 0.361) (Table 4).
Discussion
The results of the present study suggested that older paternal age at the time of a child's birth was associated with an increased risk of breast cancer in female offspring, which was enhanced after controlling maternal age. The effect appears to be stronger in premenopausal women. The results also suggested that there was no consistent pattern of association between maternal age at birth and risk of breast cancer although the risk increased up to women with mothers aged 30–34 years old and then reduced in women with maternal age older than 35 years old.
The mean age at first full-term pregnancy was 22.0 years old in marriage cohort before 1980 while 27.3 years old in the cohort of 2000–2003 in Korea [29], which shows that the age at first full-term pregnancy became progressively older in younger age groups. This study also found that the paternal and maternal age were significantly higher in the subjects born after 1950 than in those before 1950 for both case and control groups.
The results of previous studies of the relationships of the risk of breast cancer with paternal and maternal ages are summarized in Table 5 and these are not consistent. Four studies indicating that there was a increasing trend of the risk of breast cancer in women having older father, were consisted with the present study [5,7-9]. A number of previous studies found no statistically significant association between paternal age and the risk of breast cancer in population-based large scale studies. Most of negative studies did not adjust known risk factors for breast cancer (i. e., family history of breast cancer in 1st and 2nd degree relatives and reproductive factors). In contrast, Hodgson et al [5] showed that there was a positive association between breast cancer and paternal age in African-American only. Innes et al [7] and Le Marchand et al [9] also reported a significant linear trend in breast caner risk with paternal age only in young cancer cases.
Most previous studies showed that maternal age was not associated with breast cancer risk in white women. However, the finding of this study that the breast cancer risk increased with increasing maternal age up to the intermediate (30–34 years old), then reduced for older than 35 years old, is consistent with previous epidemiological studies [5,11-13,24,30]. Experimental evidences also support the association between maternal age and estrogen level during pregnancy [31-33]. Panagiotopoulou et al [32] showed the inverse U-shaped relationship of maternal age to estrogen levels; there was a peak of estrogen levels in the intermediate then there was a reduction in estrogen levels with maternal age. This pattern that maternal age was not linear relationship with estradiol level during pregnancy and/or generally not in a dose-dependent gradient with age was observed among other recent studies [31,33]. These results supported that subjects whose mothers were oldest at the time of birth were not in the highest risk group in this study and it may be biologically relevant because there is a perimenopausal reduction in estrogens.
With regard to menopausal status, most previous results between the risk of breast cancer and maternal age including ours are inconsistent [6,7,9,22]. Potential reasons for inconsistent association between maternal age and breast cancer risk include differences in other maternal factors (i.e., maternal diet, pregnancy complication, and maternal reproductive history), covariates adjusted, or ethnicity.
Data published so far shows that maternal and paternal ageing may affect offspring by different mechanisms. Higher concentration of estrogen in utero could create a fertile soil for cancer initiation with regard to maternal age [1]. There is some evidence that high prenatal estrogen level affects the morphology of the mammary gland (i.e., the number of ductal branching, the density of terminal end buds). An increased number of epithelial and stromal cells offer more targets for carcinogens and greater probability for genetic/epigenetic events that affect the susceptibility of the breast cancer [34]. Higher paternal age has been implicated to be responsible for increases in chromosomal aberrations and genetic disorders [35], of which the risk increases with paternal age, but maternal age has little or no effect on the risk after controlling for paternal age [3]. Spermatogomia undergo continuous cell divisions throughout the lifetime of the adult male, while oocytes undergo only one cell division between puberty and fertilization [2,3]. This difference may provide a greater opportunity for germ cells of the father to experience errors of DNA replication in regard to delayed parenthood. Furthermore, the ability to respond to mutagens with germ-cell apoptosis in order to avoid genetically altered spermatozoa decreased with paternal age [36], while oocytes have an efficient DNA repair system which is independent of maternal age [37].
Although maternal and paternal ages were correlated, we were able to ascribe the risk of breast cancer to the paternal age independent on maternal age. There are, however, several issues in the present results to consider whether the observed effects can be due to confounding or bias. First, potential selection bias because offspring with higher socioeconomic status may have detected their breast cancer earlier than those with lower socioeconomic status and because individuals with higher socioeconomic status may tend to have children later. In this study, however, the education levels of study subjects (offspring) were not associated with paternal or maternal ages and education levels did not change the β coefficient of paternal and maternal ages less than 15%, thus education levels were not included in the final model. Moreover, controlling for subjects' education levels in narrower categories (5 categories; at or under elementary school, at middle school, at high school, at college, and at graduate school) did not significantly alter the breast cancer risk of paternal and maternal age. However, we could not adjust parental education since we had no information about parental education levels. Seventy-nine percent of cases and 91% of controls came from the same catchment areas (Seoul and suburbs). However, it did not indicate that the subjects came from the different study bases since the hospitals participated in the study were university hospitals which covered the patients in the whole country according to the current health delivery system in Korea [38]. Even if the catchment areas were different between cases and controls, this difference may push the association toward the null since more controls came from urban area where subjects had higher education levels which was associated with increased breast cancer risk. Second, possible unadjusting confounders from other risk factors for breast cancer in adult life [39,40], but the risk of paternal and maternal ages on breast cancer remained same even after adjusting the established risk factors for breast cancer. However, other studies have mostly not found similar results and since small relative risks are involved and chance remains a likely explanation, the results of this moderately sized case-control study should be interpreted with extreme cautions.
Conclusion
This is the first and largest report about paternal and maternal ages and breast cancer in Asian women. These findings suggest that paternal age is associated with an increased risk of breast cancer in female offspring. However, further prospective study is needed to verify the present findings.
Competing interests
The author(s) declare that they have no competing interests.
Authors' contributions
J-YC conceived of the study, conducted data analysis, and drafted the manuscript. K-ML contributed to the design and management of data. SKP participated in data analysis and interpretation. D-YN, S-HA, and K-YY designed the study. DK conceived of and designed the study, obtaining funding and drafted the manuscript. All authors read and approved the final manuscript.
Pre-publication history
The pre-publication history for this paper can be accessed here:
Acknowledgements
This research was supported in part by a grant No. 01-PJ1-PG3-21900-0006 of the 2001 Good Health R&D Projects, Minister of Health and Welfare, South Korea..
Figures and Tables
Table 1 Selected characteristics for 1,011 breast cancer cases and 1,011 controls matched by year of birth and menopausal status
Risk factor Cases (%) Control (%) OR (95% CI)
n = 1,011 n = 1,011
Age (yrs) (mean ± SD) 47.7 ± 10.9 47.7 ± 11.7 p = 0.984*
Education
Under or at middle school 315 (31.3) 319 (31.7) 1.0
At high school 378 (37.5) 407 (40.5) 1.0 (0.80–1.24)
At or over college 315 (31.3) 280 (27.8) 1.2 (0.98–1.57)
FHBC‡
No 937 (92.7) 976 (96.5) 1.0
Yes 74 (7.3) 35 (3.5) 2.2 (1.47–3.38)
Age at FFTP‡ (yrs) or nulliparity
<25 325 (32.2) 349 (34.9) 1.0
25–29 477 (47.3) 471 (47.1) 1.1 (0.92–1.38)
≥30 or nulliparity 206 (20.4) 181 (18.1) 1.3 (1.04–1.76)
P for trend p = 0.027
LEE‡ (yrs) (mean ± SD) 29.1 ± 6.9 28.3 ± 7.4 1.2 (1.05–1.41)†
BMI (kg/m2)
<25.0 748 (74.4) 769 (76.8) 1.0
25.0–29.9 227 (22.6) 216 (21.6) 1.0 (0.84–1.29)
≥30.0 31 (3.1) 16 (1.6) 2.0 (1.07–3.78)
Smoking status
Nonsmoker 933 (92.4) 943 (93.4) 1.0
Smoker 77 (7.6) 67 (6.6) 1.2 (0.85–1.70)
Alcohol drinking
<1/month 756 (74.8) 709 (72.3) 1.0
1–3/month 178 (17.6) 211 (21.5) 0.8 (0.64–1.02)
≥1/week 77 (7.6) 61 (6.2) 1.2 (0.86–1.76)
OR of case vs. matched control adjusted for family history of breast cancer in 1st or 2nd degree relatives and lifetime estrogen exposure duration
*P value by student t test, †OR per 10 year, ‡FHBC, Family history of breast cancer in 1st or 2nd degree relatives; FFTP, first full-term pregnancy; LEE, lifetime estrogen exposure duration
Table 2 Association between paternal and maternal ages at birth and risk of breast cancer in daughters
Age at birth of the subjects (yrs) Cases (%) Controls (%) OR (95% CI)1 OR (95% CI)2
Paternal age
Mean ± SD 33.1 ± 7.5 32.5 ± 7.2 1.1 (1.00–1.28)†
<30 362 (35.8) 400 (39.6) 1.0 1.0
30–34 266 (26.3) 261 (25.8) 1.2 (0.92–1.45) 1.0 (0.79–1.34)
35–39 169 (16.7) 174 (17.2) 1.1 (0.83–1.40) 1.0 (0.71–1.39)
≥40 214 (21.2) 176 (17.4) 1.4 (1.07–1.76) 1.6 (1.04–2.32)
P for trend 0.025 0.090
Maternal age
Mean ± SD 28.7 ± 6.4 28.3 ± 6.5 1.1 (0.97–1.28)†
<25 285 (28.2) 330 (32.6) 1.0 1.0
25–29 302 (29.9) 297 (29.4) 1.2 (0.97–1.54) 1.2 (0.95–1.56)
30–34 237 (23.4) 192 (19.0) 1.5 (1.18–1.95) 1.4 (1.00–1.94)
≥35 187 (18.5) 192 (19.0) 1.1 (0.88–1.48) 0.8 (0.55–1.27)
P for trend 0.079 0.998
1OR adjusted for family history of breast cancer in 1st or 2nd degree relatives and lifetime estrogen exposure duration
2OR adjusted for maternal age or paternal age (mutual parental age, categorical value) in addition to the variables in the model 1
†OR per 10 year
Table 3 OR (95% CI) of combined effect of paternal age and maternal ages on risk of breast cancer [cases/controls]
Paternal age (yrs) Maternal age (yrs)
<25 25–29 30–34 ≥35
<30 1.0
[230/273] 1.3 (0.97–1.81)
[122/116] 1.1 (0.45–2.86)
[9/10] 1.1 (0.07–17.69)
[1/1]
30–34 1.2 (0.76–1.87)
[46/47] 1.3 (0.96–1.76)
[132/127] 1.3 (0.91–1.89)
[81/79] 1.1 (0.38–2.97)
[7/8]
35–39 1.6 (0.55–4.81)
[8/6] 1.1 (0.67–1.70)
[41/46] 1.4 (1.00–2.07)
[83/74] 0.9 (0.55–1.42)
[37/48]
≥40 0.3 (0.04–3.07)
[1/4] 1.2 (0.43–3.60)
[7/8] 2.8 (1.74–4.62)
[64/29] 1.3 (0.96–1.75)
[142/135]
OR adjusted for family history of breast cancer in 1st or 2nd degree relatives and lifetime estrogen exposure duration
Table 4 Association between parental ages at birth and risk of breast cancer in daughters after stratified by menopause status
Age at birth of the subjects (yrs) Cases (%) Controls (%) OR (95% CI)1 OR (95% CI)2
Premenopause
Paternal age
Mean ± SD 33.7 ± 7.0 32.9 ± 6.6 1.16 (0.98–1.38)†
<30 183 (31.6) 210 (36.3) 1.0 1.0
30–34 162 (28.0) 161 (27.8) 1.1 (0.85–1.54) 1.1 (0.76–1.50)
35–39 108 (18.7) 105 (18.1) 1.1 (0.82–1.60) 1.2 (0.77–1.85)
≥40 126 (21.8) 103 (17.8) 1.4 (1.00–1.93) 1.9 (1.12–3.26)
P for trend 0.062 0.031
Maternal age
Mean ± SD 29.1 ± 5.8 28.8 ± 6.1 1.06 (0.87–1.29)†
<25 133 (23.0) 159 (27.5) 1.0 1.0
25–29 194 (33.5) 187 (32.3) 1.3 (0.93–1.73) 1.2 (0.86–1.70)
30–34 147 (25.4) 120 (20.7) 1.5 (1.06–2.07) 1.2 (0.78–1.88)
≥35 105 (18.1) 113 (19.5) 1.1 (0.76–1.55) 0.7 (0.38–1.15)
P for trend 0.363 0.361
Postmenopause
Paternal age
Mean ± SD 32.2 ± 8.1 31.8 ± 8.0 1.11 (0.94–1.32)†
<30 179 (41.4) 190 (44.0) 1.0 1.0
30–34 104 (24.1) 100 (23.2) 1.2 (0.82–1.65) 1.0 (0.66–1.48)
35–39 61 (14.1) 69 (16.0) 1.0 (0.65–1.49) 0.8 (0.47–1.35)
≥40 88 (20.4) 73 (16.9) 1.4 (0.94–2.03) 1.2 (0.66–2.29)
P for trend 0.173 0.804
Maternal age
Mean ± SD 28.2 ± 7.1 27.7 ± 6.9 1.16 (0.95–1.41)†
<25 152 (35.2) 171 (39.6) 1.0 1.0
25–29 108 (25.0) 110 (25.5) 1.1 (0.80–1.60) 1.2 (0.80–2.29)
30–34 90 (20.8) 72 (16.7) 1.6 (1.05–2.31) 1.6 (0.97–2.69)
≥35 82 (19.0) 79 (18.3) 1.2 (0.82–1.78) 1.1 (0.58–2.03)
P for trend 0.112 0.415
1OR adjusted for family history of breast cancer in 1st or 2nd degree relatives and lifetime estrogen exposure duration
2OR adjusted for maternal age or paternal age (mutual parental age, categorical value) in addition to the variables in the model 1
†OR per 10 year
Table 5 Summary of results from previous studies of parental ages and breast cancer development
Study Country N (# cases) Age (yrs) Paternal age (PA) RR Maternal age (MA) RR Adjusting Restricted
Mutually* Covariates†
Cohort study
Colditz et al, 1991 US 118,309 (1,976) 3055 ≥39 vs. <20 0.9 (0.56–1.45) ≥39 vs. <20 0.9 (0.57–1.39) YES YES -
Zhang et al, 1995 US 2,662 (149) 29–62 ≥36 vs. <29 0.8 (0.5–1.4) 26–31 vs. <26 1.5 (1.0–2.4) - YES -
Holmberg et al1, 1995 US 384,769 (1,967) - ≥45 vs. <20 0.8 (0.48–1.28) ≥45 vs. <20 1.3 (0.85–1.98) - YES -
Hemminki et al, 1999 Sweden 3,800,000 (8,877) 15–53 40–49 vs. <25 1.1 (0.94–1.18) 40–49 vs. <20 1.1 (0.91–1.27) YES - PA in sporadic cases
Hilakivi-Clarke et al, 2001 Finland 3,447 (177) - - - - NS - - -
Study Country Cases/controls Age (yrs) Paternal age (PA) OR Maternal age (MA) OR Adjusting Remark
Mutually* Covariates†
Case-control study
Standfast et al, 1967 US 229/229 40–44 - - 29.1 vs. 28.2 2p < 0.05 - - -
Henderson et al, 1974 US 308/308 <64 31.5 vs. 30.7 NS 27.3 vs. 26.3 P < 0.01 - - -
Rothman et al, 1980 International 4339/12760 - - - 35–39 vs. <20 1.30 - YES -
Baron et al, 1984 UK 971/971 ≤50 - - 21–25 vs. ≤20 1.4 (0.92–2.18) - YES -
Le Marchand et al, 1988 US 153/461 <45 36–59 vs. 19–26 1.4 (0.81–2.41) 30–46 vs. 23–26 1.7 (0.99–2.78) - - Pts <33 yrs
Janerich et al, 1989 US 801/2647 - Per 10-yrs 1.2 (1.07–1.33) Each 10-yrs 1.2 (1.09–1.41) - - -
Thompson et al, 1990 US 2492/2687 20–54 - - 35–39 vs. <20 1.5 (1.10–1.93) - YES parous
Hsieh et al, 1991 International 927/2616 ≥35 - - Each 5-yrs 1.1 (1.01–1.10) - YES postmenopause
Ekbom et al, 1992 Sweden 458/1197 ≥35 - - Each 5-yrs 1.0 (0.92–1.12) - - -
Janerich et al, 1994 US 2414/9138 - ≥45 vs. <25 1.0 (0.76–1.28) ≥40 vs. <20 1.1 (0.87–1.37) - YES -
Sanderson et al, 1995 US 1147/1399 21–45, 50–64 - - ≥35 vs. <25 1.0 (0.7–1.4), 1.0 (0.7–1.5) - - -
Ekbom et al, 1997 Sweden 1068/2727 - - - Each 5-yrs 1.1 (0.99–1.14) - - -
Newcomb et al, 1997 US 1253/1121 - ≥40 vs. ≤24 0.9 (0.68–1.24) ≥40 vs. ≤20 0.9 (0.62–1.37) - YES -
Weiss et al, 1997 US 2173/1990 20–55 - - ≥35 vs. <20 0.9 (0.7–1.3) - YES -
Innes et al, 2000 US 484/2870 14–37 ≥40 vs. 25–29 1.5 (1.03–2.23) ≥35 vs. 20–24 1.9 (1.18–3.18) YES - -
Titus-Ernstoff et al, 2002 US 5629/5928 50–79 - - ≥40 vs. 25–29 1.3 (0.90–1.79) - - postmenopause
Mellemkjǽr et al, 2003 Denmark 881/3423 <40 - - ≥30 vs. <25 1.1 (0.90–1.36) - - -
Hodgson et al, 2004 US 280/236 18–74 35–56 vs. 23–27 1.5 (0.7–3.2) ≥23–27 vs. 19–22 3.5 (2.0–5.9) YES - PA in African American
* Adjusted for mutually parental age; †Adjusted for other risk factors of breast cancer risk including family history of breast cancer in 1st or 2nd degree relatives and reproductive history; 1Mortality of breast cancer, relative hazard; 2Paired t test
==== Refs
Trichopoulos D Hypothesis: does breast cancer originate in utero? Lancet 1990 335 939 940 1970028 10.1016/0140-6736(90)91000-Z
Crow JF The origins, patterns and implications of human spontaneous mutation Nat Rev Genet 2000 1 40 47 11262873 10.1038/35049558
Wilcox AJ Sandler DP Everson RB Using father's age to explore the role of germ cell mutation as a cause of human cancer Int J Epidemiol 1988 17 469 471 3403143
Henderson BE Powell D Rosario I Keys C Hanisch R Young M Casagrande J Gerkins V Pike MC An epidemiologic study of breast cancer J Natl Cancer Inst 1974 53 609 614 4369771
Hodgson ME Newman B Millikan RC Birthweight, parental age, birth order and breast cancer risk in African-American and white women: a population-based case-control study Breast Cancer Res 2004 6 R656 67 15535848 10.1186/bcr931
Hsieh CC Tzonou A Trichopoulos D Birth order and breast cancer risk Cancer Causes Control 1991 2 95 98 1873442 10.1007/BF00053127
Innes K Byers T Schymura M Birth characteristics and subsequent risk for breast cancer in very young women Am J Epidemiol 2000 152 1121 1128 11130617 10.1093/aje/152.12.1121
Janerich DT Hayden CL Thompson WD Selenskas SL Mettlin C Epidemiologic evidence of perinatal influence in the etiology of adult cancers J Clin Epidemiol 1989 42 151 157 2918324 10.1016/0895-4356(89)90088-7
Le Marchand L Kolonel LN Myers BC Mi MP Birth characteristics of premenopausal women with breast cancer Br J Cancer 1988 57 437 439 3390382
Standfast SJ Birth characteristics of women dying from breast cancer J Natl Cancer Inst 1967 39 33 42 6028287
Thompson WD Janerich DT Maternal age at birth and risk of breast cancer in daughters Epidemiology 1990 1 101 106 2073495
Baron JA Vessey M McPherson K Yeates D Maternal age and breast cancer risk J Natl Cancer Inst 1984 72 1307 1309 6587149
Colditz GA Willett WC Stampfer MJ Hennekens CH Rosner B Speizer FE Parental age at birth and risk of breast cancer in daughters: a prospective study among US women Cancer Causes Control 1991 2 31 36 1873432 10.1007/BF00052358
Ekbom A Hsieh CC Lipworth L Adami HQ Trichopoulos D Intrauterine environment and breast cancer risk in women: a population-based study J Natl Cancer Inst 1997 89 71 76 8978409 10.1093/jnci/89.1.71
Ekbom A Trichopoulos D Adami HO Hsieh CC Lan SJ Evidence of prenatal influences on breast cancer risk Lancet 1992 340 1015 1018 1357410 10.1016/0140-6736(92)93019-J
Hemminki K Kyyronen P Parental age and risk of sporadic and familial cancer in offspring: implications for germ cell mutagenesis Epidemiology 1999 10 747 751 10535790 10.1097/00001648-199911000-00013
Hilakivi-Clarke L Forsen T Eriksson JG Luoto R Tuomilehto J Osmond C Barker DJ Tallness and overweight during childhood have opposing effects on breast cancer risk Br J Cancer 2001 85 1680 1684 11742488 10.1054/bjoc.2001.2109
Janerich DT Thompson WD Mineau GP Maternal pattern of reproduction and risk of breast cancer in daughters: results from the Utah Population Database J Natl Cancer Inst 1994 86 1634 1639 7932828
Mellemkjaer L Olsen ML Sorensen HT Thulstrup AM Olsen J Olsen JH Birth weight and risk of early-onset breast cancer (Denmark) Cancer Causes Control 2003 14 61 64 12708726 10.1023/A:1022570305704
Newcomb PA Trentham-Dietz A Storer BE Parental age in relation to risk of breast cancer Cancer Epidemiol Biomarkers Prev 1997 6 151 154 9138656
Sanderson M Williams MA Malone KE Stanford JL Emanuel I White E Daling JR Perinatal factors and risk of breast cancer Epidemiology 1996 7 34 37 8664398
Titus-Ernstoff L Egan KM Newcomb PA Ding J Trentham-Dietz A Greenberg ER Baron JA Trichopoulos D Willett WC Early life factors in relation to breast cancer risk in postmenopausal women Cancer Epidemiol Biomarkers Prev 2002 11 207 210 11867509
Weiss HA Potischman NA Brinton LA Brogan D Coates RJ Gammon MD Malone KE Schoenberg JB Prenatal and perinatal risk factors for breast cancer in young women Epidemiology 1997 8 181 187 9229211
Zhang Y Cupples LA Rosenberg L Colton T Kreger BE Parental ages at birth in relation to a daughter's risk of breast cancer among female participants in the Framingham Study (United States) Cancer Causes Control 1995 6 23 29 7718731 10.1007/BF00051677
Holmberg L Ekbom A Calle E Mokdad A Byers T Parental age and breast cancer mortality Epidemiology 1995 6 425 427 7548354
Office KNS Office KNS Annual report on the marriage and divorce statistics (based on vial registration) 2004
Noh DY Kim JS Choe KJ Lee MS Ahn YO A clinical and epidemiological review of benign breast disease J Korean Surgical Society 1993 44 797 808
Choi JY Lee KM Park SK Noh DY Ahn SH Chung HW Han W Kim JS Shin SG Jang IJ Yoo KY Hirvonen A Kang D Genetic polymorphisms of SULT1A1 and SULT1E1 and the risk and survival of breast cancer Cancer Epidemiol Biomarkers Prev 2005 14 1090 1095 15894657 10.1158/1055-9965.EPI-04-0688
Korea Institute for Health & Social Affairs Affairs KIHS Survey on the national fertility, family health and welfare (2003) 2004
Rothman KJ MacMahon B Lin TM Lowe CR Mirra AP Ravnihar B Salber EJ Trichopoulos D Yuasa S Maternal age and birth rank of women with breast cancer J Natl Cancer Inst 1980 65 719 722 6932524
Kaijser M Granath F Jacobsen G Cnattingius S Ekbom A Maternal pregnancy estriol levels in relation to anamnestic and fetal anthropometric data Epidemiology 2000 11 315 319 10784250 10.1097/00001648-200005000-00015
Panagiotopoulou K Katsouyanni K Petridou E Garas Y Tzonou A Trichopoulos D Maternal age, parity, and pregnancy estrogens Cancer Causes Control 1990 1 119 124 2102281 10.1007/BF00053162
Potischman N Troisi R In-utero and early life exposures in relation to risk of breast cancer Cancer Causes Control 1999 10 561 573 10616825 10.1023/A:1008955110868
Shibata A Minn AY Perinatal sex hormones and risk of breast and prostate cancers in adulthood Epidemiol Rev 2000 22 239 248 11218375
Jung A Schuppe HC Schill WB Are children of older fathers at risk for genetic disorders? Andrologia 2003 35 191 199 12950402 10.1046/j.1439-0272.2003.00579.x
Brinkworth MH Paternal transmission of genetic damage: findings in animals and humans Int J Androl 2000 23 123 135 10844537 10.1046/j.1365-2605.2000.00222.x
Ashwood-Smith MJ Edwards RG DNA repair by oocytes Mol Hum Reprod 1996 2 46 51 9238657
Nam SM A study on choosing the medical care provider School of Public Health 2001 MS Seoul , Seoul National University
Baik I Becker PS DeVito WJ Lagiou P Ballen K Quesenberry PJ Hsieh CC Stem cells and prenatal origin of breast cancer Cancer Causes Control 2004 15 517 530 15286472 10.1023/B:CACO.0000036450.06092.ce
Joseph KS Kramer MS Review of the evidence on fetal and early childhood antecedents of adult chronic disease Epidemiol Rev 1996 18 158 174 9021310
|
16259637
|
PMC1291359
|
CC BY
|
2021-01-04 16:03:08
|
no
|
BMC Cancer. 2005 Oct 31; 5:143
|
utf-8
|
BMC Cancer
| 2,005 |
10.1186/1471-2407-5-143
|
oa_comm
|
==== Front
BMC Cell BiolBMC Cell Biology1471-2121BioMed Central London 1471-2121-6-381626289610.1186/1471-2121-6-38Research ArticleThe response of VEGF-stimulated endothelial cells to angiostatic molecules is substrate-dependent Addison Christina L [email protected]ör Jacques E [email protected] Huijun [email protected] Stephanie A [email protected] Peter J [email protected] Christie E [email protected] Centre for Cancer Therapeutics, Ottawa Health Research Institute, 501 Smyth Rd., Ottawa Ontario, K1H 8L6, Canada2 Department of Cariology, Restorative Sciences and Endodontics, School of Dentistry, University of Michigan, 1011 North University Ave., Ann Arbor Michigan 48109-1078, USA3 Oral Medicine, Pathology and Oncology, School of Dentistry, University of Michigan, 1011 North University Ave., Ann Arbor Michigan 48109-1078, USA2005 31 10 2005 6 38 38 30 12 2004 31 10 2005 Copyright © 2005 Addison et al; licensee BioMed Central Ltd.2005Addison et al; licensee BioMed Central Ltd.This is an Open Access article distributed under the terms of the Creative Commons Attribution License (), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
Background
The microenvironment surrounding cells can exert multiple effects on their biological responses. In particular the extracellular matrix surrounding cells can profoundly influence their behavior. It has been shown that the extracellular matrix composition in tumors is vastly different than that found in normal tissue with increased amounts of certain matrices such as collagen I. It has been previously demonstrated that VEGF stimulation of endothelial cells growing on type I collagen results in the induction of bcl-2 expression and enhanced endothelial cell survival. We sought to investigate whether this increased endothelial cell survival resulted in the failure of angiostatic molecules to inhibit angiogenesis.
Results
We now demonstrate that VEGF-induced survival on collagen I impairs the ability of three known angiostatic molecules, TSP-1, IP-10 and endostatin to inhibit endothelial cell proliferation. Apoptosis of endothelial cells, growing on collagen I, induced by TSP-1 and IP-10 was also inhibited following VEGF stimulation. In contrast, endostatin induced apoptosis in these same cells. Further analysis determined that endostatin did not decrease the expression of bcl-2 nor did it increase activation of caspase-3 in the presence of VEGF. Alternatively, it appeared that in the presence of VEGF, endostatin induced the activation of caspase-8 in endothelial cells grown on collagen I. Furthermore, only endostatin had the ability to inhibit VEGF-induced sprout formation in collagen I gels.
Conclusion
These data suggest that TSP-1, IP-10 and endostatin inhibit endothelial cells via different mechanisms and that only endostatin is effective in inhibiting angiogenic activities in the presence of collagen I. Our results suggest that the efficacy of angiostatic treatments may be impaired depending on the context of the extracellular matrix within the tumor environment and thus could impede the efficacy of angiostatic therapies.
==== Body
Background
The growth of tumors beyond 1 mm3 is dependent on the induction of angiogenesis, defined as the growth of new blood vessels from preexisting vasculature [1]. The process of angiogenesis is regulated by a number of promoters (angiogenic) and inhibitors (angiostatic), and it is the balance of expression of these opposing molecules that ultimately dictates whether or not angiogenesis proceeds. There are a number of endogenous angiostatic inhibitors that have been actively investigated including thrombospondin, interferon-inducible protein 10 (IP-10) and endostatin. Thrombospondin (TSP) was initially identified as a human platelet derived protein [2] that played a key role in platelet aggregation [3,4]. It was subsequently shown to modulate the biological responses of endothelial cells [5,6], induce apoptosis of endothelial cells via a caspase-3 dependent mechanism [7-9] and inhibit pathological angiogenesis [10-13]. IP-10 is a CXC chemokine that is secreted by a variety of different cell types in response to interferon stimulation [14]. In addition to being a T-cell chemoattractant [15,16], IP-10 has been shown to be an inhibitor of angiogenesis [17-19] and tumor growth in vivo [19-21]. Endostatin is the 20 kDa C-terminal cleavage product of collagen XVIII that has been shown to have a number of anti-angiogenic properties including inhibition of endothelial cell proliferation [22] and migration [23,24], induction of endothelial cell apoptosis [25], and inhibition of tumor growth in vivo [22,26-28]. Although these angiostatic molecules have been shown to be efficacious in pre-clinical models, their success in clinical trials has been more limited. Part of this may be due to the fact that very little is understood about the mechanisms by which these molecules exert their biological effects and how their efficacy might be altered by different tumor microenvironments.
The extracellular matrix (ECM) composition in tumors is vastly different than that found in its normal tissue counterparts. Certain ECM proteins such as collagen I, fibronectin and tenascin C are increased, while the basement membrane proteins collagen IV and laminin are decreased in neoplastic as compared to normal breast tissue [29,30]. Previous data suggested that stimulation of human dermal microvascular endothelial cells (HDMEC) grown on collagen I with vascular endothelial growth factor (VEGF) resulted in the increased survival of these cells by a mechanism involving the upregulation of bcl-2 [31]. It has also been shown that increased endothelial cell survival following overexpression of bcl-2 was associated with enhanced tumorigenesis in a xenograft model of human tumorigenesis [32]. These observations suggested that increased endothelial cell survival might be modulated by extracellular matrix components such as collagen I, and may contribute to the progression of tumor growth and metastasis as a result of enhanced angiogenic potential within these tumors. Furthermore, this increased survival may render the endothelial cells more resistant to the inhibitory effects of certain angiostatic molecules, thus limiting their efficacy in certain tumor microenvironments. As collagen I is found to be overexpressed in tumor versus normal tissue of the breast [29,30], and is itself a substrate for attachment of integrins that can profoundly influence endothelial function [33-37], we therefore examined the ability of three known angiostatic molecules, TSP-1, IP-10 and endostatin to inhibit endothelial cell proliferation, induce endothelial cell apoptosis on both plastic and collagen I coated surfaces in vitro and to inhibit VEGF-induced endothelial cell tube formation in collagen I gels in vitro. We found that although TSP-1 and IP-10 could inhibit endothelial cell proliferation and induce apoptosis when HDMEC were cultured on plastic tissue culture dishes, their ability to inhibit these processes was impaired following culture of HDMEC on collagen I. We observed similar results when cells were exposed to endostatin in that it inhibited proliferation of HDMEC cultured on plastic, but failed to inhibit proliferation of HDMEC grown on collagen I. In contrast to the results observed with TSP-1 and IP-10, we found that endostatin retained the ability to induce apoptosis of HDMEC even in the presence of collagen I. Furthermore, only endostatin could inhibit the formation of VEGF-induced sprouts on collagen I gels in vitro, while TSP-1 and IP-10 remained ineffective in this model system. These data suggest that the context of the tumor matrix microenvironment may modulate the inhibitory activity of angiostatic molecules and this may have significant impact on the ability of these molecules to function as anti-angiogenic therapeutics clinically in certain tumor types.
Results
Growth on collagen I impairs the ability of TSP-1, IP-10 and endostatin to inhibit endothelial cell proliferation
There is a significant body of literature demonstrating that the matrix microenvironment within tumors can be vastly different than that found in the normal tissue counterparts [29,30]. Furthermore, there is increasing evidence that "tumor-associated" ECM proteins may enhance the proliferation or survival of endothelial cells. Thus, tumor microenvironments may create a pro-angiogenic surrounding that impairs the ability of angiostatic molecules to inhibit endothelial cell growth and survival. We sought to determine if the angiostatic molecules TSP-1, IP-10 and endostatin could mediate the same inhibitory effects in endothelial cells when cells were cultured on plastic or on collagen I coated dishes. Collagen I was chosen as a "tumor-associated" ECM protein as its overexpression has been documented in a number of tumor types, and previous data suggested that growth of HDMEC on collagen I followed by stimulation with the angiogenic factor VEGF led to an increase in bcl-2 expression and enhanced survival of endothelial cells [31]. As these results suggested an increase in the survival potential of endothelial cells when cultured on collagen I, we initially investigated whether various angiostatic molecules could inhibit the proliferation of endothelial cells when cultured on tissue culture treated plastic dishes or collagen I coated dishes. Cells were starved overnight and then stimulated with 50 ng/ml VEGF (determined to be the optimal stimulatory dose of VEGF, data not shown) in combination with various doses of each inhibitor, and 72 h post-stimulation, cells were collected by trypsinization or collagenase treatment and the total number of cells per well was counted using a Coulter Counter. We observed a dose-dependent decrease in VEGF-induced HDMEC proliferation following addition of TSP-1 (Figure 1A), IP-10 (Figure 1B) and endostatin (Figure 1C) when cells were cultured on plastic. In contrast, all three angiostatic molecules failed to inhibit VEGF-induced proliferation when HDMEC were cultured on collagen I as a substrate (Figure 1). These results suggested that the ability of the various angiostatic molecules to inhibit VEGF-induced endothelial cell proliferation was impaired by growth of these cells on the matrix collagen I. As further confirmation, we examined the ability of the inhibitors to function following growth of HDMEC on the normal basement membrane ECM laminin, or on the tumor-associated ECM tenascin C. In these cases we saw statistically significant inhibition of cell proliferation by all three inhibitors when HDMEC were cultured on laminin while HDMEC were resistant to the inhibitory effects of the inhibitors following culture on tenascin C (Figure 1). These results suggest that ECM proteins associated with tumor growth, such as collagen I and tenascin C, enhance endothelial cell survival and impart a resistance to anti-angiostatic molecules.
Figure 1 Growth on tumor-associated matrices impairs the ability of TSP-1, IP-10, and endostatin to inhibit endothelial cell proliferation. HDMEC were seeded onto plastic or dishes coated with either collagen I, laminin or tenascin C, starved and then stimulated with either 50 ng/ml VEGF alone or in combination with 10, 50 or 250 ng/ml TSP-1 (panel A), 16, 80 or 400 ng/ml IP-10 (panel B) or 20, 100 or 500 ng/ml recombinant endostatin (panel C) for 72 h. Adherent cells were harvested by trypsinization or collagenase treatment (in the case of collagen I coated dishes) and direct cell counts were determined by counting using a Coulter Counter. All bars represent the mean and standard error of triplicate dishes from three independent experiments (n = 9). Within each experiment all samples were normalized to VEGF-induced proliferation and statistical significance (* = p < 0.05, ** = p < 0.001) was determined using unpaired t-tests as compared to VEGF.
We examined whether or not the apparent inhibition of proliferation as measured by direct cell counts was as a result of inhibition of cellular proliferation and replication or as a result of increased apoptosis of cells. To address this issue, we directly measured DNA replication by BrdU-incorporation assays. Cells were plated in 96-well microtiter plates, starved overnight in MCDB 131 supplemented with 1% fetal bovine serum (FBS), and then stimulated with VEGF either alone or in combination with various doses of TSP-1, IP-10 or endostatin. BrdU labeling mixture was then added 48 h post-stimulation and cells were incubated overnight. The amount of incorporated BrdU was then determined using the cell proliferation BrdU ELISA according to the manufacturer's directions. As was seen previously using direct cell counts, we saw statistically significant dose-dependent decreases in VEGF-induced endothelial cell proliferation as measured by BrdU incorporation following treatment with TSP-1 (Figure 2A) and IP-10 (Figure 2B). We also observed a trend in the reduction of BrdU incorporation following treatment with endostatin (Figure 2C), however this did not reach statistical significance. As seen previously, following growth of endothelial cells on collagen I, all three molecules were unable to inhibit VEGF-induced proliferation of HDMEC (Figure 2). These data suggest that one of the main mechanisms of inhibition of angiogenesis by TSP-1 and IP-10 is inhibition of endothelial cell proliferation, and that this mechanism is impaired by growth of cells on collagen I.
Figure 2 Growth on collagen I impairs HDMEC proliferation as measured by BrdU incorporation. HDMEC were seeded onto plastic or dishes coated with collagen I, starved and then stimulated with 50 ng/ml VEGF either alone or in combination with 10, 50 or 250 ng/ml TSP-1 (panel A), 16, 80 or 400 ng/ml IP-10 (panel B) or 100, 500 or 2500 ng/ml recombinant endostatin (panel C) for 48 h. BrdU labeling mixture was then added and the cells were incubated overnight at which time incorporated BrdU was quantitated using the BrdU Cell Proliferation ELISA (Roche). All bars represent the mean and standard error of triplicate samples. All samples were normalized to VEGF-induced BrdU incorporation and statistical significance (* = p < 0.05) was determined using unpaired t-tests as compared to VEGF.
Collagen I impairs endothelial cell apoptosis induced by TSP-1 and IP-10, but not by endostatin
Many angiostatic molecules have been shown to induce apoptosis of endothelial cells and this has been proposed as one of the main mechanisms by which they inhibit angiogenesis [7-9]. Since we observed that growth of HDMEC on the tumor-associated matrices collagen I and tenascin C impaired the ability of the anti-angiogenic molecules to inhibit VEGF-induced proliferation, we next examined whether growth on collagen I influenced the ability of the angiostatic molecules to induce apoptosis of endothelial cells, by examining the proportion of cells in sub-G1 following staining with propidium iodide as a measure of apoptotic cells. For both HDMEC grown on plastic or collagen I, we observed a statistically significant reduction in the number of apoptotic cells following treatment with VEGF (Figure 3), thus demonstrating its ability to enhance survival of endothelial cells as has been previously documented. It should be noted that unstimulated HDMEC have a high level of apoptosis following 62 h in culture without media changes. This is a result of serum deprivation that is overcome by the addition of VEGF to the culture media in the same time frame. Similar to our results with HDMEC proliferation, we found that TSP-1 could partially overcome the protective effects of VEGF and induced a statistically significant induction of apoptosis when cells were cultured on plastic (Figure 3A). In contrast, when HDMEC were cultured on collagen I and treated with VEGF in combination with TSP-1, we observed only a small but statistically insignificant increase in apoptosis (Figure 3A) suggesting that TSP-1 was ineffective at inducing apoptosis of endothelial cells in the presence of collagen I. Similar results were observed following treatment of HDMEC with IP-10, where we observed statistically significant induction of apoptosis using 80 ng/ml of IP-10 when cells were cultured on plastic and stimulated with VEGF, and the failure of IP-10 to induce significant apoptosis when cells were cultured on collagen I in the presence of VEGF (Figure 3B). Surprisingly, we observed that endostatin significantly induced apoptosis of VEGF-stimulated HDMEC regardless of whether they were cultured on plastic or collagen I as a substrate (Figure 3C). These data suggest that even though all three of the inhibitors tested in these experiments have the capacity to induce endothelial cell apoptosis, they likely do so through different mechanisms that may or may not be affected by the substrate upon which the cells are cultured. Taken together with our data regarding BrdU incorporation, these data suggest that endostatin has greater effects on apoptosis of endothelial cells as opposed to their proliferation.
Figure 3 Growth on collagen I impairs the ability of TSP-1 and IP-10, but not endostatin to induce apoptosis of HDMEC. HDMEC were seeded onto plastic or collagen I coated tissue culture dishes, monolayers were washed to remove nonadherent cells, and subsequently adherent cells were stimulated with either 50 ng/ml VEGF alone or in combination with 10, 50 or 250 ng/ml TSP-1 (panel A), 16, 80 or 400 ng/ml IP-10 (panel B) or 100, 500 or 2500 ng/ml recombinant endostatin (panel C) for 60 h. Non-adherent and adherent cells were then collected, washed twice with PBS, and permeabilized in 70% ethanol for a minimum of 24 h prior to staining with propidium iodide. The percentage of apoptotic cells was determined following identification of the subG1 population of cells using FACS analysis. The adherent and nonadherent populations of unstimulated cells were included as a control for the basal level of apoptosis of HDMEC (unstimulated-all) and an adherent-only population of unstimulated cells was included to demonstrate the surviving fraction of untreated HDMEC (unstim-adherent). Cells treated with 50 nM of actinomycin D were included as a positive control for apoptosis. All samples were normalized to the levels of apoptosis observed following stimulation with 50 ng/ml VEGF alone. Bars are representative of the mean and standard error of duplicate dishes from three independent experiments (n = 6). Statistical significance was determined following analysis using unpaired t-tests as compared to VEGF (* = p < 0.05).
The pro-apoptotic activity of endostatin is independent of bcl-2 expression
To further elucidate the mechanism by which the angiostatic molecules induced apoptosis, and whether these mechanisms were affected by the presence of collagen I, we examined the ability of angiostatic molecules to modulate the levels of the anti-apoptotic protein bcl-2 which we had previously found to be upregulated in HDMEC on collagen I following VEGF treatment [31], and the activation of caspase-3 as one of the main mediators of apoptosis in endothelial cells. HDMEC cultured on either plastic or collagen I substrates were stimulated with increasing concentrations of TSP-1, IP-10 or endostatin, alone or in combination with VEGF and total protein lysates were generated. Initially a time course was performed and it was determined that changes in the expression of the proteins of interest were maximal at 48 h post-stimulation (data not shown). As seen previously, we observed that VEGF stimulation of HDMEC on collagen I induced an increase in the level of bcl-2 protein as compared to unstimulated controls (Figure 4, right hand panels, lanes 1 and 2). A similar increase in bcl-2 levels was not observed following VEGF-stimulation of HDMEC cultured on plastic dishes (Figure 4, left hand panels, lanes 1 and 2). In fact, when total protein loading is taken into consideration, it appeared that the levels of bcl-2 in HDMEC grown on plastic dishes did not change following treatment with any of the angiostatic molecules either alone or in combination with VEGF (Figure 4, left hand panels). In contrast to the results observed following culture on plastic, we observed a dose-dependent decrease in the expression of bcl-2 following treatment with TSP-1 alone (Figure 4A, right hand panel, lanes 3, 5 and 7 as compared to lane 1) or IP-10 alone (Figure 4B, right hand panel, lanes 3, 5 and 7 as compared to lane 1) as compared to unstimulated HDMEC grown on collagen I. This decrease in bcl-2 expression in HDMEC grown on collagen I was also apparent for TSP-1 (Figure 4A right hand panel, lanes 4, 6 and 8 as compared to lane 2) and IP-10 (Figure 4B right hand panel, lanes 4, 6 and 8 as compared to lane 2) treatments in combination with VEGF, but only when the highest doses of angiostatic molecules were used. The observed decrease in bcl-2 expression could account for the small increases in endothelial cell apoptosis noted in HDMEC grown on collagen I following stimulation with TSP-1 or IP-10 (Figure 3A,B). With respect to stimulation of HDMEC on collagen I with endostatin, there was a slight decrease in bcl-2 expression following treatment with 100 ng/ml endostatin in combination with VEGF (Figure 4C, right hand panel, lane 4 as compared to lane 2) when HDMEC were grown on collagen I, however, no dose-dependent decrease in bcl-2 expression was observed on either plastic (Figure 4C, left hand panel) or collagen I (Figure 4C, right hand panel) in response to endostatin treatment. Thus, it is unlikely that changes in bcl-2 expression are the cause of the increase in apoptosis of HDMEC observed in response to endostatin treatment on either of these substrates at the higher doses of inhibitor. We also probed for bax expression levels in these same samples and observed that there was no change in bax protein regardless of the treatment used or upon which matrix the HDMEC were cultured (data not shown). Thus it would appear that VEGF induces increased survival of HDMEC on collagen I in part as a result of its ability to upregulate bcl-2. Moreover, the slight decreases in bcl-2 expression observed following treatment with TSP-1 or IP-10 in combination with VEGF are not sufficient to overcome this increased survival and fail to induce significant endothelial cell apoptosis when HDMEC are cultured on collagen I. In contrast, the ability of endostatin to overcome the VEGF-induced HDMEC survival on collagen I appeared to be independent of the levels of bcl-2 expression.
Figure 4 HDMEC treated with TSP-1 and IP-10, but not endostatin show decreases in bcl-2 expression that is inhibited by concomitant VEGF stimulation. HDMEC were seeded on either plastic or collagen I coated tissue culture dishes, and following removal of non-adherent cells by washing, were stimulated with either 50 ng/ml VEGF alone or in combination with 10, 50 or 250 ng/ml TSP-1 (panel A), 16, 80 or 400 ng/ml IP-10 (panel B), or 100, 500 or 2500 ng/ml recombinant endostatin (panel C) for 48 h. Cells were collected by trypsinization or collagenase treatment and total protein lysates were generated. Western blots for bcl-2 were performed as described in the methods section. Blots were stripped and reprobed for tubulin as a loading control.
Endostatin-induced endothelial cell apoptosis is independent of caspase-3 activation
The angiostatic molecules used in these studies demonstrated some capacity to induce apoptosis when HDMEC were cultured on plastic; however, only endostatin appeared able to induce apoptosis of HDMEC cultured on collagen I. To determine potential mechanisms of induction of apoptosis, we examined whether or not the angiostatic molecules in question led to the activation of caspase-3 when cells were cultured on either plastic or collagen I as a substrate. At high doses of TSP-1, we saw a decrease in procaspase-3, indicating cleavage of this protein into active caspase-3, following treatment of HDMEC cultured on plastic with 50 ng/ml TSP-1 as compared to unstimulated controls (Figure 5A, left hand panel, lane 5 as compared to lane 1). For all other treatment conditions of HDMEC cultured on plastic there did not appear to be cleavage or activation of procaspase-3, hence no reduction in the amount of procaspase-3 detected by western blot (Figure 5A, left hand panel). When HDMEC cultured on plastic were treated with various doses of IP-10 (Figure 5B, left hand panel) or endostatin (Figure 5C, left hand panel) alone or in combination with VEGF, we observed essentially no change in the levels of procaspase-3 suggesting that there is no active caspase-3 produced following HDMEC stimulation with these inhibitors. We observed no change in procaspase-3 levels when HDMEC were cultured on collagen I and stimulated with TSP-1 alone as compared to unstimulated controls (Figure 5A, right hand panel, lanes 3, 5 and 7 as compared to lane 1). We did however observe decreases in procaspase-3 with increasing concentrations of TSP-1 when HDMEC grown on collagen I were treated with TSP-1 in combination with VEGF, indicating that at higher doses of TSP-1, activation of caspase-3 following procaspase-3 cleavage occurs even in the presence of collagen I (Figure 5A, right hand panel, lanes 4, 6 and 8 as compared to lane 2). It would appear that under certain conditions, VEGF prevents the cleavage of procaspase-3 into active caspase-3 to some extent, however we did not consistently observe this phenomenon. It should also be noted that unstimulated HDMEC on plastic or collagen I have a fairly high apoptotic rate with approximately 40–50% of unstimulated cells being apoptotic after 62 h of culture (data not shown). Therefore, the baseline level of procapase-3 seen in the unstimulated condition inherently reflects a certain level of caspase-3 activation. Even though we detected a decrease in procaspase-3 indicating that caspase-3 had been activated and proteolytically processed following treatment of HDMEC cultured on collagen I with TSP-1 and VEGF as compared to treatment with VEGF alone (Figure 5A, right hand panel), we surprisingly did not observe significant increases in the apoptosis of TSP-1 and VEGF treated cells on collagen I. It is possible that the degree of procaspase-3 cleavage was insufficient to induce significant apoptosis in this system or that growth on collagen I induces a block in the apoptotic pathway downstream of caspase-3. Alternatively, it is possible that measurement of the reduction in procaspase-3 levels is not indicative of the generation of active caspase-3. In contrast to TSP-1, we saw very little changes in the levels of procaspase-3 following stimulation of HDMEC on collagen I with either IP-10 (Figure 5B, right hand panel) or endostatin (Figure 5C, right hand panel) alone or in combination with VEGF.
Figure 5 HDMEC treated with TSP-1 and IP-10, but not endostatin show slight decreases in procaspase-3 expression. HDMEC were seeded on either plastic or collagen I coated tissue culture dishes, and following removal of non-adherent cells by washing, were stimulated with either 50 ng/ml VEGF alone or in combination with 10, 50 or 250 ng/ml TSP-1 (panel A), 16, 80 or 400 ng/ml IP-10 (panel B), or 100, 500 or 2500 ng/ml recombinant endostatin (panel C) for 48 h. Cells were collected by trypsinization or collagenase treatment and total protein lysates were generated. Western blots for procaspase-3 were performed as described in materials and methods. Blots were stripped and reprobed for tubulin as a loading control.
As decreases in procaspase-3 levels may not necessarily be reflective of caspase-3 activation, we performed additional experiments to detect active caspase-3 using a fluorometric assay based on the cleavage of the caspase-3 substrate Ac-DEVD-AMC. As with previous experiments, cells were cultured on plastic or collagen I and stimulated with the indicated doses of TSP-1, IP-10 or endostatin in the presence of 50 ng/ml VEGF. Although none of the conditions reached statistical significance, we observed a trend for increased caspase-3 activity in HDMEC treated with TSP-1 (Figure 6A), IP-10 (Figure 6B), or endostatin (Figure 6C) in combination with VEGF following culture on plastic. We did not observe these same increases in caspase-3 activity following treatment of HDMEC with TSP-1, IP-10 or endostatin in combination with VEGF when cells were cultured on collagen I (Fig. 6). These results support the results obtained by western blot analysis and suggest that when HDMEC are cultured on collagen I activation of caspase-3 is inhibited and cells do not undergo apoptosis in response to inhibitors as a result.
Figure 6 Activation of caspase-3 is abrogated by VEGF stimulation in HDMEC cultured on collagen I. HDMEC were seeded on either plastic or collagen I coated tissue culture dishes, and following removal of non-adherent cells by washing, were stimulated with either 50 ng/ml VEGF alone or in combination with 10, 50 or 250 ng/ml TSP-1 (panel A), 16, 80 or 400 ng/ml IP-10 (panel B), or 100, 500 or 2500 ng/ml recombinant endostatin (panel C) for 48 h. Cells were collected following collagenase treatment and counted. Cell pellets were lysed in appropriate volumes of lysis buffer to generate extracts containing equal number of cells per unit volume. A fluorometric based assay for caspase-3 activity was performed using Ac-DEVD-AMC as a substrate as described in the methods section. Bars are representative of the mean and standard error for triplicate samples. The experiment was performed three independent times with similar results. Statistical significance was determined using unpaired t-tests as compared to unstimulated samples or to VEGF stimulated samples for TSP-1, IP-10 or endostatin alone or in combination with VEGF respectively (* = p < 0.05).
Since we observed the induction of apoptosis of HDMEC following treatment by endostatin when cells were cultured on collagen I, we were surprised to find that caspase-3 did not appear to be activated by endostatin. As both detection of procaspase forms and the fluorometric activity assays are known to be variable, we performed western blots using an antibody to the active form of caspase-3, which readily detects the 17 and 12 kDa cleavage products, in an attempt to confirm our previous findings. We found that when cultured on plastic, a slight increase in the active form of caspase-3 could be observed following treatment with 100 ng/ml of endostatin alone as compared to unstimulated cells (Fig. 7, top panel, lane 3 as compared to lane 1). However, in the presence of VEGF, there was no detectable activation of caspase-3 at any of the doses of endostatin tested (Fig. 7, top panel, lanes 4, 6 and 8 as compared to lane 2). Similarly, although some active caspase-3 could be detected following endostatin treatment of HDMEC cultured on collagen I, this activity was not increased as compared to that observed in the unstimulated HDMEC (Fig. 7, bottom panel, lanes 3, 5 and 7 as compared to lane 1). In support of our previous data, we observed no increases in the levels of the active forms of caspase-3 following stimulation of HDMEC on collagen I with VEGF in combination with endostatin at any of the doses tested (Fig. 7, bottom panel, lanes 4, 6 and 8 as compared to lane 2). The lack of active caspase-3 cleavage products following treatment with VEGF also supports the contention that VEGF can increase the survival potential of endothelial cells by reducing their apoptotic rate. Taken together, our data suggests that endostatin is the sole angiostatic molecule tested in this system that retains the ability to induce apoptosis of endothelial cells following growth on collagen I. Surprisingly however, the endostatin-induced apoptosis appears to occur via a caspase-3 independent mechanism that is insensitive to signals mediated through attachment of cells to collagen I.
Figure 7 Endostatin does not induce cleavage of caspase-3 into its active forms. HDMEC were seeded on either plastic or collagen I coated tissue culture dishes, and following removal of non-adherent cells by washing, were stimulated with either 50 ng/ml VEGF alone or in combination with 100, 500 or 2500 ng/ml recombinant endostatin for 48 h. Cells were collected by trypsinization or collagenase treatment and total protein lysates were generated. Western blots for active caspase-3 were performed as described in the methods section.
Endostatin induces the activation of caspase-8 in HDMEC grown on collagen I
It has been demonstrated that apoptosis of cells may occur through activation of "death receptors" that primarily mediate their effects via activation of caspase-8 (recently reviewed by [38]). As caspase-8 can also be activated by other mechanisms independent of death receptor activation [39,40], we sought to determine whether the mechanism of endostatin-induced apoptosis of endothelial cells was caspase-8 dependent. We thus examined the expression of caspase-8 following treatment of cells with various doses of endostatin alone or in combination with 50 ng/ml VEGF. We could detect a reduction in procaspase-8, suggesting cleavage into its active forms in HDMEC treated with 2500 ng/ml of endostatin alone or in combination with 50 ng/ml VEGF (Fig. 8A, lane 7 as compared to lane 1, and lane 8 as compared to lane 2). Endothelial cells that were not allowed to adhere to substratum and are known to undergo apoptosis via a fas-mediated mechanism in this situation were used as a positive control for procaspase-8 cleavage and activation (Fig. 8A, lane 9). We further confirmed these results following the detection of active caspase-8 using a fluorometric assay based on cleavage of the substrate Ac-IETD-AMC. HDMEC were stimulated with the various concentrations of endostatin alone or in combination with VEGF as indicated. As a positive control, an activating antibody to fas receptor was added to cultured cell supernatants to induce cleavage of caspase-8 (Fig. 8B, closed diamonds). To demonstrate the specificity of the activity in the assay, cells were also treated with anti-fas antibody in the presence of a specific caspase-8 inhibitor (Fig. 8B, open diamonds). Active caspase-8 could only be detected in HDMEC treated with 2500 ng/ml of endostatin alone or in combination with 50 ng/ml VEGF (Fig. 8B, open and closed upright triangles respectively). We did not detect active caspase-8 following stimulation with the lower doses of endostatin either by western blot or activity assays. These data suggest that endostatin mediates endothelial cell apoptosis in part by the activation of caspase-8, at least at higher doses. It is possible that endostatin induces apoptosis at the lower doses tested by activating caspase-8 at a level that is below our threshold of detection, or that at lower doses endostatin induces apoptosis via a caspase-8 independent pathway.
Figure 8 Endostatin induces the activity of caspase-8. A) HDMEC were seeded on tissue culture dishes, and following removal of non-adherent cells by washing, were stimulated with either 50 ng/ml VEGF alone or in combination with 100, 500 or 2500 ng/ml recombinant endostatin for 48 h. Cells were collected and total protein lysates were generated. Western blots for procaspase-8 were performed as described in materials and methods. Blots were stripped and reprobed for tubulin as a loading control. B) HDMEC were seeded on tissue culture dishes, and following removal of non-adherent cells by washing, were stimulated with either regular media (unstimulated), or media containing 100, 500, or 2500 ng/ml endostatin, in the presence or absence of 50 ng/ml VEGF, or stimulated with anti-fas antibody as a positive control for caspase-8 activation. Cells were collected by trypsinization and cell pellets stored at -80°C. Subsequently, pellets were lysed and total protein concentration was determined. Caspase-8 activity was then detected using a commercially available fluorometric detection kit (Sigma, Oakville, ON) as described in the methods section. An additional sample was included using the anti-fas treated cell lysate in the presence of specific caspase-8 inhibitors to demonstrate the specificity of the assay.
Endostatin but not TSP-1 or IP-10 inhibits VEGF-induced HDMEC sprout-formation on collagen I gels
An additional manner by which angiogenesis might be inhibited is via the prevention of the organization of endothelial cells into vessel-like structures. To determine the ability of TSP, IP-10 and endostatin to inhibit this process we utilized an in vitro tube-formation assay that uses collagen I gels as a substrate. In this system, HDMEC seeded on fibrillar collagen I gels do not spontaneously form tube-like structures, but only do so after stimulation with angiogenic factors such as VEGF. Cells will initially align with one another and begin to extend themselves into an initial "sprouting" cell; and then following 8–12 days in culture, they will form multicellular tube structures with lumen. HDMEC were seeded on collagen I gels and were then stimulated with 50 ng/ml VEGF alone or in combination with either 50 ng/ml TSP-1, 80 ng/ml IP-10 or 500 ng/ml endostatin. The number of multicellular tube-like structures was counted daily for a period of 12 days. We found that neither TSP-1 nor IP-10 could inhibit the VEGF-induced tube formation in this assay system (Fig. 9, triangles and diamonds respectively). In contrast, we observed that endostatin significantly inhibited VEGF-induced tube formation on collagen I (Fig. 9, circles). Interestingly, it would appear that endostatin did not interfere with the early stages of tube formation and cells were able to align with one another, however their ability to organize into tube structures with lumen was significantly impaired. These data again support the notion that the mechanisms by which TSP-1 and IP-10 inhibit endothelial cells and angiogenesis are different than that of endostatin. Furthermore, it would appear that the mechanisms of inhibition induced by TSP-1 and IP-10 are affected by growth of HDMEC on collagen I while endostatin retains its ability to inhibit endothelial cell survival and organization in the presence of signals induced by growth of cells on collagen I.
Figure 9 Endostatin, but not TSP-1 or IP-10, can inhibit VEGF-induced tube-formation on collagen I gels. HDMEC were seeded on collagen I gels, and following removal of non-adherent cells by washing, were stimulated with media alone (unstimulated) or media containing 50 ng/ml VEGF alone or in combination with 50 ng/ml TSP-1, 80 ng/ml IP-10 or 500 ng/ml endostatin. Cells were counted daily and media was replenished every second day. Graph is representative of the mean and standard error of the number of sprouts counted in 12 random fields of view at 100× magnification in triplicate dishes (n = 24). Statistical significance was determined at day 12 using unpaired t-tests as compared to VEGF treated cells (* = p < 0.05).
Discussion
We had previously observed that VEGF enhanced survival of HDMEC grown on collagen I [31], suggesting that factors such as elevated collagen I levels within tumor microenvironments could increase endothelial cell survival and thus could potentially induce resistance to angiostatic molecules in vivo. We decided to formally examine this hypothesis and determined that growth on the tumor-associated matrices collagen I and tenascin C impaired the ability of TSP-1 and IP-10 to inhibit endothelial cell proliferation and induce endothelial cell apoptosis (in the case of collagen I). The ability of TSP-1 to inhibit endothelial cell proliferation has previously been suggested to be dependent on its ability to bind to heparin and compete with heparin-binding growth factors for binding sites on cells [41]. As TSP-1 has been previously shown to be able to bind collagen I, it is quite possible that TSP-1 is sequestered from cells by binding collagen I, thus allowing the heparin-binding growth factors to stimulate the proliferation of HDMEC on collagen I [42]. Thrombospondin has also been previously shown to induce apoptosis in endothelial cells via activation of caspase-3 [8,9]. Surprisingly, although we did see a dose-dependent decrease in the levels of procaspase-3 following treatment of HDMEC on collagen I with TSP-1 in combination with VEGF, indicating activation of caspase-3, this was not associated with the induction of significant apoptosis as compared to stimulation with VEGF alone. Unlike other studies suggesting that TSP-1 induced endothelial cell apoptosis in a caspase-3 dependent manner [8,9], our studies were performed in the presence of exogenous VEGF, thus although some processing of caspase-3 occurs in response to TSP-1 in our experimental conditions, it is not enough to overcome the survival effects induced by simultaneous exogenous VEGF stimulation. Other studies have implicated caspase-8, fas and fasL as the main mediators of TSP-1-induced apoptosis [43]. We attempted to determine if caspase-8 was playing a role in TSP-1-induced apoptosis in our model system, however were unable to detect active caspase-8 using western blot assays. There are a number of differences between our study and those previously reported, including different endothelial cell types that are known to behave differently in experimental conditions [44-46] and different TSP-1 concentrations than those used in our studies. In fact, it was reported that when TSP-1 was delivered in combination with 50 ng/ml of VEGF, the cells became resistant to TSP-1 [43], thus confirming our observations.
We also observed that growth on collagen I prevented the angiostatic chemokine IP-10 from inhibiting endothelial cell proliferation and inducing endothelial cell apoptosis. The mechanisms of endothelial cell inhibition by IP-10 have not been extensively studied. Previous data has suggested that IP-10 can inhibit endothelial cell proliferation [47], however other reports suggest that IP-10 had no effect on endothelial cell growth, attachment or migration, but did however inhibit bFGF-induced tube formation in vitro and in vivo [18]. More recently, using IP-10 mutants that have impaired binding domains, the angiostatic activity of IP-10 was shown to be dependent on its binding to the g-protein coupled receptor CXCR3 [48]. It has been previously demonstrated that cell migration in response to SDF-1 via activation of CXCR4 is enhanced in the presence of fibronectin, suggesting that signaling through integrins can affect the downstream signal transduction pathway activated by the CXC receptors [49]. To date, there are no studies implicating integrins in CXCR3 signaling, however the CXCR family of receptors are very homologous so it is highly possible that ECM binding through the integrins can modulate the response following CXCR3 activation. As very little is known about the mechanism by which IP-10 inhibits angiogenesis, further investigation into the mechanism of inhibition of endothelial cell proliferation and induction of apoptosis by IP-10, and the effects of tumor-associated extracellular matrices is warranted.
We also examined the ability of endostatin to inhibit endothelial cell proliferation following growth on plastic, collagen I, laminin or tenascin C. As seen with the other inhibitors, endostatin was not able to inhibit endothelial cell proliferation when HDMEC were cultured on collagen I or tenascin C as a substrate. Previous data suggested that inhibition of proliferation of endothelial cells is one of the primary mechanisms of its anti-angiogenic capabilities [22,24]. Subsequently, endostatin has been shown to affect multiple facets of the angiogenic process including cell migration [23,24], survival [50], protease activity [51,52], and vessel stabilization [50,53]. Discrepancies in the results obtained using endostatin can be observed depending on the source of endostatin used in the experiments, which can be either bacterially derived [24], purified from murine hemangioendothelioma cells [22], or derived from Pichia pastoris. We have used the latter source of endostatin for our studies, and it has previously been shown to inhibit endothelial cell proliferation and migration and to induce a G1 arrest in cells [54]. Our observations support previous observations that endostatin can inhibit VEGF-induced endothelial cell proliferation to some extent; however, we have extended these findings to demonstrate that in the presence of the "tumor-associated" ECM proteins collagen I and tenascin C, this inhibition of proliferation by endostatin is dramatically impaired. The receptor that binds endostatin to inhibit angiogenic activities remains unclear despite numerous studies indicating its ability to bind cell surface molecules such as glypicans [55], heparin [56], tropomyosin-3 [57], VEGFR1 and VEGFR2 [58], and integrins αv and α5 [59]. It has been previously shown that α5β1 integrins cluster following endostatin binding [60], and since collagen I and tenascin C can both bind β1 integrins (in association with various α integrin chains) it is highly possible that binding of these matrices by integrins could modulate either the affinity for endostatin binding or integrin signal transduction cascades through an inside-out signal [61]. Furthermore, a peptide derived from endostatin has recently been shown to be a potent inhibitor of angiogenesis in a β1 integrin and heparin-dependent manner [62], suggesting that perhaps collagen I and tenascin C compete with endostatin for integrin-binding on cells and supporting our observations that endostatin was ineffective at inhibiting endothelial cell proliferation following growth on collagen I or tenascin C.
In contrast to our results with endothelial cell proliferation, we found that endostatin could induce apoptosis of endothelial cells regardless of whether they were cultured on plastic or collagen I substrates. In fact, we observed endothelial cell apoptosis at doses 100-fold lower than has previously been described [25]. The previous studies indicated that endostatin's ability to induce apoptosis was due to its ability to decrease bcl-2 expression and induce activation of caspase-3 [25]. In contrast to these results, we did not observe either a decrease in bcl-2 expression nor activation of caspase-3 in HDMEC following endostatin stimulation in either the presence or absence of VEGF. In these studies, bovine pulmonary artery endothelial cells were used, thus differences observed between the doses of endostatin required to induce apoptosis could again be a reflection of differences in the behavior of different endothelial cell types.
As endostatin has been previously shown to bind to a number of different integrin molecules [59], and recently it has been demonstrated that the inappropriate ligation of integrins can lead to what has been referred to as integrin-mediated death [63], it is possible that endostatin may induce apoptosis via alternative mechanisms such as integrin-mediated death. This form of apoptosis was shown to be dependent on caspase-8, and our observation of activation of caspase-8 observed at the higher doses of endostatin supports the notion that the apoptosis observed in our system following treatment with endostatin may be related to integrin-mediated death. Recently, it has also been shown that endostatin inhibited the migration on and attachment of endothelial cells to collagen I, but did not affect the proliferation of endothelial cells cultured on this matrix [64], which is in complete agreement with our results. Other evidence would also suggest that the main mechanism of inhibition of angiogenesis by endostatin is via its ability to inhibit cell migration through α5β1 integrin, heparan sulfate, and lipid raft-mediated interactions [65], supporting our observations regarding endostatin's ability to inhibit vessel formation in our tube formation assays, where endothelial cell migration is a prerequisite for structure formation, and our results would suggest that this effect is independent of signals from collagen I.
In conclusion, we have demonstrated that TSP, IP-10 and endostatin mediate inhibition of angiogenesis via different mechanisms that are affected to different degrees by growth of the endothelial cells on substrates such as collagen I. Our data suggests that the biological effects that various angiostatic molecules have on endothelial cells may be affected by the type of extracellular matrix upon which the cells are in contact. Indeed, the importance of cues from the extracellular matrix in vessel survival are also evident from transgenic animal systems where targeted deletion of the collagen I gene in mice led to embryonic death, due to rupture of blood vessels [66]. The effects of ECM on vessel survival and angiogenesis is extremely important in the context of inhibition of angiogenesis as an anti-tumor therapy as tumors usually have extensively remodeled matrix with differences in the composition of this matrix as compared to that found in normal tissues. Given the differences in the human tumor matrix microenvironment, and our results that "tumor-associated" matrices such as collagen I and tenascin C may enhance the resistance of endothelial cells to certain anti-angiogenic agents, it is imperative that we gain a clearer understanding of the mechanism of inhibition of angiogenesis by these molecules and how this may be affected by differences in the tumor microenvironment. These insights will potentially help to predict patient response to these inhibitors and elucidate targets of intervention that will be unaffected by differences in the tumor matrix microenvironment.
Methods
Endothelial cells and antibody reagents
Human dermal microvascular endothelial cells (HDMEC) derived from neonatal tissue were obtained from Cambrex Corporation (Walkersville, MD) and were propagated in EGM-2MV media (Cambrex Corp., Walkersville, MD). All experiments were performed using cells between passages 4 and 10. Rabbit anti-Bax and goat anti-Caspase-3 antibodies were from R&D Systems (Minneapolis, MN). Hamster anti-Bcl-2 antibody was purchased from BD Pharmingen (Mississauga, ON) and rabbit anti-active caspase-3 antibody was from Biovision (Palo Alto, CA). Mouse anti-caspase-8 antibody was purchased Cell Signaling Technologies (Beverly, MA).
Coating tissue culture dishes with collagen
Vitrogen 100 bovine dermal collagen (Cohesion Technologies, Palo Alto, CA) was used to coat tissue culture vessels in all experiments, and is a mixture of ~97% collagen I and ~3% collagen III matrices. The acidified collagen solution was kept on ice, diluted to a concentration of 1.5 mg/ml, and neutralized following addition of 10× PBS and 0.1 N NaOH to a pH of approximately 7.4. The appropriate volume of collagen solution was added to coat each vessel and the plates were rocked to ensure even distribution of collagen across the surface. Plates were then incubated at 37°C for 4 h to overnight to allow gelation to occur. The collagen surfaces were washed with Hanks Buffered Salt Solution (HBSS, Invitrogen, Carlsbad, CA), and incubated in EGM-2MV for a minimum of 2 h to equilibrate the collagen prior to addition of endothelial cells. For coating with laminin, human placenta laminin (Sigma, Oakville, ON) was diluted in PBS and added to dishes at a concentration of 1 μg/cm2. The dishes were allowed to dry overnight, uncovered, in a laminar-flow hood and cells were seeded on the surfaces the next day. For tenascin C, dishes were coated with tenascin C purified from a human tumor cell line (Chemicon International, Temecula, CA). For coating, tenascin-C was diluted to 0.1 μg/ml in PBS, and added to dishes which were then incubated overnight at 4°C. The next day, the solution was aspirated, and the dishes were blocked with 0.1% casein solution for at least 1 h. The dishes were then washed three times with 1× PBS, and seeded with cells.
Proliferation assays
HDMEC were seeded at a density of 15,000 cells per well in EGM-2MV into untreated or matrix coated 12-well plates as described above and allowed to incubate overnight. The next day, media was removed, cells were washed twice with HBSS and then starved overnight in MCDB 131 (Invitrogen, Carlsbad, CA) containing 1% fetal bovine serum (FBS). The following day cells were stimulated with varying concentrations of either TSP-1 (Calbiochem, San Diego, CA), IP-10 (Intergen, Purchase, NY), or endostatin (Calbiochem, San Diego, CA) together with 50 ng/ml VEGF (kind gift from the Biological Resources Branch of the National Cancer Institute) in MCDB 131 containing 5% FBS. Stimulation with 50 ng/ml VEGF in MCDB 131 containing 5% FBS was used as a positive control for proliferation, and stimulation with MCDB 131 containing only 1% FBS was used as a negative control for proliferation. Following 72 h of incubation, cells were collected by trypsinization or by collagenase treatment in the case of cells seeded on collagen substrate, and the number of cells per unit volume in each well was determined following counting in a Coulter counter. Each treatment was performed in triplicate and the entire experiment was performed a minimum of three independent times.
BrdU Cell Proliferation ELISA
HDMEC were seeded at a density of 2,500 cells per well in EGM-2MV into untreated or collagen I coated 96-well plates as described above and allowed to incubate overnight. The next day, media was removed, cells were washed twice with HBSS and then starved overnight in MCDB 131 containing 1% FBS. The following day cells were stimulated with varying concentrations of either TSP-1, IP-10, or endostatin together with 50 ng/ml VEGF in MCDB 131 containing 5% FBS. BrdU labeling solution was subsequently added 48 h post-stimulation with growth factors and inhibitors, and cells were incubated overnight. Incorporated BrdU was subsequently detected using the BrdU Cell Proliferation ELISA, kit according to the manufacturers directions (Roche, Laval, PQ), using a Fluoroskan Ascent Plate reader (Thermolabs, Franklin, MA) for luminescent detection.
Determination of apoptosis by FACS
HDMEC were seeded into untreated or collagen I coated dishes and allowed to adhere overnight. The following day, monolayers were washed with HBSS to remove non-adherent cells, and adherent cells were stimulated with MCDB 131 media containing 5% FBS (unstimulated), or MCDB 131 media with 5% FBS supplemented with 50 ng/ml VEGF alone, 10, 50 or 250 ng/ml TSP-1, 16, 80 or 400 ng/ml IP-10, or 100, 500 or 2500 ng/ml endostatin alone or in combination with 50 ng/ml VEGF. Cells were incubated for 60 h, and then harvested by collecting the non-adherent and adherent cell populations either by trypsinization or collagenase treatment for plastic and collagen I coated dishes respectively. Collected cells were pelleted by centrifugation at 300 × g, washed twice with PBS, and then resuspended in ice-cold 70% ethanol. Cell suspensions were incubated in 70% ethanol at -20°C for a minimum of 24 h. Following permeabilization, cells were washed two times with PBS, and then resuspended in 500 ul of propidium iodide solution (48 ug/ml propidium iodide, 40 ug/ml Rnase A in PBS). The percentage of apoptotic cells was determined following identification of the sub-G1 population of cells by flow cytometric analysis.
Detection of caspase-3 or caspase-8 activity
HDMEC were seeded into untreated or collagen I coated dishes and allowed to adhere overnight. The following day, monolayers were washed with HBSS to remove non-adherent cells, and cells were stimulated with EGM-2MV media alone, 50 ng/ml VEGF alone, or 50 ng/ml TSP-1, 80 ng/ml IP-10, or 500 ng/ml endostatin alone or in combination with 50 ng/ml VEGF in EGM-2MV. Cells were incubated for 48 h, and then harvested by collecting the non-adherent and adherent cell population by either trypsinization or collagenase treatment for plastic and collagen I coated dishes respectively. Collected cells were pelleted by centrifugation at 300 × g, washed with PBS, counted and then stored at -80°C. At time of assay, cells were lysed in the appropriate volume of lysis buffer (10 mM HEPES, 1 mM EDTA, 100 mM NaCl, 5 mM MgCl, 142.5 mM KCl, and 1 mM DTT) so that equal cell number/unit volume was obtained for each sample. Samples were incubated on ice for 45 minutes followed by a centrifugation at 13,000 rpm at 4°C for 30 minutes in a microfuge. Supernatants were removed and 20 ul of each sample was added to assay buffer (50 mM HEPES, 1 mM EDTA, 100 mM NaCl, 10% glycerol, 0.1% CHAPS, 10 mM DTT, pH 7.4) in a 96-well microtiter plate. Plates were allowed to equilibrate at 37°C for 10 minutes, and then 10 ul of a 1 mM stock of the fluorogenic caspase-3 substrate Ac-DEVD-AMC (Alexis Biochemicals, San Diego, CA) was added to each well. Plates were incubated at 37°C to allow the enzymatic reaction to proceed, and the fluorescence was measured at various times post-addition of substrate using a fluorescent plate reader set at 460 nm emission/360 nm excitation wavelengths. Purified caspase-3 was used as a positive control in all assays, and all samples were assayed in triplicate. Experiments were performed a minimum of three independent times.
For detection of caspase-8 activity, cells were seeded into dishes and stimulated as described above. Following incubation for 48 h, cells were isolated following trypsinization, washed once with PBS, and cell pellets were stored at -80°C. Subsequently, the pellets were lysed in provided caspase activity lysis buffer according to the manufacturer's directions (Sigma, Oakville, ON), the protein concentration for each lysate was determined, and caspase-8 activity was determined according to the manufacturer, using a fluorogenic caspase-8 activity kit (Sigma, Oakville, ON) that used Ac-IETD-AMC as a substrate and fluorometric detection in a fluorescent plate reader set at 460 nm emission/360 nm excitation wavelengths.
Western blot analysis
HDMEC were seeded on plastic or collagen I coated dishes as described above and allowed to adhere overnight. The next day, monolayers were washed twice with HBSS and then stimulated with varying concentrations of TSP-1, IP-10 or endostatin alone or in combination with 50 ng/ml VEGF. Total protein lysates were generated in boiling lysis buffer (1% SDS, 1.0 mM sodium ortho-vanadate, 10 mM Tris pH 7.4, 0.2 mM PMSF, 2 μg/ml aprotinin) following recovery of cells from dishes by trypsinization or collagenase treatment for growth on plastic or collagen I respectively. Aliquots containing 5–40 μg of total protein were subjected to SDS-PAGE electrophoresis followed by transfer to nitrocellulose membranes. Specific proteins were detected following incubation with primary and horse-radish peroxidase-conjugated secondary antibodies and visualization using chemiluminescent detection (Supersignal, Pierce, Rockford, MD).
Endothelial cell sprouting assays
Collagen I gels were prepared as described above. 2 × 105 HDMEC were seeded onto each 60 mm dish of collagen I and allowed to adhere overnight. The following day the dishes were washed twice with HBSS and then stimulated with EGM-2MV alone or supplemented with 50 ng/ml VEGF alone or in combination with 50 ng/ml TSP, 80 ng/ml IP-10 or 500 ng/ml endostatin. Cells were counted on day 0 prior to stimulation to ensure similar numbers of cells had been seeded on each dish, and were then counted daily for 12 days. Media containing supplements was replaced every 48 hours.
Authors' contributions
CLA contributed to the conception, design, acquisition and analysis of data in addition to writing the manuscript and performing the BrdU Cell proliferation ELISA and tube-formation assays. JEN contributed intellectually to the study and critical evaluation of the manuscript. SAL performed the caspase-3 activity assays and contributed to the critical evaluation of the manuscript. PJP contributed intellectually to the study and critical evaluation of the manuscript. HZ performed western blot analyses and CED performed western blot analyses, caspase-8 activity assays, and apoptosis experiments in addition to contributing to the critical evaluation of the manuscript.
Acknowledgements
We would like to thank Diana Malenica for technical assistance. This work has been supported in part by the Canadian Institute of Health Research (MOP-53076, C.L.A.) and by the National Institutes of Health/National Institute of Dental and Craniofacial Research (DE13161, P.J.P.). C.L.A. is the recipient of a CIHR New Investigator Award.
==== Refs
Gimbrone MAJ Leapman SB Cotran RS Folkman J Tumor dormancy in vivo by prevention of neovascularization J Exp Med 1972 136 261 276 5043412 10.1084/jem.136.2.261
Baenziger NL Brodie GN Majerus PW Isolation and properties of a thrombin-sensitive protein of human platelets J Biol Chem 1972 247 2723 2731 4260214
Dixit VM Haverstick DM O'Rourke KM Hennessy SW Grant GA Santoro SA Frazier WA A monoclonal antibody against human thrombospondin inhibits platelet aggregation Proc Natl Acad Sci U S A 1985 82 3472 3476 2582413
Leung LL Role of thrombospondin in platelet aggregation J Clin Invest 1984 74 1764 1772 6501568
O'Shea KS Dixit VM Unique distribution of the extracellular matrix component thrombospondin in the developing mouse embryo J Cell Biol 1988 107 2737 2748 3204123 10.1083/jcb.107.6.2737
Taraboletti G Roberts D Liotta LA Giavazzi R Platelet thrombospondin modulates endothelial cell adhesion, motility, and growth: a potential angiogenesis regulatory factor J Cell Biol 1990 111 765 772 1696271 10.1083/jcb.111.2.765
Guo N Krutzsch HC Inman JK Roberts DD Thrombospondin 1 and type I repeat peptides of thrombospondin 1 specifically induce apoptosis of endothelial cells Cancer Res 1997 57 1735 1742 9135017
Jimenez B Volpert OV Crawford SE Febbraio M Silverstein RL Bouck N Signals leading to apoptosis-dependent inhibition of neovascularization by thrombospondin-1 Nat Med 2000 6 41 48 10613822 10.1038/71517
Nor JE Mitra RS Sutorik MM Mooney DJ Castle VP Polverini PJ Thrombospondin-1 induces endothelial cell apoptosis and inhibits angiogenesis by activating the caspase death pathway J Vasc Res 2000 37 209 218 10859480 10.1159/000025733
Castle V Varani J Fligiel S Prochownik EV Dixit V Antisense-mediated reduction in thrombospondin reverses the malignant phenotype of a human squamous carcinoma J Clin Invest 1991 87 1883 1888 2040684
Castle VP Dixit VM Polverini PJ Thrombospondin-1 suppresses tumorigenesis and angiogenesis in serum- and anchorage-independent NIH 3T3 cells Lab Invest 1997 77 51 61 9251678
Rastinejad F Polverini PJ Bouck NP Regulation of the activity of a new inhibitor of angiogenesis by a cancer suppressor gene Cell 1989 56 345 355 2464438 10.1016/0092-8674(89)90238-9
Sheibani N Frazier WA Thrombospondin 1 expression in transformed endothelial cells restores a normal phenotype and suppresses their tumorigenesis Proc Natl Acad Sci U S A 1995 92 6788 6792 7624320
Luster AD Ravetch JV Biochemical characterization of a gamma interferon-inducible cytokine (IP-10) J Exp Med 1987 166 1084 1097 2443596 10.1084/jem.166.4.1084
Sarris AH Esgleyes-Ribot T Crow M Broxmeyer HE Karasavvas N Pugh W Grossman D Deisseroth A Duvic M Cytokine loops involving interferon-gamma and IP-10, a cytokine chemotactic for CD4+ lymphocytes: an explanation for the epidermotropism of cutaneous T-cell lymphoma? [see comments] Blood 1995 86 651 658 7605995
Taub DD Longo DL Murphy WJ Human interferon-inducible protein-10 induces mononuclear cell infiltration in mice and promotes the migration of human T lymphocytes into the peripheral tissues and human peripheral blood lymphocytes-SCID mice Blood 1996 87 1423 1431 8608232
Strieter RM Kunkel SL Arenberg DA Burdick MD Polverini PJ Interferon gamma-inducible protein 10 (IP-10), a member of the C-X-C chemokine family, is an inhibitor of angiogenesis Biochem Biophys Res Commun 1995 210 51 57 7537965 10.1006/bbrc.1995.1626
Angiolillo AL Sgadari C Taub DD Liao F Farber JM Maheshwari S Kleinman HK Reaman GH Tosato G Human interferon-inducible protein 10 is a potent inhibitor of angiogenesis in vivo J Exp Med 1995 182 155 162 7540647 10.1084/jem.182.1.155
Arenberg DA Kunkel SL Polverini PJ Morris SB Burdick MD Glass MC Taub DT Iannettoni MD Whyte RI Strieter RM Interferon-gamma-inducible protein 10 (IP-10) is an angiostatic factor that inhibits human non-small cell lung cancer (NSCLC) tumorigenesis and spontaneous metastases J Exp Med 1996 184 981 992 9064358 10.1084/jem.184.3.981
Feldman AL Friedl J Lans TE Libutti SK Lorang D Miller MS Turner EM Hewitt SM Alexander HR Retroviral gene transfer of interferon-inducible protein 10 inhibits growth of human melanoma xenografts Int J Cancer 2002 99 149 153 11948506 10.1002/ijc.10292
Regulier E Paul S Marigliano M Kintz J Poitevin Y Ledoux C Roecklin D Cauet G Calenda V Homann HE Adenovirus-mediated delivery of antiangiogenic genes as an antitumor approach Cancer Gene Ther 2001 8 45 54 11219493 10.1038/sj.cgt.7700278
O'Reilly MS Boehm T Shing Y Fukai N Vasios G Lane WS Flynn E Birkhead JR Olsen BR Folkman J Endostatin: an endogenous inhibitor of angiogenesis and tumor growth Cell 1997 88 277 285 9008168 10.1016/S0092-8674(00)81848-6
Yamaguchi N Anand-Apte B Lee M Sasaki T Fukai N Shapiro R Que I Lowik C Timpl R Olsen BR Endostatin inhibits VEGF-induced endothelial cell migration and tumor growth independently of zinc binding Embo J 1999 18 4414 4423 10449407 10.1093/emboj/18.16.4414
Taddei L Chiarugi P Brogelli L Cirri P Magnelli L Raugei G Ziche M Granger HJ Chiarugi V Ramponi G Inhibitory effect of full-length human endostatin on in vitro angiogenesis Biochem Biophys Res Commun 1999 263 340 345 10491294 10.1006/bbrc.1999.1342
Dhanabal M Ramchandran R Waterman MJ Lu H Knebelmann B Segal M Sukhatme VP Endostatin induces endothelial cell apoptosis J Biol Chem 1999 274 11721 11726 10206987 10.1074/jbc.274.17.11721
Sauter BV Martinet O Zhang WJ Mandeli J Woo SL Adenovirus-mediated gene transfer of endostatin in vivo results in high level of transgene expression and inhibition of tumor growth and metastases Proc Natl Acad Sci U S A 2000 97 4802 4807 10758166 10.1073/pnas.090065597
Perletti G Concari P Giardini R Marras E Piccinini F Folkman J Chen L Antitumor activity of endostatin against carcinogen-induced rat primary mammary tumors Cancer Res 2000 60 1793 1796 10766159
Blezinger P Wang J Gondo M Quezada A Mehrens D French M Singhal A Sullivan S Rolland A Ralston R Min W Systemic inhibition of tumor growth and tumor metastases by intramuscular administration of the endostatin gene Nat Biotechnol 1999 17 343 348 10207881 10.1038/7895
Lochter A Bissell MJ Involvement of extracellular matrix constituents in breast cancer Semin Cancer Biol 1995 6 165 173 7495985 10.1006/scbi.1995.0017
Noel A Kebers F Maquoi E Foidart JM Cell-cell and cell-matrix interactions during breast cancer progression Curr Top Pathol 1999 93 183 193 10339911
Nor JE Christensen J Mooney DJ Polverini PJ Vascular endothelial growth factor (VEGF)-mediated angiogenesis is associated with enhanced endothelial cell survival and induction of Bcl-2 expression Am J Pathol 1999 154 375 384 10027396
Nor JE Christensen J Liu J Peters M Mooney DJ Strieter RM Polverini PJ Up-Regulation of Bcl-2 in microvascular endothelial cells enhances intratumoral angiogenesis and accelerates tumor growth Cancer Res 2001 61 2183 2188 11280784
Feng X Clark RA Galanakis D Tonnesen MG Fibrin and collagen differentially regulate human dermal microvascular endothelial cell integrins: stabilization of alphav/beta3 mRNA by fibrin1 J Invest Dermatol 1999 113 913 919 10594730 10.1046/j.1523-1747.1999.00786.x
Kanda S Tomasini-Johansson B Klint P Dixelius J Rubin K Claesson-Welsh L Signaling via fibroblast growth factor receptor-1 is dependent on extracellular matrix in capillary endothelial cell differentiation Exp Cell Res 1999 248 203 213 10094827 10.1006/excr.1999.4400
Isik FF Gibran NS Jang YC Sandell L Schwartz SM Vitronectin decreases microvascular endothelial cell apoptosis J Cell Physiol 1998 175 149 155 9525473 10.1002/(SICI)1097-4652(199805)175:2<149::AID-JCP4>3.0.CO;2-O
Luscinskas FW Lawler J Integrins as dynamic regulators of vascular function Faseb J 1994 8 929 938 7522194
Ingber DE Extracellular matrix as a solid-state regulator in angiogenesis: identification of new targets for anti-cancer therapy Semin Cancer Biol 1992 3 57 63 1378310
Thorburn A Death receptor-induced cell killing Cell Signal 2004 16 139 144 14636884 10.1016/j.cellsig.2003.08.007
Bantel H Engels IH Voelter W Schulze-Osthoff K Wesselborg S Mistletoe lectin activates caspase-8/FLICE independently of death receptor signaling and enhances anticancer drug-induced apoptosis Cancer Res 1999 59 2083 2090 10232592
Inman GJ Allday MJ Apoptosis induced by TGF-beta 1 in Burkitt's lymphoma cells is caspase 8 dependent but is death receptor independent J Immunol 2000 165 2500 2510 10946276
Vogel T Guo NH Krutzsch HC Blake DA Hartman J Mendelovitz S Panet A Roberts DD Modulation of endothelial cell proliferation, adhesion, and motility by recombinant heparin-binding domain and synthetic peptides from the type I repeats of thrombospondin J Cell Biochem 1993 53 74 84 8227183 10.1002/jcb.240530109
Galvin NJ Vance PM Dixit VM Fink B Frazier WA Interaction of human thrombospondin with types I-V collagen: direct binding and electron microscopy J Cell Biol 1987 104 1413 1422 3571333 10.1083/jcb.104.5.1413
Volpert OV Zaichuk T Zhou W Reiher F Ferguson TA Stuart PM Amin M Bouck NP Inducer-stimulated Fas targets activated endothelium for destruction by anti-angiogenic thrombospondin-1 and pigment epithelium-derived factor Nat Med 2002 8 349 357 11927940 10.1038/nm0402-349
Jackson CJ Nguyen M Human microvascular endothelial cells differ from macrovascular endothelial cells in their expression of matrix metalloproteinases Int J Biochem Cell Biol 1997 29 1167 1177 9438380 10.1016/S1357-2725(97)00061-7
Geiger M Stone A Mason SN Oldham KT Guice KS Differential nitric oxide production by microvascular and macrovascular endothelial cells Am J Physiol 1997 273 L275 81 9252565
Mason JC Yarwood H Sugars K Haskard DO Human umbilical vein and dermal microvascular endothelial cells show heterogeneity in response to PKC activation Am J Physiol 1997 273 C1233 40 9357767
Luster AD Greenberg SM Leder P The IP-10 chemokine binds to a specific cell surface heparan sulfate site shared with platelet factor 4 and inhibits endothelial cell proliferation J Exp Med 1995 182 219 231 7790818 10.1084/jem.182.1.219
Yang J Richmond A The angiostatic activity of interferon-inducible protein-10/CXCL10 in human melanoma depends on binding to CXCR3 but not to glycosaminoglycan Mol Ther 2004 9 846 855 15194051 10.1016/j.ymthe.2004.01.010
Sbaa-Ketata E Vasse M Lenormand B Schneider P Soria C Vannier JP Fibronectin increases the migration induced by stromal cell-derived factor-1 alpha (SDF-1 alpha) in pre-B acute lymphoblastic leukemia cells Eur Cytokine Netw 2001 12 223 230 11399509
Dixelius J Larsson H Sasaki T Holmqvist K Lu L Engstrom A Timpl R Welsh M Claesson-Welsh L Endostatin-induced tyrosine kinase signaling through the Shb adaptor protein regulates endothelial cell apoptosis Blood 2000 95 3403 3411 10828022
Kim YM Jang JW Lee OH Yeon J Choi EY Kim KW Lee ST Kwon YG Endostatin inhibits endothelial and tumor cellular invasion by blocking the activation and catalytic activity of matrix metalloproteinase Cancer Res 2000 60 5410 5413 11034081
Wickstrom SA Veikkola T Rehn M Pihlajaniemi T Alitalo K Keski-Oja J Endostatin-induced modulation of plasminogen activation with concomitant loss of focal adhesions and actin stress fibers in cultured human endothelial cells Cancer Res 2001 61 6511 6516 11522648
Ergun S Kilic N Wurmbach JH Ebrahimnejad A Fernando M Sevinc S Kilic E Chalajour F Fiedler W Lauke H Lamszus K Hammerer P Weil J Herbst H Folkman J Endostatin inhibits angiogenesis by stabilization of newly formed endothelial tubes Angiogenesis 2001 4 193 206 11911017 10.1023/A:1014027218980
Dhanabal M Volk R Ramchandran R Simons M Sukhatme VP Cloning, expression, and in vitro activity of human endostatin Biochem Biophys Res Commun 1999 258 345 352 10329390 10.1006/bbrc.1999.0595
Karumanchi SA Jha V Ramchandran R Karihaloo A Tsiokas L Chan B Dhanabal M Hanai JI Venkataraman G Shriver Z Keiser N Kalluri R Zeng H Mukhopadhyay D Chen RL Lander AD Hagihara K Yamaguchi Y Sasisekharan R Cantley L Sukhatme VP Cell surface glypicans are low-affinity endostatin receptors Mol Cell 2001 7 811 822 11336704 10.1016/S1097-2765(01)00225-8
Sasaki T Larsson H Kreuger J Salmivirta M Claesson-Welsh L Lindahl U Hohenester E Timpl R Structural basis and potential role of heparin/heparan sulfate binding to the angiogenesis inhibitor endostatin Embo J 1999 18 6240 6248 10562536 10.1093/emboj/18.22.6240
MacDonald NJ Shivers WY Narum DL Plum SM Wingard JN Fuhrmann SR Liang H Holland-Linn J Chen DH Sim BK Endostatin binds tropomyosin. A potential modulator of the antitumor activity of endostatin J Biol Chem 2001 276 25190 25196 11335715 10.1074/jbc.M100743200
Kim YM Hwang S Pyun BJ Kim TY Lee ST Gho YS Kwon YG Endostatin blocks VEGF-mediated signaling via direct interaction with KDR/Flk-1 J Biol Chem 2002 277 27872 27879 12029087 10.1074/jbc.M202771200
Rehn M Veikkola T Kukk-Valdre E Nakamura H Ilmonen M Lombardo C Pihlajaniemi T Alitalo K Vuori K Interaction of endostatin with integrins implicated in angiogenesis Proc Natl Acad Sci U S A 2001 98 1024 1029 11158588 10.1073/pnas.031564998
Wickstrom SA Alitalo K Keski-Oja J Endostatin associates with integrin alpha5beta1 and caveolin-1, and activates Src via a tyrosyl phosphatase-dependent pathway in human endothelial cells Cancer Res 2002 62 5580 5589 12359771
Calvete JJ Structures of integrin domains and concerted conformational changes in the bidirectional signaling mechanism of alphaIIbbeta3 Exp Biol Med (Maywood) 2004 229 732 744 15337827
Wickstrom SA Alitalo K Keski-Oja J An endostatin-derived peptide interacts with integrins and regulates actin cytoskeleton and migration of endothelial cells J Biol Chem 2004 279 20178 20185 14973128 10.1074/jbc.M312921200
Stupack DG Puente XS Boutsaboualoy S Storgard CM Cheresh DA Apoptosis of adherent cells by recruitment of caspase-8 to unligated integrins J Cell Biol 2001 155 459 470 11684710 10.1083/jcb.200106070
Furumatsu T Yamaguchi N Nishida K Kawai A Kunisada T Namba M Inoue H Ninomiya Y Endostatin Inhibits Adhesion of Endothelial Cells to Collagen I via alpha(2)beta(1) Integrin, a Possible Cause of Prevention of Chondrosarcoma Growth J Biochem (Tokyo) 2002 131 619 626 11927001
Wickstrom SA Alitalo K Keski-Oja J Endostatin associates with lipid rafts and induces reorganization of the actin cytoskeleton via down-regulation of RhoA activity J Biol Chem 2003 278 37895 37901 12851410 10.1074/jbc.M303569200
Lohler J Timpl R Jaenisch R Embryonic lethal mutation in mouse collagen I gene causes rupture of blood vessels and is associated with erythropoietic and mesenchymal cell death Cell 1984 38 597 607 6467375 10.1016/0092-8674(84)90514-2
|
16262896
|
PMC1291360
|
CC BY
|
2021-01-04 16:31:29
|
no
|
BMC Cell Biol. 2005 Oct 31; 6:38
|
utf-8
|
BMC Cell Biol
| 2,005 |
10.1186/1471-2121-6-38
|
oa_comm
|
==== Front
BMC Clin PharmacolBMC Clinical Pharmacology1472-6904BioMed Central London 1472-6904-5-41625314110.1186/1472-6904-5-4Research ArticleDistribution of cytochrome P450 2C, 2E1, 3A4, and 3A5 in human colon mucosa Bergheim Ina [email protected] Christiane [email protected] Alexandr [email protected] Hohenheim University (140), Dep. Physiology of Nutrition, Stuttgart, Germany2 Department of Pharmacology and Toxicology, University of Louisville Health Sciences Center, Louisville, KY, USA2005 27 10 2005 5 4 4 3 5 2005 27 10 2005 Copyright © 2005 Bergheim et al; licensee BioMed Central Ltd.2005Bergheim et al; licensee BioMed Central Ltd.This is an Open Access article distributed under the terms of the Creative Commons Attribution License (), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
Background
Despite the fact that the alimentary tract is part of the body's first line of defense against orally ingested xenobiotica, little is known about the distribution and expression of cytochrome P450 (CYP) enzymes in human colon. Therefore, expression and protein levels of four representative CYPs (CYP2C(8), CYP2E1, CYP3A4, and CYP3A5) were determined in human colon mucosa biopsies obtained from ascending, descending and sigmoid colon.
Methods
Expression of CYP2C, CYP2E1, CYP3A4, and CYP3A5 mRNA in colon mucosa was determined by RT-PCR. Protein concentration of CYPs was determined using Western blot methods.
Results
Extensive interindividual variability was found for the expression of most of the genes. However, expression of CYP2C mRNA levels were significantly higher in the ascending colon than in the sigmoid colon. In contrast, mRNA levels of CYP2E1 and CYP3A5 were significantly lower in the ascending colon in comparison to the descending and sigmoid colon. In sigmoid colon protein levels of CYP2C8 were significantly higher by ~73% than in the descending colon. In contrast, protein concentration of CYP2E1 was significantly lower by ~81% in the sigmoid colon in comparison to the descending colon.
Conclusion
The current data suggest that the expression of CYP2C, CYP2E1, and CYP3A5 varies in different parts of the colon.
==== Body
Background
Throughout the last decades drug metabolism in the alimentary tract has received growing attention as part of the body's first line of defense against orally ingested harmful xenobiotica. Most xenobiotic compounds require an enzymatic activation to form a carcinogen or toxicant. The reactive intermediates resulting form enzymatic drug metabolism are often unstable and therefore are unlikely to be transported from the liver to other tissues to exert toxicity. Therefore, the chemical toxicity found in extrahepatic tissue often results from cellular metabolic activities in the organ. However, knowledge on the variability and regulation of expression of drug metabolizing enzymes in the human gastrointestinal tract and particularly the large intestine is poor in comparison with the "classical" drug metabolizing organs (e.g. liver).
Cytochrome P450 (CYP) is a multi-gene superfamily of heme-containing enzymes catalyzing the oxidative metabolism of many compounds [1]. CYP families 1, 2, and 3, which are the main CYP families participating in the metabolism of xenobiotics, are highly expressed within the liver, but are also expressed in extrahepatic tissues (for review see [2]). Members of the CYP families 2 and 3, and herein especially CYP2C and CYP3A4, are present in relative high concentrations in small intestinal epithelium [3-5], and it has been suggested that they facilitate a barrier function to protects the small intestine from toxic xenobiotics [4]. Furthermore, it has been shown that total CYP content increases slightly in the progression from duodenum to jejunum and subsequently decreases significantly in the ileum [3]. Despite the fact that it has been suggested that the absence of some of these microsomal enzymes in the colon may be involved in the comparably high incidence of carcinoma in this organ [4], information available on the expression of CYPs in the large intestine of humans is limited and some of the available data are contradictory. Therefore, the main purpose of the present study was to evaluate the expression pattern, protein concentration, and distribution of four representative CYPs (CYP2C, CYP2E1, CYP3A4, and CYP3A5) in ascending, descending and sigmoid human colon mucosa of virtually healthy subjects. Furthermore, protein levels were related to mRNA expression pattern.
Material and methods
Subjects
The study was approved by the Ethics Committee of the Medical Association Stuttgart, Germany. Informed consent was obtained from all subjects included in the study. During routinely performed coloscopies, two biopsy specimens (each 5–10 mg) of normal colon mucosa were obtained from either ascending (n = 10), descending (n = 7), or sigmoid (n = 24) colon of a total of 41 volunteers (age 30 – 80). None of the subjects displayed any macroscopic evidence of colonic neoplasia or other disease in colon. All patients completed a questionnaire concerning factors that may influence the expression of cytochrome P450 such as medication, smoking, and alcohol consumption (see Table 1). Neither anthropometrics nor lifestyle and drug intake differed significantly between donors of biopsy specimens of ascending, descending, and sigmoid colon. Furthermore, no significant correlation was observed between these parameters and the protein levels as well as mRNA expression patterns of CYPs investigated. Subjects and controls did not consume drugs known to interfere with or to induce CYPs investigated in this study.
Table 1 Anthropometic and life style data of subjects.
Parameter Ascending Colon Descending Colon Sigmoid Colon
n 10 7 24
Age 52 ± 9 54 ± 15 55 ± 9
Sex : Female/male 5/5 1/6 8/16
BMI 26.5 ± 5.7 27.8 ± 4.9 25.4 ± 4.0
Cigarette use
Yes/No 3/7 3/4 4/20
(number/d) 2.8 ± 6.2 8.3 ± 9.8 1.5 ± 4.6
Alcohol consumption
Yes/No 3/7 2/5 12/12
(g/d) 9.4 ± 14.6 8.2 ± 7.8 9.8 ± 12.6
All data are expressed as means ± SD. (BMI = body mass index)
Tissues and isolation of total RNA and protein
All colon mucosa specimens were immediately frozen in liquid nitrogen after excision and stored at -80°C. Both total RNA and protein was isolated using Trizol reagent (Invitrogene, Gaithersburg, MD, USA).
Electrophoresis and immunoblotting
Protein concentration was determined using a commercially available Bradford assay (BioRad, Munich, Germany). Twenty to 30 μg of total protein were separated by electrophoresis through a 9% SDS-polyacrylamid gel and was transferred onto a nitrocellulose membrane. To ensure equal loading of samples, membranes were stained with Ponceau red. In addition, some blots were probed for β-actin (Sigma, Munich, Germany) to ascertain identical protein loading of samples. Membranes were blocked in 5% non-fat milk in Tris-buffered saline-Tween 20 (TBST, 0.01% v/v Tween 20) and probed with dilutions of primary antibodies in TBS followed by an incubation with the secondary antibody. Anti-CYP2E1 and anti-CYP3A4 antibodies were generous gifts of Dr. M. Ingelman-Sundberg, Karolinska Institute, Stockholm, Sweden; anti-CYP2C8 and anti-CYP3A5 were purchased from Chemicon, Inc. (Frankfurt, Germany). The protein/antibody complex was visualized by enhanced chemiluminescence (SuperSignal® West Dura, Pierce, Bad Godesberg, Germany). Blots were photographed (Camera LAS 1000, Fuji, USA) and densitometric analysis was performed using the software AIDA (Raytest, Isotopenmessgeraete, Straubenhardt, Germany). As positive control and for semiquantification, serial dilutions of microsomes and Supersomes derived from cell lines and clones expressing human CYP2C8, CYP2E1, CYP3A4, and CYP3A5 (Gentest Corporation, Woburn, MA, USA) respectively were loaded on each gel.
Reverse transcription and PCR
The integrity and concentration of RNA was analyzed in a 1.2% agarose gel. First-strand complementary DNA was synthesized from 200 ng of total RNA using a First-Strand cDNA Synthesis Kit (Invitrogen, Gaithersburg, MD, USA). Sequences of primers are summarized Table 2. The PCR reaction consisted of 0.6 μl of cDNA, 10 × PCR buffer, 200 μM dNTPs (Boehringer, Mannheim, Germany), BSA (0.25 mg/ml), DMSO (2% v/v), 0.5 μM of specific primer and 0.5 U Taq-polymerase (Promega, Madison, WI, USA), and water to a final volume of 10 μl. For amplifications of the four cytochrome P450 cDNAs, PCR-conditions were as follows: 3 s at 94°C, 3 s at 45°C, 30 s at 72°C, for 32 cycles. Amplification of histone 3.3 was performed applying the following conditions: 3 s at 94°C, 3 s at 45°C, and 30 s at 72°C for 30 cycles. All PCR amplifications were carried out in triplicate in a Rapid Cycler (Idaho Tec., USA) within the linear range of the reaction. PCR products were separated in a 1.5% agarose gel, stained with ethidium bromide and photographed using a digital camera from Biometra (Goettingen, Germany). To ensure the success of PCR, human liver cDNA was used as a positive control.
Table 2 Primers used for RT-PCR analysis
Sense primer location Antisense primer location PCR product (bp) Reference
Histone 3.3 GCGTGCTAGCTGGATGTCTT CCACTGAACTTCTGATTCGC 150 [18]
CYP2C8-19 GCTAAAGTCCAGGAAGAGATTGA TCCTGCTGAGAAAGGCATGAAGT 332 [19]
CYP2E1 AGCACAACTCTGAGATATGG ATAGTCACTGTACTTGAACT 365 [19]
CYP3A4 CCAAGCTATGCTCTTCACCG TCAGGCTCCACTTACGGTGC 324 [20]
CYP3A5 TGTCCAGCAGAAACTGCAAA TTGAAGAAGTCCTTGCGTGTC 472 [20]
Statistical analysis
Mann-Whitney U test, Chi-square-test for crosstabulation tables, and analysis of variances (ANOVA) with the Posthoc test of Tukey was used for the determination of statistical significance as appropriate. A P value of less than 0.05 was selected as the level of significance before the study.
Results
Expression of CYP2C, CYP2E1, CYP3A4, and CYP3A5 mRNA in ascending, descending, and sigmoid colon
The presence of a band of the correct size in agarose gels was regarded as evidence of gene expression. Expression of the housekeeping gene histone 3.3 was detected in all samples. A total of 10 specimens obtained from the ascending colon, 6 specimens from the descending colon and 21 specimens obtained form the sigmoid colon were included in the analysis of mRNA expression. Three RNA samples obtained form the sigmoid colon had to be excluded as RNA concentrations were too low. Representative gels depicting RNA integrity and results of RT-PCR measurements are shown in Figure 1.
Figure 1 CYP3A5, CYP2E1, CYP3A4, and CYP2C mRNA expression in human colon mucosa obtained ascending, descending, and sigmoid colon. (A) Representative agarose gel of mRNA integrity. Lane 1 and 10 = positive control, lane 2–9 = RNA extracted from colonic mucosa biopsies. (B) Representative photomicrograph of RT-PCR products of three subjects. All measurements were carried out in triplicate. Lane 1 = Histone 3.3; Lane 2 = CYP3A5; Lane 3 = CYP2E1; Lane 4 = CYP3A4; Lane 5 = CYP2C, bp = base pairs, A = ascending colon, D = descending colon, S = sigmoid colon. (C) Quantitative analysis of mRNA expression of CYP2C, CYP2E1, CYP3A4, and CYP3A5 in human ascending, descending, and sigmoid colon. Data are normalized to histone 3.3 expression. Data are means ± SEM. (n.d. = not detectable, aP < 0.05 compared to ascending colon)
Expression of CYP2C was found to be higher in the ascending colon than in the descending and sigmoid colon. However, due to large interindividual variability differences in CYP2C expression were only significant between the ascending and the sigmoid colon. In the ascending colon expression of CYP2E1 was not detectable. In the descending and the sigmoid colon expression of CYP2E1 did not differ significantly, however expression of CYP2E1 was detectable. Expression of CYP3A4 did not differ between the three regions of colon investigated. In contrast, CYP3A5 mRNA expression was significantly higher in the descending and the sigmoid colon in comparison to the ascending colon. Specifically, in the descending colon CYP3A5 mRNA expression was ~2-fold and in the sigmoid colon ~3-fold higher than in the ascending colon. No differences in CYP3A5 expression were found between the descending and sigmoid colon.
Protein levels of CYP2C8, CYP2E1, CYP3A4, and CP3A5 in the descending and sigmoid colon
Due to a low protein content of some mucosal biopsies, Western blot analyses of CYP2E1, CYP3A4, and CYP3A5 were only performed in 6 tissue specimens obtained from the descending and 24 from sigmoid colon. Since higher protein concentrations were needed for the detection of CYP2C8, protein concentration of CYP2C8 was only determined in 17 specimens taken from sigmoid colon. Biopsies obtained from the ascending region of the colon were not included in the analysis as sufficient protein concentrations for Western blot analysis were only obtained from four samples. Figure 2 depicts representative Western blot and quantitative analysis of protein.
Figure 2 CYP2E1, CYP2C8, CYP3A4, CYP3A5 protein levels in normal human colon mucosa obtained from descending and sigmoid colon. (A) Representative Western blot of CYP2E1, CYP2C8, CYP3A4, CYP3A5. (B) Representative Western blot of β-actin performed in 5 different individuals. (C) Quantitative analysis of blots. Data are means ± SEM. (aP < 0.05 compared to descending colon)
The mean protein level of CYP2C8 was significantly lower in the descending colon when compared with the sigmoid colon. Specifically, protein levels of CYP2C8 were found to be ~73% lower in the descending colon in comparison to the sigmoid colon. In contrast, protein concentration of CYP2E1 was significantly higher by ~81% in the descending colon compared to the sigmoid colon. However, no significant differences were found when comparing protein levels of CYP3A4 and CYP3A5 between the descending and sigmoid colon.
Relation of protein levels and mRNA expression pattern of CYPs in sigmoid colon
To address the question whether mRNA profiles correlated CYP protein concentration in colonic mucosa, protein levels of CYPs of subjects with detectable mRNA expression were compared with those with undetectable mRNA expression of the respective CYP. Since the sample number of biopsies obtained from the ascending and descending colon was too small to perform a statistical analysis (ascending: n = 4; descending: n = 6), this comparison was only performed for the sigmoid colon. Results are summarized in Figure 3. Interestingly, no significant differences were found when comparing protein levels of CYP2C8, CYP2E1, CYP3A4, and CYP3A5 of subjects with detectable mRNA expression of these CYPs with those with undetectable mRNA levels. No correlation was found between CYP2E1 expression and individual alcohol consumption.
Figure 3 Relation of CYP2E1, CYP2C8, CYP3A4, CYP3A5 protein levels and mRNA expression pattern in normal human colon mucosa obtained from sigmoid colon. Comparison of protein levels of CYP2E1, CYP2C8, CYP3A4, and CYP3A5 of subjects with detectable mRNA expression and subjects with no detectable mRNA expression of the respective CYP. Data are means ± SEM.
Discussion
Abundance of CYPs differs between ascending, descending, and sigmoid colon and between individuals
It has been suggested that the expression of CYPs (e.g. the absence of expression of certain CYPs) in the colon might be an important factor in the susceptibility of this organ for cancer. Expression of members of the CYP2C-, CYP2E-, and CYP3A-family in normal mucosa along with substantial interindividual differences in the expression levels of CYPs in the colon has been reported by others [6-11]. However, some of the available data are contradictory and most studies either determined mRNA or protein levels. Furthermore, detailed information on the distribution of CYPs within the colon (e.g. distal vs. proximal colon) is lacking. For example, McKay et al. [6] detected CYP3A protein in two out of 13 morphologically normal colon mucosa specimens of patients with neoplasia in the colon. Similar results were also reported by Mercurio et al.[7] and McKinnon et al.[8] for CYP3A4 and CYP3A5 mRNA expression in colon. Using immunohistochemical methods, Yokose et al. [9] reported the presence of CYP2C(8–19) in healthy human colon mucosa also finding large interindividual differences. In contrast, Western blot analyses by de Waziers et al. [10] and Massaad et al. [11] applying conventional immunoperoxidase staining procedure failed to detect the expression of cytochromes P450 2C8-10 and 2E1 protein in human colon mucosa. In the present study, the expression and protein levels of four representative CYPs were determined in different regions of the large intestine. At the levels of mRNA expression CYP2C, CYP2E1, and CYP3A5 concentration was found to be significantly different between the three different regions investigated with CYP2C mRNA expression being significantly higher in proximal regions of the colon than in the distal (e.g. descending and sigmoid colon). In contrast, mRNA expression of CYP2E1 and CYP3A5 was significantly lower in the proximal colon (i.e. ascending) than in the distal part of the organ. However, in accordance with the findings of others, mRNA expression of all four CYPs was found to vary extensively between individuals, even though expression of histone 3.3, which was used as housekeeping gene, was detected in all samples. At the level of protein, CYP2C8 concentration was found to be lower in the descending colon when compared with the more distal sigmoid region of the colon. Contrary, CYP2E1 protein levels were higher in descending colon than in the sigmoid colon. Taken together, these data suggest that CYP expression in colon not only varies among individuals but also among different regions of the organ.
CYP protein levels and mRNA expression are not related
Studies of the mRNA and protein expression of CYP3A4 and CYP3A5 in rat duodenum and kidney, as well as the expression of CYP2C7 (corresponding to human CYP2C8) and CYP2E1 in rat colon mucosa, revealed a dissociation of mRNA expression and protein levels of CYPs. [12] Hakkak et al. [13] suggested that the expression of CYPs is not solely regulated at the transcriptional level. This is supported by results of in vitro investigations in rat hepatocytes, which indicated that protein level of CYP2E1 is regulated by posttranscriptional ligand-dependent stabilization of the enzyme [14]. Similar mechanisms have been described for rat and human CYP3A [4,15,16]. Furthermore, in vitro studies using cultured hepatocytes showed that only ~60–70% of mRNA encoding for CYP2E1 is translated [17]. Indeed, in the present study, no differences with respect to protein levels of subjects with detectable and undetectable mRNA expression of the CYPs were found.
Conclusion
In summary, interindividual variability seems to be a characteristic of CYP expression in colon as has been reported by others before [6-11]. However, in the present study we found significant differences of CYP2C(8), CYP2E1 and CYP3A5 mRNA expression and protein levels between different regions of the colon (e.g. ascending, descending, and sigmoid colon). Metabolic implications of this "zonification" remain to be determined. Nevertheless, differences found in the present study might result in alterations of detoxification of carcinogens or pro-carcinogens and therefore contribute to high susceptibility of this organ to carcinoma.
Competing interests
The author(s) declare that they have no competing interests.
Authors' contributions
IB has made substantial contributions to acquisition of data, the biochemical analysis, and the drafting of article. CB has made substantial contribution to conception and design as well as the interpretation of data. AP has been involved in the design, the drafting of the article, and revised it critically for intellectual content. All authors have given final approval of the version to be published.
Pre-publication history
The pre-publication history for this paper can be accessed here:
Acknowledgements
The antibodies against human CYP2E1 and CYP3A4 were kindly provided by Dr. M. Ingelman-Sundberg. This work was supported by a grant from the Deutsche Krebshilfe (70-1881-B01) to AP and CB. The authors would like to thank Dr. G.E. Arteel for proofreading the manuscript.
==== Refs
Sheweita SA Drug-metabolizing enzymes: mechanisms and functions Curr Drug Metab 2000 1 107 132 11465078 10.2174/1389200003339117
Ding X Kaminsky LS Human extrahepatic cytochromes P450: function in xenobiotic metabolism and tissue-selective chemical toxicity in the respiratory and gastrointestinal tracts Annu Rev Pharmacol Toxicol 2003 43 149 173 12171978 10.1146/annurev.pharmtox.43.100901.140251
Zhang QY Dunbar D Ostrowska A Zeisloft S Yang J Kaminsky LS Characterization of human small intestinal cytochromes P-450 Drug Metab Dispos 1999 27 804 809 10383924
Peters WH Kock L Nagengast FM Kremers PG Biotransformation enzymes in human intestine: critical low levels in the colon? Gut 1991 32 408 412 1902809
Obach RS Zhang QY Dunbar D Kaminsky LS Metabolic characterization of the major human small intestinal cytochrome p450s Drug Metab Dispos 2001 29 347 352 11181505
McKay JA Murray GI Weaver RJ Ewen SW Melvin WT Burke MD Xenobiotic metabolising enzyme expression in colonic neoplasia Gut 1993 34 1234 1239 8406161
Mercurio MG Shiff SJ Galbraith RA Sassa S Expression of cytochrome P450 mRNAs in the colon and the rectum in normal human subjects Biochem Biophys Res Commun 1995 210 350 355 7755610 10.1006/bbrc.1995.1668
McKinnon RA Burgess WM Gonzalez FJ McManus ME Metabolic differences in colon mucosal cells Mutat Res 1993 290 27 33 7694095
Yokose T Doy M Taniguchi T Shimada T Kakiki M Horie T Matsuzaki Y Mukai K Immunohistochemical study of cytochrome P450 2C and 3A in human non-neoplastic and neoplastic tissues Virchows Arch 1999 434 401 411 10389623 10.1007/s004280050359
De Waziers I Cugnenc PH Berger A Leroux JP Beaune PH Drug-metabolizing enzyme expression in human normal, peritumoral and tumoral colorectal tissue samples Carcinogenesis 1991 12 905 909 2029756
Massaad L de Waziers I Ridbrag V Janot F Beaune PH Morizet J Gouyette A Chabot GG Comparison of Mouse and Human Colon Tumors with Regard to Phase I and Phase II Drug-metabolizing Enzyme Systems Cancer Research 1992 52 6567 6575 1423302
Ronis MJ Huang J Longo V Tindberg N Ingelman-Sundberg M Badger TM Expression and distribution of cytochrome P450 enzymes in male rat kidney: effects of ethanol, acetone and dietary conditions Biochem Pharmacol 1998 55 123 129 9448734 10.1016/S0006-2952(97)00381-X
Hakkak R Korourian S Ronis MJ Ingelman-Sundberg M Badger TM Effects of diet and ethanol on the expression and localization of cytochromes P450 2E1 and P450 2C7 in the colon of male rats Biochem Pharmacol 1996 51 61 69 8534269 10.1016/0006-2952(95)02154-X
Eliasson E Mkrtchian S Ingelman-Sundberg M Hormone- and substrate-regulated intracellular degradation of cytochrome P450 (2E1) involving MgATP-activated rapid proteolysis in the endoplasmic reticulum membranes J Biol Chem 1992 267 15765 15769 1639811
Feierman DE Melnikov Z Zhang J The paradoxical effect of acetaminophen on CYP3A4 activity and content in transfected HepG2 cells Arch Biochem Biophys 2002 398 109 117 11811955 10.1006/abbi.2001.2677
Zangar RC Hernandez M Novak RF Posttranscriptional elevation of cytochrome P450 3A expression Biochem Biophys Res Commun 1997 231 203 205 9070249 10.1006/bbrc.1997.6054
Kocarek TA Zangar RC Novak RF Post-transcriptional regulation of rat CYP2E1 expression: role of CYP2E1 mRNA untranslated regions in control of translational efficiency and message stability Arch Biochem Biophys 2000 376 180 190 10729204 10.1006/abbi.2000.1704
Futscher BW Blake LL Gerlach JH Grogan TM Dalton WS Quantitative polymerase chain reaction analysis of mdr1 mRNA in multiple myeloma cell lines and clinical specimens Anal Biochem 1993 213 414 421 8238918 10.1006/abio.1993.1440
Hakkola J Pasanen M Purkunen R Saarikoski S Pelkonen O Maenpaa J Rane A Raunio H Expression of xenobiotic-metabolizing cytochrome P450 forms in human adult and fetal liver Biochem Pharmacol 1994 48 59 64 8043031 10.1016/0006-2952(94)90223-2
Kivisto KT Griese EU Fritz P Linder A Hakkola J Raunio H Beaune P Kroemer HK Expression of cytochrome P 450 3A enzymes in human lung: a combined RT- PCR and immunohistochemical analysis of normal tissue and lung tumours Naunyn Schmiedebergs Arch Pharmacol 1996 353 207 212 8717162
|
16253141
|
PMC1291361
|
CC BY
|
2021-01-04 16:27:54
|
no
|
BMC Clin Pharmacol. 2005 Oct 27; 5:4
|
utf-8
|
BMC Clin Pharmacol
| 2,005 |
10.1186/1472-6904-5-4
|
oa_comm
|
==== Front
BMC Clin PharmacolBMC Clinical Pharmacology1472-6904BioMed Central London 1472-6904-5-51626291010.1186/1472-6904-5-5Research ArticlePharmacovigilance program to monitor adverse reactions of recombinant streptokinase in acute myocardial infarction Betancourt Blas Y [email protected] María A [email protected]énez-López Giset [email protected] Carmen [email protected]ía-Iglesias Elizeth [email protected]ández-Bernal Francisco [email protected]ía Francisco [email protected]ález-López Tania [email protected]ón Leovaldo [email protected]ópez-Saura Pedro A [email protected] Cuban National Network of Pharmacoepidemiology [email protected] Clinical Trials Division, Center for Biological Research, Havana, Cuba2 National Center for Clinical Trials Coordination, Havana, Cuba3 National Coordinating Unit of Pharmacovigilance, Centre for the Development of Pharmacoepidemiology, Minister of Heath, Havana, Cuba4 A list of the National Network of Pharmacoepidemiology participating investigators appears in the Appendix2005 2 11 2005 5 5 5 5 3 2005 2 11 2005 Copyright © 2005 Betancourt et al; licensee BioMed Central Ltd.2005Betancourt et al; licensee BioMed Central Ltd.This is an Open Access article distributed under the terms of the Creative Commons Attribution License (), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
Background
Streptokinase (SK) is an effective fibrinolytic agent for the treatment of acute myocardial infarction (AMI). The objective of the present study was to assess the adverse drug reactions (ADRs) associated with intravenous recombinant SK in patients with AMI in routine clinical practice.
Methods
A national, prospective and spontaneous reporting-based pharmacovigilance program was conducted in Cuba. Patient demographics, suspected ADR description, elements to define causality, and outcomes were documented and analyzed.
Results
A total of 1496 suspected ADRs identified in 792 patients out of the 1660 (47.7 %) prescriptions reported in the program, were received from July 1995 to July 2002. Most of the patients (71.3%) were male, 67.2% were white and mean age was 61.6 ± 13.0 years. The mean time interval between the onset of symptoms and the start of the SK infusion was 4.9 ± 3.7 h. The most frequently reported ADRs were hypotension, arrhythmias, chills, tremors, vomiting, nauseas, allergy, bleeding and fever. ADR severity was 38% mild, 38% moderate, 10% severe, and 4% very severe. Only 3 patients with hemorrhagic stroke were reported. Seventy-two patients died in-hospital mainly because of cardiac causes associated with the patient's underlying clinical condition. Mortality was 3 times more likely in patients suffering arrhythmias than in those without this event (odds ratio 3.1, 95% CI: 1.8 to 5.1). Most of the reported ADRs were classified as possibly or probably associated with the study medication.
Conclusion
Recombinant SK was associated with a similar post-marketing safety profile to those suggested in previous clinical trials.
==== Body
Background
Streptokinase (SK) is an indirect fibrinolytic agent that interacts with plasminogen, forming an active complex with protease activity that converts plasminogen to plasmin [1]. SK efficacy with regard to mortality reduction in patients with acute myocardial infarction (AMI) has been demonstrated in large, placebo-controlled trials [2,3].
A SK produced by recombinant DNA techniques (rSK) has been evaluated previously in clinical trials in AMI patients. The first studies suggested that rSK produced the same benefits that were described for natural SK [4,5]. These results were further supported by the Thrombolysis with Recombinant Streptokinase in Acute Myocardial Infarct (TERIMA)-1 trial, a multicenter, randomized, comparative study of rSK vs. natural SK in 224 patients [6]. Both SKs behaved similarly regarding coronary patency at 8 days after thrombolysis and the changes induced on fibrinogen, fibrinogen-degradation-products, and thrombin time. They were also similar with respect to anti-SK antibodies titers and their anti-SK neutralizing activities [7]. The effect on AMI patients' in-hospital mortality was evaluated in a 2923-patients, multicenter, open, phase IV, clinical study (TERIMA-2) [8]. In-hospital 10.4% mortality was found which represented a 4% absolute and a 28.3% relative lethality reduction as compared to a survey made before rSK treatment was introduced.
Clinical trials are usually conducted on a limited number of patients with restrict selection and handling criteria and thus only the most frequent reactions are identified. Postmarketing monitoring is an important procedure to detect some reactions that can become apparent only when the drug is used in a large and varied population. Therefore, an active surveillance program provides a vital service to the heath care system to identify and assess early warning signals, and when appropriate, to take preventive actions to minimize the deleterious effects of drugs.
An adequate safety profile for rSK was observed in clinical trials but it was unknown whether similar results would be found in routine clinical practice. This concern was particularly important for rSK given its nature as a fibrinolytic drug, the potential antigenic capacity due to its bacterial origin, the novelty of the product production process, and its use in emergency situations. Therefore, the further extension of this treatment in Cuba was monitored through an appropriate pharmacovigilance program with the aim to assess rSK safety in AMI patients in routine practice. This paper reports the analysis of data collected after 7 years of such surveillance system implementation.
Methods
A national, prospective and spontaneous reporting-based pharmacovigilance study was conducted in Cuba. This study was extended to the 14 provinces throughout the different levels of the healthcare system, including municipal, provincial, and clinical-surgical hospitals as well as third-level specialized cardiology units. An adverse drug reaction (ADR) report form was designed. It explicitly asked for the most commonly rSK-associated ADRs and also requested for other reactions. The Clinical Trials Division of the Center for Biological Research and the National Network of Pharmacoepidemiology (NNP) were responsible for the distribution to the hospital services, gathering and quality assessment of ADR report forms. This network is coordinated by the Centre for the Development of Pharmacoepidemiology and its National Coordinating Unit of Pharmacovigilance, as previously described [9]. Since 1994 Cuba is member of the WHO International Drug Monitoring Program.
The rSK (Heberkinasa, Heber Biotec, Havana) studied in this pharmacovigilance system is indicated in AMI as soon as possible, before 12 hours of symptoms onset, in patients without contraindications for thrombolysis. The recommended dose is 1 500 000 IU during 1 hour, through a peripheral vein infusion. Suspected ADRs were identified by physicians through a direct patient evaluation during and after SK administration, as part of the AMI patient assessment. Patient demographics, time between symptom onset and SK infusion, concomitant therapy, outcome (if death and cause of death) and ADR description, including duration, severity, treatment and need for discontinuation, were documented.
An ADR was defined as any noxious, unintended, and undesired effect of a drug that was observed at doses usually administered therapeutically in humans. The medical terminology such as signs and symptoms definition as well as disease or syndrome diagnostic criteria was left to the attending physicians' judgment so that it could reflect routine clinical practice. A temporal or possible association was sufficient for a report to be made.
A qualitative assessment was used to classify the causal relationship as definite, probable, possible or doubtful [10]. According to this method, a reaction was classified as definite if (1) followed a reasonable temporal sequence after drug administration; (2) followed a known pattern of response to the suspected drug; (3) could not be explained by concurrent disease or other drugs; and (4) was confirmed by improvement upon removal of the drug and by reappearance on rechallenge. It was considered as probable if it had the criteria (1), (2), (3) and was confirmed on suspension of the drug but not on rechallenge. A reaction was defined as possible if followed a reasonable time sequence to administration of the drug, but could also be explained by concurrent disease or other drugs. Finally, a reaction that was more likely related to factors other than the suspected drug was classified as doubtful.
The severity of ADRs was classified in four levels: (1) mild if no therapy was necessary, (2) moderate if needed specific treatment, (3) severe when hospitalization or its prolongation was required, and (4) very severe if a reaction was potentially life-threatening or contributed to patient's death.
The causality assessment, as well as the severity analysis, were undertaken by members of the NNP taking into account the reporting physician's criteria, and were further reviewed by the responsible investigators.
The study was sanctioned by the Scientific Committee of the Center for Biological Research. As the study did not involve any intervention apart from the usual medical procedures, and confidentiality of the subjects was maintained, ethics approval and patients' informed consent were not required. The analysis of all ADRs was reported to the National Regulatory Authority.
Data management and statistical analyses
Databases were double-entered and validated on Microsoft Visual FoxPro version 5.0 and then imported into SPSS version 11.5 for further analysis. Continuous variables are given as means ± standard deviations (SD), whereas discrete variables are given as frequency distributions. Univariate analyses were performed to identify the variables that could influence on adverse reactions and mortality. The analyses were assessed by the χ2 or Fisher's exact tests, depending on the minimum expected values of the tables. A logistic regression analysis was done with variables that resulted statistically significant. The level of significance chosen was p = 0.05.
Results
A total of 1660 notifications of SK prescriptions was received from July 1995 to July 2002 from 50 hospitals. Of them, 792 (47.7%) reported suspected ADRs. The baseline characteristics of the patients are shown in Table 1. Most of the subjects (71.3%) were male, 67.2% were white and the mean age was approximately 62 years. Age was gender-related: women were 4.4 (95% confidence interval [CI]: 2.9 to 5.9) years older. The mean time interval between the onset of symptoms and the start of the SK infusion was 4.9 ± 3.7 hours. Concomitant acetylsalicylic acid was administered in 81% of the patients.
Table 1 Baseline characteristics of the patients (N = 1660)
Characteristic n (% of total)*
Gender
Male 1183 (71.3%)
Female 430 (25.9%)
Not specified 47 (2.8%)
Ethnicity
White 1116 (67.2%)
Mestizo 300 (18.1%)
Black 154 (9.3%)
Chinese 8 (0.5%)
Not specified 82 (4.9%)
Age (yr) total (183 not specified) 61.6 ± 13.0
Onset-infusion time (hours) 4.9 ± 3.7
Previous SK
< 1 year 5 (0.3%)
> 1 year 9 (0.5%)
Concomitant therapy
Aspirin 1341 (80.8%)
Beta-blockers 943 (56.8%)
Nitrosorbide 258 (15.5%)
Lidocaine 66 (3.9%)
Pentaerythritol tetranitrate 45 (2.7%)
Others 641 (38.6%)
*Continuous variables are presented as Mean ± SD
We found 1496 suspected ADRs identified in 792 patients (1.9 ADRs per patient). The most frequent rSK-related ADRs were hypotension, arrhythmias, chills, tremors, vomiting, and nauseas (Table 2). Allergy was recorded in 47 patients, 34 of them (72%) only with skin rash. Hemorrhage was notified only in 43 patients. ADRs present in less than 5 patients each, included dizziness, dyspnea, arterial hypertension, unspecified pain, paleness, rubor, headache, diarrhea and others.
Table 2 Suspected adverse reactions to streptokinase
ADR n %*
Hypotension 285 36.0
Arrhythmias 281 35.5
Chills 212 26.8
Tremors 197 24.9
Vomiting 187 23.6
Nauseas 103 13.0
Allergy 47 5.9
Bleeding 43 5.4
Fever 41 5.2
Lumbar pain 10 1.3
Sweating 9 1.1
Muscular cramp 5 0.6
Others 76 9.6
* % of total patients with ADR, n = 792.
The causality assessment showed that 1419 (94.9%) of the total ADRs were classified as possibly associated with the study medication. Probable and doubtful associations were considered in 63 (4.2%) and 6 (0.4%) of suspected reactions, respectively. Definite reactions were only confirmed in one patient with allergic reactions. The information about eight adverse events was not sufficient for a proper causality assessment.
Of the overall number of suspected ADRs, severity was classified as mild in 564 (37.7%), moderate in 573 (38.3%), severe in 149 (10.0%), and 63 (4.2%) were very severe. Severity was not classified in 147 (9.8%) reactions. Considering the maximal severity to which the patient was exposed, 53 (6.7%) subjects had at least one very severe reaction, which was described for hypotension, arrhythmias and hemorrhage. Of the remaining patients, 104 (13.1%) were exposed to at least one severe reaction.
Definitive discontinuation due to ADRs was observed in 54 (6.8% of those with ADRs) patients. Hypotension, arrhythmias and vomiting were the main causes of treatment withdrawal. Hypotension was also the ADR most frequently leading to temporal SK infusion discontinuation, which was required for the adequate control of 27.7% of the patients experiencing this reaction. The appropriate management of this event included infusion rate reduction, infusion discontinuation, Trendelenburg's position, volume expansion and vasopressors.
ADR occurrence was related to the time between symptoms onset and SK infusion, both in univariate and multivariate analyses. Patients who received SK ≤ 3 hours were more likely to have any ADR as compared to those receiving the infusion after 3 hours (odds ratio 1.5, 95% CI: 1.2 – 1.8). Individual analysis for each suspected ADR showed that arrhythmias were the most time-influenced, being more frequent this event in patients who received SK ≤ 3 h (95% CI: 1.26 – 2.18). The same was true for hypotension (1.08 – 1.85), chills (1.11 – 2.08), and vomiting (1.13 – 2.18). Multivariate analyses confirmed these associations for arrhythmias and hypotension, but for chills and vomiting none of the models obtained were reliable.
Out of the 1660 patients reported, 72 (4.3%) died. Death causes are summarized in Table 3. Cardiac causes (pump failure, arrhythmias, wall rupture), associated mainly with the patient's clinical state, were the most frequent (79.2%). Hemorrhagic stroke was reported as fatal in 3 cases and was the only cause of death that could be directly explained by the use of the study medication.
Table 3 Causes of death
Cause n % of total deaths
Cardiac causes
Pump failure 32 44.4
Arrhythmias 15 20.8
Wall rupture 10 13.9
Non cardiac causes
Hemorrhagic stroke 3 4.2
Pulmonary embolism 1 1.4
Hypoxic encephalopathy 1 1.4
Sepsis 1 1.4
Unknown 9 12.5
Total 72 100
Univariate analyses with baseline variables showed that mortality was directly related to older age (p = 1.0 × 10-7) and female gender (p = 1.7 × 10-4). Patients recorded as dead were 10.1 (95% CI: 7.0 to 13.2) years older than survivors. None of the logistic regression models for the multivariate analysis of the effect of age and gender on mortality was adequate, probably due to the small number of deaths in the sample. Arrhythmias was the only suspected ADR significantly related with death (p = 4.2 × 10-5). Mortality was 3 times more probable in patients suffering arrhythmias than in those without this event (odds ratio 3.1, 95% CI: 1.8 to 5.1).
Discussion
This pharmacovigilance study was performed to further investigate the rSK safety profile in the treatment of AMI patients in clinical practice. Although a national, spontaneous reporting scheme for monitoring all market drugs exists in Cuba, rSK surveillance operated as a separate national system. Using a different ADR report form, this system provided specific information about a thrombolytic drug that otherwise was not possible to obtain (e.g. symptom-infusion time) and also contributed to encourage the number of reports. Results of this study showed a post-marketing safety profile for rSK similar to those suggested in previous clinical trials.
Spontaneous reporting is the most widely used method for pharmacovigilance, but is associated with considerable under-reporting rate [11]. In fact 29,753 rSK doses were distributed throughout the country during the period studied (Heber Biotech, personal communication). Apart from other minor rSK uses, this represents approximately 5.6% of the number of doses used throughout the country in AMI patients. This figure is similar to those showed in other spontaneous reporting schemes [12]. Various possible reasons for not reporting have been identified, including uncertainty as to whether the reaction was caused by a drug, trivial or well known ADRs, unawareness of the need to report ADRs, not enough time and the thought that it is too bureaucratic [13]. All of the above mentioned factors could have influenced the under-reporting observed in this study. In accordance with other spontaneous reporting systems [11], it was not possible to estimate ADRs incidence rates because the numerator was affected by under-reporting and the accurate number of subjects treated (denominator) was unknown.
Time between symptoms and infusion is a critical variable for success of thrombolysis [2,3]. Since the results obtained up to 1995 in the national extension study [8], more experience with the procedures and organizational measures in the heath care system could have improved the time to thrombolysis. However, in the present study this time did not differ significantly from the previous period. This was probably influenced by the under-reporting and do not reflect what actually happened in practice. Currently, the implementation of new organizational actions for the nationwide extension of thrombolysis to the pre-hospital level will surely have a greater impact on this aspect.
Some deviations from the approved indications were observed in routine clinical practice. Although rSK has been approved for routine use within 12 h, 18 patients were reported as receiving the drug after 12 h of symptoms onset. On other hand, the proportions of patients that received aspirin and beta blockers were smaller than previously reported in TERIMA-2 study, where a significant benefit was demonstrated for both drugs.
The most frequent suspected ADRs associated to rSK therapy were hypotension and arrhythmias. This is consistent with the safety profile observed in previous clinical trials. Both events can also be explained by AMI physiopathology and sometimes by other concomitant drugs, so it was difficult to establish a casual relationship with the study medication and all were classified as possible. Hypotension has been described as a transient event that can be related to SK infusion rate [14,15]. Thus, it is recommended to reduce the infusion rate as its initial approach. Arrhythmias have been considered as reperfusion signs after thrombolytic therapy [16], but are also important complications in the course of AMI. Despite the use of thrombolytic therapy, in-hospital ventricular arrhythmias have been associated with higher 30-day and 1-year mortality rates [17,18]. Patients treated earlier after coronary occlusion, when risk is greatest, are more likely to develop ventricular fibrillation [19], which can explain why arrhythmias were more frequent in patients who received SK ≤ 3 h in this study.
Bleeding was notified in 43 patients, of whom only 8 were considered as serious. Hemorrhagic stroke, which is the most severe bleeding complication of fibrinolytic therapy, was reported only in 3 patients (all fatal). This is consistent with the low hemorrhagic stroke rate observed previously in the TERIMA-2 trial (0.3%).
Because SK is an antigenic bacterial protein, physicians need to be aware that some patients experience allergic reactions. Following rSK administration, a significant increase of anti-SK antibodies binding and neutralization titers persisting for 6 months or more have been described [7]. Therefore, concern about SK re-administration exists since it is possible that these antibodies cause allergic reactions or neutralize a further SK dose with the consequent ineffectiveness. Interestingly, 14 patients (5 < 1 year) with antecedent of exposure to SK were reported in this pharmacovigilance and none of them experienced allergic reactions nor account for in-hospital death.
According to the definitions used in this study, at least 19.8% of patients with ADR had a serious one (sum of severe and very severe ADRs). This relatively high proportion is a tendency of spontaneous reporting schemes, since under-reporting involves mainly the less severe and the well known effects [11] so that severe reactions are more likely reported. However, few patients withdrew due to ADRs, emphasizing that prompt recognition and appropriate symptoms handling result in adequate treatment compliance. The in-hospital lethality rate was low but under-reporting could also affect the interpretation of this variable.
Definite reaction assessment was difficult because SK is administered once and rechallenge information is usually not available. Also, many of the SK associated reactions can be explained by concurrent diseases or other drugs as well. Ideally, both, the causality assessment and severity analysis should have been performed by two individuals at the investigation site, separately and then agreement determined. This was not done for practical reasons, since there was only one NNP member at each hospital and thus the evaluation had to be done by this person considering the criteria from the reporting physician, with the subsequent bias derived from application of subjective scales.
Recombinant tissue plasminogen activator (tPA) or its derivatives are recommended by ACC/AHA guidelines to be used in patients who present early after onset of chest pain or symptoms and in those with previous administration of streptokinase and at low risk of intracranial hemorrhage [20]. However, this product is much more expensive than SK and not all countries can afford its general use, especially those with a general health reimbursement policy. Furthermore, a meta-analysis of large trials comparing SK vs. tPA did not demonstrate any clear differences in net clinical outcome because the beneficial cardiovascular effects of tPA is abated by an excess of hemorrhagic stroke [21].
The favorable safety profile achieved with rSK in this and previous studies, together with its demonstrated efficacy and its adequate cost, justifies the extension of this treatment in clinical practice as a reperfusion therapy for AMI patients. Recent reports about deviations from quality of different SK preparations [22,23], raise the importance of a stringent regulation for this product. These reports should be backed up by clinical data [24]. Therefore, results from previous clinical trials and this pharmacovigilance study, along with an appropriate quality control system [25], support the recognized impact of Cuban rSK on AMI mortality in the country.
As in other countries, ischemic heart disease is the major cause of death in Cuba (18.8% of all deaths in 2004; 135.4 per 100000 inhabitants) [26]. The possibility to extend thrombolytic therapy with a safe homemade product is an example of how local biotechnology can influence on one of the main public health problems with a favorable cost/benefit balance, not possible in a developing country with an imported drug, which would had meant a significant economic burden for the public health system.
Conclusion
In this study, no new signals of previously unknown adverse events were reported, and the nature and severity of the reported events were consistent with information known about rSK from clinical trials. New directions about thrombolysis are emerging in the country, including the use of a new albumin-free formulation [27,28], pre-hospital extension of the procedure, and new indications approval. Therefore, the continuity of this pharmacovigilance would be of inestimable value for safety monitoring of the product.
Appendix
The National Network of Pharmacoepidemiology participating investigators: Dr. Frank Ravelo González, Dra. Déborah Rodríguez Piñeiro, Dr. Juan C. Mesa Hernández, Lic. Diana R. Fernández Ruiz, Lic. Susana Gómez Pentón, Dr. Rubén Escalante Guardarramos, Lic. Yaneisi C. Perdomo Espinosa, Dr. Ramón Suarez Ramírez, Dra Zaida Herrera López, Lic. Lídice Pérez Incola, Dra. Arlette Linares Borges, Dra. Ismary Alfonso Orta.
Competing interests
Authors BYB, CVS, EGI, TGL, LAF, and PLS are employees of the Center for Biological Research, which is part of the Center for Genetic Engineering and Biotechnology, Havana network, where recombinant streptokinase is produced. MAMM, FHB were members of the same staff when this work was done. The rest of the authors have no competing interests at all.
Pre-publication history
The pre-publication history for this paper can be accessed here:
Acknowledgements
The authors wish to thank all the personnel at each of the participating hospitals which contributed to the reports.
==== Refs
Young KC Shi GY Chang YF Chang BI Chang LC Lai MD Chuang WJ Wu HL Interaction of streptokinase and plasminogen. Studied with truncated streptokinase peptides J Biol Chem 1995 270 29601 29606 7494004 10.1074/jbc.270.49.29601
Gruppo Italiano per lo Studio della Streptochinasi nell'Infarto Miocardico (GISSI) Effectiveness of intravenous thrombolytic treatment in acute myocardial infarction Lancet 1986 1 397 402 2868337
ISIS-2 (Second International Study of Infarct Survival) Collaborative Group Randomised trial of intravenous streptokinase, oral aspirin, both, or neither among 17,187 cases of suspected acute myocardial infarction: ISIS-2 Lancet 1988 2 349 360 2899772
Toruncha A Torres W Mora E León L Padrón N Castañeda M Martínez M Pardo A Zamora R Galhardo M Pérez A Banasco J Jiménez R Fariñas H Mainet D Sorell L Estrada MP de la Fuente J Comparative clinical trial in patients with acute myocardial infarction using recombinant and natural streptokinase Biotecnología Aplicada 1992 9 289 291
Llerena L Llerena R Rodríguez-de la Vega A Valdés-Rucabado JA Fernández-Felipe R Rodríguez-Bereau C Trombolisis intracoronaria con Estreptoquinasa Recombinante (SK-R) en el infarto agudo del miocardio XVII Congreso Nacional de Cardiología, Guadalajara, Jalisco (México) Septiembre 29–Octubre 3, 1991
The TERIMA Group Investigators Multicenter, randomised, comparative study of recombinant vs. natural streptokinases in acute myocardial infarct Thromb Haemost 1999 82 1605 1609 10613642
Mainet D del Rosario M Toruncha A Prats P Valenzuela C Lopez-Saura P Similar, more than 6-months persisted, antibody and neutralizing activity responses in patients with acute myocardial infarction treated with recombinant or natural streptokinase Fibrinolysis & Proteolysis 1998 12 301 309
The TERIMA Group of Investigators TERIMA-2: national extension of thrombolytic treatment with recombinant streptokinase in acute myocardial infarct in Cuba Thromb Haemost 2000 84 949 954 11330244
Debesa F Jimenez G Figueras A Diogene E Perez Peña J Avila J Laporte JR Spontaneous reporting of adverse drug reactions in Cuba: integrating continuous education, training and research in a network approach Br J Clin Pharmacol 2002 54 335 336 12236857 10.1046/j.1365-2125.2002.01635.x
Naranjo CA Shear NH Busto U Kalant H, Roschlau WHE Adverse drug reactions Principles of medical pharmacology 1998 6 Oxford University Press: New York 791 800
Alvarez-Requejo A Carvajal A Begaud B Moride Y Vega T Arias LH Under-reporting of adverse drug reactions. Estimate based on a spontaneous reporting scheme and a sentinel system Eur J Clin Pharmacol 1998 54 483 488 9776440 10.1007/s002280050498
Pirmohamed M Breckenridge AM Kitteringham NR Park BK Adverse drug reactions BMJ 1998 316 1295 1298 9554902
Eland IA Belton KJ van Grootheest AC Meiners AP Rawlins MD Stricker BH Attitudinal survey of voluntary reporting of adverse drug reactions Br J Clin Pharmacol 1999 48 623 627 10583035 10.1046/j.1365-2125.1999.00060.x
Lew AS Laramee P Cercek B Shah PK Ganz W The hypotensive effect of intravenous streptokinase in patients with acute myocardial infarction Circulation 1985 72 1321 1326 4064276
Lateef F Anantharaman V Hypotension in acute myocardial infarction patients given streptokinase Singapore Med J 2000 41 172 176 11063182
Carabello BA Spaan JF Comerota AJ Thrombolytic therapy in acute myocardial infarction Thrombolytic Therapy 1988 Grune & Stratton Inc.: Orlando 165 87
Newby KH Thompson T Stebbins A Topol EJ Califf RM Natale A Sustained ventricular arrhythmias in patients receiving thrombolytic therapy: incidence and outcomes. The GUSTO Investigators Circulation 1998 98 2567 2573 9843464
Al-Khatib SM Stebbins AL Califf RM Lee KL Granger CB White HD Armstrong PW Topol EJ Ohman EM Sustained ventricular arrhythmias and mortality among patients with acute myocardial infarction: results from the GUSTO-III trial Am Heart J 2003 145 515 521 12660676 10.1067/mhj.2003.170
Wilkinson J Foo K Sekhri N Cooper J Suliman A Ranjadayalan K Timmis AD Interaction between arrival time and thrombolytic treatment in determining early outcome of acute myocardial infarction Heart 2002 88 583 586 12433884 10.1136/heart.88.6.583
Antman EM Anbe DT Armstrong PW Bates ER Green LA Hand M Hochman JS Krumholz HM Kushner FG Lamas GA Mullany CJ Ornato JP Pearle DL Sloan MA Smith SC Jr ACC/AHA guidelines for the management of patients with ST-elevation myocardial infarction – executive summary. A report of the American College of Cardiology/American Heart Association Task Force on Practice Guidelines (Writing Committee to revise the 1999 guidelines for the management of patients with acute myocardial infarction) J Am Coll Cardiol 2004 44 671 719 15358045 10.1016/j.jacc.2004.07.002
Collins R Peto R Baigent C Sleight P Aspirin, heparin, and fibrinolytic therapy in suspected acute myocardial infarction N Engl J Med 1997 336 847 860 9062095 10.1056/NEJM199703203361207
Longstaff C Thelwell C Whitton C The poor quality of streptokinase products in use in developing countries J Thromb Haemos 2005 3 1092 1093 10.1111/j.1538-7836.2005.01271.x
Hermentin P Cuesta-Linker T Weisse J Schmidt KH Knorst M Scheld M Thimme M Comparative analysis of the activity and content of different streptokinase preparations Eur Heart J 2005 26 933 940 15637085 10.1093/eurheartj/ehi093
Llerena LD Caceres-Loriga FM Betancourt BY Recombinant streptokinase: evidences from clinical use Eur Heart J 2005 26 1448 1449 15905301 10.1093/eurheartj/ehi319
Hernandez L Martinez Y Quintana M Besada V Martinez E Heberkinasa: recombinant streptokinase Eur Heart J 2005 26 1691 15946953 10.1093/eurheartj/ehi377
Ministry of Public Health of Cuba Health Statistics Yearbook 2004
López M González LR Reyes N Sotolongo J Pujol V Stabilization of a freeze-dried recombinant streptokinase formulation without serum albumin J Clin Pharm Ther 2004 29 367 363 15271104 10.1111/j.1365-2710.2004.00573.x
Llerena LD Betancourt BY Quirós JJ Sainz B Valdés JA Zorio B Díaz-Rojo G García AI Villanueva LH Filgueiras CE Cabrera F Echarte JC Pérez del Todo JM Guerrero I López L García EJ Nadal B López-Saura P Angiographic patency study of an albumin-free recombinant streptokinase formulation in acute myocardial infarction J Pharm Pharm Sci 2004 7 372 377 15576019
|
16262910
|
PMC1291362
|
CC BY
|
2021-01-04 16:27:54
|
no
|
BMC Clin Pharmacol. 2005 Nov 2; 5:5
|
utf-8
|
BMC Clin Pharmacol
| 2,005 |
10.1186/1472-6904-5-5
|
oa_comm
|
==== Front
BMC Fam PractBMC Family Practice1471-2296BioMed Central London 1471-2296-6-451627113610.1186/1471-2296-6-45Research ArticleThe epidemiology of suicide and attempted suicide in Dutch general practice 1983–2003 Marquet Richard L [email protected] Aad IM [email protected] Ad JFM [email protected] François G [email protected] der Zee Jouke [email protected] Netherlands Institute of Health Services Research (NIVEL), PO Box 1568, 3500 BN, Utrecht, The Netherlands2 Department of Clinical Psychology, Free University Amsterdam, van der Boechorststraat 1, 1081 BT Amsterdam, The Netherlands; and Research Institute Psychology and Health, Utrecht, The Netherlands2005 4 11 2005 6 45 45 16 3 2005 4 11 2005 Copyright © 2005 Marquet et al; licensee BioMed Central Ltd.2005Marquet et al; licensee BioMed Central Ltd.This is an Open Access article distributed under the terms of the Creative Commons Attribution License (), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
Background
Many patients attempting or committing suicide consult their general practitioner (GP) in the preceding period, indicating that GPs might play an important role in prevention. The aim of the present study was to analyse the epidemiology of suicidal behaviour in Dutch General Practice in order to find possible clues for prevention.
Method
Description of trends in suicide and suicide attempts occurring from 1983–2003 in the Dutch General Practice Sentinel Network, representing 1% of the Dutch population. The data were analysed with regard to: 1) suicidal behaviour trends and their association with household situation; 2) presence of depression, treatment of depression and referral rate by GPs; 3) contact with GP before suicide or suicide attempt and discussion of suicidal ideation.
Results
Between 1983 and 2003 the annual number of suicide and suicide attempts decreased by 50%. Sixty percent of the patients who committed or attempted suicide were diagnosed as depressed, of whom 91% were treated by their GP with an antidepressant. Living alone was a risk factor for suicide (odds ratio 1.99; 95% CI 1.50 to 2.64), whereas living in a household of 3 or more persons was a relative risk for a suicide attempt (odds ratio 1.81; 95% CI 1.34 to 2.46). Referral to a psychiatrist or other mental health professionals occurred in 65% of the cases. GPs recalled having discussed suicidal ideation in only 7% of the cases, and in retrospect estimated that they had foreseen suicide or suicide attempts in 31% and 22% of the cases, respectively, if there had been contact in the preceding month.
Conclusion
With regard to the prescription of antidepressants and referral of suicidal patients to a psychiatrist, Dutch GPs fulfil their role as gatekeeper satisfactorily. However, since few patients discuss their suicidal ideation with their GP, there is room for improvement. GPs should take the lead to make this subject debatable. It may improve early recognition of depressed patients at risk and accelerate their referral to mental health professionals.
==== Body
Background
As in most Western countries, suicide and suicide attempts are major public health problems in the Netherlands [1]. Suicide is the 14th leading cause of death and the fifth leading cause of years of potential life lost [[2], Statistics Netherlands]. Attempted suicide, being an important risk factor for suicide, is estimated to be as high as 130/100,000 which represents a significant burden to the individuals affected, their families and the health services [3]. Many of these patients consult their general practitioner (GP) in the period preceding their suicide or suicide attempt, indicating that GPs might play an important role in prevention [4]. However, knowing that a Dutch GP may lose a patient to suicide only once every 4–5 years, it is clear that early recognition of suicidal patients is difficult [5]. Still, GPs may have ways to prevent suicide and suicide attempts: 1) by addressing patients' emotional problems, 2) screening for depression, 3) probing for suicide ideation and 4) making proper referrals to mental health professionals [6,7]. Our question was: to what extent do Dutch GPs fulfil these requirements? Ever since 1979, suicide and suicide attempts occurring in Dutch general practice are registered by GPs participating in the Dutch Sentinel Network [5]. In order to find clues for early recognition of suicide-endangered patients we analysed relevant data from the Network over the period 1983–2003.
Methods
The data were derived from the Dutch Sentinel Practice Network, which has been operational since 1970; it constitutes a sample of about 65 GPs, covers about 1% of the Dutch population, and is representative of the total population (16.3 million in 2003), with regard to age, sex, geographical distribution, and level of urbanisation. Over 95% of non-institutionalised citizens in the Netherlands are registered with a GP. The Dutch Sentinel Network structure takes account of the geographical distribution of the population and its spread over areas with different degrees of urbanisation. A census is held every two years among the practice populations involved, in order to determine the size and composition of the population to which the gathered data is to be related. The Dutch Sentinel Network has been participating in many national [5,8,9] and international projects. The oldest international project is the European Influenza Surveillance Scheme [EISS; [10]]. Using specific forms, GPs participating in the network report weekly to our Institute on the incidence of various diseases, types of diagnosis and events, including requests for euthanasia, and suicide and suicide attempts. For every case of suicide or suicide attempt reported anonymously by the sentinel GPs, additional data are provided with regard to: 1) age and sex, 2) previous attempts, 3) method and place, 4) number of persons in the household, 5) previous contact with the GP, 6) nature of this contact, 7) presence of depression (defined according to guidelines by the Dutch GP Association [NHG, [11]], and its treatment, 8) referral to mental health professionals, and 9) whether the GP had foreseen the suicide or suicide attempt.
Results
Suicides and suicide attempts: numbers, age distribution and relationship with household
From 1983–2003 a total of 248 suicides (169 men and 79 women) and 1,231 suicide attempts (409 men and 822 woman) were registered. As far as GPs were informed, most suicides (total 75%; 80% for men, 65% for women) were committed at first attempt. Likewise, 67% of male non-fatal suicidal behaviour and 59% of female non-fatal behaviour were considered to be first attempts.
In figure 1 the 3-year averages of the number of suicides and suicide attempts per 100,000 inhabitants are depicted. It can be observed that both rates gradually decreased over time. Compared to the 1983–1985 period the average number of suicides and suicide attempts in the 2001–2003 period was almost 50% lower, both for men and women. A similar trend has been identified nationally, indicating that the sample from the sentinel network is representative [2]. The total number of suicides decreased from 11.2 to 5.7/100,000; the number of suicide attempts decreased from 53.5 to 27.0/100,000.
Figure 1 Suicide (S) and attempted suicide (A) in Dutch general practice 1983–2003; 3-year averages/100,000 inhabitants.
The age distribution of suicides and suicide attempts is given in figure 2. About 50% of all suicides and 70% of all suicide attempts were committed between the age of 20 and 50 years. Prominent peaks occurred in the age group of 30–39 years, except for suicide attempts in women, which reached their highest proportion between 20 and 29 years. It is noteworthy that in persons older than 60 years the proportion of attempted suicides decreased, whereas suicide in elderly men and women did not. Figure 3 shows the percentage of persons who committed suicide or attempted suicide in the context of their domestic situation. Suicides were more common in persons living alone (odds ratio 1.99; 95% CI 1.50 to 2.64), whereas suicide attempts occurred more frequently in households consisting of 3 or more persons (odds ratio 1.81; 95% CI 1.34 to 2.46).
Figure 2 Age distribution of suicides (S) and attempted suicides (A) of men an women 1983–2003; % of total.
Figure 3 Suicides and suicide attempts; relationship with number of persons in household
Depression
More than 50% of the patients (60% of total; 52% men, 68% women) who committed or attempted suicide were diagnosed by their GP as being depressed. On average, the prevalence of depression was about 8% higher in patients who committed suicide than in those who attempted suicide. The characteristics and symptoms of suicidal patients suffering from depression are presented in figure 4. Most of the characteristics were similar for those who died by suicide and for those who attempted suicide. Nearly all depressed patients (91%) were treated with an antidepressant. From 1983 to 1993 the first drug of choice prescribed by the GP was a tri-cyclic antidepressant (TCA); selective serotonin reuptake inhibitors (SSRIs) became increasingly popular thereafter and comprised 83% of the prescriptions for depression in 2003. An important finding was that in only 7% of cases GPs recalled to have discussed suicidal ideation with depressed patients who committed or attempted suicide.
Figure 4 Characteristics and symptoms of suicidal patients suffering from depression.
Referral of patients; did the GP foresee suicide or the suicide attempt?
Sixty-five percent of the patients who ultimately committed suicide and 63% of those who attempted suicide had been referred by their GP to a psychiatrist or an institution for ambulatory mental health care. In the Netherlands, mental health care professionals at these institutions include psychiatrists, psychologists and social workers. The referral frequency steadily increased over time, from 70% in 1983 to 84% in 2003.
Fifty-three percent of the patients who committed suicide and 54% of those who attempted suicide had contacted their GP in the 30-day period preceding the event. In the case of contact GPs mentioned retrospectively that suicide was foreseen in 31% of the cases; when there had been no contact suicide was foreseen in 7% of the cases. For suicide attempts these percentages were 22% and 3%, respectively.
Discussion
A few characteristics of a person at risk of suicide or a suicide attempt do emerge from the present study. Somewhat overstated, the typical profile of a person at risk of suicide in the Netherlands is a young or elderly depressed patient, frequently living alone, most likely to be a man, treated by the GP with a modern antidepressant and rarely speaking of suicidal ideation. Living alone as a risk factor has been cited in many epidemiological studies on suicide. (12, 13, 14). Strong associations between suicide/suicide ideation and different aspects of loneliness, either subjective (feeling lonely) or objective (living alone and being without friends) have been observed (13). The profile of a person at risk for a suicidal attempt is almost the same, except that this depressed patient is more frequently a younger woman not living alone. Most of these characteristics are certainly not exclusively Dutch, they have been cited before by several investigators from different countries [12-19].
Typical for the Dutch situation is that for the first time a more precise estimate can be made about the incidence of attempted suicides in the Netherlands. Earlier estimates were based on a four-year inventory (1989–1991) among general hospitals, psychiatric hospitals and GPs in a defined area around the city of Leiden (3). From these data an incidence of 143/100,000 was calculated. Our results indicate that the mean yearly incidence of attempted suicides reported to GPs in the period 1983–2003 is 41/100,000. In the Leiden study it was observed that about 40% of all suicide attempts was reported by GPs. If this percentage also applies for the whole country it can be calculated that the incidence of attempted suicides in the Netherlands is lower than estimated in the Leiden study, namely about 100/100,000, an incidence almost similar to that recently calculated for France [20]. Interestingly, the number of attempted suicides reported by GPs in the period 2001–2003 was 50% less than reported in 1983–1985, both for men and women. This coincides with the increased prescription of SSRIs, however there is no proof that this coincidence, which also applies for the reduced number of suicides, has causal aspects. We do not know whether this decreasing trend fully represents the decline in the incidence of (attempted) suicides, or is also due to a shift from GP to secondary mental health care and psychiatric hospitals. This factor of "missed patients" might also explain why the official trend line for suicides calculated by Statistics Netherlands lies below the trend line calculated from the GP data [[2], Statistics Netherlands].
Depression is the most common psychiatric disorder in patients who attempt or commit suicide [21,22]. The role of the GP in the treatment of depression therefore is of considerable clinical importance. We found that 60% of patients who attempted or committed suicide were diagnosed by their GP as being depressed. This is fairly consistent with percentages cited by others [20]. In the year 2000 the Dutch Sentinel Network registered 440/100,000 new cases of depression [9]. If we relate this number to our data, while realising that being depressed is not the same as being diagnosed as depressed, it can be calculated that about 1% of depressed patients commit suicide and about 5% of depressed patients will attempt suicide. These figures are comparable with those of Khan et al. in the USA who calculated that 0.8% of patients participating in antidepressant trials committed suicide and 2.9% attempted suicide [23]. How does the Dutch GP cope with depressed patients who ultimately commit or attempt suicide, considering that a percentage of 6% is still tantamount to a needle in a haystack? From our survey the following picture emerges: Dutch GPs prescribe antidepressants fairly readily; they have switched almost unanimously to prescribing SSRIs, and refer depressed patients readily to a mental health professional. This sounds like an ideal situation, which is in keeping with the gatekeeper's role of GPs in the Netherlands [24]. With regard to serious depression this policy means that the GP provides the basic medication and leaves the in- depth treatment and responsibility to a psychiatrist and allied mental health caregivers. Still, there is room for improvement. The Dutch GP is either not a great communicator with regard to discussing suicidal ideation, or simply lacks the time to address this emotive subject: GPs recalled to have discussed suicidal ideation with only 7% of the depressed patients that ultimately committed or attempted suicide, which is a low percentage as far as risk assessment is concerned. However, it should be noted that this percentage is only an estimate. It is based on GP's recollection of having discussed suicidal ideation after the patient had attempted or committed suicide. Similar scepticism is warranted on the retrospective estimate whether the GP might have foreseen the suicide or suicide attempt. The relatively high percentages of 'foreseeing' in our analysis if there had been contact in the month preceding the suicide or suicide attempt (31% and 22%, respectively), are also based on GPs' recollection and therefore liable to bias.
Conclusion
With regard to prescribing antidepressants and referring depressed patients to a psychiatrist, Dutch GPs fulfil their role as gatekeeper satisfactorily. However, since few patients discuss their suicidal ideation with their GP, there is room for improvement. Our results suggest that GPs should ask more persistently for suicidal ideation in depressed patients, especially in those living alone. Making this issue debatable may enhance early recognition of high-risk patients and accelerate their referral to mental health professionals.
Competing interests
The author(s) declare that they have no competing interests.
Authors' contributions
RLM and AIMB performed the data analysis. AIMB also developed the study concept and headed its coordination. JvdZ and FGS participated in the design and coordination of the study. AJFMK contributed to the design and draft of the manuscript.
All authors read and approved the final manuscript.
Pre-publication history
The pre-publication history for this paper can be accessed here:
==== Refs
Chishti P Stone DH Corcoran P Williamson E Petridou E EUROSAVE Working Group Suicide mortality in the European Union Eur J Public Health 2003 3 108 14 12803408
Statistics Netherlands (Centraal Bureau voor de Statistiek)
Arensman E Kerkhof AJFM Hengeveld MW Mulder JD Medically treated suicide attempts: a four years monitoring study of the epidemiology in The Netherlands J Epidemiol Community Health 1995 49 285 9 7629465
Hamilton NG Suicide prevention in primary care Postgrad Med 2000 108 81 7 11098260
Diekstra RFW van Egmond M Suicide and attempted suicide in general practice, 1979–1986 Acta Psychiatr Scand 1989 79 268 75 2711854
Luoma JB Martin CE Pearson JL Contact with mental and primary care providers before suicide: a review of the evidence Am J Psychiatry 2002 159 909 16 12042175 10.1176/appi.ajp.159.6.909
Freedenthal S Primary care and suicide prevention Am J Psychiatry 2003 160 1012 3 12727716 10.1176/appi.ajp.160.5.1012-c
Marquet RL Bartelds A Visser GJ Spreeuwenberg P Peters L Twenty five years of requests for euthanasia and physician assisted suicide in Dutch general practice: trend analysis BMJ 2003 327 201 2 12881262 10.1136/bmj.327.7408.201
Donker GA Spreeuwenberg P Bartelds AIM van der Velden K Foets M Hormone replacement therapy: changes in frequency and type of prescription by Dutch GPs during the last decade of the millennium Fam Pract 2000 17 508 13 11120723 10.1093/fampra/17.6.508
Fleming DM Zambon M Bartelds AIM De Jong JC The duration and magnitude of influenza epidemics: A study of surveillance data from sentinel general practices in England, Wales and the Netherlands Eur J Epidemiol 1999 15 467 73 10442473 10.1023/A:1007525402861
Verhaak PFM Bartelds AIM Schellevis FG How do GPs treat new cases of depression? (Article in Dutch, abstract in English) Huisarts en Wetenschap (General Practitioner and Science) 2002 45 722 6
Middleton N Whitley E Frankel S Dorling Sterne J Gunnell D Suicide risk in small areas in England and Wales, 1991–1993 Soc Psychiatry Psychiatr Epidemiol 2004 39 45 52 15022046 10.1007/s00127-004-0707-y
Crawford MJ Prince M Increased rates of suicide in young men in England during the 1980s: the importance of social context Soc Sci Med 1999 49 1419 23 10509831 10.1016/S0277-9536(99)00213-0
Stravynski A Boyer R Loneliness in relation to suicide ideation and parasuicide: a population-wide study Suicide Life Threat Behav 2001 31 32 40 11326767 10.1521/suli.31.1.32.21312
Van Casteren V Van der Veken J Tafforeau J Van Oyen H Suicide and attempted suicide reported by general practitioners in Belgium 1990–1991 Acta Psychiatr Scand 1993 87 451 5 8356898
Hawton K Houston K Sheppard R Suicide in young people. Study of 174 cases, aged under 25 year based on coroners' and medical records Br J Psychiatry 1999 175 271 6 10645330
Waern M Rubenowitz E Wilhelmson K Predictors of suicide in the old elderly Gerontology 2003 49 328 34 12920354 10.1159/000071715
Wilkinson D Gunnel D Comparison of trends in method-specific suicide rates in Australia and England & Wales Aust N Z J Public Health 2000 24 153 7 10790934
Joseph HB Reznik I Mester R Suicidal behaviour of adolescent girls: profile and meaning Isr J Psychiatry Relat Sci 2003 40 209 19 14619680
Le Pont F Letrilliart L Massari V Dorleans Y Thomas G Flahault A Suicide and attempted suicide in France: results of a general practice sentinel network Br J Gen Pract 2004 54 282 4 15113496
Houston K Haw C Townsend E Hawton K General practitioners contacts with patients before and after deliberate self-harm Br J Gen Pract 2003 53 365 70 12830563
Rihmer Z Can better recognition and treatment of depression reduce suicide rates? A brief review Eur Psychiatry 2001 16 406 9 11728853 10.1016/S0924-9338(01)00598-3
Khan A Warner HA Brown WA Symptom reduction and suicide risk in patients treated with placebo in antidepressant clinical trials: an analysis of the Food and Drug Administration database Arch Gen Psychiatry 2000 57 311 7 10768687 10.1001/archpsyc.57.4.311
Kulu-Gascow I Delnoij D de Bakker D Self-referral in a gate-keeping system: Patient's reason for skipping the general practitioner Health policy 1998 45 221 38 10338953 10.1016/S0168-8510(98)00045-1
|
16271136
|
PMC1291363
|
CC BY
|
2021-01-04 16:29:13
|
no
|
BMC Fam Pract. 2005 Nov 4; 6:45
|
utf-8
|
BMC Fam Pract
| 2,005 |
10.1186/1471-2296-6-45
|
oa_comm
|
==== Front
BMC Fam PractBMC Family Practice1471-2296BioMed Central London 1471-2296-6-471627448110.1186/1471-2296-6-47Research ArticleReasons for and consequences of missed appointments in general practice in the UK: questionnaire survey and prospective review of medical records Neal Richard D [email protected] Mahvash [email protected] Victoria L [email protected] Debbie A [email protected] Owen [email protected] North Wales Clinical School Department of General Practice, University of Cardiff, Wrexham Technology Park, Wrexham LL13 7YP, UK2 Clinical Trials Research Unit, University of Leeds, 17 Springfield Mount, Leeds LS2 9NG, UK3 Centre for Research in Primary Care, University of Leeds, 71–75 Clarendon Road, Leeds LS2 9PL, UK4 Department of Social Medicine, University of Bristol, Canynge Hall, Whiteladies Road, Bristol, UK5 Lockwood Research Practice, 3 Meltham Road, Lockwood, Huddersfield, West Yorkshire, HD1 3XH, UK2005 7 11 2005 6 47 47 8 4 2005 7 11 2005 Copyright © 2005 Neal et al; licensee BioMed Central Ltd.2005Neal et al; licensee BioMed Central Ltd.This is an Open Access article distributed under the terms of the Creative Commons Attribution License (), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
Background
Missed appointments are a common occurrence in primary care in the UK, yet little is known about the reasons for them, or the consequences of missing an appointment. This paper aims to determine the reasons for missed appointments and whether patients who miss an appointment subsequently consult their general practitioner (GP). Secondary aims are to compare psychological morbidity, and the previous appointments with GPs between subjects and a comparison group.
Methods
Postal questionnaire survey and prospective medical notes review of adult patients missing an appointment and the comparison group who attended appointments over a three week period in seven general practices in West Yorkshire.
Results
Of the 386 who missed appointments 122 (32%) responded. Of the 386 in the comparison group 223 (58%) responded, resulting in 23 case-control matched pairs with complete data collection. Over 40% of individuals who missed an appointment and participated said that they forgot the appointment and a quarter said that they tried very hard to cancel the appointment or that it was at an inconvenient time. A fifth reported family commitments or being too ill to attend. Over 90% of the patients who missed an appointment subsequently consulted within three months and of these nearly 60% consulted for the stated problem that was going to be presented in the missed consultation. The odds of missing an appointment decreased with increasing age and were greater among those who had missed at least one appointment in the previous 12 months. However, estimates for comparisons between those who missed appointments and the comparison group were imprecise due to the low response rate.
Conclusion
Patients who miss appointments tend to cite practice factors and their own forgetfulness as the main reasons for doing so, and most attend within three months of a missed appointment. This study highlights a number of implications for future research. More work needs to be done to engage people who miss appointments into research in a meaningful way.
==== Body
Background
Missed appointments are important for patients and staff, and have a prevalence of 4.5–6.5% of booked consultations in the UK [1-3]. Previous studies have identified a number of socio-demographic and health factors including mental illness, associated with missing an appointment, but have had important methodological problems such as lack of control groups, cross-sectional design and small numbers [4-8]. We have recently demonstrated how health professionals in the UK blame missed appointments on patients' mistakes and forgetfulness, and how patients who miss appointments are viewed negatively [9]. Only one study has investigated patients' issues related to missing appointments; participants in this study did not feel obliged to keep an appointment in part because they felt disrespected by the health care system, an effect which was compounded by participants' lack of understanding of the scheduling system [10]. No studies have reported explanations as to why socio-demographic variables should be associated with missed appointments, or data regarding the outcomes of missed appointments. Hence there is an insufficient evidence base to deal with missed appointments in primary care [11,12]. Furthermore, research from secondary care in the UK or primary care in the US is unlikely to be generalisable to UK primary care [4].
The aims of this study are to: (a) determine the reasons for missed appointments; (b) determine whether patients who miss an appointment subsequently consult their GP for the problem or symptom that was going to be presented in the missed appointment and (c) to compare psychological morbidity, and the previous number of missed appointments between individuals who miss an appointment and a comparison group who did not.
Methods
Main study design
The study was conducted over a three week period in 2001 in seven practices in West Yorkshire. These differed in terms of size, location, annual consulting rates, organisation of clinical workload, and missed appointment rate. Whilst all were members of a research network, it has recently been demonstrated that research practices do not differ from others in terms of their demography or morbidity [13].
All adult patients who missed an appointment and patients in the comparison group (the next adult seen in the same surgery session to control for: doctor-related factors, climate factors and practice factors) were sent a pack within 24 hours of the appointment, comprising a short and neutral invitation letter, and a questionnaire with open and closed questions about the missed appointment (cases only) and appointments in general. Patients with severe mental impairment, learning difficulties, severe emotional distress, or terminal illness were excluded. Two reminders were sent to initial non-responders.
A second mailing was sent to responders comprising a GHQ-12 [14], and a consent form to review medical records. Multi-centre and local ethical approval was received.
Questionnaire
The first part of the questionnaire aimed to empower and engage the patients by asking their views on how practices could help patients keep appointments (data not reported in this paper). The second part (cases only) asked directly the reason for making the appointment, and the reasons for missing it. The third section (cases only) explored the reasons behind the missed appointment by asking their agreement with sixteen statements.
GHQ-12 and data from medical records
Returned GHQ-12s were scored according to the manual [14]. Patients scoring three or more were sent a letter explaining that they may be suffering from mental health problems and were advised to seek help from their GP. For those providing written consent, data were abstracted from medical records three months after the index appointment, relating to whether they consulted for the stated reason of the missed appointment, the number of days from the missed appointment to the next consultation, and the number of missed and kept appointments in the twelve months prior to the index appointment.
Sample size
The sample size calculation was determined by the difference in GHQ-12 scores between cases and the comparison group. In order to identify a high and low GHQ score difference of 20% between the groups at 5% significance level and 80% power, 200 pairs of patients and controls would be required. To allow for a likely response rate of approximately 60%, we estimated a sample size of 360 pairs.
Data analysis
Free text data relating to reasons for missing appointments were read independently by three of the researchers (MH-G, OD and RDN) and the main themes discussed. It was apparent that for each patient, there was one main reason. A categorisation emerged that was applied across the data. This was then refined further and the final categories agreed. Comparisons between patients who missed appointments and their controls were assessed using conditional multiple logistic regression for matched pairs. Data relating to the numbers of missed and kept appointments for patients who were not registered for the entire year prior to their missed appointment were adjusted to allow for the proportion of the year that they were registered and rounded to the nearest integer.
Results
Reasons for missing an appointment
Of the 386 patients who missed appointments 122 (31.6%) returned a valid questionnaire. Table 1 shows the responses to statements provided in the questionnaire concerning factors contributing to the missed appointment. Over 40% said that they forgot the appointment and a quarter that they tried very hard to cancel the appointment or that it was at an inconvenient time. A fifth reported family commitments or being too ill to attend. Table 2 shows the categorisation of free text reasons for each of the 122 respondents. The largest categories were 'misunderstandings and mistakes', 'illness or personal circumstances', 'forgetting', and 'other commitments'. Of the misunderstandings and mistakes, the largest sub-category was 'by the practice', and included 'being unable to get through', 'had cancelled', 'did not have an appointment', and 'told wrong time or date'.
Table 1 Responses to statements about the specific missed appointment
Item: N n Yes % (95% CI) answering Yes
I forgot about the appointment 89 44 49.4 (38.7, 60.3)
I tried very hard to cancel the appointment 81 24 29.6 (20.2, 40.8)
The appointment was at an inconvenient time 81 24 29.6 (20.0, 40.8)
I had family commitments 80 20 25.0 (16.0, 35.0)
I was too ill to attend 84 20 23.8 (15.2, 34.3)
The appointment wasn't with the doctor of my choice 76 12 15.8 (8.4, 2.6)
My GP asked me to come back on that day 78 11 14.1 (7.3, 2.4)
I was unable to get transport 78 11 14.1 (7.3, 2.4)
I overslept 79 11 13.9 (7.2, 2.4)
I did cancel before the appointmenta 82 10 12.2 (6.0, 21.3)
My symptoms were better / problems resolved 79 9 11.4 (5.3, 20.5)
I was unable to get time off work 77 6 7.8 (2.9, 16.2)
I was there and did not miss the appointmenta 80 5 6.3 (2.1, 14.0)
I was unable to get there because of the weather 78 2 2.6 (0.3, 8.9)
I was in hospital at the time 77 2 2.6 (0.3, 9.1)
I couldn't be bothered 76 1 1.3 (0.0, 7.1)
N: Number answering particular item; n Yes: number answering yes; CI: confidence interval
a No evidence for this when medical records reviewed for those who provided consent.
Table 2 Analysis of free text comments about the specific missed appointment
N
1. Misunderstanding and mistakes 36
By family member 2
By patient 14
By practice 20
2. Illness or personal circumstances 28
Self 17
Family 11
3. Forgot 26
No reason/excuse 17
Waited too long 2
Pre-occupied/distracted 7
4. Other commitments 9
5. Other 23
Overslept/late 6
Resolution of symptom 5
Not doctor of choice 2
Travel problems 2
No idea/don't know 3
Unclassifiable or blank 5
N: Number answering particular item; n Yes: number answering yes; CI: confidence interval a No evidence for this when medical records reviewed for those who provided consent.
Consulting after a 'missed' appointment
Of the 122 patients who missed an appointment and responded, 57 (47%) consented to medical records review. Age and gender distributions between those who gave consent and those who did not were similar (both p values > 0.2). Of these, 52 (91.2%) patients subsequently consulted within three months of the index consultation. One third were in the first week and over half were within three weeks. 33 (63.5%, 95% CI 49.0% to 76.4%) consulted for the stated problem that was going to present in the missed consultation. Nineteen patients (36.5%, 95% CI 23.6% to 51.0%) had no record of consulting for the same reason, and it was unclear for five patients.
Comparisons between cases and controls
The overall response rate among the comparison group (223, 58%) was greater than for cases, and there was a tendency for a greater proportion of the comparison group who responded, to also respond to the GHQ-12 (65.4% versus 50.8%), and to consent for medical record review (61.0% versus 45.1%). There were 27 case-control pairs who responded and provided data on the GHQ-12 and 23 pairs who responded and provided consent to medical records.
The odds of missing an appointment was lower for women compared to men (odds ratio (95% confidence interval) 0.67 (0.19, 2.36)), and the odds of missing an appointment decreased with increasing age (odds ratio (95% confidence interval) for an increase of 1 year in age 0.95 (0.91, 0.99)) though due to small numbers these differences were imprecise and the gender difference did not reach statistical significance at the conventional 5% level. The odds of missing an appointment were greater among those who had missed at least one appointment in the previous 12 months (5.88 (0.83, 41.36)), though again due to low response, this did not reach statistical significance at the conventional 5% level. This association was not substantively changed with adjustment for age, sex and the number of appointments made in the previous 12 months. There was a slight tendency for the odds of missing an appointment to be greater among those with a high GHQ-12 score (≥ 3), though because of small numbers the estimates were imprecise. With adjustment for age and gender the odds ratio (95% confidence interval) for missing an appointment comparing those with a high to a low GHQ-12 score was 1.14 (0.41, 3.15).
Discussion
We found that the commonest reasons that patients give for missing appointments are mistakes and misunderstandings (frequently by the practice) and forgetfulness. The majority of patients who missed an appointment consulted their GP within three months, with most of these consulting for the original problem that they were originally going to present. The low number of matched case-control pairs led to the comparisons between those who missed appointments and those who did not having wide confidence intervals which included 1.0, and despite some large effect sizes, for example a 33% reduced odds for a woman compared to a man of missing an appointment, these data cannot exclude the role of chance. However, reporting these results is important since the low response rate has important implications for future work in this area and highlights the importance of developing ways of engaging the disengaged in primary care research.
Study strengths and limitations
The age distribution of the comparison group was in keeping with national figures for those attending general practice, indicating that the practices participating in this study are representative of general practice as a whole [15]. One of the main strengths of this study is the investigation of an area which is important to primary care but which has been rarely investigated, perhaps because of anticipated difficulties. Further, we have attempted to improve upon previous studies with a prospective design, comparisons with a comparison group and use of a neutral place for correspondence. The main weakness of the study is the low response and the differential response between patients and the comparison group. A selection bias may exist between those who responded with respect to the outcomes of interest; this may be because of both social desirability bias and post-hoc rationalisations in their written responses. Responders may be over-represented by those whose missed appointment was a genuine practice mistake, and those who were less obliged to operate within the practices' booking systems being less likely to respond. Assuming that all 68% of non-responders did not miss their appointment because of forgetfulness or practice misunderstandings, then interventions aimed at dealing with these problems would only be able to reduce the overall missed appointment rate by a small amount.
Reasons for missing appointments
Within the category 'misunderstandings and mistakes', the largest sub-category was 'by the practice'. Further, in the response to the pre-defined categories, 30% stated that they had tried to cancel their appointment. A small number of patients reported that they had either cancelled their appointment or indeed kept their appointment. These findings were not verified by the medical records, therefore it is difficult to know whether these represent practice mistakes of post-hoc rationalisations of behaviour. Whilst this may reflect post-hoc rationalisations or may be exaggerated by selection bias, there is consistency between the two questions and the results suggest that improvements in practice communication systems could reduce missed appointments. Forgetfulness was another important cause. One practice strategy that may reduce this problem would be the use of aide-memoirs. However, intervention studies are required to demonstrate their effect. In a parallel study we have assessed the perceptions of primary care health professionals of the reasons for missing appointments [9]. Interestingly, practice staff tended to blame patients for missed appointments and dismissed the idea that practice factors may contribute. The results of this study suggest that practice factors contribute to some missed appointments suggesting that considering practice level interventions (e.g. making it easier to cancel appointments, and greater convenience of appointment times) may be useful.
Consequences of missed appointments
The finding that virtually all patients who missed an appointment did subsequently consult within a three month period has not been previously demonstrated. In fact, over half of those who responded consulted within the following two weeks of the missed appointment. Again selection bias may have influenced this result if those who did not respond were different to those who did and were less likely to subsequently attend. This study was unable to determine whether those with psychiatric morbidity are more likely to miss appointments; and whilst further research in this area is required to determine this, a recent paper has found that psychological morbidity did affect the likelihood of missing appointments [7].
Engaging with the disengaged
There are several implications of these results for future research. This work adds to the literature confirming that more work needs to be done to engage people who miss appointments with research in a more meaningful way. Creative approaches, in terms of accessing and engaging patients who have missed appointments in research, and of using appropriate methodologies to answer important questions need developing; qualitative methods may be best at exploring the phenomenon from user perspectives and would help to unravel the complex relationships surrounding consultation etiquette and patients' and practitioners' behaviour. Closer involvement of users might inform this process [16], as may the use of inducements. One option would be to recruit patients when they next consult in primary care, as we have shown the vast majority do consult again soon. If research was then conducted in a neutral environment, the use of practice members to recruit participants need not bias the responses. The lessons that we have learned have implications for future research for others who are developing work with other 'disengaged' or potentially vulnerable groups of patients. There is still a need for health professionals to maintain their concern for the health of patients who miss appointments, since this study does not exclude the possibility that they may have unmet mental health problems.
Conclusion
Patients who miss appointments tend to cite practice factors and their own forgetfulness as the main reasons for doing so, and most attend within three months of a missed appointment. This study highlights a number of implications for future research. More work needs to be done to engage people who miss appointments into research in a meaningful way.
List of abbreviations used
GP – General practitioner; GHQ – General Health Questionnaire
Competing interests
The author(s) declare that they have no competing interests.
Authors' contributions
The study was designed by RDN, DAL, VLA and OD. MH-G conducted the fieldwork under the direction of RDN. Data analysis was conducted by DAL and VLA. RDN and MH-G drafted the paper, which has been seen and critically revised by all co-authors.
Pre-publication history
The pre-publication history for this paper can be accessed here:
Acknowledgements
We would like to thank patients and staff at the following practices for their co-operation with this work: The Ridge Medical Practice, Bradford; Drs Eisner, Goldman and Ship at Shipley Health Centre; Mixenden Stones Surgery, Halifax; Wilsden Health Centre, Bradford; Fieldhead Surgery, Huddersfield; Meanwood Group Practice, Leeds; Woodhouse Health Centre, Leeds; and Kilmeny Surgery, Keighley. We would also like to thank other members of YReN, who contributed to the study at various points. The study was funded by the NHS Northern and Yorkshire R&D Directorate. DAL is funded by a UK Department of Health Career Scientist Award. The views expressed in this publication are those of the authors and not necessarily those of the funding bodies.
==== Refs
Neal RD Lawlor DA Allgar VL Colledge M Ali SM Hassey GA Wilson A Portz C Missed appointments in general practice: retrospective data analysis from four practices Br J Gen Pract 2001 51 830 2 11677708
Waller J Hodgkin P Defaulters in general practice: who are they and what can be done about them? Fam Pract 2000 17 252 253 10846145 10.1093/fampra/17.3.252
Eve R Hodgkin P Kilner K Waller J Did not attends: who are they and what can we do about them? Practice data comparison project 1999 Sheffield: Centre for Innovation in Primary Care
Oppenheim GL Bergman JJ English EC Failed appointments: a review J Fam Pract 1979 8 789 796 429996
Inglesfield J Non-attendance and mental health problems in primary care Br J Gen Pract 1999 49 488 10562759
Cosgrove MP Defaulters in general practice: reasons for default and patterns of attendance Br J Gen Pract 1990 40 50 52 2107849
Cashman SB Savageau JA Lemay CA Ferguson W Patient health status and appointment keeping in an urban community health centre J Health Care Poor & Underserved 2004 15 474 88 15453182
Weingarten N meyer DL Scheid JA Failed appointments in residency practices: who misses them and what providers are most affected? J Am Board Fam Pract 1997 10 407 11 9407481
Hussain-Gambles M Neal RD Dempsey O Lawlor DA Hodgson J Missed appointments in primary care: questionnaire and focus group study of health professionals Br J Gen Pract 2004 54 108 113 14965389
Lacy NL Paulman A Reuter MD Lovejoy B Why we don't come: patient perceptions on no-shows Ann Fam Med 2004 2 541 545 15576538 10.1370/afm.123
Sharp DJ Hamilton W Non-attendance at general practices and outpatient clinics BMJ 2001 323 1081 2 11701560 10.1136/bmj.323.7321.1081
George A Rubin G Non-attendance in general practice: a systematic review and its implications for access to primary health care Fam Pract 2003 20 178 184 12651793 10.1093/fampra/20.2.178
Hammersley V Hippisley-Cox J Wilson A Pringle M A comparison of research general practices and their patients with other practices – a cross-sectional survey in Trent Br J Gen Pract 2002 52 463 468 12051210
Goldberg DP Williams P A user's guide to the General Health Questionnaire 1988 Windsor: NFER-Nelson
Royal College of General Practitioners Office of Population Censuses and Surveys, and Department of Health. Morbidity statistics from general practice – fourth national study 1995 London: HMSO
Hanley B Involving consumers in research and development in the NHS 2000 Winchester: Consumers in the NHS Research Support Unit
|
16274481
|
PMC1291364
|
CC BY
|
2021-01-04 16:29:13
|
no
|
BMC Fam Pract. 2005 Nov 7; 6:47
|
utf-8
|
BMC Fam Pract
| 2,005 |
10.1186/1471-2296-6-47
|
oa_comm
|
==== Front
BMC GenomicsBMC Genomics1471-2164BioMed Central London 1471-2164-6-1541627115010.1186/1471-2164-6-154Research ArticleCharacterization of SR3 reveals abundance of non-LTR retrotransposons of the RTE clade in the genome of the human blood fluke, Schistosoma mansoni Laha Thewarach [email protected] Nonglack [email protected] Alex [email protected] Paul J [email protected] Department of Parasitology, Faculty of Medicine, Khon Kaen University, Khon Kaen 40002, Thailand2 Division of Infectious Diseases & Immunology, Queensland Institute of Medical Research, Brisbane, Queensland, 4029, Australia3 Department of Tropical Medicine, and Center for Infectious Diseases, Tulane University, Health Sciences Center, New Orleans, Louisiana, 70112, USA2005 4 11 2005 6 154 154 13 5 2005 4 11 2005 Copyright © 2005 Laha et al; licensee BioMed Central Ltd.2005Laha et al; licensee BioMed Central Ltd.This is an Open Access article distributed under the terms of the Creative Commons Attribution License (), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
Background
It is becoming apparent that perhaps as much as half of the genome of the human blood fluke Schistosoma mansoni is constituted of mobile genetic element-related sequences. Non-long terminal repeat (LTR) retrotransposons, related to the LINE elements of mammals, comprise much of this repetitive component of the schistosome genome. Of more than 12 recognized clades of non-LTR retrotransposons, only members of the CR1, RTE, and R2 clades have been reported from the schistosome genome.
Results
Inspection of the nucleotide sequence of bacterial artificial chromosome number 49_J_14 from chromosome 1 of the genome of Schistosoma mansoni (GenBank AC093105) revealed the likely presence of several RTE-like retrotransposons. Among these, a new non-LTR retrotransposon designated SR3 was identified and is characterized here. Analysis of gene structure and phylogenetic analysis of both the reverse transcriptase and endonuclease domains of the mobile element indicated that SR3 represented a new family of RTE-like non-LTR retrotransposons. Remarkably, two full-length copies of SR3-like elements were present in BAC 49-J-14, and one of 3,211 bp in length appeared to be intact, indicating SR3 to be an active non-LTR retrotransposon. Both were flanked by target site duplications of 10–12 bp. Southern hybridization and bioinformatics analyses indicated the presence of numerous copies (probably >1,000) of SR3 interspersed throughout the genome of S. mansoni. Bioinformatics analyses also revealed SR3 to be transcribed in both larval and adult developmental stages of S. mansoni and to be also present in the genomes of the other major schistosome parasites of humans, Schistosoma haematobium and S. japonicum.
Conclusion
Numerous copies of SR3, a novel non-LTR retrotransposon of the RTE clade are present in the genome of S. mansoni. Non-LTR retrotransposons of the RTE clade including SR3 appear to have been remarkably successful in colonizing, and proliferation within the schistosome genome.
==== Body
Background
Schistosomiasis is considered among the most important of the tropical diseases in terms of morbidity and mortality, ranking only behind malaria [1]. International efforts are underway to sequence the entire genomes of two of the three major schistosome species, S. mansoni and S. japonicum [2]. It is anticipated that an enhanced understanding of the schistosome genome will aid in the control of this disease, including the development of vaccines and new anti-parasite medications [3]. Up to half of the schistosome genome may be composed of repetitive sequences, including LTR and non-LTR retrotransposons, mobile genetic elements that transpose through an RNA intermediate (reviewed by Brindley et al. [4]). Mobile genetic elements are drivers of genome evolution [5,6]. In addition to this role, from a practical perspective mobile genetic elements offer potential as transgenesis vectors [7]. Problematically, however, their interspersed, repetitive nature can impede progress during genome sequencing using shotgun sequencing approaches through the confounding effects of their repetitions on sequence assembly algorithms [8,9]. For these and other reasons, we and others have been characterizing the retrotransposons of the schistosome genome [10-15]. Here we report a novel non-LTR retrotransposon termed SR3, a member of the RTE clade of non-LTR retrotransposons, from the genome of S. mansoni. Based on the multi-copy, interspersed nature of SR3, and the presence of other RTE elements characterized previously from the genomes of schistosomes, it appears that members of the RTE clade may be the most common and successful of the non-LTR retrotransposons to have colonized the genomes of these metazoan parasites.
Results and Discussion
New retrotransposons identified in bacterial artificial chromosome 49_J_14 from the genome of S. mansoni
BLASTn searches revealed the presence of reverse transcriptase (RT)-encoding sequences in the S. mansoni bacterial artificial chromosome (BAC) number 49_J_14 [16], the entire sequence of which has been deposited in GenBank with accession number AC093105 by El Sayed and co-workers [3]. Annotation provided with GenBank AC093105 indicated that the sequence included in BAC 49_J_14 is from chromosome 1 of the genome of S. mansoni. Inspection of the nucleotide sequence of BAC 49_J_14, of ~123 kb in length, indicated the presence of a number of discrete retrotransposons. One of these encodes a novel long terminal repeat (LTR) retrotransposon, which we have described in a recent report [11] (Fig. 1). In addition, at least three non-LTR retrotransposons appeared to be located in BAC 49_J_14. One of these appeared to be a degenerate copy of an SR2 element. SR2 elements are non-LTR retrotransposons of the RTE clade [17] which are present in high copy (estimated at up to 10,000 copies) in the genome of S. mansoni [10,18]. This fragment of SR2 was located between nucleotide residues numbers 11,176 and 13,119 of BAC 49_J_14 and, more specifically appeared to be located within intron number 1 of the gene encoding cytosolic Zn/Cu superoxide dismutase [19]. As illustrated in Fig. 1, the Cu/Zn superoxide dismutase gene is present in BAC 49_J_14 between residues 8,020 and 16,898 of BAC 49_J_14. The copy of SR2 in the intron of the Cu/Zn superoxide dismutase gene is ~1,830 nucleotides (nt) in length, and included regions encoding the retrotransposon RT domain (Fig. 1). Over the putative RT-encoding region, the sequence was 47% identical to the RT sequence of SR2. At only ~1.8 kb in length, and since full-length copies of SR2 are ~3.9. kb in length [18], this appears to be a truncated copy of SR2 that is unlikely to be autonomously mobile. In like fashion to the location of this truncated copy of SR2, copies of other SR2 elements (and indeed other retrotransposons) have been identified previously in introns of other S. mansoni protein encoding genes [20,21].
Figure 1 Schematic diagram of the location, size and structure of copies of the SR3 non-LTR retrotransposon in bacterial artificial chromosome number 49_J_14 from chromosome 1 of the genome of Schistosoma mansoni. The location of copies of the SR2 and fugitive retrotransposons is also presented. The arrows indicate the direction of transcription of the mobile elements. The numbers on each bar indicate the nucleotide position of the elements within the bacterial artificial chromosome. A degenerate copy of the SR2 element was evident within intron 1 of Zn-Cu superoxide dismutase (SOD) gene, in particular within intron 1 of the gene: three exons of the SOD are indicated in red, with the position and structure of the degenerate copy of SR2 indicated within intron 1. On the bottom right, a schematic of the length and domain structure of the SR3-right copy of the SR3 retrotransposon is presented. RT, reverse transcriptase; EN, endonuclease.
SR3 represents a new family of non-LTR retrotransposon from the genome of S. mansoni
In addition to the fugitive LTR retrotransposon [11], and the truncated copy of SR2, at least two other retrotransposons were readily identifiable in BAC 49_J_14. The first of these was located between nt 346 and 3,552 (i.e., 3,207 bp in length), and the second between nt 97,832 and 101,042 (3,211 bp in length). Comparison of the sequences of these two prospective retrotransposons revealed that they were closely related to one another and appeared to represent discrete copies of a novel family of retrotransposons. We have termed the new retrotransposon SR3, whose phylogenetic analysis indicated a new family of the RTE clade of non-LTR retrotransposons (see below). (SR3 stands for Schistosome Retrotransposon 3 because two other non-LTR retrotransposons described previously from S. mansoni are termed SR1 and SR2 [18,22]). (A recent article, published after this present report was submitted for publication, identified a SR3-like element in the S. mansoni transcriptome, termed Perere-3, and also identified several other novel retrotransposons [15].) For convenience of description, we refer here to the copy of SR3 resident between nt 346 and 3,552 of BAC 49_J_14 as SR3-left and the other copy between nt 97,842 and 101,042 as SR3-right, because they are located on the left and right sides of the BAC as in Figure 1. The full-length SR3-left and SR3-right elements were comprised of a single, read through open reading frame (ORF) encoding two functional domains similar to apurinic-apyrimidic (AP) endonuclease (EN) and RT, in that order. The element terminated with a short repeat sequence, (TAAG)4 or (TAAG)5 (Fig. 1). The nucleotide and deduced amino acid sequences of the SR3-left and SR3-right copies are provided in Additional files 1 and 2, respectively.
The sequence of 3,211 bp long SR3-right element translated into a single, deduced open reading frame (ORF) of 922 amino acid residues that did not include any apparent frameshift or stop codon mutations (Additional file 2). By contrast, the deduced ORF of SR3-left was interrupted by stop codons at amino acid positions 719 and 913 of the ORF (Additional files 1, 3). SR3-right has a longer terminal repeat unit than SR3-left, (TAAG)5 compared with (TAAG)4, which accounts for the difference in total lengths of the two copies (3,207 and 3,211 bp). (By contrast, comparison of the ORFs of Perere-3 (Accession CAJ00236.1) and the SjR2 retrotransposon (AY027869) of S. japonicum, with the deduced ORFs of both SR3-left and SR3-right revealed that the similarity extends well beyond the predicted ORF of 922 deduced amino acids of SR3-right [not shown]. Whereas this suggests the possibility of premature stop codon in the SR3 copies presented here, it may also simply reflect phylogenetic relatedness in the carboxy-terminal encoding regions and 3'UTRs of these elements.) Nonetheless, SR3-left and SR3-right are very similar to each other in sequence, with the ORFs region exhibiting 94 % identity and 97 % similarity over the predicted ORF of 922 residues (Additional file 3). Together, these findings suggest that both SR3-left and SR3-right are full-length copies and, moreover, that SR3-right is an intact, putatively functional and active copy, capable of autonomous retrotransposition activity. It was remarkable not only that two copies (SR3-left and SR3-right) of this retrotransposon reside in close proximity to each other in the region of the S. mansoni genome represented by BAC 49_J_14, but also that both copies are full-length and intact or close to intact. Most copies of non-LTR retrotransposons are 5'-truncated, due to deficits in their elongation processes, and generally include deletions or insertions (indels), and are thereby rendered inactive [6,23,24].
Four other non-LTR retrotransposons have been reported from the genome of S. mansoni. These are SR1 and Perere, discrete members of the CR1 clade, and SR2 and Perere-3, members of the RTE-1 clade [14,15,18,22]. SR3 was dissimilar to these non-LTR retrotransposons reported previously from the genome of S. mansoni: when compared with the deduced amino acid sequence of the ORF of SR3, SR1 shared 23 %/ 38 % amino acid sequence identity/similarity with SR3, Perere shared 22 %/35 % identity/similarity, SR2 shared 39 %/55 % identity/similarity and Perere-3 shared 78 %/88 % amino acid sequence identity/similarity with SR3 (not shown). Together, these differences indicated that SR3 was a novel element distinct from these other schistosome non-LTR retrotransposons.
SR3 represents a new member of a family of the RTE-1 non-LTR retrotransposons
The predicted RT domain of SR3 was aligned with orthologous domains of numerous other non-LTR retrotransposons including representatives from 11 clades of non-LTR retrotransposons, as defined by Eickbush and colleagues [25,26]. Phylogenetic comparison of the RT domains of these diverse elements revealed that the closest relatives of SR3 were ShR3 from S. haematobium and Perere-3 from S. mansoni, with close identity also to AC150430 element from Branchiostoma floridae, SR2 from S. mansoni, SjR2 from S. japonicum and also to RTE-1 from Caenorhabditis elegans (Figure 2; and Additional file 4), placing SR3 in the RTE-1 clade of non-LTR retrotransposons. In like fashion, a phylogenetic tree was constructed based on the EN domain of eight clades of non-LTR retrotransposons. The topography of the EN tree, and the position of SR3 within the RTE clade, was similar to the topography represented on the RT-based tree, confirming both the inclusion of SR3 as an RTE clade element and that SR3 and SR2 were discrete families of RTE-like retrotransposons (Figure 3; and Additional file 5). Indeed, in the EN tree, SR3 was more closely related to RTE-1 of C. elegans than to SR2 of S. mansoni (Figs. 2, 3).
Figure 2 Phylogram constructed using the neighbor-joining method to compare the relationships among reverse transcriptases of SR3 and of representative elements belonging to the major clades of non-LTR retrotransposons [25] from a range of host genomes. Representatives of 11 clades of non-LTR retrotransposons including the RTE, CR1, L1, R1 and Jockey clades were included in the analysis. Bootstrap values, where 500 or greater from a maximum of 1,000 replicates, are presented at the nodes.
Figure 3 Phylogram constructed using the neighbor-joining method to compare the relationships among endonucleases of SR3 and of representative elements belonging to the major clades of non-LTR retrotransposons [25] from a range of host genomes. Representatives of eight major clades of non-LTR retrotransposons including the RTE, L1, CR1, Jockey, and I clades were included in the analysis. Bootstrap values of 500 or greater from 1,000 replicates are presented at the nodes.
Structure of SR3
Youngman et al. [27] provided the first report of a RTE retrotransposon, from the genome of C. elegans. RTE clade elements display a broad host range, having been described from numerous invertebrate and vertebrate taxa, and from algae and flowering plants [14,15,17,18]. RTE-1 encodes a 1,066-amino-acid ORF containing both apurinic-apyrimidic endonuclease and reverse-transcriptase domains. A possible first ORF of only 43 amino acids overlaps with the larger ORF and may be the site of translation initiation. Members of the RTE clade are characterized by unusually short 3' untranslated regions that are predominantly composed of AT-rich trimer, tetramer, and/or pentamer repeats [17]. RTE-derived SINE elements are also found in mollusc and flatworm genomes.
In addition to the demonstration by phylogenetic analyses targeting both the RT and EN domains that SR3 is an RTE like element, we compared the structural motifs and domains of SR3 with RTE-1 of C. elegans and SR2 of S. mansoni in order to confirm the identity of SR3 as an RTE clade non-LTR retrotransposon. First, the three elements were of generally similar length; 3,291 bp for RTE-1 of C. elegans [17], 3,913 bp for SR2 [18], and 3,211 kb for SR3-right. Second, the length of the ORF was somewhat similar; 1066, 1016, and 922 amino acids for RTE-1, SR2, and SR3 respectively. The RTE-1 and SR2 elements may also contain a short ORF upstream of the major ORF, although this has not been confirmed by functional analysis [17,18,25]. Third, the 3'-UTRs of RTE clade elements are usually short in length and terminate in several tetrameric or pentameric, A-rich repeats. SR3 conformed to RTE-1 in this regard, with SR3 exhibiting a short 3'-UTR of 177 bp in length and terminating with several copies of the tetramer, TAAG (Fig. 1; Additional files 1, 2).
A schematic comparison of the structures of RTE-1, SR2, SR3, CR1, and an SR1-like element, Perere-5 [15,22], is presented in Figure 4. In summary, the SR3 elements of S. mansoni conform in all respects to the generalized structure of the RTE clade of non-LTR retrotransposons. Moreover, as with other RTE elements, SINE-like elements reported from schistosomes may be derived from SR3-like elements [4].
Figure 4 Schematic representation of the structure of non-LTR retrotransposons of the RTE and CR1 clades, including RTE-1 from Caenorhabditis elegans, SR2 and SR3 from Schistosoma mansoni, CR1 from Gallus gallus, and Perere-5, and SR1-like retrotransposon from S. mansoni. (The structure of a full-length copy of SR1 has not been reported [15, 22].) The sequence motifs of the 3'-termini are shown, along with positions of enzymatic domains, EN (endonuclease) and RT (reverse transcriptase). The RTE-1 element (3291 bp) illustrated here includes the 3'-UTR so that it is longer than the 3258 bp, described in Malik and Eickbush [17] (which included only the region between the 5'-end and the termination codon).
SR3 is present in genomes of other schistosome species
Investigation of SR3 sequences in the genomes of other human schistosomes by BLAST search analysis revealed many sequences similar to SR3 in the transcriptomes of S. japonicum (e.g., GenBank AY810372, AY915175, AY813885 and AY915893). In addition, when the nucleotide sequence of SR3-right was employed as the query in BLASTx analysis against the GenBank non-redundant database, SR3-like sequences were identified within introns 1 and 6 of the gene encoding S. haematobium acetylcholinesterase (AChE) (GenBank AY167025) [28]. The two copies are similar in sequence (~70% identical), both copies are 5' truncated, and both include regions encoding the RT domain of the retrotransposon (not shown). The fragment within intron 1 was located between nt 1,023–2,474, and the fragment in intron 6 was located between nt 18,742–20,658. The predicted RT domain of the SR3 like element from S. haematobium (termed ShR3) was included in the phylogenetic tree presented in Fig. 2 and was found to be phylogenetically similar to SR3 from S. mansoni. The presence of SR3 elements in other schistosome species can be explained by vertical transmission from a progenitor schistosome species [29], given that vertical transmission is the expected route of transmission of non-LTR retrotransposons [24].
Numerous copies of SR3 are interspersed throughout the genome of S. mansoni
Southern hybridization analysis revealed that multiple bands of digested genomic DNA of S. mansoni hybridized to the SR3 specific probe, indicating the presence of numerous copies of SR3 in the S. mansoni genome (Fig. 5, lanes 1 and 2). Hybridization to the gDNA fragments released by double enzyme digestions revealed an even more smeared pattern (Fig. 5, lanes 3, 4), clearly suggesting that SR3 elements have interspersed throughout the genome of S. mansoni. In addition, a bioinformatics analysis using the approach of Copeland et al. [13] was used to estimate copy number of SR3 by comparisons with reference copy number estimates of other mobile genetic elements and genes reported previously. BLASTn searches were undertaken using the nucleotide sequences of these reference genes and the complete nucleotide sequence of SR3-right. Because the construction of the S. mansoni BAC library (from which BAC 49_J_14 was isolated) involved partial digestion of the genomic DNA with Hind III [16], genes without Hind III sites will be underrepresented in the BAC end sequences. Accordingly, since sequenced BAC ends from this library constitute a large proportion of the genomic S. mansoni sequences in the public domain, we used only genes containing Hind III sites as reference sequences. As shown in Table 1, the number of hits for SR3, 110, was higher than the number of hits for the single-copy cathepsin D gene (0 hits) and for three high copy number retrotransposons Boudicca (100 hits, 1,000–10,000 reported copies), SR2 (102 hits, 1,000–10,000 copies), and SR1 (104 hits, 200–2,000 reported copies) but lower than that for the multiple-copy 28S ribosomal RNA gene (157 hits) (100–200 copies). Although it is difficult with these available data to obtain a good estimate of the number of copies, however a comparison with the other 3 retrotransposons would give a tentative copy number for SR3 of between 1,000 and 10,000.
Figure 5 Southern hybridization analysis of genomic DNA of S. mansoni probed with a SR3 retrotransposon specific probe. Genomic DNA was cleaved with endonucleases Hind III (lane 1), BamH I (lane 2), EcoR I plus Xba I (lane 3) and Hind III plus Xho I (lane 4). Molecular size standards in kilobase pairs (kb) are shown at the left. (Lanes 1 and 2 were exposed to film longer than lanes 3 and 4.)
Table 1 Estimation by bioinformatics approaches of gene copy number of the SR3 non-LTR retrotransposon in the genome of Schistosoma mansoni.
Gene GenBank Accession Query Length (bp) Number of hits (Expect 0.000001) Copy number Key references
Cathepsin D, Intron 4 AY309267 1636 0 1 [20]
28S rRNA Z46503 1694 157 100–200 [46]
Sinbad AY506538 6288 38 50 [13]
Boudicca AY662653 5858 100 1,000–10,000 [12]
SR3 3211 110 >1,000 This study
SR2 AF025672 3913 102 1,000–10,000 [18]
SR1 U66331 2337 104 200–2,000 [22]
Saci-2 BK004069 4946 107 85–850* [14]
Saci-1 BK004068 5980 133 70–700* [14]
SR3 is transcribed in all developmental stages of S. mansoni
The nucleotide sequences of the full length of SR3-left and SR3-right elements were employed as query sequences for BLAST searches of the GenBank EST database of S. mansoni sequences. The database includes more than 160,000 EST sequences from six developmental stages of S. mansoni – egg, miracidium, cercaria, germball (= sporocyst), schistosomulum, and mixed sex adults [30]. Significant hits were found to ESTs from all six of these stages (not shown). Representative accession numbers of the positive matches are presented in Additional files 6 and 7, along with brief details of the regions where the matches were located and statistical significance of the matches. In brief, positive ESTs spanning all of the 5'UTR, 3'UTR, EN and RT were located in most of these developmental stages. Based on these findings, it appeared that SR3 was expressed in developmental stages throughout the life cycle of S. mansoni.
SR3 integration sites
In order to investigate the nature of integration sites or target sequences of the new retrotransposon within the schistosome genome, five kilobases of nucleotide sequences flanking the 5'- and 3'-UTRs of both SR3-left and SR3-right were employed as queries to search the GenBank non-redundant nucleotide and protein databases, and the GSS and EST entries for S. mansoni. These searches revealed no significant matches to any sequences encoding genes of Schistosoma species (not shown). However, they did reveal that SR3 elements appear to target AT-rich sites, as indicated in Figure 6, a similar preference to L1 retrotransposons within the human genome [31,32]. More specifically, the average AT content of the integration sites of the 21 copies of SR3 shown in Figure 6 was 68 % AT. Whereas target site specificity does not appear to be stringent for SR3, it can be expected to reflect the recognition sequence of the SR3 endonuclease. For example, L1 elements apparently integrate at numerous sites in the genome because the endonuclease of L1 preferably cleaves DNA at the short consensus sequence, 5'-TTTT/A-3', where/designates the cleavage site [31,33].
Figure 6 Multiple sequence alignment of the nucleotide sequences flanking the insertion sites of copies of the SR3 non-LTR retrotransposon within the genome of Schistosoma mansoni. Sequences flanking the 5'UTR of SR3-left and SR3-right are aligned in the top panel, while those flanking the 3'-terminus of SR3-left and SR3-right are presented in the bottom panel. Target site duplications are evident at the sites of SR3-left and SR3-right integration, and are indicated with bold font. Conservation of residues is indicated by the shading of boxes. Target sequences were identified among entries in the GSS database of S. mansoni sequences at GenBank or the Sanger Institute [39, 40].
To propagate, non-LTR retrotransposons employ their EN and RT enzymes respectively to nick a genomic target site and reverse transcribe the retrotransposon, integrating the element into a new genomic locus [33-35]. This process is termed target-site-primed reverse transcription. For the L1 elements in the human genome, a new L1 insertion is usually flanked by short direct repeats derived from the target DNA locus upon L1 integration [32,36]. These repeats are called target site duplications (TSDs), and can range from several to several hundred nucleotides in length [32,37]. Interestingly, both SR3-left and SR3-right are flanked by TSDs of 10 and 12 bp, respectively; TAGTGGCTAATCT for SR3-right and CGCTCTTAAA for SR3-left (Fig. 6). The presence of these TSDs provides further indication, along with their intact structure, of recent activity of these two copies of SR3 localized in BAC 49_J_14 [see [32]]. Apparently unlike SR3, and certainly unlike L1, some other clades of non-LTR retrotransposons exhibit extreme target site specificity, the well-known examples being the R2 and R4 elements which are found exclusively in the ribosomal RNA genes of insects (e.g., Bombyx mori, Drosophila melanogaster) and nematodes (e.g., Ascaris lumbricoides) or in simple repeats (e.g., the Dong element from B. mori) [25].
Nonetheless, as noted above, we have detected the presence of SR3 of S. haematobium within introns of the AChE gene [28], and in addition, other RTE elements have been reported from gene-rich sites of the schistosome genome. The degenerate copy of a non-LTR retrotransposon, SR2 [18] in BAC clone BAC 49_J_14 has integrated into intron 1 of the Zn-Cu superoxide dismutase (SOD) gene of S. mansoni (Figure 1). SR2 from schistosomes has been recorded from several other target genes including 28 kDa glutathione S transferase [18], cathepsin D [20] and the UTR of heat shock protein 70 [10]. Furthermore, the RTE-1 retrotransposon of C. elegans was found inserted in the intron of pim related kinase-1 (prk-1) gene [27]. Thus, although SR3 and other RTE clade retrotransposons do not exhibit tight target site specificity, they seem to prefer to integrate into AT-rich sites and, in addition, are frequently found in introns and other-non coding areas of protein encoding gene loci.
Conclusion
A new non-LTR retrotransposon, SR3, is reported from the genome of the human blood fluke Schistosoma mansoni. Numerous copies of SR3 are interspersed throughout the S. mansoni genome, and given the apparently intact sequence of the SR3-right copy of SR3 located in BAC 49_J_14 and the presence of transcripts from at least six developmental stages of S. mansoni, SR3 appears to be an active or recently active retrotransposon. This element is also present in the related human schistosomes, S. haematobium and S. japonicum. Based on phylogenetic comparisons of both the reverse transcriptase and endonuclease domains, SR3 represents a distinct family of RTE elements, discrete from the SR2 family described previously from schistosomes [18]. While there are numerous non-LTR retrotransposons in the schistosome genome, most elements so far described belong either to the RTE clade or CR1 clades [4], both of which are considered to be more advanced clades of non-LTR retrotransposons with progressive features including lack of target site specificity and an ORF encoding endonuclease and reverse transcriptase, respectively [25]. The presence of these and the apparent absence of some other clades of non-LTR retrotransposons should be informative in understanding the influence of mobile genetic elements in shaping the schistosome genome and its evolution and in studies of the phylogeny of schistosomes and related taxa. Finally, for studies with transgenesis of schistosomes, it may be possible to adapt an active copy of SR3 – such as SR3-right – for the introduction of transgenes into the schistosome genome in similar fashion to the adaptation of L1 elements of humans for studies on the movement of LINE elements in cultured human cell lines [23,37,38].
Methods
Bioinformatics approaches for detection of mobile sequences in the schistosome genome
The keyword phrase <Reverse Transcriptase> was used as the query to search the EST_others and GSS databases at GenBank for novel schistosome sequences associated with mobile genetic elements. Schistosome RT-like sequences that were retrieved were employed subsequently to search for matches in the GenBank non-redundant sequence database using BLASTn, BLASTx and/or tBLASTn [39]. Sequences of the previously characterized schistosome retrotransposons including Gulliver, pido, SjR2 of S. japonicum [4] and Boudicca [12] also were employed as queries. In addition, retrotransposon integration sites were investigated by interrogation of the S. mansoni genome survey sequences (GSS) at the Sanger Institute, Hinxton, U.K [40].
Parasites and parasite DNA
The life cycle of Schistosoma mansoni (NMRI strain, of Puerto Rican origin) was maintained at the Queensland Institute of Medical Research, Brisbane, Australia using experimentally infected mice and albino Biomphalaria glabrata snails. Genomic DNAs (gDNAs) of adult mixed sex parasites perfused from mice and cercariae (shed from snails) of S. mansoni were extracted using Qiagen's Genomic Tip-100 system according to the manufacturer's instructions.
Southern hybridization
Thirty micrograms of S. mansoni gDNA was cleaved with restriction enzymes Hind III, EcoR I, BamH I and Xho I. Digested gDNA was fractionated through 0.8% agarose gel and then was transferred to nylon membrane (Hybond-N+, Amersham Biosciences) by capillary action [41]. Southern hybridization analysis was performed using a horseradish peroxidase labelled probe and the ECL detection system (Amersham Biosciences). The membrane was incubated in hybridization medium (provided with kit) supplemented with the labeled probe overnight at 42°C, after which the membrane was washed in 0.4% SDS, 0.5× SSC at 42°C (two washes, 20 min. each) and subsequently in 2× SSC at room temperature (two washes, 5 min. each). The retrotransposon-like gene probe was amplified by polymerase chain reaction (PCR) with specific primers using S. mansoni gDNA as a template. Specific primers targeting the amplification of the RT domain of SR3 were SR3-forward, 5'-GAAGATTTGGGAAGAGGAACA and SR3-reverse, 5'-AACGATGCTCCCCAGATAAT (spanning nucleotides 1,809–2,622, Additional file 1). The SR3-right gene probe was amplified using specific primer SR3 forward (same as for the SR3-left probe) and SR3-right reverse 5'-CAACGATGCTCCCCAGGTACTTG (nt 1,809–2,622). Probes were sized in gels, isolated and purified before use. These probe sequences have been assigned GenBank accession numbers DQ008120 and DQ008121 for the SR3-left- and SR3-right-based probes, respectively.
Sequence analysis and phylogenetic analysis of new retrotransposons
The amino acid sequences of the functional domains of both RT and EN of both copies of the new non-LTR retrotransposon were aligned to other non-LTR retrotransposons by the ClustalW method [42] using BioEdit software [43] and optimized gaps and errors were referenced to conserved domains defined by Malik et al. [25]. Edited sequence alignments of the RT and EN domains were analyzed for phylogenetic relationships using the PHYLIP package [44]. Phylograms were generated and assessed for bootstrap values of 1,000 replicates using the neighbor-joining method with assistance from SEQBOOT and NEIGHBOR in the PHYLIP software suite [44]. Trees were displayed by TreeView [45]. Sequences used in the phylogenetic analyses were obtained from the GenBank, EMBL and PIR databases. They included family representatives from 11 major clades of non-LTR retrotransposons [25]. RT sequences of Group II introns from bacteria and EN sequences from bacteria were used as outgroups for the RT and EN trees, respectively. The names and accession numbers of the aligned sequences were: SR1 (U66331), SR2 (AF025672), Perere (BK004067) and Perere 3 (BN000794) from S. mansoni, SjR2 (AY027869) and pido (AY034003) from S. japonicum, ShR3 (AY167025) from S. haematobium, RTE1 from C. elegans (AF054983), JAM1 (Z86117) and Lian (U87543) from Ae. aegypti, Bov-B LINE from Vipera ammodytes (AF332697), Branchiostoma floridae clone CH302-99K22 (AC150430), BDDH from Bos taurus (AC150753), BCNT from Tragulus javanicus (AB191483), ENSANGP00000028171 from Anopheles gambiae strain PEST (XM556470), Tx1 from Xenopus laevis (M26915), Swimmer from the medaka fish, Oryzias latipes (AF055640), L1 from the rat (U83119), L1 from the mouse (AF081114), L1 from the human (U93574), R4 from Ascaris lumbricoides (U29445), R2 from Bombyx mori (M16558), R2 from the earwig, Forficular auricularia (AF015819), R2 from Drosophila melanogaster (X51967), CZAR from Trypanosoma cruzi (M62862), CRE2 from Crithidia fasciculata (U19151), CRE1 from C. fasciculata (M33009), CR1 from the turtle Platymys spixii (AB005891), CR1 from the chicken (U88211), Q from Anopheles gambiae (U03849), Tad1 from Neurospora crassa (L25662), CgT1 from the fungal phytopathogen, Colletotrichum gloeosporioides (L76169), R1 from B. mori (M19755), R1 from D. melanogaster (X51968), Tart from D. melanogaster (U14101), Juan from Ae. aegypti (M95171), Jockey (M22874), Doc (X17551), and I (M14954) from D. melanogaster, Group II intron-encoding maturase from Symbiobacterium thermophilum (BAD41717), Group II intron protein from Streptococcus pneumoniae (CAI33690), AP1 endonuclease from Paracentrotus lividus (AAY37515), AP endonuclease from Pseudomonas syringae (AAY37515), and exonuclease III from Escherichia coli (NP288182).
Copy number estimation
Estimates of the copy number of the SR3 retrotransposon were established by a comparative bioinformatics approach [12-14] wherein BLAST analysis of the BAC-end database of S. mansoni genomic sequences targeted more well-characterized retrotransposable elements from S. mansoni, and some other reference genes, for which copy numbers have been reported. These included the Boudicca and Sinbad LTR retrotransposons [12,13], the non-LTR retrotransposons SR1 and SR2 [18,22], the 18S ribosomal RNA genes, a middle repetitive element [46], and cathepsin D, a single copy gene [20]. The NCBI database was searched by BLAST using the sequences of these mobile genetic elements and some other genes of S. mansoni, all of which included at least one Hind III site. Specifically, the Advanced BLAST function was used, set to search only the S. mansoni sequences in the GSS database (Limit by Entrez Query: <Schistosoma mansoni [organism]>), and with the E (Expect) value at 0.000001. This stringent cutoff value was used to minimize the chance of counting other RTE-1-like elements in the total copy number of SR3. Since the formula for E is based not only on the bit scores of the local alignment of each pair of sequences, but also on the lengths of the subject and query [47], no additional correction was made for the length of the query sequence. Only hits with a Blast score of ≥100 were counted.
Investigation of integration sites
Five kilobases of the sequence flanking both 5'- and 3'-termini of SR3-left and SR3-right were employed as queries in BLAST searches of the non-redundant and dbEST GenBank databases limited by the organism [Schistosoma mansoni]. Sequences flanking additional copies of SR3 identified in other GenBank entries were also used as queries in BLAST searches to investigate the target site of SR3 integration. Multiple sequence alignments of integration sites were assembled and examined for target site preferences.
List of abbreviations
LTR, long terminal repeat; RT, reverse transcriptase; EN, endonuclease; UTR, untranslated region; SR3, schistosome retrotransposon 3; ORF, open reading frame; BAC, bacterial artificial chromosome; AChE, acetylcholineesterase; GSS, genome survey sequence; TSD, target site duplication.
Authors' contributions
TL carried out the sequence analysis, multiple sequence alignments, phylogenetic trees and drafted the manuscript. NK performed the Southern hybridization and assisted with the bioinformatics analyses. AL contributed to the experimental designs, sequence alignments, bioinformatics, and with drafting the manuscript. PJB oversaw the project, carried out copy number and other bioinformatics analyses, and drafted the manuscript. All authors read and approved the final manuscript.
Supplementary Material
Additional File 1
Nucleotide and deduced amino acid sequence of the SR3-left retrotransposon.
Click here for file
Additional File 2
Nucleotide and deduced amino acid sequence of the SR3-right retrotransposon.
Click here for file
Additional File 3
Sequence alignment of deduced open reading frames of SR3-left and SR3-right.
Click here for file
Additional File 4
Multiple sequence alignment of the reverse transcriptase domain of SR3 and related non-LTR retrotransposons.
Click here for file
Additional File 5
Multiple sequence alignment of the endonuclease domain of SR3 and related non-LTR retrotransposons.
Click here for file
Additional File 6
Table of representative GenBank accessions to show the presence in the Schistosoma mansoni transcriptomes of messenger RNAs encoding SR3 expressed in six developmental stages of the parasite.
Click here for file
Additional File 7
Table of representative GenBank accessions to show the presence in the Schistosoma mansoni transcriptomes of messenger RNAs encoding SR3 expressed in six developmental stages of the parasite.
Click here for file
Acknowledgements
We thank Mary Duke for maintenance of the Schistosoma mansoni life cycle, and the anonymous reviewers for helpful suggestions. Additional S. mansoni parasites were supplied by Dr. Fred Lewis through NIAID-NIH supply contract NO1-A1-55270. This investigation received financial support from the UNICEP/UNDP/World Bank/WHO Special Programme for Research and Training in Tropical Diseases (TDR) (ID A20723). PJB is a recipient of a Burroughs Wellcome Fund scholar award in Molecular Parasitology and AL is a recipient of an R. Douglas Wright Biomedical Career Development Award from the National Health and Medical Research Council of Australia.
==== Refs
Chitsulo L Loverde P Engels D Schistosomiasis Nat Rev Microbiol 2004 2 12 13 15035004 10.1038/nrmicro801
Hu W Brindley PJ McManus DP Feng Z Han ZG Schistosome transcriptomes: new insights into the parasite and schistosomiasis Trends Mol Med 2004 10 217 225 15121048 10.1016/j.molmed.2004.03.002
El-Sayed NM Bartholomeu D Ivens A Johnston DA LoVerde PT Advances in schistosome genomics Trends Parasitol 2004 20 154 157 15099549 10.1016/j.pt.2004.02.002
Brindley PJ Laha T McManus DP Loukas A Mobile genetic elements colonizing the genomes of metazoan parasites Trends Parasitol 2003 19 79 87 12586476 10.1016/S1471-4922(02)00061-2
Charlesworth B Sniegowski P Stephan W The evolutionary dynamics of repetitive DNA in eukaryotes Nature 1994 371 215 220 8078581 10.1038/371215a0
Kazazian HH Jr Mobile elements: drivers of genome evolution Science 2004 303 1626 1632 15016989 10.1126/science.1089670
Plasterk RH Izsvak Z Ivics Z Resident aliens: the Tc1/mariner superfamily of transposable elements Trends Genet 1999 15 326 332 10431195 10.1016/S0168-9525(99)01777-1
Venter JC Adams MD Myers EW Li PW Mural RJ Sutton GG Smith HO Yandell M Evans CA Holt RA Gocayne JD Amanatides P Ballew RM Huson DH Wortman JR Zhang Q Kodira CD Zheng XH Chen L Skupski M Subramanian G Thomas PD Zhang J Gabor Miklos GL Nelson C Broder S Clark AG Nadeau J McKusick VA Zinder N Levine AJ Roberts RJ Simon M Slayman C Hunkapiller M Bolanos R Delcher A Dew I Fasulo D Flanigan M Florea L Halpern A Hannenhalli S Kravitz S Levy S Mobarry C Reinert K Remington K Abu-Threideh J Beasley E Biddick K Bonazzi V Brandon R Cargill M Chandramouliswaran I Charlab R Chaturvedi K Deng Z Di Francesco V Dunn P Eilbeck K Evangelista C Gabrielian AE Gan W Ge W Gong F Gu Z Guan P Heiman TJ Higgins ME Ji RR Ke Z Ketchum KA Lai Z Lei Y Li Z Li J Liang Y Lin X Lu F Merkulov GV Milshina N Moore HM Naik AK Narayan VA Neelam B Nusskern D Rusch DB Salzberg S Shao W Shue B Sun J Wang Z Wang A Wang X Wang J Wei M Wides R Xiao C Yan C Yao A Ye J Zhan M Zhang W Zhang H Zhao Q Zheng L Zhong F Zhong W Zhu S Zhao S Gilbert D Baumhueter S Spier G Carter C Cravchik A Woodage T Ali F An H Awe A Baldwin D Baden H Barnstead M Barrow I Beeson K Busam D Carver A Center A Cheng ML Curry L Danaher S Davenport L Desilets R Dietz S Dodson K Doup L Ferriera S Garg N Gluecksmann A Hart B Haynes J Haynes C Heiner C Hladun S Hostin D Houck J Howland T Ibegwam C Johnson J Kalush F Kline L Koduru S Love A Mann F May D McCawley S McIntosh T McMullen I Moy M Moy L Murphy B Nelson K Pfannkoch C Pratts E Puri V Qureshi H Reardon M Rodriguez R Rogers YH Romblad D Ruhfel B Scott R Sitter C Smallwood M Stewart E Strong R Suh E Thomas R Tint NN Tse S Vech C Wang G Wetter J Williams S Williams M Windsor S Winn-Deen E Wolfe K Zaveri J Zaveri K Abril JF Guigo R Campbell MJ Sjolander KV Karlak B Kejariwal A Mi H Lazareva B Hatton T Narechania A Diemer K Muruganujan A Guo N Sato S Bafna V Istrail S Lippert R Schwartz R Walenz B Yooseph S Allen D Basu A Baxendale J Blick L Caminha M Carnes-Stine J Caulk P Chiang YH Coyne M Dahlke C Mays A Dombroski M Donnelly M Ely D Esparham S Fosler C Gire H Glanowski S Glasser K Glodek A Gorokhov M Graham K Gropman B Harris M Heil J Henderson S Hoover J Jennings D Jordan C Jordan J Kasha J Kagan L Kraft C Levitsky A Lewis M Liu X Lopez J Ma D Majoros W McDaniel J Murphy S Newman M Nguyen T Nguyen N Nodell M Pan S Peck J Peterson M Rowe W Sanders R Scott J Simpson M Smith T Sprague A Stockwell T Turner R Venter E Wang M Wen M Wu D Wu M Xia A Zandieh A Zhu X The sequence of the human genome Science 2001 291 1304 1351 11181995 10.1126/science.1058040
Fischer C Bouneau L Coutanceau JP Weissenbach J Volff JN Ozouf-Costaz C Global heterochromatic colocalization of transposable elements with minisatellites in the compact genome of the pufferfish Tetraodon nigroviridis Gene 2004 336 175 183 15246529 10.1016/j.gene.2004.04.014
Laha T Brindley PJ Smout MJ Verity CK McManus DP Loukas A Reverse transcriptase activity and untranslated region sharing of a new RTE-like, non-long terminal repeat retrotransposon from the human blood fluke, Schistosoma japonicum Int J Parasitol 2002 32 1163 1174 12117499 10.1016/S0020-7519(02)00063-2
Laha T Loukas A Smyth DJ Copeland CS Brindley PJ The fugitive LTR retrotransposon from the genome of the human blood fluke, Schistosoma mansoni Int J Parasitol 2004 34 1365 1375 Corrigendum, Int J Parasitol 2005, 35: 461 15542097 10.1016/j.ijpara.2004.08.007
Copeland CS Brindley PJ Heyers O Michael SF Johnston DA Williams DL Ivens AC Kalinna BH Boudicca, a retrovirus-like long terminal repeat retrotransposon from the genome of the human blood fluke Schistosoma mansoni J Virol 2003 77 6153 6166 12743272 10.1128/JVI.77.11.6153-6166.2003
Copeland CS Mann VH Morales ME Kalinna BH Brindley PJ The Sinbad retrotransposon from the genome of the human blood fluke, Schistosoma mansoni, and the distribution of related Pao-like elements BMC Evol Biol 2005 5 20 15725362 10.1186/1471-2148-5-20
DeMarco R Kowaltowski AT Machado AA Soares MB Gargioni C Kawano T Rodrigues V Madeira AM Wilson RA Menck CF Setubal JC Dias-Neto E Leite LC Verjovski-Almeida S Saci-1, -2, and -3 and Perere, four novel retrotransposons with high transcriptional activities from the human parasite Schistosoma mansoni J Virol 2004 78 2967 2978 14990715 10.1128/JVI.78.6.2967-2978.2004
DeMarco R Machado AA Bisson-Filho AW Verjovski-Almeida S Identification of 18 new transcribed retrotransposons in Schistosoma mansoni Biochem Biophys Res Commun 2005 333 230 240 15939396 10.1016/j.bbrc.2005.05.080
Le Paslier MC Pierce RJ Merlin F Hirai H Wu W Williams DL Johnston D LoVerde PT Le Paslier D Construction and characterization of a Schistosoma mansoni bacterial artificial chromosome library Genomics 2000 65 87 94 10783255 10.1006/geno.2000.6147
Malik HS Eickbush TH The RTE class of non-LTR retrotransposons is widely distributed in animals and is the origin of many SINEs Mol Biol Evol 1998 15 1123 1134 9729877
Drew AC Minchella DJ King LT Rollinson D Brindley PJ SR2 elements, non-long terminal repeat retrotransposons of the RTE-1 lineage from the human blood fluke Schistosoma mansoni Mol Biol Evol 1999 16 1256 1269 10486981
Mei H Hirai H Tanaka M Hong Z Rekosh D LoVerde PT Schistosoma mansoni: cloning and characterization of a gene encoding cytosolic Cu/Zn superoxide dismutase Exp Parasitol 1995 80 250 259 7895835 10.1006/expr.1995.1031
Morales ME Kalinna BH Heyers O Schulmeister A Mann VH Copeland CS Loukas A Brindley PJ Genomic organization of the Schistosoma mansoni aspartic protease gene, a platyhelminth orthologue of mammalian lysosomal cathepsin D Gene 2004 338 99 109 15302411 10.1016/j.gene.2004.05.017
Foulk BW Pappas G Hirai Y Hirai H Williams DL Adenylosuccinate lyase of Schistosoma mansoni: gene structure, mRNA expression, and analysis of the predicted peptide structure of a potential chemotherapeutic target Int J Parasitol 2002 32 1487 1495 12392914 10.1016/S0020-7519(02)00161-3
Drew AC Brindley PJ A retrotransposon of the non-long terminal repeat class from the human blood fluke Schistosoma mansoni. Similarities to the chicken-repeat-1-like elements of vertebrates Mol Biol Evol 1997 14 602 610 9190061
Han JS Boeke JD A highly active synthetic mammalian retrotransposon Nature 2004 429 314 318 15152256 10.1038/nature02535
Laha T Brindley PJ Verity CK McManus DP Loukas A pido, a non-long terminal repeat retrotransposon of the chicken repeat 1 family from the genome of the Oriental blood fluke, Schistosoma japonicum Gene 2002 284 149 159 11891056 10.1016/S0378-1119(02)00381-5
Malik HS Burke WD Eickbush TH The age and evolution of non-LTR retrotransposable elements Mol Biol Evol 1999 16 793 805 10368957
Malik HS Eickbush TH NeSL-1, an ancient lineage of site-specific non-LTR retrotransposons from Caenorhabditis elegans Genetics 2000 154 193 203 10628980
Youngman S van Luenen HG Plasterk RH Rte-1, a retrotransposon-like element in Caenorhabditis elegans FEBS Lett 1996 380 1 7 8603714 10.1016/0014-5793(95)01525-6
Bentley GN Jones AK Agnew A Mapping and sequencing of acetylcholinesterase genes from the platyhelminth blood fluke Schistosoma Gene 2003 314 103 112 14527722 10.1016/S0378-1119(03)00709-1
Hirai H Taguchi T Saitoh Y Kawanaka M Sugiyama H Habe S Okamoto M Hirata M Shimada M Tiu WU Lai K Upatham ES Agatsuma T Chromosomal differentiation of the Schistosoma japonicum complex Int J Parasitol 2000 30 441 452 10731567 10.1016/S0020-7519(99)00186-1
Verjovski-Almeida S DeMarco R Martins EA Guimaraes PE Ojopi EP Paquola AC Piazza JP Nishiyama MY JrKitajima JP Adamson RE Ashton PD Bonaldo MF Coulson PS Dillon GP Farias LP Gregorio SP Ho PL Leite RA Malaquias LC Marques RC Miyasato PA Nascimento AL Ohlweiler FP Reis EM Ribeiro MA Sa RG Stukart GC Soares MB Gargioni C Kawano T Rodrigues V Madeira AM Wilson RA Menck CF Setubal JC Leite LC Dias-Neto E Transcriptome analysis of the acoelomate human parasite Schistosoma mansoni Nat Genet 2003 35 148 157 12973350 10.1038/ng1237
Jurka J Sequence patterns indicate an enzymatic involvement in integration of mammalian retroposons Proc Natl Acad Sci USA 1997 94 1872 1827 9050872 10.1073/pnas.94.5.1872
Szak TS Pickeral OK Makalowski W Boguski MS Landsman D Boeke JD Molecular archeology of L1 insertions in the human genome Genome Biol 2002 3 10 10.1186/gb-2002-3-10-research0052
Feng Q Moran JV Kazazian HH JrBoeke JD Human L1 retrotransposon encodes a conserved endonuclease required for retrotransposition Cell 1996 87 905 916 8945517 10.1016/S0092-8674(00)81997-2
Luan DD Korman MH Jakubczak JL Eickbush TH Reverse transcription of R2Bm RNA is primed by a nick at the chromosomal target site: a mechanism for non-LTR retrotransposition Cell 1993 72 595 605 7679954 10.1016/0092-8674(93)90078-5
Feng Q Schumann G Boeke JD Retrotransposon R1Bm endonuclease cleaves the target sequence Proc Natl Acad Sci U S A 1998 95 2083 2088 9482842 10.1073/pnas.95.5.2083
Fanning TG Singer MF a mammalian transposable element Biochim Biophys Acta 1987 910 203 212 2445384
Moran JV Holmes SE Naas TP DeBerardinis RJ Boeke JD Kazazian HH Jr High frequency retrotransposition in cultured mammalian cells Cell 1996 87 917 927 8945518 10.1016/S0092-8674(00)81998-4
Perepelitsa-Belancio V Deininger P RNA truncation by premature polyadenylation attenuates human mobile element activity Nat Genet 2003 35 363 366 14625551 10.1038/ng1269
Altschul SF Gish W Miller W Myers EW Lipman DJ Basic local alignment search tool J Mol Biol 1990 215 403 410 2231712 10.1006/jmbi.1990.9999
The Sanger Institute: S. mansoni blast server
Southern EM Detection of specific sequences among DNA fragments separated by gel electrophoresis J Mol Biol 1975 98 503 517 1195397
Thompson JD Gibson TJ Plewniak F Jeanmougin F Higgins DG The CLUSTAL_X windows interface: flexible strategies for multiple sequence alignment aided by quality analysis tools Nucleic Acids Res 1997 25 4876 4882 9396791 10.1093/nar/25.24.4876
Hall T BioEdit: a user friendly biological sequence alignment editor and analysis program for Windows 95/98/NT Nucl Acids Symp Ser 1999 41 95 98
Felsenstein J PHYLIP (Phylogeny Inference Package) version 3.5c Distributed by the author 1993 Department of Genetics, University of Washington, Seattle
Page RD TreeView: an application to display phylogenetic trees on personal computers Comput Appl Biosci 1996 12 357 358 8902363
Littlewood DT Johnston DA Molecular phylogenetics of the four Schistosoma species groups determined with partial 28S ribosomal RNA gene sequences Parasitology 1995 111 167 175 7675531
The statistics of sequence similarity scores
|
16271150
|
PMC1291365
|
CC BY
|
2021-01-04 16:32:47
|
no
|
BMC Genomics. 2005 Nov 4; 6:154
|
utf-8
|
BMC Genomics
| 2,005 |
10.1186/1471-2164-6-154
|
oa_comm
|
==== Front
BMC Health Serv ResBMC Health Services Research1472-6963BioMed Central London 1472-6963-5-691627115110.1186/1472-6963-5-69Research ArticleCost and efficiency of public sector sexually transmitted infection clinics in Andhra Pradesh, India Dandona Lalit [email protected] Pratap [email protected] TLN [email protected] Elliot [email protected] Rao M [email protected] A Anod [email protected] SG Prem [email protected] YK [email protected] Mead [email protected] M [email protected] James G [email protected] Health Studies Area, Centre for Human Development, Administrative Staff College of India, Hyderabad, India2 Andhra Pradesh State AIDS Control Society, Hyderabad, India3 Institute for Health Policy Studies and AIDS Research Institute, University of California, San Francisco, USA4 Development Research Group, World Bank, Washington DC, USA2005 5 11 2005 5 69 69 14 4 2005 5 11 2005 Copyright © 2005 Dandona et al; licensee BioMed Central Ltd.2005Dandona et al; licensee BioMed Central Ltd.This is an Open Access article distributed under the terms of the Creative Commons Attribution License (), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
Background
Control of sexually transmitted infections (STIs) is an important part of the effort to reduce the risk of HIV/AIDS. STI clinics in the government hospitals in India provide services predominantly to the poor. Data on the cost and efficiency of providing STI services in India are not available to help guide efficient use of public resources for these services.
Methods
Standardised methods were used to obtain detailed cost and output data for the 2003–2004 fiscal year from written records and interviews in 14 government STI clinics in the Indian state of Andhra Pradesh. The economic cost per patient receiving STI treatment was calculated, and the variations of total and unit costs across the STI clinics analysed. Multivariate regression technique was used to estimate incremental unit costs. The optimal number of STIs that could be handled by the clinics was estimated.
Results
18807 STIs were diagnosed and treated at the 14 STI clinics in fiscal year 2003–2004 (range 323–2784, median 1199). The economic cost of treating each STI varied 5-fold from Indian Rupees (INR) 225.5 (US$ 4.91) to INR 1201.5 (US$ 26.15) between 13 clinics, with one other clinic having a very high cost of INR 2478.5 (US$ 53.94). The average cost per STI treated for all 14 clinics combined was INR 729.5 (US$ 15.88). Personnel salaries made up 76.2% of the total cost. The number of STIs treated per doctor full-time equivalent and cost-efficiency for each STI treated had a significant direct non-linear relation (p < 0.001, R2 = 0.81; power function). With a multiple regression model, apart from the fixed costs, the incremental cost for each STI detected and cost of treatment was INR 55.57 (US$ 1.21) and for each follow-up visit was INR 3.75 (US$ 0.08). Based on estimates of optimal STI cases that could be handled without compromising quality by each doctor full-time equivalent available, it was projected that at 8 of the 14 clinics substantially more STI cases could be handled, which could increase the total STI cases treated at the 14 clinics combined by 38% at an additional cost of only 3.5% for service provision.
Conclusion
There is un-utilised capacity in the public sector STI clinics in this Indian state. Efforts to facilitate utilisation of this capacity would be useful, as this would enable more poor patients with STIs to be served at minimal additional cost, and would also reduce the cost per STI treated leading to more efficient use of public resources.
==== Body
Background
India has one of the highest number of persons living with HIV in the world [1,2]. The presence of sexually transmitted infections other than HIV (referred to as STIs in this paper) increases substantially the risk of acquiring/transmitting HIV, and therefore, the control of STIs can be an important part of the effort to control emerging HIV epidemics [3]. The state of Andhra Pradesh with 80 million population has one of the highest estimated burden of HIV among the Indian states based on antenatal sentinel surveillance, and also one of the highest HIV prevalence (16–30%) among public sector STI clinic attendees included in the sentinel surveillance during 1998–2004 [4].
STI clinics in the public sector hospitals run by the government provide care predominantly to the poor segment of society in India, and therefore, their role in controlling STIs and HIV is particularly important. Recognising the important role that the public sector STI clinics could play in HIV control, the number of formally designated STI clinics in the government hospitals in Andhra Pradesh was increased from 28 to 85 in the year 2002 (oral communication, Andhra Pradesh State AIDS Control Society, Hyderabad, India, 14 December 2004).
With HIV/AIDS control having become a major public health issue in India [5], the funding available for this has been increasing [6,7]. However, assessment of the cost and efficiency of various strategies to control HIV in India is not readily available [8]. This information is needed for efficient utilisation of resources available for HIV control. As part of a study to assess the cost and efficiency of various HIV prevention strategies in Andhra Pradesh, we assessed the cost and efficiency of STI clinics in the public sector.
Methods
This study was part of a larger multi-country effort to study cost and efficiency of HIV prevention in India, Mexico, Russia, South Africa and Uganda by the Prevent AIDS Network for Cost-Effectiveness Analysis (PANCEA) [9]. Details of the methods for the overall multi-country study are described elsewhere [10]. Description of the methods relevant for this paper follows. This study was approved by the Ethics Committee of the Administrative Staff College of India, Hyderabad, India and the Committee on Human Research of the University of California, San Francisco, USA.
Selection of STI clinics
We included STI clinics in the government-run public sector hospitals for this study. These clinics provide services predominantly to the poor at no or minimal fee, and many persons receiving services here have relatively advanced STIs. At the time of starting data collection for this study in mid-2004, 85 public sector clinics were functioning. Of these, 28 STI clinics had been functioning since the 1960's in the state capital and the district headquarters as part of the medical college hospitals and other major public hospitals of Andhra Pradesh. In addition, 57 STI clinics were formally designated in mid-2002 in public hospitals, in district headquarter hospitals and in area hospitals located in smaller jurisdictions. Generally, the STI care in medical college hospitals and district headquarter hospitals is provided in distinct separate clinics, whereas in area hospitals it is provided as part of the general outpatient.
Andhra Pradesh has three geographic regions: the northern Telangana region has the state capital and nine other districts, the eastern Coastal region has nine districts, and the southern Rayalseema region has four districts with a population of nearly half that in the other two regions. In order to obtain a broad sample of STI clinics in public hospitals, we used the three geographic regions of Andhra Pradesh and the old-new clinics as the two strata for sampling. Fifteen STI clinics were randomly sampled to obtain six clinics each in the Telangana and Coastal regions and three clinics in the Rayalseema region, which included three each of old and new clinics in Telangana and Coastal regions and two old and one new clinics in Rayalseema region.
Data collection procedures
The initial versions of the data collection instruments from the global PANCEA study were reviewed and adapted to suit the context of Andhra Pradesh. The data collection team, consisting of six researchers with background in economics or finance, was involved with the adaptation of the instruments and received extensive training to ensure a standardised approach to data collection. A pilot study was done to make final refinements in the data collection format and approach.
Data were collected for the April 2003 – March 2004 fiscal year at the 15 sampled STI clinics during June – August 2004. Data collection included a history of the evolution of the STI clinic, detailed cost and output data by month, and patient exit interviews. Formal informed consent to collect data was obtained from the senior-most person responsible for each STI clinic, generally the superintendent of the hospital in which the clinic was located. The STI clinic medical officer(s), medico-social worker, counsellor, and laboratory technician were interviewed and available written records reviewed to obtain data. Each visit started with an interview containing structured open-ended questions on the history of the STI clinic, and operational or community factors that may have affected the demand for or supply of services. Data collection then proceeded to cost and output data, in parallel with exit interviews of patients. Data collection at an STI clinic by three investigators lasted one week. Data were recorded in the field on laptop computers in MS Excel and MS Word files, which after review were entered into an MS Access database.
Cost data
The cost of the STI clinic was divided into five categories: salaries, recurrent goods, capital goods, recurrent services, and rentals. These cost data were collected for each month, as far as possible. Economic cost was computed, i.e. the true resource cost incurred rather than just the financial cost. The inpatient cost was also included for the inpatient facilities that were being utilised for some of the STI clinic patients. Similar costing methods were used across the clinics for the five cost categories.
Salary cost was computed for all personnel contributing to the work of the outpatient and inpatient services of the STI clinic, which included the medical officer(s), medico-social worker/counsellor, nurse(s), laboratory technician, and in some cases attender. If a particular person was contributing part time to the STI clinic work, only that proportion of salary was included in the STI clinic salary cost. For example, if the medical officer was contributing only half time to the STI clinic and the other half to teaching then only half his/her salary was included in the personnel cost. The salary of the personnel was noted from the official records of hospital of which the STI clinic was a part. If fringe benefits were paid in addition to the regular salary, these were included.
The major component of recurrent goods was medicines. The medicines prescribed to patients at these public clinics are provided free. In order to calculate the cost of medicines, first the lowest market retail rates of the medicines usually prescribed at these public clinics for each STI diagnosis during the 2003–2004 were obtained. This was discounted by 30%, which is the estimated discount for bulk purchase of medicines by the government as suggested by the procurement agency. These rates were then applied to the number of each type of STI reported by a STI clinic. The experience of the STI incharge at the Andhra Pradesh State AIDS Control Society suggested that complicated cases needed further treatment and that there were variations in treatment regimens, which would enhanced the overall medication cost by 25%. Accordingly, this enhancement was applied to obtain the total medicine cost for each STI clinic. The other recurrent goods utilised at the STI clinics included inpatient food expenses, test kits for STIs, male condoms, behavioural change communication materials, needles and syringes, gloves, spirit, sodium hypochlorite solution, cotton, soap, dettol, phenyl, distilled water, test tubes, antigen, blotting paper, alcohol, liquid paraffin, buffer solution, stationery and some other laboratory materials and miscellaneous items. The inpatient food expenses were calculated based on cost estimates for items on the menu and the number of inpatients. The test kits for STIs were costed by applying 30% discount to the lowest market rate for bulk purchase by government, as for the medicines. The market price of the condoms supplied to the STI clinic is subsidized by 70% by the government. We considered as the economic cost of the condom what it would have been without the subsidy. Information was obtained from the Andhra Pradesh State AIDS Control Society about the cost of the behavioural change communication materials supplied to the STI clinic, which was used for our analysis. Attempts were made to get the cost of the other goods from the STI clinic records, which if available were used for analysis. If not available, three quotations for these goods were obtained from the market for the 2003–2004 fiscal year and the average of these taken as the cost.
Capital goods used for the work of STI clinics included outpatient and inpatient furniture, electrical fixtures, refrigerator, centrifuge, microscope, needle and syringe destroyer, water bath, slide rotator, water filter, sterilizer, and weighing machine. As for recurrent goods, if information about the cost of capital goods was not available from the STI clinic or its parent hospital, the market price was determined from retail sellers of these goods. The life of the capital goods was assumed to be five years, and therefore, one-fifth of the cost was allocated to the 2003–2004 fiscal year if the good was used for the full year. If a good was purchased in the middle of this fiscal year and used only for half the year, the cost allocated to this good was half of the yearly cost. If a capital good, for example refrigerator or centrifuge, was also being used for work other than that of the STI clinic, a determination was made from the STI clinic staff about the proportion of use for STI work and that proportional cost was allocated to the STI clinic.
Recurrent services included cleaning and building maintenance, electricity, water, telephone, gas/oil, waste disposal, the occasional training of staff during that fiscal year, and some miscellaneous items. The cost of building maintenance was calculated based on the space occupied by the outpatient and inpatient components of the STI clinic. Electricity and water costs were based on applying the market rates to the estimated usage. The estimated usage of electricity was calculated from the electrical fixtures and the number of hours of use per day and for water from the daily estimated use by clinic staff and patients. Telephone and other recurrent services costs were calculated based on actual expenditure.
All STI clinics were located in a parent public hospital, and no rent was being paid. In order to calculate economic rental cost, the actual floor area occupied by the outpatient and inpatient services of the STI clinic was determined, rent rates obtained from three sources in that area for health facilities for the 2003–2004 period, and the average of these rates applied to this area.
For area hospitals, where STI care is generally provided as part of the general outpatient, the costs for STI care were calculated based on estimates of the proportion of personnel time, goods, and services used for STI care, and the rental cost was apportioned based on the ratio of STI cases to all outpatient cases. This estimation did not pose substantial difficulties.
The average exchange rate of Indian Rupees (INR) 45.95 to a US$ for the 2003–2004 fiscal year was used to convert the INR cost to US$ [11].
Output data
Detailed data were obtained from the written monthly summary records of the STI clinics regarding the services provided every month. These included the number and type of STIs detected and treated, total patient visits, type of tests done, treatment given for STIs, and number of patients receiving inpatient services for STIs and inpatient days. For area hospitals, where STI care is generally provided as part of the general outpatient, data on visits related to STIs were taken from the outpatient records that specify the diagnoses for the visits. Characteristics of programme operation were ascertained, including client characteristics and whether there were any hurdles to the provision of services.
At each STI clinic, 20 patients using the STI services were interviewed regarding their perceptions about these services. The first patients available to the investigators at the STI clinic during the week of data collection were selected for interview, after obtaining verbal informed consent that explained the purpose of the interview and assured anonymity of the respondent. The interviews were conducted in a quiet corner out of the hearing range of others in order to encourage honest responses.
Quality control
Quality control measures included a thorough pilot study before commencing formal data collection, comprehensive training of a qualified data collection team including their conceptual understanding of all data issues, full back-up and justification for any data recorded, supervision of data collection at each STI clinic by the project coordinator, thorough review by the study team of the data obtained at each STI clinic, and contacting the STI clinics again to obtain information about data issues that needed clarification after the review.
Data analysis
Analysis of the data was done using SPSS statistical software. The average economic cost per patient diagnosed and treated for STI was calculated as the measure of cost-efficiency of each STI clinic. The relation of this measure of efficiency with the cost components was assessed through regression analyses. Incremental costs for initial and follow-up visits were assessed using a multiple regression model. The personnel, capital goods and rental costs were relatively fixed for each STI clinic irrespective of the volume of services provided. We therefore treated these as relatively fixed costs, and treated the recurrent goods and recurrent services as variable costs for each clinic. In addition to using the number of initial and follow-up visits as independent variables for each STI clinic, we also used the fixed costs as a variable in the right side of the equation as the constant alone may not estimate the entire fixed costs [12]. We used this regression model to assess the total cost function. A high R2 value would be expected for this model, since the left side of the equation (total cost) was determined in our costing largely based on information present in the right side of the equation. That is, fixed costs are clinic-specific and a major portion of variable costs (medicines) was calculated assuming standard inputs and costs per visit across clinics. This is consistent with the objective of our model, which was to determine the incremental costs for initial and follow-up visits.
The optimal number of patients that could have STI detected and treated in a year was calculated using the following assumptions, which were based on input of STI clinic staff and persons familiar with the working of these STI clinics:
1. The number of STIs detected and treated depended mainly on the availability of full-time equivalents of the doctor in the outpatient of the STI clinic. Since the majority of the patients attending these clinics have STIs, the proportion of initial visits not related to STIs is small. Therefore, the latter would not have a substantial impact on the number of STIs that could be detected and treated per full-time equivalent of doctor.
2. An average of two follow-up visits for each STI, i.e. a total of three visits for each STI diagnosis, would be useful for quality care for patients attending these clinics.
3. During the usual 5 working hours of the STI clinic outpatient in a day, a doctor could satisfactorily see 5 patients with new STI diagnosis and 10 follow-ups, without compromising quality.
4. Considering 250 working days in a year, for each doctor full-time equivalent available in the outpatient clinic, 1250 new STIs could be diagnosed and treated without compromising quality.
The economic cost for providing services to the optimal number of STI patients was then estimated.
Reliable data on the number of STIs detected and treated at one of the sampled STI clinics located in an area hospital could not be obtained. Therefore, this clinic was excluded from the analysis and data are presented for 14 of the 15 sampled STI clinics.
Results
A total of 18807 diagnoses and associated treatment for STIs were reported by the 14 STI clinics in the 2003–2004 fiscal year, with a median value of 1199 and mean of 1343 (Table 1). This number was generally highest at STI clinics located in the tertiary hospitals of medical colleges, followed by the STI clinics in district headquarter hospitals, and least in the STI clinics in area hospitals. This trend was largely due to the differences in the relative sizes of the catchment populations for these three categories. Of the total STIs detected and treated, 59.2% were in males. The male to female ratio was highest at 3:2 for medical college hospital STI clinics but was reverse in area hospital STI clinics (Table 1). The ratio of total out-patient visits to the cases of STIs detected and treated was reported to be highest by the medical college hospital STI clinics (3.33) as compared with the district headquarter hospital STI clinics (1.40) and for the area hospital STI clinics (1.76), suggesting that the number of follow-up visits is relatively higher in the medical college STI clinics. Data regarding the exact reason for each visit were not available. However, it is estimated that the majority of the initial visits in these clinics result in STI diagnoses, and the majority of other visits are follow-up visits related to STI treatment and counselling.
Table 1 Number of STIs treated at STI clinics in the fiscal year 2003–2004.
STI clinic STIs diagnosed and treated Total visits Ratio of total visits to STIs treated
Total number Male Female
Number Percent Number Percent
Medical college hospitals
MC1 2784 1827 65.6 957 34.4 9093 3.27
MC2 1078 792 73.5 286 26.5 7260 6.73
MC3 1921 1189 61.9 732 38.1 7204 3.75
MC4 2186 1514 69.3 672 30.7 7064 3.23
MC5 1743 1333 76.5 410 23.5 4638 2.66
MC6 2223 1255 56.5 968 43.5 4543 2.04
All MC 11935 7910 66.3 4025 33.7 39802 3.33
District headquarter hospitals
DHQ1 1759 851 48.4 908 51.6 2058 1.17
DHQ2 1320 788 59.7 532 40.3 1365 1.03
DHQ3 1012 447 44.2 565 55.8 1530 1.51
DHQ4 323 257 79.6 66 20.4 1220 3.78
All DHQ 4414 2343 53.1 2071 46.9 6173 1.40
Area hospitals
AH1 551 76 13.8 475 86.2 650 1.18
AH2 582 253 43.5 329 56.5 1215 2.09
AH3 558 220 39.4 338 60.6 631 1.13
AH4 767 331 43.2 436 56.8 1842 2.40
All AH 2458 880 35.8 1578 64.2 4338 1.76
Total 18807 11133 59.2 7674 40.8 50313 2.68
All sampled medical college hospitals and most district headquarters hospitals had specialist(s) trained in STIs available for outpatients in their STI clinics, whereas most area hospitals did not have a specialist for their STI clinics. The diagnosis of STIs was most frequently made based on clinical assessment without complete laboratory investigations. Investigations were done most often in STI clinics in medical college hospitals, infrequently in district headquarter hospitals, and almost never in area hospitals. The diagnoses of STIs reported by these STI clinics are shown in Table 2. Of the classic STIs, the diagnosis of herpes was reported to be the highest. Considering all diagnoses together, the highest reported diagnosis was the miscellaneous category that included mostly scabies and some pediculosis and others.
Table 2 Distribution of reported STI diagnoses.
No. Type of STI Number Percent
1 Herpes 2507 13.3
2 Chlamydial infection 2212 11.8
3 Gonorrhea 1389 7.4
4 Nonspecific genital ulcer 1249 6.6
5 Nonspecific vaginal discharge 1139 6.1
6 Candidiasis 1123 6.0
7 Syphilis 1007 5.4
8 Genital warts 808 4.3
9 Trichomonas 643 3.4
10 Chancroid 586 3.1
11 Lymphogranuloma venereum 228 1.2
12 Bacterial vaginosis 225 1.2
13 Donovanosis 160 0.9
14 Miscellaneous including scabies and pediculosis 5532 29.4
Total 18807 100
The total economic cost of services provided by the 14 STI clinics during the 2003–2004 fiscal year was INR 13,719,992 (US$ 298,585), of which 9.9% was for inpatient services provided for 1106 (5.9%) of the total 18807 STIs treated. Of the total cost, personnel made up 76.2%, recurrent goods 10.2% (58.5% of this was for STI medicines), rentals 9.9%, recurrent services 2.1%, and capital goods 1.7%. There were modest variations in these proportional costs among some of the STI clinics (Table 3). If the financial costs were considered, by excluding the costs for rentals and condom subsidy, these would be 10.4% less than the economic costs for the 14 STI clinics combined.
Table 3 Economic cost of STI clinics in the 2003–2004 fiscal year.
STI clinic* Total economic cost Percent of economic cost Percent cost due to inpatient services† Number of STIs treated Cost per STI treated
INR US$ Personnel Recurrent goods Rentals Recurrent services Capital goods INR US$
Medical college hospitals
MC6 1283694 27937 76.5 12.2 6.2 2.7 2.4 8.3 2223 577.5 12.6
MC5 1345786 29288 82.7 7.4 6.7 1.7 1.6 2.6 1743 772.1 16.8
MC4 1746668 38012 75.0 10.0 11.3 1.8 1.9 9.8 2186 799.0 17.4
MC3 1569840 34164 80.6 9.7 6.0 2.0 1.7 5.3 1921 817.2 17.8
MC2 1061341 23098 66.8 8.8 19.2 3.0 2.1 10.2 1078 984.5 21.4
MC1 3345020 72797 79.8 8.0 9.9 1.2 1.1 17.2 2784 1201.5 26.1
All MC 10352349 225296 77.7 9.1 9.6 1.9 1.6 10.4 11935 867.4 18.9
District headquarter hospitals
DHQ1 396627 8632 63.9 25.3 6.4 2.9 1.4 11.0 1759 225.5 4.9
DHQ2 349816 7613 62.5 22.7 8.2 3.4 3.3 15.0 1320 265.0 5.8
DHQ3 612111 13321 73.2 9.8 14.1 1.8 1.1 9.1 1012 604.9 13.2
DHQ4 800543 17422 75.3 4.9 16.7 1.8 1.3 5.2 323 2478.5 53.9
All DHQ 2159097 46988 70.5 12.9 12.7 2.2 1.6 9.0 4414 489.1 10.6
Area hospitals
AH2 198736 4325 60.2 18.4 11.9 5.9 3.6 14.2 582 341.5 7.4
AH4 329889 7179 71.5 14.4 9.0 2.8 2.2 9.7 767 430.1 9.4
AH3 340348 7407 78.7 11.2 5.4 3.0 1.7 6.2 558 609.9 13.3
AH1 339572 7390 76.0 14.2 5.0 3.0 1.7 0.6 551 616.3 13.4
All AH 1208545 26301 72.9 14.1 7.4 3.4 2.2 6.9 2458 491.7 10.7
Total 13719992 298585 76.2 10.2 9.9 2.1 1.7 9.9 18807 729.5 15.9
INR is Indian Rupee
1 US$ = INR 45.95 (average exchange rate in the 2003–2004 fiscal year) [11]
*STI clinics arranged within each hospital type category in decreasing order of cost-efficiency, i.e. increasing order of cost per STI treated.
†This inpatient cost does not include cost of medicines for treatment of STIs, which would have been incurred even if these patients were treated as outpatient.
The economic cost for each patient detected to have STI and treated for it varied 5-fold between 13 of the 14 STI clinics from INR 225.5 (US$ 4.91) to INR 1201.5 (US$ 26.15), with a median of INR 609.9 (US$ 13.27) (Table 3). The remaining STI clinic had an unusually high cost per STI detected and treated (INR 2478.5, US$ 53.94). The average cost of detecting and treating each STI for patients at all the 14 STI clinics combined was INR 729.5 (US$ 15.88). The average cost of detecting and treating each STI was 77% higher at the STI clinics in the medical colleges compared with those at district headquarter and area hospitals (Table 3).
Since personnel accounted for the major portion of cost, and doctors had the highest relative salary among personnel, there was a significant direct non-linear relation between the number of STI cases treated per doctor full-time equivalent in a year and the cost-efficiency for each STI case treated at the 14 STI clinics, the best fit for which was obtained with a power function (p < 0.001, R2 = 0.81) as shown in Figure 1. If data for the one clinic with an exceptionally high cost per STI treated were excluded (DHQ4), considering it as an outlier, the best fit for this relation was obtained with an exponential function (p < 0.001, R2 = 0.81) as shown in Figure 2.
Figure 1 Relationship between STIs treated per doctor full-time equivalent and the cost per STI treated (p < 0.001, power function). FTE is full-time equivalent, INR is Indian Rupee.
Figure 2 Relationship between STIs treated per doctor full-time equivalent and the cost per STI treated after excluding one extreme value that could be considered an outlier (p < 0.001, exponential function). FTE is full-time equivalent, INR is Indian Rupee.
Considering the 2003–2004 fiscal year data from each STI clinic as a data point, the multiple regression model explained in the methods section was used to assess total economic cost as a function of fixed costs, cost of initial visits in which STI cases were detected (including treatment cost), and cost of other visits (mostly follow-up visits), which revealed the following relation:
= 3036.35 + 1.04 X + 55.57 Y + 3.75 Z
where is total economic cost in INR, X is fixed costs, Y is the number of initial visits in which STI cases were detected and treatment/medicines given, and Z is the number of other visits. As expected, this model had a high R2 value of almost 1.00. The fit of the model was significant at p <0.001 (F = 15906; degrees of freedom: 3 for regression, 10 for residual, 13 total). In this model, two variables (fixed costs and initial visits) were statistically significant whereas one variable (follow-up visits) was not (Table 4). This model suggests that apart from the constant/fixed costs, the additional cost for each initial visit in which STI was detected and treatment given (including cost of treatment) was INR 55.57 (US$ 1.21) and the additional cost for each follow-up visit was very small at INR 3.75 (US$ 0.08).
Table 4 Coefficients in the multiple regression model and their significance.
Variable Coefficient (βi) Standard error t Significance
Constant 3036.35 8049.64 0.377 0.714
Fixed costs 1.04 0.01 95.761 0.000
Visits in which STI was detected 55.57 8.28 6.712 0.000
Other visits (mostly follow-up) 3.75 2.88 1.300 0.223
Dependent variable: Total cost
Personnel at 13 of the 14 (92.9%) STI clinics responded that more patients could be served by their STI clinics with the available personnel and infrastructure if there were more demand. Based on the assumptions mentioned in the methods section about the number of STI cases that could be detected and treated without compromising quality by each doctor in a year if there were optimal demand, 4 of the 6 medical college STI clinics, 2 of the 4 district headquarter STI clinics, and 2 of the 4 area hospital STI clinics could increase the number of STIs detected and treated per year substantially by 54–287% with the available personnel and infrastructure (Table 5). If this were achieved, the overall number of STI cases detected and treated by the 14 STI clinics could increase by 38% from 18807 to 25916. The total cost for this would increase only by 3.5% from the 2003–2004 cost of INR 13,719,992 (US$ 298,585) to INR 14,203,222 (US$ 309,102), using the equation = 3036.35 + 1.04 X + 55.57 Y + 3.75 Z from the multiple regression model mentioned previously for incremental costs.
Table 5 Optimal number of STIs that could be treated at the STI clinics.
STI clinic* Doctor full-time equivalents available Number of STIs treated in fiscal year 2003–2004 Optimal number of STIs that could be treated† Percent increase if optimal number treated
Total Male (%) Female (%)
Medical college hospitals
MC6 1.20 1.20
(100) 0
(0) 2223 1500
MC5 1.50 1.25
(83.3) 0.25
(16.7) 1743 1875 8
MC4 3.00 1.50
(50.0) 1.50
(50.0) 2186 3750 72
MC3 2.46 1.96
(79.7) 0.50
(20.3) 1921 3075 60
MC2 1.50 1.50
(100) 0
(0) 1078 1875 74
MC1 5.50 5.00
(90.9) 0.50
(9.1) 2784 6875 147
All MC 15.16 12.41
(81.9) 2.75
(18.1) 11935 18950 59
District headquarter hospitals
DHQ1 0.40 0.40
(100) 0
(0) 1759 500
DHQ2 0.33 0
(0) 0.33
(100) 1320 413
DHQ3 1.25 0.25
(20.0) 1.00
(80.0) 1012 1563 54
DHQ4 1.00 1.00
(100) 0
(0) 323 1250 287
All DHQ 2.98 1.65
(55.4) 1.33
(44.6) 4414 3725
Area hospitals
AH2 0.20 0.20
(100) 0
(0) 582 250
AH4 0.58 0.33
(56.9) 0.25
(43.1) 767 725
AH3 1.00 0.70
(70.0) 0.30
(30.0) 558 1250 124
AH1 0.81 0.31
(38.5) 0.50
(61.5) 551 1016 84
All AH 2.59 1.54
(59.5) 1.05
(40.5) 2458 3241 32
Total 20.73 15.60
(75.3) 5.13
(24.7) 18807 25916 38
*STI clinics arranged within each hospital type category in decreasing order of cost-efficiency, i.e. increasing order of cost per STI treated.
†Based on the assumption that one full-time equivalent doctor could satisfactorily treat 1250 STIs in a year, as described in the methods section.
In addition to inadequate demand as the major hurdle to provision of services, two STI clinics mentioned inadequate supplies and two clinics mentioned inadequate staffing as hurdles. The two clinics that mentioned inadequate supplies as hurdle (AH1 and AH3) had the highest cost per STI treated among the area hospitals (Table 3) but this was close to the median cost per STI treated for all clinics considered together. One of the clinics that mentioned inadequate staffing as hurdle (DHQ4) had the highest cost of all per STI treated and the other (AH2) had one of the lowest costs of all per STI treated (Table 3). The small proportion of clinics reporting hurdles other than inadequate demand prevent generalisations about the relation of these hurdles to efficiency.
The proportion of full-time equivalents for female doctors available in the STI clinics of medical colleges was much lower than in the STI clinics of district headquarters and area hospitals (Table 5), which would partly be responsible for the relatively lower proportion of female patients treated in the medical college STI clinics (Table 1).
The patient interviews at the STI clinics revealed 70.4% were very satisfied with the services provided, and the remaining were somewhat satisfied, with no one mentioning that he/she was not satisfied (Table 6).
Table 6 Satisfaction of patients with services provided by STI clinics.
STI clinic Total number of patients interviewed Very satisfied Some what satisfied Not satisfied
No. % No. % No. %
Medical college hospitals 120 79 65.8 41 34.2 0 0
District headquarter hospitals 80 57 71.2 23 28.8 0 0
Area hospitals 80 61 76.2 19 23.8 0 0
Total 280 197 70.4 83 29.6 0 0
All medical college hospital STI clinics and two district headquarter STI clinics (DHQ2 and DHQ4) were old STI clinics (established since 1960s), and the remaining were newly designated STI clinics in 2002. Because major differences in the services and cost-efficiency were observed between STI clinics located in the three categories of hospitals, and the only category that had both old and new clinics was the district headquarter hospitals, separate analysis for old and new clinics was not done.
Discussion
In the Andhra Pradesh state of India, analysis of 14 public sector STI clinics revealed that the average cost of diagnosing and treating each STI was INR 729.5 (US$ 15.88), and this cost ranged many-fold between the 14 clinics. Fixed costs of the STI clinics made up the predominant proportion, with personnel costs exceeding three-fourths of the total cost. Although it is unusual to think of personnel as fixed costs, this concept applies here as the personnel were employed at the STI clinics regardless of the amount of services delivered. Based on estimates of optimal workload that could be handled by the available personnel, we estimated that a large proportion of these STI clinics could deal with more STI cases without compromising quality, which would increase the number of STIs treated at the 14 clinics by over one-third the number treated in the 2003–2004 fiscal year at a minimal additional cost of about 3.5% for provision of services. This does not include the additional costs that may be involved in increasing the demand for these services. These are difficult to assess and need to be assessed over a period of time, and may or may not be large. Since the public sector STI clinics predominantly serve poor patients in India at almost no direct cost to the patients, it is important that mechanisms be developed to utilise the unused capacity in these clinics. This would result in more poor patients to be served and more efficient use of the public resources available for STI and HIV control.
Substantial differences were found in the services and cost-efficiency between the STI clinics at the three different type of locations, i.e. medical college, district headquarter and area hospitals. The number of STIs treated per full-time doctor equivalent, as well as the cost-efficiency per STI treated, was least in the medical college hospital STI clinics. However, it is important to note that medical college clinics also had the highest number of follow-up visits per STI, availability of STI specialists, and level of investigations for diagnosis, suggesting that the thoroughness of STI care would be higher at these clinics. In addition, the tertiary medical college clinics are estimated to get more complicated patients and also serve a vital role in the teaching of residents and medical students. Therefore, the relatively higher cost per STI treated in medical college clinics is by itself not a negative thing, if it is associated with provision of higher quality care and care to more complicated patients. Attempts at improving cost-efficiency should not be at the expense of quality of care. We calculated the optimal workload that could be handled by the personnel at the STI clinics taking into account the need to maintain quality. The highest relative unused capacity was in the medical college hospital STI clinics. On the other hand, the low number of follow-up visits at the district headquarter and area hospital STI clinics suggests that emphasis on encouraging follow-ups and quality of care is needed. In addition, the exceptionally high number of STI patients treated at two district headquarter hospital STI clinics (DHQ1 and DHQ2) and the high number at one area hospital STI clinic (AH2) per full-time doctor equivalent available suggests that the need for ensuring quality of care must be emphasised, assuming that quality is impaired by an excessive caseload.
The relatively low proportion of female patients treated at medical college hospital STI clinics can perhaps be attributed to the lower proportion of female doctors available at these clinics as compared with district headquarter and area hospital STI clinics, and also to the fact that gynaecology clinics are available at medical college hospitals that are usually staffed by female doctors and many females go to these clinics for STI-related symptoms. Another factor for the relatively higher proportion of female STI patients seen at the area hospitals is the fact that the STI clinics in these hospitals run as part of the general outpatient and not as a distinctly separate clinic, which encourages more females to utilise these services.
Inpatient services are used by the STI clinics to admit patients of very advanced STIs who are quite poor and have a low chance of coming back for follow-up. Although these inpatient services make up 10% of the total cost for 6% of the total STIs treated, this is generally utilised for the STI patients who are most in need and who are at high risk of not completing the treatment and follow-up in the absence of these services, and therefore, may be justified.
The reported distribution of STI diagnoses should be interpreted cautiously in the background that investigations are done almost never at area hospital clinics and infrequently at district headquarter hospital clinics, and that STI specialists are mostly not available at area hospital clinics. The very high proportion of scabies reported by these public sector STI clinics is generally suggestive of the very low socioeconomic status of the patients utilising the services of these clinics.
An encouraging finding in this study was that of the patients interviewed at the STI clinics over two-third were very satisfied with the clinic services and the remaining somewhat satisfied, with no respondent reporting dissatisfaction. This implies that once the patients, who are mostly poor, reach the STI clinic and use its services, they are generally satisfied with the services. An important factor in this perception is likely to be the fact that free medicines are given for treatment.
Consideration of the above mentioned issues would have to be kept in mind while planning methods to enhance utilisation of the unused capacity in the public sector STI clinics in this Indian state such that the quality of services is also adequate. One major issue that comes in the way of more demand for services of these clinics is the stigma that is attached with attending an STI clinic, which is commonly perceived by the society to be associated with poor personal character or other negative attributes. An option that could be explored to reduce this stigma would be to give a more innocuous name to these clinics, such as, family health clinics, and provide STI services as part of a comprehensive sexual health package. Such an approach would require some restructuring, but may be worth considering as a long-term solution for better utilisation of STI services in the public sector hospitals. Another issue that needs to be addressed to increase the utilisation of services at public sector STI clinics is the development of systematic linkages with non-governmental organisations working with high-risk groups such as sex workers, such that they can easily avail the services of these STI clinics.
The majority of outpatient health care in India, including STI treatment, is provided by the private sector [13]. It is also widely believed that the patients who use the public sector STI clinics are predominantly poor and with relatively advanced STIs that have often been unsuccessfully treated elsewhere. In this background, the cost and efficiency analysis presented in this paper pertains to this most vulnerable group. The high rate of 16–30% HIV in the public sector STI clinic attendees who participated in the sentinel surveillance in Andhra Pradesh over the past few years [4] supports the interpretation that these patients are highly vulnerable. It is possible that the cost-efficiency trends for the private sector STI care in India may be different from those in the public sector. However, it may not be unreasonable to assume that the cost-efficiency estimates of public sector STI care apply to those who are at most risk of acquiring or transmitting HIV, and therefore, are important for assessing and comparing the cost-efficiency and cost-effectiveness of the various HIV prevention strategies.
Estimates of cost and efficiency of STI services in India, and their effectiveness in preventing HIV, have not been previously readily available from India [14]. In fact, the cost estimates of STI treatment from the Mwanza study in Tanzania [15] were used previously by the World Bank to estimate the STI care resources needed in India while preparing the loan agreement for supporting India's AIDS control programme [16]. The cost-efficiency estimates presented in this paper could be used to estimate the resources needed for STI care by the public health care system in Andhra Pradesh, and these data also point to some issues which, if addressed, could lead to enhancement in the provision of these services and also their cost-efficiency. These cost and efficiency estimates of STI treatment could also be used to estimate cost-effectiveness of STI treatment for HIV prevention in India either by using published estimates of effectiveness from other parts of the world or by conducting studies on effectiveness of STI treatment in preventing HIV in India. In addition, these cost-efficiency estimates of STI treatment would be compared with similar estimates from other countries in the PANCEA study that are using similar methodology [10], and also with estimates for other HIV prevention strategies in Andhra Pradesh, such as HIV voluntary counselling and testing [17] and HIV prevention programmes for female sex workers [18] using similar methodology. Such standardised cross-country and country-specific cost-efficiency estimates would be very useful in estimating the resources needed for HIV control in specific countries as well as globally with more confidence than has been possible so far with the available data. This is important in the background of the extensive debate in recent times about the resources needed to control HIV.
Conclusion
In the Indian state of Andhra Pradesh, there is un-utilised capacity in the public sector STI clinics that provide services predominantly to the poor. Efforts to facilitate utilisation of this capacity would be useful, as this would enable more poor patients with STIs to be served at minimal additional cost, and would also reduce the cost per STI treated leading to more efficient use of public resources. Comprehensive and dynamic analysis of the efficiency of HIV prevention services using standardised methods is necessary to make optimal use of the increasing resources that are becoming available for this purpose in India and other parts of the developing world.
Competing interests
The author(s) declare that they have no competing interests.
Authors' contributions
LD led the PANCEA study in India, guided the design, data collection and analysis, and wrote the initial draft of this paper. PS contributed to the design, data collection and analysis. TLNP and EM contributed to the design and advised on data collection, analysis and presentation. MCR, AAK, SGPK, YKR and SM contributed to data collection and analysis. MO contributed to the design and advised on analysis. JGK oversaw the PANCEA design and contributed to the analytic design and presentation. All authors read and approved the final manuscript.
Presentation at meeting
This paper was presented at the Annual Meeting of the Global Forum for Health Research – Forum 9, Mumbai, India, 12–16 September 2005.
Pre-publication history
The pre-publication history for this paper can be accessed here:
Acknowledgements
We thank the Andhra Pradesh State AIDS Control Society for facilitating this study, and the staff and clients of the STI clinics/hospitals for participating in this study. This study was supported by the World Bank contract 7130248 and by the U.S. National Institutes of Health through Task Order #7 contract 282-98-0026 and grant R01 DA15612. The views expressed in this paper are those of the authors and do not necessarily reflect the views of the funding agencies, organisations that facilitated this study, or the institutions employing the authors.
==== Refs
Joint United Nations Programme on HIV/AIDS (UNAIDS) 2004 Report on the Global AIDS Epidemic 2004 Geneva
National AIDS Control Organisation (NACO) HIV Estimates 2003 New Delhi: NACO, Ministry of Health & Family Welfare, Government of India
Sangani P Rutherford G Wilkinson D Population-based interventions for reducing sexually transmitted infections, including HIV infection Cochrane Database Syst Rev 2004 2 CD001220 15106156
National AIDS Control Organisation (NACO) Observed HIV prevalence levels state wise: 1998–2004 New Delhi: NACO, Ministry of Health & Family Welfare, Government of India
Rao JVRP Ganguly NK Mehendale SM Bollinger RC India's response to the HIV epidemic Lancet 2004 364 1296 1297 15474121 10.1016/S0140-6736(04)17205-X
National AIDS Control Organisation (NACO) National AIDS control programme phase II (1999–2006) New Delhi: NACO, Ministry of Health & Family Welfare, Government of India
Bill & Melinda Gates Foundation Avahan: India AIDS Initiative
Dandona L Enhancing the evidence base for HIV/AIDS control in India Natl Med J India 2004 17 160 166 15253404
Prevent AIDS Network for Cost-Effectiveness Analysis Home page
Marseille E Dandona L Saba J McConnel C Rollins B Gaist P Lundberg M Over M Bertozzi S Kahn JG Assessing the efficiency of HIV prevention around the world: methods of the PANCEA project Health Serv Res 2004 39 1993 2012 15544641 10.1111/j.1475-6773.2004.00329.x
Reserve Bank of India Exchange rate of the Indian Rupee: Table 145
Klein LR An Introduction to Econometrics 1977 Englewood Cliffs, USA: Prentice-Hall
Mahal A Singh J Afridi F Lamba V Gumber A Selvaraju V Who benefits from public health spending in India: results of a benefit incidence analysis for India 2002 New Delhi: National Council of Applied Economic Research
Walker D Cost and cost-effectiveness of HIV/AIDS prevention strategies in developing countries: is there an evidence base? Health Policy Plan 2003 18 4 17 12582104 10.1093/heapol/18.1.4
Gilson L Mkanje R Grosskurth H Mosha F Picard J Gavyole A Todd J Mayaud P Swai R Fransen L Mabey D Mills A Hayes R Cost-effectiveness of improved treatment services for sexually transmitted diseases in preventing HIV-1 infection in Mwanza Region, Tanzania Lancet 1997 350 1805 1809 9428251 10.1016/S0140-6736(97)08222-6
World Bank Project appraisal document on a proposed credit in the amount of SDR 14082 million to India for second national HIV/AIDS project 1999 Washington, DC
Dandona L Sisodia P Ramesh YK Kumar SGP Kumar AA Rao MC Someshwar M Hansl B Marshall N Marseille E Kahn JG Cost and efficiency of HIV voluntary counselling and testing centres in Andhra Pradesh, India Natl Med J India 2005 18 26 31 15835489
Dandona L Sisodia P Kumar SGP Ramesh YK Kumar AA Rao MC Marseille E Someshwar M Marshall N Kahn JG HIV prevention programmes for female sex workers in Andhra Pradesh, India: outputs, cost and efficiency BMC Public Health 2005 5 98 16181491 10.1186/1471-2458-5-98
|
16271151
|
PMC1291366
|
CC BY
|
2021-01-04 16:31:49
|
no
|
BMC Health Serv Res. 2005 Nov 5; 5:69
|
utf-8
|
BMC Health Serv Res
| 2,005 |
10.1186/1472-6963-5-69
|
oa_comm
|
==== Front
BMC Health Serv ResBMC Health Services Research1472-6963BioMed Central London 1472-6963-5-701628196710.1186/1472-6963-5-70Research Article"Saying no is no easy matter" A qualitative study of competing concerns in rationing decisions in general practice Carlsen Benedicte [email protected] Ole Frithjof [email protected] Health Economics, Bergen, Stein Rokkan Centre for Social Studies, The University of Bergen, Nygårdsgaten 5, 5015 Bergen, Norway2 Professor, The Department of Public Health and Primary Health Care, Section for General Practice, The University of Bergen, Kalfarveien 31, 5018 Bergen, Norway2005 9 11 2005 5 70 70 27 4 2005 9 11 2005 Copyright © 2005 Carlsen and Norheim; licensee BioMed Central Ltd.2005Carlsen and Norheim; licensee BioMed Central Ltd.This is an Open Access article distributed under the terms of the Creative Commons Attribution License (), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
Background
The general practitioner in Norway is expected to ensure equity and effectiveness through fair rationing. At the same time, due to recent reforms of the Norwegian health care sector, both the role of economic incentives and patient autonomy have been strengthened. Studies indicate that modern general practitioners, both in Norway and in other countries are uncomfortable with the gatekeeper role, but there is little knowledge about how general practitioners experience rationing in practice.
Methods
Through focus group interviews with Norwegian general practitioners, we explore physicians' attitudes toward factors of influence on medical decision making and how rationing dilemmas are experienced in everyday practice.
Results
Four major concerns appeared in the group discussions: The obligation to ration health care, professional autonomy, patient autonomy, and competition. A central finding was that the physicians find rationing difficult because saying no in face to face relations often is felt uncomfortable and in conflict with other important objectives for the general practitioner.
Conclusion
It is important to understand the association between using economic incentives in the management of health care, increasing patient autonomy, and the willingness among physicians to contribute to efficient, fair and legitimate resource allocation.
==== Body
Background
The demand for health care services is rising in all Western countries and governments are concerned with controlling costs and ensuring a fair allocation of resources [1,2]. The general practitioner (GP) is increasingly regarded as holding a key role in securing equity and effectiveness [3]. In the wake of government concern, a number of studies of rationing and the role of gatekeepers in health care have appeared [4-6].
Rationing can be defined as the allocation of scarce resources between patients with competing needs [7]; it implies the withholding of potentially beneficial health care through deliberate choice or through financial or organisational features of the healthcare system in question [8]. Related to studies of rationing is the concept of opportunity cost, which can be defined as the value of the best alternative which is forgone in order to get or produce more of the commodity under consideration [9]. In health care decision making this signifies that any use of health care resources should be viewed as denying the opportunity for some other patient to use the money for potentially greater benefit [10].
The view of physicians as rationing agents, commonly accepted by health economists and health administrators, can be contrasted with a more traditional patient advocacy view, where the physician is seen first of all as representing the best interests of each individual patient. Thus, there is often said to be a conflict between the role of physicians as gatekeepers and as patients' advocates. This simple dichotomy is, however, misleading. Many commentators have pointed out that it is possible to hold multiple roles in the physician-patient relationship and that what role one adopts is highly context sensitive [11-13]. In fact, GPs are expected to both take care of the individual patient's need and at the same time take account of common resource use. In Norway, for example, this dual responsibility is clearly stated in the Norwegian Medical Association's ethical guidelines [14].
The gatekeeper role is one of many dilemmas. A Norwegian study from 1993 showed that among a selection of general practitioners, 93% had experienced a conflict between responsibility towards the individual patient and the requirement to manage the health budget [15]. Both British and North American studies of physicians' attitudes to rationing in their practices indicate that GPs are increasingly uncomfortable with the gatekeeper role [5,16,17]. In the case of Norway, we know that Norwegian GPs are reluctant to function as gatekeepers [18], referral rates are increasing [19], and that guidelines for rationing seldom are adhered to [20].
The Norwegian government has expanded the use of economic incentives and market mechanisms in the health sector, while still relying on GPs to filter access to a number of specialised services including both drug and non-drug treatments. In some areas, e.g. the prescription of some common drugs covered by the National Insurance Scheme, the government provides guidelines for rationing. Yet in most instances (e.g. referrals), the GPs are expected to use their best professional judgement to secure an effective and fair allocation of resources.
In 2001 a list-based system was introduced in general practice covering practically the whole population. In addition the payment system was altered from a combination of fee-for-service payment and a practice allowance component, to the current system where the fixed component is replaced by a capitation part. Every GP is granted a fixed remuneration for each person listed, while the rest of the income comes from activity based remunerations and a small out of pocket fee. The Norwegian GPs have no budget responsibility (as e.g. in the UK fund-holding system). The introduction of the patient-list system was seen as a step towards increased patient rights and adequate access to health care. In 1999 The Patients' Rights Act was passed which accentuate the patients' autonomy, e.g. the right to free choice of hospitals and the right to be informed and involved in medical decision making [21]. We have described elsewhere how the new incentive structure enhances competition and motivation for fulfilling patients' expectations and thereby weakening the GPs commitment to the gatekeeper role [22].
There is an ongoing debate among scholars about how to understand GP behaviour in clinical decision making. In principal-agent theory, the GP is seen as a rational actor who maximizes utility, but there has been considerable divergence in which motivators that are considered relevant in the utility function [3,23]. Alternative models criticise the principal-agent theory for ignoring the influence of norms and altruism in physicians' discretionary choices. The bottom line of empiric research seems to be that economic incentives do have an effect on GPs discretionary choices, but other factors, especially medical considerations and professional norms seem to be more influential [23,24]. Some authors have warned that the use of economic incentives in the management of physicians may backfire and ultimately undermine professional norms [25,26].
In order to achieve the goals of efficient and fair resource allocation, it is vital to understand how the agents of rationing, in this case GPs, experience and manage their gatekeeper role. If the gatekeepers find it impossible to act according to common principles for rationing and priority setting, there will be no rationing or rationing will be haphazard and unfair. However, we know little about physicians' experience of and attitude to these reported rationing dilemmas in general practice [16,18].
The aim of this article is to explore how Norwegian GPs experience rationing dilemmas in everyday practice. In the context that our study was conducted, we anticipated economic incentives and patient claims to be of importance in GPs decision making, but we were open for other suggestions from the interview participants and were aiming at finding out what factors where important in the view of the GPs themselves. We were also interested in understanding how one motivating factor is balanced against another.
Materials and methods
This empirical study relies on data collected in 11 focus group interviews with a total of 81 participants, all of whom were Norwegian primary care physicians. The interviews took place in spring 2002 as part of a national evaluation study in connection with the introduction of a capitation-based patient list system in primary health care.
A purposive sample was recruited from tutorial groups and specialists' continuous education groups in the counties of Hordaland and Oslo. Both rural and urban municipalities are represented (67.5% of the participants had their practice in an urban municipality). The sample was chosen to represent well-known differences and typical practices. In the final sample, 58% were male and length of experience as a GP varied between 0.5 and 27 years with a mean of 11 years. We invited 23 groups to participate by sending a letter by mail to the group co-ordinators, and followed up the invitations with telephone calls. Of these, 13 groups wished to participate, two declined and eight were difficult to reach by telephone. Among the 13 positive answers, 11 groups were interviewed. The groups were homogenous with respect to age and work experience and, with the exception of one group, balanced with respect to gender.
The data were collected through semi-structured focus group interviews and a short questionnaire given to all participants at the start of each interview. The interview guide was tested in the first group interview, after which a number of changes in the wording were decided. After this, following the recommended procedure in the framework approach [27], the questions were not changed again. The interviews were conducted by the authors; a social scientist trained in social anthropology and a GP and professor in medical ethics. Interviews lasted between one and two hours, and the discussion was audio taped and transcribed for subsequent analysis. Each interview began with an introduction by one of the researchers who clarified the study's focus and central concepts after which discussion between the participants was encouraged.
The introduction presented our underlying assumption about discretionary choices [28]. The main message was as follows: "Even when practice is based on the methods of evidence-based medicine, numerous grey areas and uncertain indications for treatment remain. Physicians are therefore key actors in rationing decisions and the scope for discretion is wide. It is in these grey zones of discretionary choices that everyday decision making shapes the content and volume of medical practice."
We focused on three types of decisions where there are substantial grey areas and where the scope for discretion is known to be wide:
- The prescription of reimbursed drugs
- The referral of patients to secondary care
- The issuing of sickness certificates
The physicians then discussed how they perceived that different factors influence their professional discretion. We used an interview guide with 12 questions focusing on the informants' view of the gatekeeper role, how to balance the obligation to serve the individual patient vs. rationing and their experience with difficult rationing decisions. We asked specifically of how the relationship with patients and economic incentives affect how they make rationing decisions. As the focus of the study was rationing dilemmas, the participants were encouraged to consider the problematic cases.
Analysis
The transcripts were coded by hand. Initially the statements in the transcripts of the first four interviews were condensed as keywords and short phrases based on both the original framework and new themes introduced by the informants. This was conducted separately by the two authors. Then the two sets of coded texts were compared and the researchers agreed upon a system of categorisation based on recurrent subjects in the keywords. The analytical steps were inspired both by Kvale's eminent description of coding and categorisation [29] and by Ritchie and Spencer's analytical steps in the framework approach [27].
The quotes presented here are not necessarily the best formulated or the most striking. Rather we have tried to select representative statements that were not too long and that can be understood when separated from the context. We have translated the statements from Norwegian to English.
Ethical considerations
The project has been reviewed and approved by the Norwegian Social Science Data Service against the privacy and license requirements of the Personal Data Registers Act and the guidelines for research ethics in the social sciences, law and the humanities according to the National Committee for Research Ethics in the Social Sciences and the Humanities. All respondents were informed about anonymity issues, their right to withdraw from the study and the purpose of the study.
Results
When describing discretionary choices, respondents' answers centred around four major topics:
- The obligation to ration health care
- Professional autonomy
- Patient autonomy
- Competition
In the following section, we report the results of our analysis of the four major issues of special concern. Many of the citations are related to more than one of the major topics. This is quite illustrative of how the different factors seem to be constantly weighed against one another but also intertwined when the informants are describing incidents of making discretionary choices.
The obligation to ration health care
Participants described many cases of how demanding patients induce them to act as patients' advocate and neglect the gatekeeper role, although some of them stressed that in most instances they have no problems balancing the two roles, for instance through informing and reasoning with the patient. Others stated that they experience a predicament between rationing on society's behalf and offering the best service available to the consulting patient. In the words of one of the participants: Saying no is no easy matter.
Below three informants are discussing reasons for giving in to patients' claims for referrals:
Informant 1: The patients who want a referral to a specialist have made up their mind beforehand that that's what they want. And then I find it quite difficult to explain to them that they don't need it.
Informant 2: And it takes twice as much time.
Informant 3: I feel in a way that I have to offer good service. I don't want anyone to leave my list. I don't want a bad reputation, do I? I live in the same place that I work and I want to please people.
(Three participants in a spesialists' group, Oslo)
Many times informants exclaimed that they just can not bring themselves to say no when patients are insistent. Instead of saying no, other strategies are used to evade conflict as this story illustrates:
I often find it hard and very time consuming to get people to change their minds. The most recent example that comes to mind was an elderly lady. I finally ended up saying, "Yes, I think that before we go any further I should refer you to a rheumatologist who can assess your need for this treatment, and then we can talk." It was perhaps a bit cowardly of me. I could have told her, "You won't get that treatment from me and if you want me to be your doctor that's the way it's going to be." But I couldn't face doing that so I referred her; so now she has to go to the rheumatologist and then come back to me and then we'll see. Hopefully he agrees with me.
(Male GP specialist, 26 years of experience as a GP)
Another finding, however, was a lack of awareness among several of the respondents about the need for rationing and the reasons behind government guidelines and regulations. Some said that they do not see the point in rationing when the costs involved are small.
I believe that as GPs we very rarely consider the aspect of public finances, that we for instance refrain from making a referral because of high public costs. We usually use more rational arguments related to what is best for the patient, then we argue that the patient will not benefit much from a certain referral and therefore we will not refer. We try not to think too much that this implicitly means less public expenditure, so that only to a small degree do we make our decisions based on the common good.
(Male GP, 4 years of experience as a GP)
Many were unsure of what are expected of them as gatekeeper. Several seemed to have solved the quandary by concluding that when a GP spends time and effort in trying to convince the patient that there is no indication for a requested intervention, the GP is an active gatekeeper even if the result is that the patient gets his or her own way in the end.
But aren't we gatekeepers when we put up barriers for people by discussing things and putting forward arguments, even when the conclusion is that the person gets what he wants? Isn't that another way of being a gatekeeper?
(Male GP, 3 years experience as a GP)
Professional autonomy
Most of the informants were concerned about the degree of professional autonomy and freedom in their work, both in rationing decisions and in how to run their practice. Hence the concept of professional autonomy as it is used here is dual, incorporating autonomy in relation to patients and autonomy in relation to the health authorities.
Some stated that the relative freedom of GPs compared to for example hospital physicians was an important reason for their initial career choice. During interviews, some informants initially claimed that their professional autonomy is not at all complicated by patients' demands and government guidelines for rationing, but during the discussions it became evident that many informants are fearful that they are gradually losing their traditional freedom and autonomy in clinical decision making. Many illustrative examples were given of how difficult it is to maintain an autonomous position.
Early in the interviews several informants claimed that, in situations where there is room for discretionary choice, they are influenced neither by patients nor by health authorities.
I would never compromise my own judgement and give patients what they want all the time just to keep them on my list. [...] I have to say that professional concerns always come first, before that sort of selfish need.
(Female GP specialist, 20 years of experience as a GP)
In contrast a few informants explained how they refer to government rules and guidelines because it makes it easier to refuse demands from patients.
If they come in and say that they're depressed and so on, it can be difficult, but it's tidier if you follow the rules. So I prefer to tell them that these are the rules instead of making exceptions in individual cases, because that's a dangerous path.
(Female GP specialist, 16 years of experience as a GP)
There were several informants who conveyed a feeling of conflict between professional autonomy and rules and guidelines set by central health authorities. One example much discussed was the use of statins and anti-hypertensives:
Question: Do you find that the limits the government has set are clear when it comes to anti-hypertensives and statins?
Informant: No. I tend to stick to my own, let's say, professional judgement and not to any official limitations. That is definitely the case.
(Female GP specialist, 15 years of experience as a GP)
Another matter is the kind of public finance decision about which level of spending to choose. How much money does Norway as a society want to spend on anti-hypertensives, and how much does Russia want to spend? It's a political decision. The reason why it's sensible to have a spending limit in a country is of course that if all physicians have different limits and some of them are treating patients too aggressively while others are not treating aggressively enough, then the cost-benefit of the use of medicines is less, and it becomes more or less random which group of patients gets help. If there had been an absolute limit, and the limit was set at a reasonable level, maybe not a cholesterol level of 8.2 but let's say 7.5, then the physicians would have to respect the rule and the patients who did not fulfil this criterion but still were eager to get treatment could pay for it themselves. [The levels as they are set today] are not respected by GPs, and that may be because the criteria are unrealistic, and we are trying to satisfy our patients.
(Male GP specialist, 4 years of experience as a GP)
Others said that the rules for remunerated prescribing are perfectly clear and that they never bend them. However, such statements tended to be moderated or even contradicted later in the discussion. Sometimes one of the participants provoked this openness by admitting to bending the rules for remunerated prescriptions or to referring without sufficient medical indications.
A fear of losing professional autonomy because of economic incentives and competition was expressed by many of our informants, although not always in a straightforward or personal manner. Below are two quotes from the same informant, the first early in the interview and the second later as the discussion had accelerated:
1: Our decisions are always based on our professional judgement. The reform [the new capitation- and list-based system] has had no influence whatsoever.
2: A relatively serious drawback [of the reform] is that you are forced to become a peddler. At the start it made the hairs at the back of my neck stand on end. It feels very unpleasant. Once you've gone along with this premise I think it's a terribly sad situation for our health services as a whole. It has huge consequences ...
(Male GP, 6 years of experience as a GP)
Patient autonomy
There were repeated claims that patients have become better informed about their rights as patients, and that they appear increasingly demanding. When discussing why they give in to patients' demands even when the claims are not clearly within what would be defined as medically necessary; many referred to the ideal of patient autonomy and patient centred care which implies respect for patients' subjective experience and sharing decisions with patients.
What we see as trivial and marginal can be very important for the patient for some reason or other that is not immediately clear. When it comes to referrals for instance, it's incredible how manywomen want to see a gynaecologist for no apparent reason. Most are persuaded not to, but some have had an aunt who had ovarian cancer and they tell you that they're worried because of that. One woman told me that she had found a whole new life after she started visiting the gynaecologist once a year. Then it is not marginal after all.
(Female GP, 4 years of experience as a GP)
I have not experienced much conflict when it comes to saying "no", maybe because we manage to rationalise it so that it becomes reasonable to say "yes", i.e. you refer a patient for a CT scan, or something like that, even though it is obvious to you that medically the patient doesn't need it. But they can become quite anxious, and this is one of your regular patients, and you know that one way of getting rid of the problem is to do it even though you know that medically it is absolutely unlikely that they will find anything. And of course there is always a risk that they do find something. I have been asking myself whether I am able to give a categorical "no" to a demanding patient with whom I expect to have a long-term relationship.
(Male GP specialist, 21 years of experience as a GP)
A common explanation for why it is difficult to say no was the need for social approval.
What's quite typical for the way some colleagues work, and maybe sometimes for the way we work too, is that we're interested in getting the patient admitted to the right place, to the right hospitals, to the right specialists, because then we get happy and satisfied patients.
(Male specialist GP, 20 years of experience as a GP)
The participants were also concerned about avoiding conflicts with patients. The physicians described in detail how they manage to avoid conflict by explaining their medical opinion and sharing decisions with patients. The majority claimed that talking with and convincing the patient virtually kept them out of conflict with patients.
Most often, to avoid conflict, I try to get the patient to share my view.
(Male GP, 3 years of experience as a GP)
However, they often give in when the patient is not convinced:
You might call it gatekeeping if you could explain to him [a patient] that he could manage without the CT scan that it ended up with. But it was quite obvious that to him having the scan was a reasonable way of getting rid of the worry. So that is probably the way it works. You could define gatekeeping as pushing it until you see that there is a real need that is reasonable. Then you rarely get into that conflict. But you would get into that conflict if you received a message from the radiology department telling you that the monthly budget was up, but we don't, you know.
(Male specialist GP, 21 years of experience as a GP)
Competition
The physicians generally agreed that they find themselves increasingly drawn into a health care market, where patients act as demanding consumers and they as physicians compete to please these consumers.
Some patients put pressure on the physician to write out a sick leave certificate or prescribe when it perhaps isn't necessary. And if you don't do it, you might be pressured and they might leave for another GP.
(Male GP, 2 years of experience as a GP)
For others it was problematic to admit to being influenced by economic incentives. One of the informants initially answered "no" to our question whether the economic and social incentives in the patient list system influenced his discretionary choices, and he continues:
1: It can't be that one suddenly should treat patients differently in consultations. That would be strange, wouldn't it? Because I believe that would mean that we are driven by organisational moves rather than our own medical understanding.
(Male GP, 3 years of experience as a GP)
Later in the interview session, the same informant comments on the same question:
2: I think maybe that GPs are behaving differently [after new incentives were introduced]. Why on earth one would do that is of course a question, but ...
A general impression from the interviews was willingness to departure from adherence to government rules for e.g. reimbursed prescriptions, in order to satisfy the patients:
I suppose we have all prescribed cortisone creams and asthma medication and things like that as standard prescriptions [i.e. covered by the National Insurance Scheme]. Isn't that right? I mean, the combination of regular patients and the demands made by consumers in the health care market, they won't wait or take no for an answer, right? And I am not sure we are good at setting limits to control it.
(Male specialist GP, 21 years of experience as a GP)
One reason given for not adhering to the remuneration guidelines was that it is pointless to ration if patients can get what they want from somebody else. This argument clearly also touches on the element of competition for patients.
A new patient came in who had several things she wanted me to look at. I had just given her three prescriptions and then she adds that she just wanted a prescription she had had before, which I didn't know about as she was a completely new patient. So I wrote out a private prescription [i.e. not covered by the National Insurance Scheme] for her, and then she says, "But I usually get a standard prescription!" Then her time was up, and I thought "I can't stand listening to all the details and discussing them," so I just said "Fine" and wrote her the standard prescription.
(Female specialist GP, 15 years of experience as a GP)
Competing and compatible motivators
A key finding is that the obligation to ration health care is not generally embraced by GPs. Professional norms and respect for patient autonomy on the other hand, are strongly emphasized as integrated and at the very core of their professional judgement. When it comes to the role of economic incentives, the statements are more ambiguous. The idea of economic incentives is sometimes dismissed as an external factor without power to influence GP decision making whatsoever, while some informants state that they are worried that market mechanisms are gradually undermining professional autonomy.
General practitioners in our study find themselves caught in a web of conflicting concerns presented here as the four main topics. Some of these motives are easy to combine and some are often experienced as conflicting. In the paragraphs about rationing and patient autonomy we reported the participants descriptions of how a firm professional autonomy in relation to patients is experienced as a necessary requirement to fulfil the rationing role, because in the current system with universal coverage by a third party payer and free choice of healthcare provider combined with capitation and competition for patients, it is tempting to go along with patients' wishes. At the same time rationing is often presented as in conflict with respect for patient autonomy. On the other hand, as the findings in the paragraph about competition suggest, respect for patient autonomy is easily combined with the situation of competition and the economic incentives of fee-for-service and capitation. So a central line of conflict seems to be how to balance the obligation to ration health care combined with professional autonomy in relation to patients on the one side and the demand for patient autonomy combined with concerns about competition for patients on the other.
Discussion
Summing up, the central finding was that when the physicians discuss why rationing can be difficult, they emphasize that saying no in face to face relations often is felt uncomfortable and in conflict with other important objectives.
In a health care system with universal coverage and gate keeping through general practice, rationing presupposes a relatively high level of professional autonomy in relation to patients, but this is often in conflict with other ideals and motives faced by the modern physician, such as respect for patient autonomy. This finding supports other studies in Norway and abroad, describing a dilemma between the roles of gatekeeper and the patient's advocate [15]. In addition, there seems to be a lack of understanding of the rationale behind rationing accompanied by a low degree of adherence to government guidelines. There is apparently a lack of clarity in the signals from the health authorities on how principles for priority setting and rationing should be applied in primary care, but part of the explanation seem to lie in the GPs' concern for upholding their professional autonomy in relation to the health authorities.
It is interesting to ask whether physicians' avoidance of explicit rationing is related to the evolving shift in power from GP to patient. According to Daniels [30], already twenty years ago physicians in the US found it difficult to say no to patients. In the NHS/Scandinavian type of health care system patients have been used to GPs acting as gatekeepers, but now the Norwegian system is changing and patients seem to be beginning to demand their rights as consumers. Physicians, both as a group and in their meeting with patients, are experiencing a decline in status and a loss of autonomy. This is a widely recognised phenomenon and a recurring concern in medical professional discourse [31-34]. At a theoretical level the problem is elegantly solved – the modern physician should seek partnership with patients, ground decisions in evidence-based medicine and surrender to transparency and accountability in their practice [2,34-36] – but our results suggest that things are not so simple in practice. When the interviewees discuss how it is sometimes experienced as very difficult to say no to insistent patients, it implies that rationing in practice and especially in face-to-face relationships is quite another task than priority setting and rationing at the macro level. For many GPs, it is problematic even to talk about such divergent and conflicting motives.
Interestingly, there are studies, especially those focusing on patients' experiences that find that patients seldom challenge the physician's autonomy [2]. Some studies of the patient's role claim that patients do not experience themselves as powerful in meetings with the physician [2]. This is also demonstrated in a Norwegian study of the illness experience of Norwegian chronic back pain sufferers [37]. However, there is not necessarily any contradiction in divergent views between physicians and patients. In fact, it has been shown in several studies of doctor-patient communication that whether the physician experiences pressure from the patient is a more powerful predictor of the physician's actions than the actual preferences of patients [38-40]. This might imply, as suggested by Rogers [41], that physicians' stories of demanding patients are sometimes an excuse when the actual explanation is to be found in private economic motives or low awareness of the need for rationing. Elston [36], in her criticism of the deprofessionalisation thesis, also supports this view. She claims that the so-called deprofessionalisation and alleged loss of power of the medial profession is exaggerated. It is not so much a question of whether physicians are capable of carrying out rationing as whether they want to. Likewise Frankel and colleagues [42] suggest that part of the claims of unsatisfied demand in the NHS spring from professional self interest of the NHS-employees.
The ethics of the profession still emphasise that to be a physician is not like other jobs, and the altruistic professional is contrasted with the amoral and therefore untrustworthy expert governed by market forces [43,44]. In this context, being a physician is still presented as a vocation to do good, as an art of caring for the individual patient's needs. The physician as rationing agent is less discussed. The introduction of "anti-professional" incentives such as external surveillance and control as well as the incentives of consumerism can be viewed as the start of a negative spiral [36], as some of the interviewees also indicate. At the same time the modern physician is expected to respect patients' or consumers' autonomy in clinical decision making and to practise patient-centred medicine, which is associated with more satisfied patients and a better health outcome [44,45]. According to this influential research, in clinical decision making the physician should also consider the patient's experience of the symptoms and include such subjective factors in addition to the biomedical facts. Keeping this context in mind it is not difficult to comprehend the informants' ambiguous attitudes toward increasing consumerism in health care and why professional autonomy might be experienced as a competing concern to patient autonomy.
Strengths and limitations of the study
Focus groups were chosen in preference to other qualitative methodologies because they have been shown to be a useful method of revealing attitudes to complex and sensitive topics [46], partly because the more reserved participants are encouraged by participants who are less inhibited [53]. Participants should have a similar background in relation to the topic to be discussed. In a group of peers where the participants are in the majority, it appears that they are more able and willing to clarify and express thoughts that are difficult to reveal in one-to-one interviews [47,48]. This enhances the common ground and clarifies the prevailing opinions within the specific group under study, as well as the rationale behind them. Thus the focus group interview has been reported to have advantages when the purpose is to reveal common norms in groups of peers, as for example professional groups [53]. On the other hand, focus groups may not be suitable for exploring individual opinions that diverge from group norms [53]. The group context will sometimes scare participants from voicing disagreement.
The main purpose of this study was to reveal common experiences and professional culture regarding the gatekeeper role rather than to investigate individual cases or register the whole spectre of opinions. We therefore decided that focus groups would be more appropriate than for instance individual interviews. we did, however, decide to register how much each participant spoke during the focus group meetings to get an indication of the level of dissent voiced in the groups. We found that the proportion of speech varied extensively between participants, and that those who expressed the strongest opinions spoke the most. This picture is in agreement with our general impressions of the interviews, and it appears that the participants were not afraid of confronting each other's opinions. While it would also have been an advantage to know what was left unspoken, The GPs seem to have a strong feeling of professional loyalty that might have prevented interviewees from revealing opinions that are at odds with the collegial culture. It is therefore doubtful whether this type of data would have emerged if we had had the opportunity to interview all the participants individually afterwards.
Focus groups have a limited value in drawing conclusions on the distribution of different opinions in the population. The sample of GPs participating in these interviews is similar to the general populations of Norwegian GPs according to a few measurable characteristics, as we have described elsewhere [22]. However this is not a randomised sample and we are not aiming at calculating the proportion of Norwegian GPs that would agree with one or all of the arguments presented here. Rather it is reasonable to expect that our interpretations shed light on some of the important elements that influence the countless daily rationing decisions made by modern GPs.
During the interviews there were several indications that we had struck on a sensitive subject with our focus on discretionary choices and how physicians are influenced by patients and economic incentives. In some sense we were questioning their professional autonomy. A fairly recent study by Pearson and Hyams suggests that physicians are reluctant to discuss conflicts of interest in medical decision making [49]. It might be argued that our role in the groups was adding to the complications of a difficult topic. Though not explicitly stated, our introduction and the context of the study, (we were funded by the Ministry of Health through the Norwegian research council to assess how the influence of the patient-list system affects rationing – decisions), gave the participants reason to believe that we could be promoting the need to ration health care and would expect the GPs to be rationing agents in accordance with governmental policy and guidelines.
One indication that the topic was felt to be uncomfortable appears when contrasting the initial answers with the statements made as the participants engaged in group discussion. Several times participants started out by rejecting the focus of the study as irrelevant to their own practice experience. Others said that they did not experience any form of predicament themselves as they were simply practising medicine and were not influenced by economic or other external factors in their discretionary choices. In spite of the slow start of the interviews, we found that attitudes changed throughout each interview from at times quite robust and categorical negative responses to the question of possible dilemmas in medical decision making to a more nuanced and positive response and enthusiastic discussion as the interview proceeded. An equivalent response is reported by Marshall et al. in a focus group study of quality assessment in primary care [50]. Marshall et al. term this "initial vs. considered response". It may well be argued that what we observe is simply a sign of the participants considering the questions in depth for the first time and thus changing their answers. This may also be a characteristic quality of the group process that, together with the comfort of being among peers and outnumbering the researchers, explains why group interviews are an appropriate tool for studying sensitive topics.
We interpret the contradictions and initial unwillingness to discuss discretionary choice as a sign of conflicting ideals, of which the participants are either conscious or not. It is apparent that it is not unproblematic morally to be influenced by economic incentives, and even being influenced by patients' wishes seems to be difficult to acknowledge.
A limitation of the study is that it does not give us any certain knowledge about the actual behaviour of the physicians. Informants may not describe their actions correctly, either because they are unaware of or do not remember their own pattern of actual choices or because they do not wish to reveal them [51] and according to Fernadez et al. [52] GPs consider it legitimate to be influenced by professional standards while to be influenced by financial incentives is not considered legitimate. In this Spanish survey, the responding GPs considered that they were most influenced by the most legitimate factors and least influenced by the least legitimate factors.
A final point to keep in mind is that since rationing as a subject was introduced by the researchers as part of the stated starting point for the group discussions, it is not possible on the basis of these data to estimate how much of the physicians' consciousness is devoted to rationing in their daily practice.
Conclusion
Based on the discussions within our selection of Norwegian general practitioners, we extract some competing concerns that complicate rationing decisions in primary care and make it difficult to say no to patients who insist on receiving services even when their demands are in conflict with a commitment to legitimate and fair resource allocation. Two central dilemmas are:
• It is often experienced as difficult to make rationing decisions within the context of patient centred medicine
• The current economic incentives do not combine well with making rationing decisions.
Although we do not claim that the informants in this study are representative of GPs everywhere and the findings in this study must be interpreted within the special organisational context of the current Norwegian primary care system, we still believe that it yields interesting knowledge about how different concerns are balanced against one another. The main factors we found influencing these GPs are quite similar to what other studies find both in the Norway and abroad. This and other studies suggest that in a publicly financed health care system we need a better understanding of the relationship between incentives used in the management of primary care, increasing patient autonomy and the willingness among physicians to contribute to efficient, fair and legitimate resource allocation.
Competing interests
The author(s) declare that they have no competing interests.
Authors' contributions
BC designed the study. BC and OFN collected and analysed the data. BC drafted the manuscript. BC and OFN revised and finally approved the manuscript.
Pre-publication history
The pre-publication history for this paper can be accessed here:
Acknowledgements
The study was financed by the Research Council of Norway.
==== Refs
H Hope T Hicks N Reynolds DJM Crisp R Griffiths Sope T Rationing and the health authority BMJ 1998 317 1067 1069 9774299
Coulter A Partnerships with patients: the pros and cons of shared clinical decision-making Journal of Health Services Research and Policy 1997 2 112 21 10180362
Scott A Culyer AJ, Newhouse JP Economics of General Practice Handbook of Health Economics 2000 Amsterdam, Elsevier Science 1175 1200 10.1016/S1574-0064(00)80035-9
Baines DL Tolley KH Whynes DK The ethics of resource allocation: The views of general practitioners in Lincolnshire, UK Social Science & Medicine 1998 47 1555 64 9823051 10.1016/S0277-9536(98)00256-1
Forrest CB Primary care in the United States: Primary care gatekeeping and referrals: effective filter or failed experiment? BMJ 2003 326 692 5 12663407 10.1136/bmj.326.7391.692
Willems DL Balancing rationalities: gatekeeping in health care Journal of Medical Ethics 2001 27 25 9 11233373 10.1136/jme.27.1.25
Mechanic D Health Care, Rationing of International Encyclopedia of the Social & Behavioral Sciences 2001 Oxford, Elsevier Science 6534 7
Norheim OF Tingle J, Foster C Clinical guidelines. Health care rationing and accountability for reasonableness Clinical Guidelines: Law, Policy and Practice 2002 London, Cavendish Publ 161 80
Folland S Goodman AC Stano M The economics of health and health care 1997 New Jersey: Prentice-Hall 603
Drummond M Stoddart G Torrance G Methods for the economic evaluation of health care programmes 1997 Oxford, Oxford University Press
Emanuel EJ Emanuel LL Four models of the physician-patient relationship JAMA 1992 267 2221 6 1556799 10.1001/jama.267.16.2221
Rogers WA Whose autonomy? Which choice? A study of GPs' attitudes towards patient autonomy in the management of low back pain Fam Pract 2002 19 140 5 11906978 10.1093/fampra/19.2.140
Szasz TS Hollender MH A Contribution to the Philosophy of Medicine. The Basic Models of the Doctor-Patient Relationship Archives of Internal Medicine 1956 97 585 92
The Norwegian Medical Association. Etiske regler for leger (Ethical guidelines for physicians) 1961, 2002
Arnesen T Fredriksen S Coping with obligations towards patient and society: an empirical study of attitudes and practice among Norwegian physicians Journal of Medical Ethics 1995 21 158 61 7674280
Ayres PJ Rationing health care: Views from general practice Social Science & Medicine 1996 42 1021 5 8730907 10.1016/0277-9536(95)00213-8
Bindman AB Majeed A Primary care in the United States: Organisation of primary care in the United States BMJ 2003 326 631 4 12649238 10.1136/bmj.326.7390.631
Gulbrandsen P Førde R Aasland OG Hvordan har legen det som portvakt? (What does it feel like to be a gatekeeper?) Tidsskr Nor Laegeforen 2002 122 1874 9 12362710
Helsedepartementet. Hovedtrekk i fastlegeordningen etter to og et halvt år – foreløpig oppsummering av statusrapport. (The Ministry of Health: Summarising two-years' report of the Regular General Practitioner Scheme) 2002
Treweek S Flottorp S Fretheim A Håvelsrud K Kristoffersen DT Oxman A Retningslinjer for allmennpraksis – blir de lest og blir de brukt? (Guidelines in general practice: are they read and are they used?) Tidsskr Nor Laegeforen 2005 125 300 3 15702152
Pasientrettighetsloven (The Patients' Rights Act. Act of 2. July 1999 no. 63 relating to Patients' Rights) 1993. Oslo: Royal Ministry of Health and Social Affairs, 1993
Carlsen B Norheim OF Introduction of the patient-list system in general practice: Changes in Norwegian physicians' perception of their gatekeeper role Scandinavian Journal of Primary Health Care 2003 21 209 13 14695070 10.1080/02813430310004155
Kristiansen IS Factors affecting doctor's decision making PHD thesis 1996 Tromsø, University of Tromsø, Institute of Community Medicine
Arnesen TM Counting health? A critical analysis of Cost Utility Analysis as a tool for setting priorities in health PhD thesis 2004 Oslo, The University of Oslo, Faculty of Medicine
Lian AS Røttingen JA Legen – homo economicus eller homo sociologicus? (The physician – homo economicus or homo sociologicus?) Tidsskr Nor Laegeforen 2002 122 1682 5 12555612
Kassirer JP Managed care and the morality of the marketplace N Engl J Med 1995 333 50 2 7776996 10.1056/NEJM199507063330110
Ritchie J Spencer L Bryman A, Burgess RG Qualitative data analysis for applied policy research Analyzing qualitative data 2001 London, Routledge 173 194
Naylor CD Grey zones of clinical practice: some limits to evidence-based medicine The Lancet 1995 345 840 2 7898234 10.1016/S0140-6736(95)92969-X
Kvale S Interviews An introduction to qualitative research interviewing 1996 Thousand Oaks: SAGE Publications
Daniels N Why saying no to patients in the United States is so hard. Cost containment, justice, and provider autonomy N Engl J Med 1986 314 1380 3 3702945
Wilsford D States facing interests: struggles over health care policy in advanced, industrial democracies Journal of Health Politics, Policy and Law 1995 20 571 613 8530769
Steinhart B Patient autonomy: evolution of the doctor-patient relationship Haemophilia 2002 8 441 6 12010448 10.1046/j.1365-2516.2002.00614.x
Lupton D Doctors on the medical profession Sociology of Health & Illness 1997 19 480 97
Light DW Aasland OG Den nye legerollen – kvalitet, åpenhet og tillit (Towards a new professionalism in medicine: Quality, value and trust) Tidsskr Nor Laegeforen 2003 123 1870 3 12830270
Little P Everitt H Williamson I Warner G Moore M Gould C Observational study of effect of patient centredness and positive approach on outcomes of general practice consultations BMJ 2001 323 908 11 11668137 10.1136/bmj.323.7318.908
Elston MA Gabe J, Calnan M, Bury M The politics of professional power: medicine in a changing health service The Sociology of the Health Service 1991 London, Routledge 58 88
Glenton C Chronic back pain sufferers – striving for the sick role Social Science & Medicine 2003 57 2243 52 14512253 10.1016/S0277-9536(03)00130-8
Britten N Patients' expectations of consultations BMJ 2004 328 416 17 14976072 10.1136/bmj.328.7437.416
Little P Dorward M Warner G Stephens K Senior J Moore M Importance of patient pressure and perceived pressure and perceived medical need for investigations, referral, and prescribing in primary care: nested observational study BMJ 2004 328 444 14966079 10.1136/bmj.38013.644086.7C
Cockburn J Pit S Prescribing behaviour in clinical practice: patients' expectations and doctors' perceptions of patients' expectations – a questionnaire study BMJ 1997 315 520 523 9329308
Rogers WA Whose autonomy? Which choice? A study of GPs' attitudes towards patient autonomy in the management of low back pain Fam Pract 2002 19 140 5 11906978 10.1093/fampra/19.2.140
Frankel S Ebrahim S Davey Smith G New B The limits to demand for health care Commentary: An open debate is not an admission of failure BMJ 2000 321 40 45 10875836 10.1136/bmj.321.7252.40
Koehn D The ground of professional ethics 1994 London, Routledge
Mead N Bower P Patient-centredness: a conceptual framework and review of the empirical literature Social Science & Medicine 2000 51 1087 1110 11005395 10.1016/S0277-9536(00)00098-8
Street RLJ Krupat E Bell RA Kravitz RL Haidet P Beliefs about control in the physician-patient relationship: effect on communication in medical encounters Journal of General Internal Medicine 2003 18 609 16 12911642 10.1046/j.1525-1497.2003.20749.x
Basch CE Focus group interview: An underutilized research technique for improving theory and practice in health education Health Education Quarterly 1987 14 411 48 3319971
Kahan JP Focus groups as a tool for policy analysis Analysis of Social Issues and Public Policy 2001 129 146 10.1111/1530-2415.00007
Farquhar C Barbour R, Kitzinger JI Are focus groups suitable for "sensitive" topics? Developing focus group research: politics, theory and practice 1999 London: Sage
Pearson SD Hyams T Talking about money: How primary care physicians respond to a patient's question about financial incentives Journal of General Internal Medicine 2002 17 75 80 11903778 10.1046/j.1525-1497.2002.10503.x
Marshall MN Hiscock J Sibbald B Attitudes to the public release of comparative information on the quality of general practice care: qualitative study BMJ 2002 325 1278 12458248 10.1136/bmj.325.7375.1278
Bernard HR Killworth P Kronenfeld D Sailer L The Problem of Informant Accuracy: The Validity of Retrospective Data Annual Review of Anthropology 1984 13 495 517 10.1146/annurev.an.13.100184.002431
Fernandez LAL Martin JMJ del Castillo JdDL Gaspar OS Millan JIM Lozano MJP Sources of influence on medical practice J Epidemiol Community Health 2000 54 623 630 10890875 10.1136/jech.54.8.623
Kitzinger J Qualitative Research: Introducing focus groups BMJ 1995 311 299 302 7633241
|
16281967
|
PMC1291367
|
CC BY
|
2021-01-04 16:31:49
|
no
|
BMC Health Serv Res. 2005 Nov 9; 5:70
|
utf-8
|
BMC Health Serv Res
| 2,005 |
10.1186/1472-6963-5-70
|
oa_comm
|
==== Front
BMC Med EducBMC Medical Education1472-6920BioMed Central London 1472-6920-5-371625964010.1186/1472-6920-5-37Research ArticleStakeholder views regarding cultural diversity teaching outcomes: a qualitative study Dogra Nisha [email protected] Olivia [email protected] University of Leicester Greenwood Institute of Child Health, Leicester, UK2 University of Maryland School of Medicine, Baltimore, MD, USA2005 1 11 2005 5 37 37 21 4 2005 1 11 2005 Copyright © 2005 Dogra and Carter-Pokras; licensee BioMed Central Ltd.2005Dogra and Carter-Pokras; licensee BioMed Central Ltd.This is an Open Access article distributed under the terms of the Creative Commons Attribution License (), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
Background
Cultural diversity teaching is increasingly present in both undergraduate and postgraduate training programmes. This study explored the views of stakeholders in medical education about the potential outcomes of cultural diversity teaching and how they thought cultural diversity programmes might be effectively evaluated.
Methods
A semi-structured interview was undertaken with 61 stakeholders (including policymakers, diversity teachers, students and users). The data were analysed and themes identified.
Results
Many participants felt that clinical practice was improved through 'cultural diversity teaching' and this was mostly as a result of improved doctor-patient communication. There was a strong view that service users need to participate in the evaluation of outcomes of cultural diversity teaching.
Conclusion
There is a general perception, rather than clear evidence, that cultural diversity teaching can have a positive effect on clinical practice. Cultural diversity teaching needs to be reviewed in undergraduate and postgraduate medicine and better evaluation tools need to be established.
==== Body
Background
Cultural diversity teaching has been advocated for over forty years in the US and more recently in the UK [1,2]. One of the major justifications for it has been that it will help reduce healthcare disparities. The Sainsbury Centre for Mental Health reported a case in which a young black man died after being inappropriately restrained. It accused the National Health Service of being racist, and advocated cultural awareness training for healthcare professionals [3]. This paper reports on stakeholders' views about the perceived outcomes of cultural diversity teaching. The discussion considers the implications for undergraduate and postgraduate medical education.
The use of 'cultural diversity' in this paper is broadly consistent with the definition of culture adopted by the Association of American Medical Colleges (AAMC) Task Force [4] in its report on Spirituality, cultural issues and end of life care. AAMC noted that:
"Culture is defined by each person in relationship to the group or groups with whom he or she identifies. An individual's cultural identity may be based on heritage as well as individual circumstances and personal choice. Cultural identity may be affected by such factors as race, ethnicity, age, language, country of origin, acculturation, sexual orientation, gender, socioeconomic status, religious/spiritual beliefs, physical abilities, occupation, among others. These factors may impact behaviours such as communication styles, diet preferences, health beliefs, family roles, lifestyle, rituals and decision-making processes. All of these beliefs and practices, in turn can influence how patients and heath care professionals perceive health and illness and how they interact with one another" [4].
One of the attitudinal objectives outlined in the first Tomorrow's Doctors [2] was that:
"At the end of the course of undergraduate medical education the student will have acquired and will demonstrate attitudes essential to the practice of medicine including respect for patients and colleagues that encompasses, without prejudice, diversity of background and opportunity, language, culture and way of life". [GMC, 1993: 15]
By publishing Tomorrow's Doctors, the GMC set the framework within which it expected medical education in the UK to develop. Although the GMC and the Quality Assurance Agency (QAA) regularly visit medical schools to monitor standards of medical education, medical schools remain free to develop curricula as they see fit. The drive behind Tomorrow's Doctors was a desire to overhaul medical school curricula. There was, in part, an acknowledgement of the growing evidence that curricula could not continue to expand at the same rate as medical knowledge [5-7]. As well as a reduction of emphasis on acquiring fact-based knowledge, Tomorrow's Doctors [2] placed increased emphasis on human skills and the ability to communicate with patients and colleagues. The GMC recognised changes in society and the need for changes towards improving equality [8]. This was maintained in the second edition [9].
Changes in undergraduate education have now been followed by the establishment of a Postgraduate Medical Education and Training Board (PMETB [10]) to regulate postgraduate education. Modernising Medical Careers (MMC [11]) is a relevant document in that it places similar expectations on postgraduate education as Tomorrow's Doctors did on undergraduate education. Although MMC does not explicitly mention diversity, the expectation is that learning outcomes will be set against the attributes in Good Medical Practice [12] to shape the structure for appraisal and revalidation of doctors. GMC clearly expects doctors to respect diversity and cultural diversity learning is seen as a lifelong objective..
Evaluation of specific cultural diversity programmes
When considering the need for an evidence-based approach to medical education, it is clear that evaluation of new programmes needs attention. Few cultural diversity teaching programmes have been subject to evaluation beyond subjective student feedback [13,14]. There are exceptions, which used pre- and post-teaching questionnaires [15-20]. All reported some degree of short term 'positive' changes in student attitudes, but none of the authors assessed long-term change in student attitudes Evaluating the effectiveness of cultural diversity teaching is difficult so the issue of evaluation is perhaps sidestepped.
Webb and Sergison [21] reported that the participants on their cultural diversity program found the training useful and staff commented on how they thought their own behaviour had changed at follow-up. However, there were no objective measures of change. Examples of changes of practice included using more culturally appropriate pictures for the ward and not using minors as interpreters.
The U.S. Task Force for Preventive Services conducted a systematic review of five interventions to improve cultural competence in healthcare systems including cultural competency training for healthcare providers [13]. The Task Force identified only one study that they felt had a fair quality of execution, and therefore concluded that the evidence was insufficient to determine the effectiveness of cultural diversity training for healthcare providers. A more extensive review by the Agency for Healthcare Research and Quality found excellent evidence for improvement in provider knowledge, good evidence for improvement in provider attitudes and skills, and good evidence for improvement in patient satisfaction [14]. Both reviews found a lack of consistency in intervention methods and measured outcomes [13,14].
This report is part of a PhD thesis [22] on the views held by key stakeholders in medical education on the teaching and learning of cultural diversity. This paper reports on the findings which explored the views of stakeholders in medical education about the potential outcomes of cultural diversity teaching and how they thought cultural diversity programmes might be effectively evaluated.
Method
Devising the interview schedule
Semi-structured in-person or telephone interviews were conducted with mostly open-ended questions [23]. It was concluded some structure was necessary to answer the specific research questions relating to the understanding of cultural diversity, its teaching and assessment.
The interview schedule drew on the literature base in sociology, medical education, education and intercultural studies [22]; previous research [20]; clinical, educational and personal experience; earlier interviews with members of the General Medical Council (GMC) Education Committee responsible for the first Tomorrow's Doctors [2]; and an Internet search of all UK medical school websites.
The interview was conducted in three parts. Part I collected basic demographic data (age and gender), as well as roles and experience. Part II began with four open-ended questions and participants talked unprompted, with the interviewer clarifying if necessary. The four questions were:
• How do you understand the term "cultural diversity"?
• What do you think should be taught at undergraduate level about cultural diversity?
• What main topics do you think that cultural diversity teaching should encompass at undergraduate level?
• How do you think cultural diversity should be taught?
The interview then continued open-ended but focused on specific aspects of teaching such as learning outcomes, delivery methods, assessment, and influence on clinical practice and student perspectives. Part III asked for the ways in which participants used or understood key terms such as race, ethnicity and multiculturalism. The interview concluded by asking participants of their experience and/or training in cultural diversity.
The interview was piloted with two policymakers and one diversity teacher and minor modifications were then made to the schedule. A flexible design was used when delivering the interviews. Modifying the interviews was part of the research process [22]. The research topic and themes remained the same, but the exact wording was modified as the interviews progressed enabling more themes to be explored as necessary. For example, the question about how learning outcomes might be phrased, was not asked after the first three interviews because participants struggled to respond with appropriate objectives immediately. They knew what they wanted to cover but could not be specific in voicing them. As the exact wording is less important than the overall aim and purpose of the teaching, this question was omitted: persisting with it would have been time-consuming and would not have yielded useful data. It may also have been frustrating for participants. The interview schedule is included as an Appendix One.
Sample and sample size
There were two stages of sampling to ensure a wide range of participants: selection of different groups of stakeholders followed by selection of individuals from these groups. Four UK medical schools from which staff could be interviewed were also identified for the study (see below). Selection was not random as key individuals were targeted. A minimum target of 50 interviews was set at the start of the study due to available time and resources, and to allow valid comparisons between different stakeholders. The sample group included:
• Communication teachers (n = 6): teachers responsible for communication skills training;
• Curriculum heads (n = 7): heads of medical education and curriculum committee members who implement policy;
• Diversity teachers (n = 7): teachers are responsible for developing and delivering cultural diversity;
• Policymakers (n = 18): members of organisations that decide or influence policy on medical education (e.g. General Medical Council);
• Researchers (n = 2): included researchers in 'cultural diversity' and associated areas actively teaching on ethnicity;
• Students (n = 7): medical students;
• Users (n = 7): included patients, patient representatives and advocates.
Table 1 details how medical schools included in the sample were identified.
Table 1 'Cultural expertise' and 'cultural sensibility' [24]
Cultural expertise: Expertise is defined as expert skill, knowledge or judgement with an expert being defined as having special skill at a task or knowledge in a subject. There is a notion that through learning knowledge about 'other' cultures, one can develop cultural expertise. In learning terms this means learning about the modus operandi of different cultural groups.
Cultural sensibility: Sensibility is defined as openness to emotional impressions, susceptibility, and sensitiveness. It relates to a person's moral, emotional or aesthetic ideas or standards. If one is open to the outside, one might reflect and change because of that experience. There is no notion of acquiring expertise about others but a recognition that we need to be aware of our perspectives and how they affect our ability to have an openness about other perspectives. This leads to a generic openness to diversity of all kinds and specific knowledge is not taught.
On this basis the following types of medical schools were included:
Where the cultural diversity programmes were more consistent with 'cultural sensibility 'than 'cultural expertise';
Where the programmes had elements of 'cultural sensibility' and elements of 'cultural expertise' model;
Where the programmes were more consistent with 'cultural expertise' but had some aspects of 'cultural sensibility'; and
Where no specified programme regarding cultural diversity was taught (this was based on information available about the curriculum and direct contact with staff at the school).
The sampling strategy ensured that interviews continued until saturation was achieved. Using purposive sampling and 'snowballing', a total of 61 individuals were interviewed. With snowballing, individuals who have been interviewed are asked after their interviews to identify other members of their group who may usefully act as informants [17]. Formal association with a medical school was defined as being employed by the medical school (including clinical NHS staff appointed as honorary teachers and external examiners), or being a student at a UK medical school. Individuals from 14 of the 26 established medical schools in the UK were involved (two schools have campuses at two sites; therefore, 12 curricula were effectively covered). These 14 schools were not specifically selected although we did ensure that schools within the above categories were part of the overall sample. The schools are representative of the UK overall. Members from eleven policymaking organisations and six medical disciplines were interviewed. Other 'clinical' perspectives included pharmacy, social work, community youth work and nursing. Non-clinical participants came from sociology, anthropology, accountancy, research and advocacy work.
Procedure
Interviews took place face-to-face as a first preference and by telephone, if the former was not possible. Initial contact was through a formal introductory letter that invited the participants to contact the researcher if there were any queries. The letter also stated that the interviews would be confidential and that the local research NHS ethics committee had approved the project. If there was no response, the initial letter was followed-up by email or by a second letter until the target number was achieved. No one was contacted more than twice if they failed to respond. Most participants replied by letter or email to agree to take part; copies of these correspondences were considered written consent to participate and kept. Twelve individuals responded but declined to participate for a variety of reasons including not being an appropriate person, too busy and no longer in post. Ten of these were policymakers and two were curriculum leads. A further 17 individuals did not respond; seven were user representatives or organisations, four policymakers, three were teachers, two researchers and one students. As we had recruited enough participants only one of these received a reminder.
Interviews were audiotaped and transcribed verbatim. Of the 61 participants, teachers and policymakers comprised the largest two groups, 45 were affiliated with medical schools, 39 were men, 31 were clinically active, 50 were White British, and 42 were between 41–60 years of age (see Table 2). At the outset, the researcher explained how the term 'cultural diversity' was being used in the study but interviewees were able to use the term as they thought appropriate.
Table 2 Summary demographics of participants
Number Ethnicity Gender (M= male F = female) Mean Age Range in years Medical school connection Currently in Clinical Practice
Communication teachers (n = 6) 5 White British
1 mixed race 3 M
3 F 46–50 6 4
Curriculum heads (n = 7) 7 White British 4 M
3 F 46–50 7 5
Diversity teachers (n = 14) 13 White British
1 Indian 5 M
9 F 46–50 14 6
Policymakers (n = 18) 13 White
3 Indian
1 Bangladeshi
1 Pakistani 17 M
1 F 51–55 9 13
Researchers (n = 2) 2 White British 2 M 41–45 2 0
Students (n = 7) 6 White British
1 Indian 3 M
4 F Under 30 7 In training
Users (n = 7) 4 White British
1 Black British
1 Indian
1 Pakistani 4 M
3 F 41–45 0 0
TOTAL (n = 61) 50 British 6 Indian
2 Pakistani
1 Bangladeshi
1 Black
1 Mixed Race 39 M
22 Female 46–50 45 28
The researcher's part in the research and interview process
This research was undertaken by ND, a female, of Indian origin, aged forty, brought up and educated in the UK, who works as a senior clinical academic in child and adolescent psychiatry at an East Midlands medical school. Having undertaken the development of a module in 'cultural diversity', she had some professional familiarity and experience with the topic. As Robson [25] has highlighted all these factors may influence the research.
Analysis
The analysis in this study was a combination of the quasi-statistical and template qualitative methodology [25] and followed a series of systematic steps. Responses were also counted to enable comparison between groups [26]. The content of the interviews was analysed manually to identify word and phrase frequencies and inter-correlations. Key themes were identified from the texts as a whole and from collations of responses to specific themes. Key themes were categorised under origins, organisation, contents, delivery and outcomes of the curriculum. The justification of this was the need to relate the findings to existing theory. The process of analysis for this research study took into account the steps outlined by Miles and Huberman [27]. The themes were organised under the major topics of origins and development, organisation, contents and delivery; and outcomes.
Results
Although this was essentially a qualitative study, some quantitative data are inevitably presented to illustrate the spread of responses. These are included in the text as appropriate and summarised in Tables 3, 4, 5, 6. Direct quotes are presented in the findings to illustrate points made. Quotes are also integrated into the discussion to highlight themes identified through qualitative analysis. Where no difference between the different stakeholders' perspectives is mentioned, the views were found across the range of stakeholders. In general, the findings showed no discernable pattern between sections of the sample, although students and service users were most optimistic about the outcomes of cultural diversity teaching.
Table 3 Does 'cultural diversity' teaching affect practice?
Improved Practice Made a Difference Did Not Know Unsure Too Early to Say
Communication Teachers (n = 6) 2 1 2 1
Curricular Heads (n = 7) 3 2 1 1
Diversity Teachers (n = 14) 9 1 3 1
Policymakers (n = 18) 9 4 3 2
Researchers (n = 2) 1 1
Students (n = 7) 6
Users (n = 7) 5 1 1 1
TOTAL (n = 61) 27 10 9 5 2
Table 4 How might 'cultural diversity' teaching affect practice?
Improve Doctor/Patient communication Improve Student Awareness Improve Patient Satisfaction Improve Sensitivity to Patients Better Health Care Better Access to Care
Communication Teachers (n = 6) 2 1
Curricular Heads (n = 7) 2 2 1
Diversity Teachers (n = 14) 5 6 1
Policymakers (n = 18) 12 4 1 1
Researchers (n = 2) 1 1
Students (n = 7) 2 2 1
Users (n = 7) 4 2
TOTAL (n = 61) 27 16 5 4 2 1
Table 5 How do you think programmes that endeavour to teach cultural diversity might be evaluated?
Observe Students in Practice Assessment Longer-term evaluations Relate to Learning Outcomes Student evaluation Ask Users Unsure – Probably Difficult Research Needed Ask Employers Reflective Portfolios Focus Groups
Communication Teachers (n = 6) 1 3 1 1
Curricular Heads (n = 7) 4 1 4 1
Diversity Teachers (n = 14) 5 4 3 1 1 1 2 1 1
Policymakers (n = 18) 3 3 1 3 2 1 2 1
Researchers (n = 2) 2 1
Students (n = 7) 1 1 1 1 1
Users (n = 7) 1
TOTAL (n = 61) 15 9 8 8 5 4 3 3 2 2 1
Table 6 How might the impact on practice be measured?
User Involvement Observed in Practice Long Term Outcomes Research Needed
Communication Teachers (n = 6) 4 1
Curricular Heads (n = 7) 2 1 1
Diversity Teachers (n = 14) 8 4 2 3
Policymakers (n = 18) 8 3 2
Researchers (n = 2) 1
Students (n = 7) 6 1
Users (n = 7) 4
TOTAL (n = 61) 32 9 4 6
Other responses were: ask employers (3); unsure (3); self-refection (3); complaints register (3); OSCES (2); ask students if felt prepared (2); patient care improves (1); audit (1); improved compliance (1) and comparing services with the National Service Framework (1).
The key questions about the outcomes were:
• Does cultural diversity teaching make a difference?
• How does it make a difference?
• How might programmes teaching cultural diversity be effectively evaluated?
Does cultural diversity teaching make a difference?
The majority of participants (35) felt that clinical practice was improved through cultural diversity teaching and that this was mostly through improved doctor-patient communication. The outcomes in clinical practice were related to a process as opposed to better knowledge with which to serve patients. This is important as this highlighted inconsistencies between what participants thought should be taught and what the anticipated outcomes were.
Only five participants stated that patient satisfaction might be improved through patients feeling better heard, valued and understood and another five thought there might be improved sensitivity to the patient's concerns and values. Others did not comment as to how cultural diversity teaching might improve the doctor-patient relationship; that is whether it was through improved knowledge of the patient's background or through a collaborative partnership with the patient. Students and users were more confident than other participants that cultural diversity teaching did improve clinical practice.
Despite the lack of evidence of any clear effect on clinical practice [13,14], there are many advocates for the inclusion of cultural diversity in the curriculum, which is contrary to an evidence-based approach. The sample interviewed was more likely to be positive about cultural diversity as most of them had involvement or responsibility in this area. If there is a belief that cultural factors affect the way health is perceived in general, then the logical step is to believe that learning about the subject is important. The work is undermined if any doubt is expressed regarding its effectiveness. Politically it may be unacceptable to say there is very little evidence that the teaching of cultural diversity makes any difference, especially if organisations have invested money on its teaching. However, the views that it does make a difference are made in the absence of firm evidence to support this perspective although there is also no clear evidence to indicate that training is in any way detrimental.
When asked if participants thought cultural diversity teaching made a difference to clinical practice, no one thought it did not. Thirty-five participants felt that it improved practice with two of these saying only if it was taught properly. Another thought it might, but had some reservations. Another felt that success was dependent on the ethos of the organisation.
"It is inevitably bound to have, both in terms of how you relate to colleagues as well as how you relate to patients. I think if it is well done it spins off beyond the obvious interface. You get to start seeing people as individuals, not as white or black, or whatever else, but actually as real people and not classify them as Catholics or Irish or whatever. Yes, I think it changes your perspective" (R30: Policymaker)
One participant gave an affirmative response based on her experience of delivering a postgraduate programme.
"I can only talk about our pack because that's the only one I have got experience and have evaluated. People reported quite a lot of things that they changed as a result of being on that. For example a lot of people said that they would find an interpreter, whereas before they might have muddled through... A couple of other people said they felt confident to challenge racism when they saw it. Someone else said he or she had changed the language they used to describe black clients, which was quite interesting. I think if you do win people over all kinds of things can happen really" (R27: Diversity teacher)
This suggests minor behavioural changes may have significant clinical implications even without any change in attitudes. Ten participants hoped it made a difference, nine did not know and five were unsure if teaching cultural diversity made any difference. One diversity teacher and one student thought it was too early to say if teaching made a difference.
How might cultural diversity affect clinical practice?
Participants were also asked how cultural diversity teaching might affect practice.
In terms of how cultural diversity teaching could affect practice, 51 gave a single response, 6 gave several responses in which cultural diversity teaching might make an impact and 4 were unsure. The number of responses is noted in parentheses below. Responses made included:
• Improved doctor/patient communication (27)
"Well I would hope that it would mean that patients would feel that they were valued, that they were listened to, and that they were understood. Hopefully when those things happen patients respond better to whatever advice or however they are treated... I think it would have a huge impact on clinical practice" (R5: Communication teacher)
• Student awareness would be improved about issues related to patients with one stating that this would lead to better communication (16)
"Yes, by making students aware. There is absolutely no doubt about that. You can see from the realisation in student's eyes when you tell them something about a particular cultural or ethnic group" (R20: Diversity teacher)
How might cultural diversity programmes be evaluated?
The variety of responses and the lack of clarity of most of the responses to methods for evaluation of cultural diversity teaching programmes indicate that more work needs to be undertaken. Interviewees were asked how cultural teaching programmes might be evaluated and how the impact of cultural diversity teaching on practice might be measured. Participants made the following responses:
• Thought that students needed to be observed in practice to effectively evaluate whether learning outcomes had been successfully met (15).
"Ultimately they have got to be evaluated by whether you get the outcomes that were specified, but of course those outcomes are expressed in practice... I would suspect that the main focus would have to be on senior SHO type of positions where people have had a chance to reflect, and try to get at whether there has been an actual change. Of course you can't do any of the things you want to do; you can't do any trials, or anything. It's very difficult to try and identify groups who have not been exposed to intervention or randomising in any way, but that's what you would be looking for... The difficulty here is that senior SHO level is quite a late point at which to offer remedial help" (R10: Curriculum head)
"They could be evaluated by people from culturally diverse backgrounds evaluating them as well. I think it would be important to have their input into the development, rather than just the evaluation" (R11: Curriculum head)
This statement implies that because patients come from the non-majority they are able to effectively decide whether care is appropriate for a range of people. It does not acknowledge that patients may bring their own biases and prejudices.
"I think that measuring of outcomes is so difficult when all the starting points are going to be different so I think we have to be looking at it from the perspective of portfolio learning and building up a portfolio. It's a very problematic area and one that needs to be examined more" (R26: Diversity teacher)
This reflects a need for outcomes to be measured in several ways and then triangulated to see whether the initial outcomes have been met.
When asked how the impact on practice might be measured, 32 participants thought that user involvement was important. This might be through the use of questionnaires, focus groups, or in-depth interviews. One participant felt it important that user input be contextualised, and that we needed to be careful, as there may be a difference between what patients and doctors regard as effective. Nine participants thought that students should be observed in practice and a further two linked this to the process of revalidation. Four participants felt that longer-term outcomes were needed, and six felt that effective research was needed. Other responses were: health outcomes but none were specified (4); ask employers (3); unsure (3); self-refection (3); complaints register (3); Objective Structured Clinical Examinations (OSCES) (2); ask students if they felt prepared for diversity issues in practice (2); patient care improves (1); audit (1); improved compliance (1) and comparing services with the National Service Framework (1).
One participant linked undergraduate teaching with the pre-registration house officer years and made suggestions on the effectiveness of teaching in practice, but also noted that consultants willing to learn could present excellent role modelling opportunities.
"You could say, 'right, regarding Doctor X, how do you think they perform in these particular aspects', so that you would actually have 360° assessment of every single member of the team by everybody else. You would have multiple observations, so that everybody would be feeding back about me to somebody else, albeit anonymously." (R36: Policymaker)
The response below indicates some consistency with an evidence-based approach, and suggests that there is a need for different teaching approaches to be implemented and compared.
"Ideally to have trial evidence for any teaching that we do, but that's obviously difficult. We can have comparative trials, non-randomised, with one half a year taking a particular model and the other group taking the Model B, and seeing what the outcomes were" (R17: Diversity teacher)
The key findings were that over half the participants felt that cultural diversity teaching had a positive influence on clinical practice and that this was mostly through improved doctor-patient communication and student awareness. Just over half of the participants felt it was necessary to involve users in deciding whether practice was improved. Several different types of evaluation were suggested but there was a lack of clarity about this.
Discussion
Many of those interviewed had a specific interest in cultural diversity teaching. It is perhaps unsurprising that despite the lack of evidence of any clear effect on clinical practice [5], they felt that cultural diversity should be included in the curriculum. However, there was also consensus that programmes had not been effectively evaluated. The implication here is that unless such teaching is shown to be detrimental, it must be a good thing but we need to be clear about what we teach and how we evaluate it.
The responses from this study indicate that a multifaceted approach to evaluation may be the most appropriate way forward; that is a range of different approaches. Comprehensive evaluation could involve the following:
1. Subjective student feedback on the usefulness and relevance of the teaching program
2. Objective measures of changes in student behaviour and attitudes using survey methods
3. User feedback
4. Staff perspectives of whether students had changed (clinical and non-clinical staff).
Each of these features of evaluation was identified by the study sample. There was a view that none of these on their own would be sufficient to demonstrate the effectiveness of a programme. Such an approach would enable triangulation of the different perspectives. Changes in student practice may be identified by observing students in clinical contexts (videotape, patient feedback, observer feedback), and noting whether they acknowledge the possibility that differences between themselves and the patients (e.g., ethnicity, gender, socioeconomic status, language) could influence the consultation. Their method of addressing this potential difficulty could also be assessed. Another format for evaluating changes in student practice may be the discussion of cases in peer groups. Two students may discuss a case while a third student observes their conversation. Observing students will be asked to note whether the students discussing the case made assumptions about the patient or disregarded the patient's perspectives. In this way, students may be supported in questioning their own and their peers' practice on a day-to-day level, thus making the issue of diversity an integral part of the clinical context.
Clinical staff could be asked to make specific observations about students who are placed with them and to comment in general terms on whether they feel students are aware of how to manage differences between themselves and patients in a way that optimises management options. However, approaches depend on the staff's level of training. Tang et al [28] found that students are often better equipped and more willing to manage diversity than their senior colleagues. However, this type of evaluation requires commitment and resources and a more rigorous research base, which has yet to be developed. In addition, the subject has little credibility when clinicians themselves have not been trained and so cannot support students once they are in clinical practice. The ambivalence that clinicians may express about cultural diversity training [28,29] may be related to the lack of a strong evidence base. It may also relate either to the fact that clinicians may not receive cultural diversity training or perceive the training they received as poor. Either way there is little evidence of a positive effect on clinical practice.
Long-term follow-up of students has to date been lacking and was commented on by some participants. This would not only allow educators to assess whether learning outcomes are met, but would also show how learning outcomes are applied in practice. However, this approach is not always practical and would again be time-consuming. There is though a need to develop an evidence base in this field.
To assist medical schools in the United States with the development, implementation and assessment of cultural competence education programs, the Association of American Medical Colleges, supported by a grant from the Commonwealth Fund, has convened experts to develop a Tool for the Assessment of Cultural Competence Training [30]. Principles and recommended standards for cultural competence education for healthcare professionals have also been developed through a consensus process funded by the California Endowment [31].
Limitations of the study
It is important to acknowledge the constraints of trying to select the 'right' people to meet the research objectives. It may be that some staff not involved in 'cultural diversity' but involved in medical education, who do not value diversity, were unlikely to be identified or participate. Given that interviews rely on the relationship established between the researcher and the participants, there is always the limitation that the research can be contaminated by the characteristics of the researcher. In this project the researcher could be viewed by some as an insider as she was both a teacher and doctor. In contrast, other interviewees, such as students and users who were perhaps the least empowered of the groups interviewed, may have perceived her as an outsider. As the policymaker group is perhaps the most influential and fairly broad, more members of this group were invited to participate than from other groups. It is possible that their views are overrepresented. However, the range of views they cover is not dissimilar to the range of views of other groups.
The interview was an effective research tool, but it is possible that it may have been too structured for some participants. The clear focus on education may also have deterred non-educationalists from expressing their thoughts in case they were perceived to be 'wrong' or 'politically incorrect'. A broader sample may have provided a more representative picture of what happens within organisations, especially medical schools. This might have revealed the rivalries existing between different subject teachers in an ever-expanding curriculum. The interview data might have been usefully triangulated with a questionnaire survey for additional breadth and depth. Whilst the data may have been limited more perspectives might have been explored. The participants were offered confidentiality and assured that the responses would be presented in such a way that no link could be made to any individual or organisation. Despite, this assurance, the lack of opportunity for completely anonymous comment may have meant that participants only gave what they perceived to be acceptable responses.
Conclusion
There is a general perception rather than clear evidence that cultural diversity teaching can have a positive effect on clinical practice. This is probably appropriate given the lack of strong evidence that currently exists about the effectiveness of such teaching. There is an urgent need to develop effective tools by which the effects of teaching on clinical practice can be measured including follow up of participants into their clinical practice. There is also a need to critically review cultural diversity programmes and question whether they are delivering what they set out to do. There is a great opportunity to consider approaches across disciplines and devise strategies to improve such education and the effect it has.
Competing interests
The author(s) declare that they have no competing interests.
Authors' contributions
ND conducted the study and analysed the data. OCP contributed to the writing of the paper.
Appendix one: Summary of interview schedule
I would like to discuss your personal views or views in your role regarding what you think should happen about teaching "Cultural diversity" to medical students
1. How do you understand the term "cultural diversity"?
2. What do you think should be taught at undergraduate level about cultural diversity?
3. What main topics do you think that cultural diversity teaching should encompass at undergraduate level?
4. How do you think cultural diversity should be taught?
5. Other areas of human diversity – from your perspective where do they fit in?
6. Should they be taught with cultural diversity or should cultural diversity be a separate course?
I am now going to focus on specific questions about teaching of cultural diversity (seek justifications for response)
7. At which stage of the medical student's career should this teaching take place?
8. How much time do you think needs to be spent in this area?
9. What kinds of learning outcomes would you like to see established for this area?
10. What teaching strategies might be usefully employed?
11. Who do you think should teach cultural diversity?
I would now like to move on to student assessment and student perspectives
12. Should students be assessed about cultural diversity?
13. How might they be assessed?
14. Should student feedback be gathered?
a. If so how might this be done?
b. How might student feedback be effectively used?
c. What might be your perspective if students said that they did not feel this kind of teaching was necessary?
15.
a. Would it be helpful to have guidelines on what should be taught?
b. What form might these take and who might develop them?
I would now like to move on to specific programmes you may be aware of
16. What specific training programmes to teach cultural diversity are you aware of?
17. In your view, could these form models of best practice. (Prompt: have they used an evidence-based approach/been subject to critical evaluation?
18. In your opinion, how do you think programmes that endeavour to teach cultural diversity might be evaluated?
19. In your opinion, how might their impact on clinical practice be measured?
20.
a. Are you aware of the GMC perspective on this issue? If so, what is your understanding of it?
b. What is your perspective on the GMC including this issue in Tomorrow's Doctors?
21. In your view, does the teaching of cultural diversity have an impact on clinical practice?
22. If no, can you think of reasons why this might be the case?
23. If yes, can you think of how it impacts on practice?
PART III
I would now like to move on to ask you your understanding of some key terms in this area. I should say that there is no right or wrong answer as such. I am just interested in your views?
What is your understanding of the following terms?
24. Culture
25. Ethnicity
26. Multiculturalism
27. Race
28. Cultural diversity
29. How do you think that the way that these terms are used and understood might influence medical education?
30. Do you have any personal training/experience in cultural diversity issues?
31. How would you classify your own ethnicity?
32. Is there anything else that you would like to add – either more about what we have covered or anything you feel I may have left out?
33. Finally, is there anyone else you think it would be useful for me to meet?
Thank you very much for you help with this project.
Pre-publication history
The pre-publication history for this paper can be accessed here:
Acknowledgements
ND would like to acknowledge the support of the University of Leicester through a semester's sabbatical to complete her PhD and also all those who participated in the study. She would also like to acknowledge the contribution of Giles Emerson who helped with some editorial support. OCP would like to acknowledge grant support from the National Heart Lung and Blood Institute (5K07HL079255).
==== Refs
Ludmerer KM Time to Heal: American Medical Education from the turn of the Century to the Era of Managed Care 1999 New York: Oxford University Press
General Medical Council Tomorrow's Doctors 1993 London: General Medical Council
The Sainsbury Centre for Mental Health Breaking the circles of fear: a review of the relationship between mental health services and African and Caribbean communities 2002 London: Sainsbury Centre for Mental Health
Association of American Medical Colleges Report III Contemporary Issues In Medicine Communication in Medicine: Spirituality, Cultural Issues and End of Life Care Medical School Objectives Project 1999
Lowry S Medical Education 1993 London: British Medical Association Publishing
Richards P Stockhill S The new learning medicine 1997 14 London: BMJ Publishing Group
Sinclair S Making doctors: an institutional apprenticeship 1997 Oxford: Berg
Irvine D The changing relationship between the public and medical profession Lloyd Roberts Lecture to the Royal Society of Medicine 16 January 2001
General Medical Council Tomorrow's Doctors 2002 2 General Medical Council, London
Postgraduate Medical Education Training Board The PMETB 2005
Department of Health Modernising medical careers: the response of the four UK Health Ministers to the Consultation on Unfinished Business: Proposals for reform of the Senior House Officer Grade Department of Health, London
General Medical Council Good Medical Practice 2001 General Medical Council, London
Anderson LM Scrimshaw SC Fullilove MT Fielding JE Normand J the Task Force on Community Preventive Services Culturally Competent Healthcare Systems: A Systematic Review Am J Prev Med 2003 24 68 79 12668199 10.1016/S0749-3797(02)00657-8
Beach MC Cooper LA Robinson KA Price EG Gary TL Jenckes MW Gozu A Smarth C Palacio A Feuerstein CJ Bass EB Powe NR Review of effectiveness of cultural competency training: Strategies for Improving Minority Healthcare Quality Evidence Report/Technology Assessment No 90 (Prepared by the John Hopkins University Evidence-based Practice Center, Baltimore, MD) AHRQ Publication No 04-E008-02 Rockville, MD 2004 Chapter 3
Mao C Bullock CS Harway EC Khalsa SK A workshop on ethnic and cultural awareness for second year students Journal of Medical Education 1988 63 624 628 3398017
Copeman RC Medical students, Aborigines and migrants: evaluation of a teaching programme Med J Aus 1989 150 84 87
Rubenstein HL Bonnie JD O'Connor B Nieman LZ Gracely EJ Introducing students to the role of folk and popular health belief systems in patient care Academic Medicine 1992 67 566 568 1520410
Majumdar B Keystone JS Cuttress A Cultural sensitivity training among foreign medical graduates Medical Education 1999 33 177 184 10211237 10.1046/j.1365-2923.1999.00291.x
Culhane-Pera KA Reif C Egli E Baker NJ Kassekert R A curriculum for multicultural education in family medicine Family Medicine 1997 29 719 23 9397362
Dogra N The development and evaluation of a programme to teach cultural diversity to medical undergraduate students Medical Education 2001 35 232 241 11260446 10.1046/j.1365-2923.2001.00734.x
Webb E Sergison M Evaluation of cultural competence and antiracism training in health services Archives of Diseases in Childhood 2003 88 291 4 12651748 10.1136/adc.88.4.291
Dogra N The teaching and learning of cultural diversity in undergraduate medical education in the UK PhD thesis 2004 University of Leicester, Department of Sociology
Strah M Burton D Qualitative interviewing Research training for social scientists 2000 London: Sage Publications 196 214
Dogra N Cultural competence or cultural sensibility? A comparison of two ideal type models to teach Association of cultural diversity to medical students International Journal of Medicine 2003 5 223 231
Robson C Real world research 2002 2 Oxford: Blackwell Publishers
Seale C The quality of qualitative research: Introducing qualitative methods 1999 London: Sage 119 139
Miles M Huberman AM Qualitative data analysis: an expanded sourcebook 1994 2 London: Sage
Tang TS Bozynski MEA Joyce M Mitchell MD Haftel HM Vanston SA Anderson RM Are residents more comfortable than faculty members when addressing sociocultural diversity in medicine? Academic Medicine 2003 78 629 633 12805044
American Medical Association Enhancing the "cultural competence" of Physician, Council on Medical Education Report 5-A-98Cultural Competence Compendium Washington 1999
American Association of Medical Colleges Tool for assessing cultural competente training (TACCT)
California Endowment Principles and Recommended Standards for Cultural Competence Education of Health Care Professionals
|
16259640
|
PMC1291368
|
CC BY
|
2021-01-04 16:30:57
|
no
|
BMC Med Educ. 2005 Nov 1; 5:37
|
utf-8
|
BMC Med Educ
| 2,005 |
10.1186/1472-6920-5-37
|
oa_comm
|
==== Front
BMC Musculoskelet DisordBMC Musculoskeletal Disorders1471-2474BioMed Central London 1471-2474-6-521626289810.1186/1471-2474-6-52Research ArticleBack pain reporting in young girls appears to be puberty-related Wedderkopp Niels [email protected] Andersen Lars [email protected] Karsten [email protected] Charlotte [email protected] The Back Research Center, BackCenter Funen, Lindevej 5, 5750 Ringe, Denmark2 Norwegian University of Sport and Physical Education, Sognsveien 220, 0806 Oslo, Norway3 Institute of Sportscience and Clinical Biomechanics, University of Southern Denmark, Campusvej 55, 5230 Odense M, Denmark2005 1 11 2005 6 52 52 17 2 2005 1 11 2005 Copyright © 2005 Wedderkopp et al; licensee BioMed Central Ltd.2005Wedderkopp et al; licensee BioMed Central Ltd.This is an Open Access article distributed under the terms of the Creative Commons Attribution License (), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
Background
There is a large increase in back pain reporting in the early teens. In no previous study has the prevalence of low back pain been investigated in relation to the onset of puberty. The objective of this study was to establish whether the onset of puberty is associated with back pain reporting in young girls.
Methods
A subsample of 254 girls aged 8–10 years and 165 girls aged 14–16 years from a cross-sectional survey of 481 children aged 8–10 years and 325 adolescents aged 14–16 years of both sexes.
Main outcome measures were back pain defined as low back pain, mid back pain, and/or neck pain in the past month.
Other variables of interest were Puberty (five different stages), age, body mass index, and smoking. Independent information on onset of puberty was obtained through a physical examination and on back pain through an individual structured interview. The association was studied between onset of puberty and the outcome variable (the one month period prevalence of back pain), controlling for overweight, and smoking. Odds ratios with 95% confidence intervals were used to describe bivariate associations, logistic regression with robust standard errors was used for multivariate analyses.
Results
There is a highly significant trend for increased back pain reporting with increasing level of puberty until maturity is reached. The biggest leap appears between the second level (beginning of puberty) and the third level (mid puberty) and the findings remain after controlling for the covariates. These results emanate from the low back, whereas pain in the mid back and neck do not seem to be linked with pubertal stage.
Conclusion
In girls, the reporting of low back pain increases in frequency during puberty until maturity, regardless of age. Why some girls are susceptible to back pain in the early stage of puberty is unknown.
==== Body
Background
Back pain has been shown to commence early in life [1,2]. In a previous report, it was shown that the one-month period prevalence of back pain (neck, mid back and/or low back pain) was 33% (95%CI: 29–37) for children aged 8–10 years and 47% (95% CI: 42–52) for adolescents aged 14–16 years [3]. Also the frequency of back pain has been shown to increase with age, particularly around the period of puberty [4].
Attempts have been made to identify the age, when back pain becomes common and, at least in the case of pain for the lower back, there seems to be a rapid increase in the back pain reporting around the age of puberty [5]. If the onset of back pain were a function of time, the consequence of the cumulative effects of injuries and repeated microtrauma, then it would be expected that back pain reporting commenced in boys, the gender most likely to incur injuries. However, it has previously been shown that that back pain reporting is more frequent in young girls than in young boys but that boys catch up within a year or two [5]. The reason for this should be sought in some sex (biological) or gender (behavioural) differences in that age group. Girls reach puberty earlier than boys. Therefore, to establish whether back pain reporting among the young is puberty-related, the prevalence of back pain was studied for girls attending school in grades 3 and 9 in relation to their stages of puberty and after controlling for smoking, and overweight.
Method
The study design and methods have been extensively described elsewhere (3) and have been summarized in Appendix 1. A subsample of 254 female children and 165 female adolescents was selected from the participants in a randomly selected sample of Danish school-children attending elementary school (i.e. children aged 8–10 years and adolescents aged 14–16 years). These had been interviewed with help of a semi-structured questionnaire by one of the authors (NW) for the presence of back pain.
The one-month period prevalence of back pain was established by asking the pupils to point to any area of discomfort in the back (low back, mid back and/or neck region) reported to have occurred on the day of the study, in the week preceding the interview, or in the month preceding the interview, i.e. the one-month period prevalence of back pain.
Start of puberty was assessed according to Tanner [6] based on breast development. Pubertal development was graded in relation to breast development from stage 1 (not started puberty), through stage 2 (just starting), to stages 3–5, where 5 denotes that puberty has ended. Six of the girls refused to have their breasts examined resulting in a sample size of 413. The spread of data is shown in Table 1. Among the children, 29% of the young girls had started puberty vs. 99% of the adolescent girls.
Table 1 Stage of puberty by pain location according to sampling scheme, showing significant trend towards more low back pain and back pain anywhere with more advanced pubertal stage df = 1 for both calculations, chi-square = 42.9 and 17.29, p < 0.0001 for both calculations.
3rd grade Pain location Puberty stage 1 Puberty stage 2 Puberty stage 3 Puberty stage 4 Puberty stage 5 Total (%)
No pain 126 46 1 0 0 173 (68.7%)
Low back pain 3 1 0 1 0 5 (2.0%)
Mid back pain 30 15 1 1 0 47 (18.7%)
Neck pain 19 5 2 0 0 26 (10.3%)
Pain in two locations 0 1 0 0 0 1 (0.4%)
Back pain anywhere 52 22 3 2 0 79 (31.4%)
Total 178 68 4 2 0 252 (100%)
9th grade
No pain 1 0 7 36 36 80 (49.7%)
Low back pain 0 0 2 18 15 35 (21.7%)
Mid back pain 0 0 3 12 14 29 (18.0%)
Neck pain 0 0 0 6 4 10 (6.2%)
Pain in two locations 0 0 0 2 5 7 (4.3%)
Back pain anywhere 0 0 5 38 38 81 (50.3%)
TOTAL 1 0 12 74 74 161
Data on back pain and puberty were collected by two independent examiners who were naïve in relation to the possible link between puberty and back pain, as the idea for the present project arose after the publication of a first set of articles, and the purpose of the main study was to determine various biological risk factors for cardiovascular disease.
One at the time, the four different definitions of back pain (back pain, low back pain, mid back pain, and neck pain) were cross-tabulated against the pubertal stage variable and reported as odds ratios with 95% confidence intervals.
These associations were tested for trends across ordered groups with a score test for trend of odds [7]. The score test for trend performs a test for linear trend of the log odds against the numerical code used for the exposure variable [7]. The association between overall back pain and puberty were, thereafter, studied using logistic regression, adjusting for overweight defined by BMI cut-points as specified by Cole et al.[8], and smoking, as these variables could be suspected to be associated to both back pain and puberty. Both the odds ratio of the individual puberty stages and the trend across the puberty stages were assessed.
To take into account the possible effect of cluster sampling (independence of observations across groups but not necessarily independence within groups), "robust standard errors" were used in the logistic regression. These have the ability to relax the assumption of independence of the observations, i.e. they can produce "correct" standard errors (in the measurement sense) even if the observations are correlated [9]. We chose to do multivariate analysis only on the back pain anywhere group because this was the only group with enough subjects for this type of analysis (Table 1).
To test the goodness of fit, the Hosmer Lemeshaw goodness of fit statistics was applied to logistic regression model [10].
Results
The spread of data by pubertal stage is shown in Table 1. As can be seen in Table 2 there is a bivariate trend for increase for reporting of back pain anywhere (p < 0.0001) and low back pain (p < 0.0001) with increasing level of puberty until maturity is reached. The same trend is found after having adjusted for smoking and overweight. Both in the bivariate analysis and after having controlled for possible confounders, the biggest leap appears between the second level (beginning of puberty) and the third level (mid puberty). The association was tested for back pain anywhere controlling for the potential confounders (overweight and smoking), and it was noted that the above results held (Table 3.). Test for goodness of fit showed good fit of the model, with p = 0.97. In addition logistic regression was used to test for trend when adjusting for the possible confounders, resulting in a significant odds ratio of 1.2 (95% CI 1.2 ; 1.4).
Table 2 The bivariate associations between pubertal stage and a) back pain anywhere b) low back pain, c) mid back pain, and d) neck pain reported as odds ratios and 95% confidence intervals.
Definition of back pain Puberty stage 1 (index group) Puberty stage 2 Puberty stage 3 Puberty stage 4 Puberty stage 5
Back pain anywhere 1 1.1
0.6–2.0 2.3
0.8–6.5 2.6
1.5–4.7 2.4
1.4–4.3
Low back pain 1 0.9
0.1–8.5 8.2
1.2–55.6 19.6
5.0–76.3 14.7
3.8–56.6
Mid back pain 1 1.4
0.7–2.8 1.6
0.5–5.4 1.0
0.5–2.1 1.1
0.6–2.3
Neck pain 1 0.7
0.2–1.8 1.2
0.2–5.6 0.7
0.3–1.9 0.5
0.2–1.4
Table 3 The multi variable association between pubertal stage and back pain anywhere, adjusting for overweight and smoking using logistic regression, reported as odds ratios with 95% confidence intervals.
Definition of back pain Puberty stage 1 (index group) Puberty stage 2 Puberty stage 3 Puberty stage 4 Puberty stage 5
Back pain anywhere 1 1.1
0.7–2.0 1.6
0.5–4.6 2.0
1.3–3.5 2.1
1.1–4.1
The associations between back pain anywhere and the covariates expressed as odds ratios were: 2.9 (95% CI 1.3 ; 6.5) for smoking and 0.7 (95% CI 0.3 ; 1.7) for overweight.
Discussion
To our knowledge, this is the first time that the presence of back pain is studied in relation to pubertal stage. The results indicate that back problems and back pain might be related to puberty stage.
Two possibilities related to puberty spring to mind. First, the growth spurt initiated during puberty could be the initiating factor of back pain and back problems, as has been suggested by others [11,12]. For example Salminen et al. found a relationship between rapid growth, degenerative changes in the spine, and back pain in adolescents [11] and, in a Canadian study a significant association could be found between high growth spurt and the development of adolescent musculoskeletal pain over a 1-yr period [13]. This could also explain why the increase in back pain in our population occurs at puberty stages 3 and 4, as it is at these stages that the growth spurt occurs, which might make the back more susceptible to mechanical injuries. On the other hand, over a period of one year, increase in height in 4th grade Finnish children was not found to be associated with self-reported low back pain [14]. Obviously it should be further investigated what circumstances in the early stages of puberty can induce increased back pain reporting in girls.
We found one study, in which low back- and neck/shoulder pain was studied in relation to whether the menarche started early, average or late in 14, 16 and 18 year old Finnish adolescents [14]. A weak association was found with early onset of puberty after having controlled for age and sex. Unfortunately the study design did not allow for an assessment of pain in relation to the different stages of puberty.
Second, hormone-induced changes at puberty might have an effect on the attitudes to or perception of pain. We could find no information as to any evidence for or against such suppositions. Nonetheless, the perception of experimentally induced pain was noted to fluctuate across the menstrual cycle in females [15] and pain perception in women has been shown to be different to that in men [16].
We decided not to include physical activity, age and height in the analysis although we have access to that information in our dataset. The reasons were that we have shown earlier that in this cohort objectively measured physical activity was not related to back pain [17], and preliminary analyses revealed that both age and height were closely related to puberty level and created collinearity problems.
The results should and have been interpreted with caution for two reasons. First, it is a cross sectional study and a secondary analysis. Our title: "Back pain reporting in young girls appears to be puberty-related", illustrates that the results derive from a hypothesis generating analysis rather than confirmatory analysis. To be able to perform a confirmatory analysis, a prospective study, including both genders from age 8 to 18 with frequent follow-ups, would be required. Second, when stratifying on the different spinal regions and puberty the number of subjects in the calculations gets small, and the odds ratios could therefore be inflated. But as the trend in the results is the same, in the analyses of back pain anywhere alone and of low back pain across the puberty stages, we believe that the association is true.
Ideally, we should have made the same analyses on the boys, as data on these were collected. However, we found a systematic error, as the distribution of boys having started puberty was illogical, 85% being classified as having reached end of puberty, in comparison with only 43% of the girls. As girls generally start and end puberty before boys and the end of puberty in boys usually is around age 18, the data on puberty in boys could not be included in the study. It therefore now remains to find out if back pain reporting in boys is also linked with puberty and whether back pain is simply an inevitable aspect of growing up.
Appendix 1. Description of study design and method
Participants
Target population
8–10-yr old children and 14–16-yr old adolescents, attending state schools in Odense, Denmark, and their parents were invited to participate in a cardiovascular study.
Sampling
Schools were stratified according to type (age ranking, selection procedures, single/mixed sex), location (urban, suburban, rural), and socio-economic uptake of area. Proportional, two-stage cluster sample of children from each stratum. Primary units (clusters): schools; secondary units: children in 3rd and 9th grades. Children were randomly selected using random number tables.
Study sample
Agreed to participate: 25/28 schools (89%) and 1020/1356 child-parent pairs (75%). Of these, 806 were included in the present study.
Representativeness
The distribution of parental income and educational levels were representative of the national distribution (Source for comparison: Information from the Danish Central Statistical Office, Copenhagen).
Data collection
Design and ethics
Questionnaire data were collected from parents, interviews conducted with children, and physical measurements made of children in a cross-sectional survey. The study was approved by the local ethics committee.
Questionnaire
Questionnaires were handled through teachers and by mail for non-responders. They contained: information on parental health, cardiovascular disease, cardiovascular risk factors, and socioeconomic status, plus information on children's birth weight and chronic diseases.
Interview
Back pain interviews were conducted by NW and tested for face validity. Children were asked if they had any back or neck pain within the preceding month and whether the back pain had any consequences.
Physical examination
Height, weight, body fatness and stage of puberty were measured but not all used in the present report.
Validity
None of the variables was validated against a golden measure.
Manipulation of data
Data entry and quality control
Data entry was manual, checked by a second person. There was an additional verification and, if necessary, correction of outliers but none was removed from the data set.
Competing interests
The author(s) declare that they have no competing interests.
Authors' contributions
NW and CLY developed the aim and analytical approach for this paper. NW, LBA and KF were responsible for elements of the study design specific to this analysis. KF, NW and LBA obtained funding, co-ordinated and performed data collection. NW undertook the statistical analyses and NW and CLY wrote the first draft of the paper. All authors contributed to the final version.
NW and CLY act as guarantors.
Pre-publication history
The pre-publication history for this paper can be accessed here:
Acknowledgements
The authors are very grateful to the participants and their families who gave their time to the study. We would also like to acknowledge all members of the European Youth Heart Study Group not listed as co-authors of this paper. In addition we are thankful of the help of M.Sc. Ph.D. (biostatistics) Lars Korsholm regarding the analysis of the study.
This study was economically supported by the following grants: Danish Medical Research Council, Health Foundation, Danish Council for Sports Research, Foundation of 17-12-1981, Foundation in Memory of Asta Florida Bolding nee Andersen, and Faculty of Health Sciences, University of Southern Denmark. The views expressed in this paper are those of the authors and not necessarily any funding body.
==== Refs
Balague F Dutoit G Waldburger M Low back pain in schoolchildren. An epidemiological study Scand J Rehabil Med 1988 20 175 179 2976526
Burton AK Clarke RD McClune TD Tillotson KM The natural history of low back pain in adolescents Spine 1996 21 2323 2328 8915066 10.1097/00007632-199610150-00004
Wedderkopp N Leboeuf-Yde C Andersen LB Froberg K Hansen HS Back pain reporting pattern in a Danish population-based sample of children and adolescents Spine 2001 26 1879 1883 11568698 10.1097/00007632-200109010-00012
Petersen S Bergstrom E Brulin C High prevalence of tiredness and pain in young schoolchildren Scand J Public Health 2003 31 367 374 14555373 10.1080/14034940210165064
Leboeuf-Yde C Kyvik KO At what age does low back pain become a common problem? A study of 29,424 individuals aged 12-41 years Spine 1998 23 228 234 9474731 10.1097/00007632-199801150-00015
Tanner JM Whitehouse RH Clinical longitudinal standards for height, weight, height velocity, weight velocity, and stages of puberty. Archives of diseases in childhood Archives-of-disease-in-childhood 1976 51 170 179 952550
StataCorp. Epitab - Tables for epidemiologists. Stata Survival analysis and epidemiological tables Stata manual release 8 2003 College Station, Stata Corporation 20 70
Cole TJ Bellizzi MC Flegal KM Dietz WH Establishing a standard definition for child overweight and obesity worldwide: international survey BMJ 2000 320 1240 1243 10797032 10.1136/bmj.320.7244.1240
Obtaining robust variance estimates Stata User's guide Stata manual release 8 2003 College Station, Stata Corporation 270 275
D.W.Hosmer S.Lemeshow Applied Logistic Regression 1989 John Wiley & Sons Inc
Salminen JJ Erkintalo M Laine M Pentti J Low back pain in the young. A prospective three-year follow-up study of subjects with and without low back pain Spine 1995 20 2101 2107 8588166
Fairbank JC Pynsent PB van Poortvliet JA Phillips H Influence of anthropometric factors and joint laxity in the incidence of adolescent back pain Spine 1984 9 461 464 6238421
Feldman DE Shrier I Rossignol M Abenhaim L Risk factors for the development of low back pain in adolescence Am J Epidemiol 2001 154 30 36 11427402 10.1093/aje/154.1.30
Nissinen M Heliovaara M Seitsamo J Alaranta H Poussa M Anthropometric measurements and the incidence of low back pain in a cohort of pubertal children. Spine 1994 19 1367 1370 8066517
Riley JLIII Robinson ME Wise EA Price DD A meta-analytic review of pain perception across the menstrual cycle Pain 1999 81 225 235 10431710 10.1016/S0304-3959(98)00258-9
Hellstrom B Lundberg U Pain perception to the cold pressor test during the menstrual cycle in relation to estrogen levels and a comparison with men Integr Physiol Behav Sci 2000 35 132 141 11021338
Wedderkopp N Leboeuf-Yde C Andersen LB Froberg K Hansen HS Back pain in children: no association with objectively measured level of physical activity Spine 2003 28 2019 2024 12973151 10.1097/01.BRS.0000083238.78155.31
|
16262898
|
PMC1291369
|
CC BY
|
2021-01-04 16:32:05
|
no
|
BMC Musculoskelet Disord. 2005 Nov 1; 6:52
|
utf-8
|
BMC Musculoskelet Disord
| 2,005 |
10.1186/1471-2474-6-52
|
oa_comm
|
==== Front
BMC Musculoskelet DisordBMC Musculoskeletal Disorders1471-2474BioMed Central London 1471-2474-6-541627114910.1186/1471-2474-6-54Study ProtocolThe effect of motor control exercise versus placebo in patients with chronic low back pain [ACTRN012605000262606] Maher Chris G [email protected] Jane [email protected] Paul W [email protected] Kathryn M [email protected] G Lorimer [email protected] Robert D [email protected] Leonardo OP [email protected] James [email protected] Back Pain Research Group, School of Physiotherapy, The University of Sydney, PO Box 170, Lidcombe, NSW, 1825, Australia2 Division of Physiotherapy, The University of Queensland, Brisbane Qld 4072, Australia3 Department of Human Anatomy & Genetics, Oxford University, South Parks Rd, Oxford, OX1 3QX, UK2005 4 11 2005 6 54 54 29 9 2005 4 11 2005 Copyright © 2005 Maher et al; licensee BioMed Central Ltd.2005Maher et al; licensee BioMed Central Ltd.This is an Open Access article distributed under the terms of the Creative Commons Attribution License (), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
Background
While one in ten Australians suffer from chronic low back pain this condition remains extremely difficult to treat. Many contemporary treatments are of unknown value. One potentially useful therapy is the use of motor control exercise. This therapy has a biologically plausible effect, is readily available in primary care and it is of modest cost. However, to date, the efficacy of motor control exercise has not been established.
Methods
This paper describes the protocol for a clinical trial comparing the effects of motor control exercise versus placebo in the treatment of chronic non-specific low back pain. One hundred and fifty-four participants will be randomly allocated to receive an 8-week program of motor control exercise or placebo (detuned short wave and detuned ultrasound). Measures of outcomes will be obtained at follow-up appointments at 2, 6 and 12 months after randomisation. The primary outcomes are: pain, global perceived effect and patient-generated measure of disability at 2 months and recurrence at 12 months.
Discussion
This trial will be the first placebo-controlled trial of motor control exercise. The results will inform best practice for treating chronic low back pain and prevent its occurrence.
==== Body
Background
The problem of chronic low back pain
Low back pain is the main cause of work absence and disability in industrialised societies. Approximately 10–20% of patients with low back pain develop chronic pain, defined as pain persisting for more than 3 months. Additional to their pain these patient's health problems typically include reduced physical function and psychological distress[1]. These patients use more than 80% of health care resources for back problems, and treatment has a low success rate [2].
In 2002, arthritis and musculoskeletal disorders were announced as the new National Health Priority Area in recognition of the major health and economic burden that these diseases place on the Australian community [3]. Amongst this group of diseases back pain is both the most prevalent and most costly single disease [4]. The 2001 National Health Survey revealed that chronic back pain is the most prevalent illness from the seven National Health Priority Areas [5].
The severity of chronic pain can be described with four hierarchical grades, Grades I–IV, that consider the pain intensity and the degree of disability associated with the pain [6]. An Australian population-based survey, noted that 22% of respondents reported chronic pain with 39% of respondents classed as Grade I (least severe), 35% as Grade II, 14% as Grade III and 13% as Grade IV (most severe) [7]. The most common cause of chronic pain was low back pain (45% of cases).
Effectiveness of treatments for chronic low back pain
While there are a myriad of treatment options for chronic low back pain, there is only one clinical practice guideline for chronic non-specific low back pain: The European Guideline[8]. This guideline and the relevant Cochrane reviews [9] provide the most reliable sources of evidence on treatment for this condition. Unfortunately the Cochrane reviews provide fairly bleak reading for both clinicians and patients. Most of the reviews (7/13) concluded that the treatment under review was of unknown value. Five of the thirteen reviews concluded that there was some evidence for the treatment under review however significant limitations for each treatment were noted. These limitations included: no long term effect (e.g. back school); serious side effects (e.g. muscle relaxants); small effect size (e.g. massage); treatment improves outcomes other than pain (e.g. work conditioning) and no information available on patient or dose selection (e.g. behavioural treatment). The European Guideline produced similar conclusions [8]. In only one Cochrane review, the review of multidisciplinary rehabilitation/functional restoration, did the reviewers conclude that there was strong evidence for the therapy. However the reviewers also noted that these programs were only effective when they included >100 hours of therapy. Because these programs are multidisciplinary they are typically provided in a tertiary setting and because of the amount of time involved they are also very expensive. Accordingly functional restoration is usually reserved for the most severe cases of chronic low back pain.
The majority of patients with chronic low back pain has less severe pain (i.e. Grades I–III) and are typically managed in primary care. Not surprisingly clinicians find managing chronic low back pain difficult with qualitative research reporting that therapists' inability to identify effective treatment choices for their patients makes them state clinicians perhaps feel 'helpless' 'disillusioned' and 'pessimistic' [10]. Studies of patients reveal similar negative feelings and emotions [11].
To address this major problem, we plan to begin a coordinated program of research in which treatments that seem most promising are rigorously evaluated in randomised controlled trials. We define 'most promising treatments' as those that (i) appear to have clinically important effects that are maintained in the long term, (ii) are readily available and of modest cost and (iii) there is biological plausibility for the effect. Exercise therapy is our first candidate for evaluation in this program of research because it satisfies each of these three criteria, however at present trials have reported conflicting results.
While some trials of exercise therapy have reported large, durable and clinically important effects of treatment [12,13] others have not [14]. The uncertainty is reflected in the conclusion of the Cochrane review of exercise therapy: '...there is conflicting evidence on the effectiveness of exercise therapy...' [15]
Many factors are likely to have contributed to the inconsistent results across trials. Importantly, interpretation of the results of exercise trials is difficult because most trials have been pragmatic trials, comparing two active treatments delivered in routine practice (e.g. exercise vs. usual medical care [12]; exercise vs. physiotherapy [16]) These comparisons cannot provide a clear estimate of the effects of exercise treatment because most of the comparison treatments are also of unknown efficacy. Secondly, there has been insufficient appreciation by researchers conducting trials and by reviewers summarising trials of the wide variety of forms exercise can take and also trials do not control the quality of exercise intervention. While exercise is typically regarded as a single class of treatment we believe that this level of conception is inappropriate and analogous to not distinguishing between different classes and doses of drugs when prescribing medication. The types of exercise programs for chronic low back pain vary widely from land-based exercise versus exercise in water to isolated trunk exercise versus a walking program and it is unlikely that all programs are equally effective for all patients. Lastly, methodological quality varies greatly across previous exercise trials, for example in the Cochrane review [15] the least sound trial attended to none of the nine methodological criteria while the best attended to seven of the nine. Because methodological quality has been shown to affect the results of trials in other areas of health care [17] it is likely that a lack of rigor has contributed to the inconsistent results.
It is not sensible to talk about evaluating the efficacy of exercise without specifying the type of exercise. We have chosen to measure the efficacy of motor control exercise (sometimes called specific spinal stabilisation exercise) for chronic low back pain, rather than other forms of exercise, because it is a widely used form of exercise and there is an extensive body of literature that provides a rationale for the mechanism of action. The only way to clearly establish the value of motor control exercise in the management of chronic low back pain is to evaluate the efficacy of this form of exercise therapy in a methodologically sound randomised placebo-controlled trial. Prior to conducting a placebo-controlled trial of exercise we felt that it was prudent to identify the most promising form of exercise that would subsequently be evaluated in the placebo-controlled trial. To do this we conducted a randomised controlled trial where 160 patients were randomised to an 8 week program of either motor control exercise or general exercise[18].
The trial demonstrated that both programs were accompanied by large improvements in pain and disability. Motor control exercise produced significantly better outcomes in the short term, and there was a trend for motor control exercise to produce better outcomes at 6 month follow-up. Accordingly we have chosen to evaluate motor control exercise in the proposed trial. Our choice coincides with the research agenda set by the 2004 European Guideline: "The effectiveness of specific types of exercise therapy needs to be further evaluated. This includes the evaluation of spinal stabilisation exercises..." [8] p 7.
Motor control exercise: treatment rationale
The use of motor control exercise is based on research that has shown that:
(i) People with low back pain have changes in the strategy for control of the trunk muscles in that activity of the deep muscles is impaired (delayed, less tonic) and these muscles are atrophied[19,20].
(ii) Although all muscles contribute to control of movement and stability of the spine, the deep muscles have a critical role for control of intervertebral motion [21-25], but with the potential advantage of allowing dynamic control of the spine.
(iii) Evidence that people with back pain tend to adopt a strategy for increased stiffness and stability at the expense of spinal function [26].
(iv) Non-resolution of changes in the deep muscle system is linked to recurrences of low back pain [27].
The evidence above underpins the primary aim of motor control exercise, which is to re-establish normal control of the deep spinal muscles, reducing the activity of more superficial muscles that tend to stiffen the spine and have increased activity in low back pain, and then maintain normal control during progressively more demanding physical and functional tasks[28].
The key feature of the motor control exercise approach is the training of the deep trunk muscles in isolation before progressing to demanding tasks that train coordination of the deep and the superficial trunk muscles [28]. However, unlike functional restoration approaches, training the deep trunk muscles in isolation from the superficial trunk muscles is difficult. In order to teach patients how to contract the deep muscles of the spine, in addition to clinical skills of palpation and observation [29] physiotherapists need to use technical devices such as pressure monitors, electromyography and ultrasound imaging to provide feedback to the patient.
The premise of the motor control approach is that simple functional exercise alone does not re-establish coordination of the trunk muscles. This premise is supported by the finding that the adaptation of these muscles to pain is still present following recovery from an episode of low back pain, when patients have returned to normal functional levels [19,20]. Furthermore, recent data confirm that coordination of the abdominal muscles can be restored with training of specific activation of the trunk muscles, but not a simple activation during a sit up task [30]. Notably, non-resolution of muscle dysfunction is associated with increased back pain recurrence [27]. Also, asymptomatic people with normal activity levels who are unable to perform a task that is thought to reflect voluntary activation of the deep trunk muscles, are ~6 times more likely to develop back pain than asymptomatic people who are able to perform the same task [31].
Motor control exercise: level I and II evidence
At present there is no evidence for the efficacy of motor control exercise in the treatment of chronic low back pain. No systematic review of motor control exercise has been published, although one is being completed by our group. While the majority of trials (5 of 8) report that motor control exercise is effective in the management of chronic or recurrent low back pain most (7 of 8) have permitted co- intervention so that the contribution of motor control exercise is unclear. Additionally, all of these previous trials have used other treatments of unknown efficacy as the comparison intervention and so treatment efficacy cannot be measured. For example the earliest trial [12] reported that motor control exercise is more effective than usual medical care however this result provides an ambiguous estimate of treatment effectiveness because other trials have reported that sham physiotherapy treatments are more effective than usual medical care [32].
We will evaluate the efficacy of motor control exercise in a placebo-controlled randomised controlled trial. The results of our study will be invaluable for more efficacious evidence-based management of patients with non-specific chronic low back pain. Once efficacy is established, we will be able to progress to measuring whether there are additive or multiplicative effects of other treatments that are commonly administered as co-interventions with motor control exercise and thus to being able to make valid recommendations for their use.
Methods
Overview of research design
The study will be a randomised, blinded, placebo-controlled trial of a motor control exercise program for patients with chronic low back pain. The exercise program will consist of 12 individually supervised half-hour sessions over an 8-week period with treatment outcomes measured at 2 months, 6 months and one year.
Hypotheses
(i) An 8-week motor control exercise program designed to restore control of the trunk muscles improves pain, disability and global perceived effect in participants with chronic low back pain at 2 months follow-up.
(ii) The improvements in pain, disability and global perceived effect following motor control exercise are maintained at 6 and 12 months follow-up.
(iii) At 12 month follow up recurrence is less in the motor control exercise group.
Subject recruitment
A total of 154 participants will be recruited into the study. Participants will be screened for suitability for motor control exercise according to usual clinical practice. The screening instruments identify participants who are unsuitable for exercise management of their low back pain because of significant co- morbidity (serious spinal pathology, contraindication to exercise). A clinical assessment will identify patients who we expect would best be managed by a motor control exercise program rather than some other form of exercise or physiotherapy management.
Screening
To screen for serious pathology, the physiotherapist will conduct a diagnostic triage [33]. Participants in whom serious spinal pathology is suspected will be excluded from the trial and referred to their medical practitioner for review. Potential participants will be screened for contraindications to exercise using the Physical Activity Readiness Questionnaire [34]. If a volunteer provides a positive response to items 1, 2, 3, 4, 6 or 7, the trial physiotherapist will discuss the case with the referring medical practitioner and if necessary a medical review will be undertaken to exclude any contraindication to exercise as listed in the ACSM guidelines [34].
The clinical assessment used to ensure that the motor control approach is indicated is based on the key text [28] and is a normal part of clinical assessment of low back pain. The assessment involves evaluation of the motor control strategy during a specific trunk muscle task – drawing in of the lower abdomen while maintaining an isometric contraction of the medial back muscles. The following criteria constitute correct performance of the task:
1. Moderate and sustained activation (> 10 seconds) of transversus abdominis
2. Moderate and sustained activation (> 10 seconds) of the lumbar multifidus muscles
3. Little or no activation of the global trunk muscles
4. No spinal or rib cage movement.
5. Normal breathing
Evaluation of task performance including satisfaction of the above criteria is dependent on the clinical skills of the physiotherapist. Patients who are unable to perform this task correctly will be considered suitable for motor control exercise.
Participants will be included if they meet all of the following inclusion criteria:
• Non-specific low back pain +/- leg pain of at least 3 months duration
• Currently seeking care for low back pain
• Aged greater than 18 and less than 80 years
• Comprehends English
• Clinical assessment indicates that the subject is suitable for motor control exercise
• Expects to continue residing in SWSAH region for study duration.
Participants will be excluded if they have any of the following:
• Suspected or confirmed serious spinal pathology (fracture, metastatic, inflammatory or infective diseases of the spine, cauda equina syndrome/widespread neurological disorder)
• Suspected or confirmed pregnancy
• Unable to speak English
• Nerve root compromise (2 of strength, reflex or sensation affected for same nerve root)
• Spinal surgery
• Scheduled for major surgery during treatment or follow-up period
• Any of the contraindications to exercise listed on page 42 of the ACSM guidelines [34]
• Any contraindication to pulsed ultrasound or pulsed shortwave.
Specific spinal pathology or contraindication to treatment may be suspected based on the results of the screening questionnaire and the Physical Activity Readiness Questionnaire. If the assessor suspects the presence of any pathology or contraindication to treatment, these subjects should be further investigated and medical clearance obtained, if necessary.
Assessment and allocation
Outcome measures
Measures of outcomes will be obtained at follow-up appointments at 2, 6 and 12 months after randomisation. To maximise attendance at these follow-ups, appointments will be made by phone and then a letter will be sent confirming appointment and a reminder phone call will be made 24 hrs before the appointment. Every attempt (within ethical constraints) will be made to obtain outcome data, regardless of subject's compliance with trial protocols. Follow-up measures will be scored by an investigator who is blinded to group allocation. At 2 months, information about side effects of treatment will be collected from all participants using open-ended questioning.
Following the screening consultation, personal characteristics (age, gender, ethnicity, religion, weight, height, level of education, employment status, doctor's details and contact information) and information about symptoms of low back pain will be collected (eg DASS 21 [35]; Chronic Pain Grade Questionnaire) The following treatment efficacy variables will be measured at baseline, 2, 6 and 12 months.
1. Average pain intensity over last week (0–10 scale) [36-38]
2. Patient-generated measure of disability (Patient-Specific Functional Scale) [36-38]
3. Global perceived effect (Global Perceived Effect Scale) [36-38]
4. Condition-specific measure of disability (Roland Morris Disability Questionnaire) [36-38]
5. Recurrence at 12 months
The primary outcomes are pain, GPE and PSFS at 2 months and recurrence at 12 months.
Randomisation
Participants will be allocated to treatment group using sealed opaque envelopes. The allocation sequence will be generated by author CM. Participants will be scheduled to receive their first treatment within one week of randomisation.
Interventions
Contemporary physiotherapy practice in exercise prescription is to assess each patient and to implement the form of exercise that is most relevant to the particular clinical presentation. At present this widely accepted approach relies primarily upon the clinical expertise of the therapist. We have elected to evaluate motor control exercise delivered in this manner because this approach is regarded as contemporary best practice.
Participants in each group will receive 12 half hour treatments over an 8-week period, i.e. 2 sessions/week in the first month and 1 session/week in the second month. The treatment sessions are designed to become less frequent over time to encourage independence and continuation of exercise when therapy is complete. This is consistent with current clinical practice.
The motor control exercise program is based on the treatment approach reported by O'Sullivan et al [12], Richardson et al [28], and Moseley [39]. A brief description is provided below.
At the first session, participants will be comprehensively assessed and then will be prescribed exercises aimed at improving function of specific muscles of the low back region to be conducted in sessions 2–11. Stage 1 involves the most commonly prescribed exercise aimed at retraining multifidus (a back muscle) and transversus abdominus (a deep abdominal muscle); these exercises will be supplemented with exercises for the pelvic floor muscles, breathing control and control of spinal posture. Participants will be taught how to contract these muscles independently from the superficial trunk muscles [28,40]. Physiotherapists will use real-time ultrasound biofeedback to enhance learning of the tasks. When participants are able to perform these exercises, they will be gradually upgraded until the patient is able to maintain isolated contractions of these muscles for 10 seconds, up to a maximum of 10 repetitions, during normal respiration [28]. When this level of competence has been achieved, patients will be considered ready to progress to Stage 2.
Stage 2 of the approach involves increasing the complexity of the exercise by progressing through a range of functional tasks and exercises targeting coordination of trunk and limb movement and maintenance of trunk stability. The range and progression of exercises is well set-out in clinical texts [28] and is individualised to the patient based on this presentation. Participants require the ongoing support of a trained physiotherapist to ensure correct performance of the exercises. Session 12 is a discharge session where the patient's progress will be reviewed and patients will be prescribed exercises to continue at home.
The placebo intervention is 20 minutes of detuned short wave diathermy and 5 minutes of detuned ultrasound for 12 sessions over an eight week period. This attention control will be used because there is no known treatment effect from the detuned machines, but it has been established in previous trials (including one of our own [37]) that participants view this as a credible treatment. To increase the perceived credibility of the attention control, participants will undergo an examination including routine screening for contraindications at the first consultation and the normal clinical reassessment that would occur with the active forms of these interventions at each subsequent treatment. Each placebo treatment session will be 30 minutes in duration to match the active treatment sessions.
Participants in both treatment groups will be asked not to seek other treatments for their chronic low back pain and where possible not to change current medications during the treatment period. Several mechanisms will be used to ensure that the trial protocol is consistently applied. Protocol manuals will be developed and staff will be trained to ensure that screening, assessment, randomisation and treatment procedures are conducted according to protocol. To ensure standardisation across sites we will hold regular meetings with site visits and teleconferencing. An independent researcher will monitor a randomly chosen subset to ensure adherence to assessment, randomisation and treatment procedures.
If a participant is concerned about his or her condition during the study, the physiotherapist will screen for potentially serious pathology and, where appropriate, refer the patient to a medical practitioner. The medical practitioner will be asked not to request the participant's group allocation unless it is deemed necessary for medical care. At the completion of the exercise program, patients will be encouraged to continue the home exercise routine demonstrated at the discharge session. Participants will be free to seek other treatment after the experimental period.
After the first treatment session the patient will complete a treatment credibility scale [41]. At the 8 week follow-up information about side-effects of treatment will be collected using open-ended questioning. At the 12 month follow-up the participants will be asked to rate the helpfulness, understanding and friendliness of therapist and helpfulness of treatment and to nominate which treatment they thought they received. Additionally information about other treatment received for their low back pain during the study period will be sought using open-ended questioning.
Data integrity
The integrity of trial data will be monitored by regularly scrutinising data sheets for omissions and errors. Data will be double entered and the source of any inconsistencies will be explored and resolved.
Data analysis
Treatment efficacy
In our primary analysis, we will use a regression model to test for the effect of treatment on outcome at 2, 6 and 12 months follow-up with the baseline value of the outcome entered as a covariate. A treatment effect size will be calculated for each of the follow-up time points and, if there is a statistically significant treatment effect at any time point, we will also calculate number needed to treat (NNT) to achieve pain recovery (pain < 1 out of 10: [42]) and 95% CI. The recurrence outcome will be analysed with logistic regression.
Predictor of response to treatment
We will include an interaction term baseline DASS-21 depression score × group to the regression analysis to see if the effect of motor control exercise is influenced by the baseline DASS-21 depression score.
Sample size calculations
We have designed the study to detect a clinically important difference of 1 unit on the 0–10 pain intensity scale (estimate for SD = 2.00), 1 unit on the 0–10 patient specific functional scale (estimate for SD = 1.8); 1 unit on the 0–10 Global Perceived Effect Scale (estimate for SD = 1.65) and 4 units on the 24 item Roland Morris Disability Questionnaire (estimate for SD = 4.9). We have taken the SD estimates from a trial we completed that recruited a similar patient cohort[37] With specifications of alpha = 0.05, power = 0.80 a sample size of 77 participants per group is required to detect an effect size of 0.50 SD (the smallest effect size we have specified for the four outcomes). Based on the results of the same trial [37] we have allowed for 15% non-compliance with treatment, 15% loss to follow-up, and assumed a correlation between baseline and change scores of outcomes of 0.5. Accordingly we will recruit 77 participants per group or 154 participants in total.
Justification of study design
Placebo
Designing an appropriate placebo treatment that mimics a physiotherapy exercise program is challenging. The sham interventions used in previous exercise trials do not satisfy the criteria of being both inert (e.g. the use of hot packs) and credible (e.g. allocation to a treatment waiting list). Accordingly, we will use sham electrotherapy as a control. This sham is clearly inert and is regarded as a credible treatment by participants. [37] To ensure that participants remain unaware of study group, it is necessary to carefully describe the study to patients. In the previous trial where we used sham electrotherapy as a control for exercise, we used the following description:
'In this trial normal physiotherapy treatment and placebo physiotherapy treatment will be provided. A placebo treatment is a harmless treatment delivered at less than the effective dose. We will not tell you which type of treatment you will receive and it is unlikely that you could distinguish them.'
Trial staff described the placebo intervention as 'pulsed ultrasound' and 'pulsed shortwave' and explained to patients that they would probably not feel any sensation during treatment.
Controlling bias
The trial has been designed to include key methodological features that have been recognised as minimising bias in clinical trials. These features include: true randomisation, concealed allocation, specification of eligibility criteria, blind outcome assessment, patient blinding, blind analysis and intention-to-treat analysis. The nature of the treatments precludes blinding of treatment provider. Trial staff will be trained to ensure consistency of screening, assessment, randomisation and treatment procedures. Participant's perception of the credibility of treatment will be determined after the first treatment [41]; and at the completion of treatment both assessors and participants will be asked to identify what treatment they think the participant received.
Outcomes
Measures of pain symptoms, disability and generic health status will be taken from the 'core set' of outcome measures for clinical research recently advocated by an international panel of back pain researchers [43]. The panel considered factors such as reliability, validity and responsiveness before recommending a measure. We have supplemented the back-related disability measure advocated in the core set (Roland Morris) with a patient-generated measure of disability (Patient-Specific Functional Scale) because there is evidence that patient-generated measures of disability are more responsive than condition-specific measures [37,44].
Conclusion
We have presented the rationale and design of a randomized controlled clinical trial evaluating the effect of motor control exercise versus placebo in patients with chronic LBP. The results of this trial will be published as soon as they are available.
Competing interests
All author(s) declare that they have no competing interests.
Authors' contributions
CGM, JL, PWH, KMR, GLM, RDH and LOPC were responsible for the design of the study. LOPC and JM will act as trial coordinators. All authors have read and approved the final manuscript.
Pre-publication history
The pre-publication history for this paper can be accessed here:
==== Refs
Bogduk N Management of chronic low back pain Medical Journal of Australia 2004 180(2) 79 83 14723591
Waddell G The clinical course of low back pain. The Back Pain Revolution 1998 Edinburg, Churchil Livingstone
Commonwealth Department of Health and Ageing National Health Priorities and Quality 2003
Australian Institute of Health and Welfare Australia's health 2000: the seventh biennial health report of the Australian Institute of Health and Welfare 2000 Camberra, AIHW
Australian Bureau of Statistics National Health Survey. Summary of Results. 2002 Canberra, Australian Bureau of Statistics
Von Korff M Ormel J Keefe FJ Dworkin SF Grading the severity of chronic pain Pain 1992 50(2) 133 149 1408309 10.1016/0304-3959(92)90154-4
Blyth F March L Cousins M Chronic pain-related disability and use of analgesia and health services in Sydney community Medical journal of Australia 2003 179 84 87 12864718
Cost B13 working group European Guidelines for the Management of Chronic Non-Specific Low back Pain 2004
Cochrane Collaboration Back Review Group Reviews 2004
Chew CA May CR The benefits of back pain. Family Practice 1997 14(6) 461 465 9476077 10.1093/fampra/14.6.461
Hansson M Bostrom C Harms-Ringdahl K Living with spine-related pain in a changing society--a qualitative study Disability & Rehabilitation 2001 23(7) 286 295 11354581
O'Sullivan P Twomey L Allison G Evaluation of specific stabilizing exercise in the treatment of chronic low back pain with radiologic diagnosis of spondylolysis or spondylolisthesis Spine 1997 22 2959 2967 9431633 10.1097/00007632-199712150-00020
Kankaanpaa M Taimela S Airaksinen O Hanninen O The efficacy of active rehabilitation in chronic low back pain. Effect on pain intensity, self-experienced disability, and lumbar fatigability Spine 1999 24(10) 1034 1042 10332798 10.1097/00007632-199905150-00019
Yelland MJ Glasziou PP Bogduk N Schluter PJ McKernon M Prolotherapy injections, saline injections, and exercises for chronic low-back pain: a randomized trial Spine 2004 29(1) 9 16 14699269 10.1097/01.BRS.0000105529.07222.5B
van Tulder M Malmivaara A Esmail R Koes B Exercise therapy for low back pain: a systematic review within the framework of the cochrane collaboration back review group. Spine 2000 25(21) 2784 2796 11064524 10.1097/00007632-200011010-00011
Hemmila HM Keinanen-Kiukaanniemi SM Levoska S Puska P Long-term effectiveness of bone-setting, light exercise therapy, and physiotherapy for prolonged back pain: a randomized controlled trial Journal of Manipulative & Physiological Therapeutics 2002 25(2) 99 104 11896377 10.1067/mmt.2002.122329
Egger M Juni P Bartlett C Holenstein F Sterne J How important are comprehensive literature searches and the assessment of trial quality in systematic reviews? Empirical study Health Technology Assessment 2003 7(1) 1 76 12583822
Ferreira PH Effectiveness of specific stabilisation exercises for chronic low back pain Physiotherapy 2004 , The University of Sydney
Hodges PW Richardson CA Inefficient muscular stabilization of the lumbar spine associated with low back pain. A motor control evaluation of transversus abdominis Spine 1996 21(22) 2640 2650 8961451 10.1097/00007632-199611150-00014
Hides JA Stokes MJ Saide M Jull GA Cooper DH Evidence of lumbar multifidus muscle wasting ipsilateral to symptoms in patients with acute/subacute low back pain Spine 1994 19(2) 165 172 8153825
Hodges P Kaigle Holm A Holm S Ekstrom L Cresswell A Hansson T Thorstensson A Intervertebral stiffness of the spine is increased by evoked contraction of transversus abdominis and the diaphragm: in vivo porcine studies Spine 2003 28(23) 2594 2601 14652476 10.1097/01.BRS.0000096676.14323.25
Wilke HJ Wolf S Claes LE Arand M Wiesend A Stability increase of the lumbar spine with different muscle groups. A biomechanical in vitro study. Spine 1995 20(2) 192 198 7716624
Solomonow M Zhou BH Harris M Lu Y Baratta RV The ligamento-muscular stabilizing system of the spine Spine 1998 23(23) 2552 2562 9854754 10.1097/00007632-199812010-00010
Macintosh J Bogduk N The biomechanics of the lumbar multifidus Clinical Biomechanics 1986 1 205 213 10.1016/0268-0033(86)90147-6
Kaigle AM Holm SH Hansson TH Experimental instability in the lumbar spine Spine 1995 20(4) 421 430 7747225
van Dieen JH Selen LP Cholowicki J Trunk muscle activation in low back pain patients, an analysis of literature Journal of Eletromyography Kinesiology 2003 13 333 351 10.1016/S1050-6411(03)00041-5
Hides JA Jull GA Richardson CA Long-term effects of specific stabilizing exercises for first-episode low back pain Spine 2001 26(11) E243 8 11389408 10.1097/00007632-200106010-00004
Richardson CA Jull GA Hodges PW Livingstone C Therapeutic Exercise for Spinal Segmental Stabilization in Low Back Pain 1999 Edinburg,
Urquhart DM Hodges P Allen TJ Story IH Abdominal muscles recruitment during a range of voluntary exercises Manual Therapy 2005 10 144 153 15922235 10.1016/j.math.2004.08.011
Tsao H Hodges P Specific abdominal retraining alters motor coordination in people with persistent low back pain: ; Sydney. 2005
Moseley GL Impaired trunk muscle function in sub-acute neck pain: etiologic in the subsequent development of low back pain? Manual Therapy 2004 9(3) 157 163 15245710 10.1016/j.math.2004.03.002
Koes BW Bouter LM van Mameren H Essers AH Verstegen GM Hofhuizen DM Houben JP Knipschild PG The effectiveness of manual therapy, physiotherapy, and treatment by the general practitioner for nonspecific back and neck complaints. A randomized clinical trial.[see comment] Spine 1992 17(1) 28 35 1531552
Waddell G Feder G McIntosh A Lewis M Hutchison A Low back pain evidence review. 1996 London, Royal College of General Practitioners
ACSM ACSM's Guidelines for Exercise Testing and Prescription. 1995 Baltimore, Willams and Wilkins
Henry JD Crawford JR The short-form version of the Depression Anxiety Stress Scales (DASS-21): construct validity and normative data in a large non-clinical sample. British Journal of Clinical Psychology 2005 44 227 239 16004657 10.1348/014466505X29657
Scrimshaw SV Maher C Responsiveness of visual analogue and McGill pain scale measures Journal of Manipulative & Physiological Therapeutics 2001 24(8) 501 504 11677548 10.1067/mmt.2001.118208
Pengel L Outcome of recent onset low back pain (PhD Thesis) Physiotherapy 2004 Sydney, The University of Sydney
Maher C Latimer J Refshauge K Atlas of Clinical Tests and Measures for Low Back Pain. 2001 Melbourne, Australian Physiotherapy Association
Moseley L Combined physiotherapy and education is efficacious for chronic low back pain Australian Journal of Physiotherapy 2002 48(4) 297 302 12443524
Twomey L Taylor J Livingstone C Physical therapy for low back pain. Clinics in Physical Therapy. 1994 New York,
Borkovec T Nau S Credibility of analogue therapy rationales Journal of Behaviour Therapy and Experimental Psychiatry 1972 3 257 260 10.1016/0005-7916(72)90045-6
McGuirk B King W Govind J Lowry J Bogduk N Safety, efficacy, and cost effectiveness of evidence-based guidelines for the management of acute low back pain in primary care. Spine 2001 26(23) 2615 2622 11725244 10.1097/00007632-200112010-00017
Deyo RA Battie M Beurskens AJ Bombardier C Croft P Koes B Malmivaara A Roland M Von Korff M Waddell G Outcome measures for low back pain research. A proposal for standardized use. Spine 1998 23(18) 2003 2013 9779535 10.1097/00007632-199809150-00018
Stratford P Gill C Westaway M Binkley J Assessing disability and change on individual patients: a report of a patient specific measure Physiotherapy Canada 1995 47 258 263
|
16271149
|
PMC1291370
|
CC BY
|
2021-01-04 16:32:05
|
no
|
BMC Musculoskelet Disord. 2005 Nov 4; 6:54
|
utf-8
|
BMC Musculoskelet Disord
| 2,005 |
10.1186/1471-2474-6-54
|
oa_comm
|
==== Front
BMC NephrolBMC Nephrology1471-2369BioMed Central London 1471-2369-6-121627448610.1186/1471-2369-6-12Research ArticleHypothyroidism attenuates protein tyrosine nitration, oxidative stress and renal damage induced by ischemia and reperfusion: effect unrelated to antioxidant enzymes activities Tenorio-Velázquez Verónica M [email protected] Diana [email protected] Martha [email protected] Edilia [email protected]ández-Pando Rogelio [email protected] Omar Noel [email protected] José [email protected] Facultad de Química, Departamento de Biología, Edificio B, Segundo Piso, Laboratorio 209, Universidad Nacional Autónoma de México (UNAM), Ciudad Universitaria, 04510, México, D.F., México2 Departamento de Nefrología, Instituto Nacional de Cardiología "Ignacio Chávez", Juan Badiano #1, Col. Sección XVI, 14080, Tlalpan, México, D.F., México3 Facultad de Medicina, Departamento de Farmacología, Universidad Nacional Autónoma de México (UNAM), Ciudad Universitaria, 04510, México, D.F., México4 Instituto Nacional de Ciencias Médicas y Nutrición "Salvador Zubirán", Departamento de Patología, 14000, México, D.F., México2005 7 11 2005 6 12 12 25 6 2005 7 11 2005 Copyright © 2005 Tenorio-Velázquez et al; licensee BioMed Central Ltd.This is an Open Access article distributed under the terms of the Creative Commons Attribution License (), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
Background
It has been established that hypothyroidism protects rats against renal ischemia and reperfusion (IR) oxidative damage. However, it is not clear if hypothyroidism is able to prevent protein tyrosine nitration, an index of nitrosative stress, induced by IR or if antioxidant enzymes have involved in this protective effect. In this work it was explored if hypothyroidism is able to prevent the increase in nitrosative and oxidative stress induced by IR. In addition the activity of the antioxidant enzymes catalase, glutathione peroxidase, and superoxide dismutase was studied. Control and thyroidectomized (HTX) rats were studied 24 h of reperfusion after 60 min ischemia.
Methods
Male Wistar rats weighing 380 ± 22 g were subjected to surgical thyroidectomy. Rats were studied 15 days after surgery. Euthyroid sham-operated rats were used as controls (CT). Both groups of rats underwent a right kidney nephrectomy and suffered a 60 min left renal ischemia with 24 h of reperfusion. Rats were divided in four groups: CT, HTX, IR and HTX+IR. Rats were sacrificed and samples of plasma and kidney were obtained. Blood urea nitrogen (BUN) and creatinine were measured in blood plasma. Kidney damage was evaluated by histological analysis. Oxidative stress was measured by immunohistochemical localization of protein carbonyls and 4-hydroxy-2-nonenal modified proteins. The protein carbonyl content was measured using antibodies against dinitrophenol (DNP)-modified proteins. Nitrosative stress was measured by immunohistochemical analysis of 3-nitrotyrosine modified proteins. The activity of the antioxidant enzymes catalase, glutathione peroxidase, and superoxide dismutase was measured by spectrophotometric methods. Multiple comparisons were performed with ANOVA followed by Bonferroni t test.
Results
The histological damage and the rise in plasma creatinine and BUN induced by IR were significantly lower in HTX+IR group. The increase in protein carbonyls and in 3-nitrotyrosine and 4-hydroxy-2-nonenal modified proteins was prevented in HTX+IR group. IR-induced decrease in renal antioxidant enzymes was essentially not prevented by HTX in HTX+IR group.
Conclusion
Hypothyroidism was able to prevent not only oxidative but also nitrosative stress induced by IR. In addition, the antioxidant enzymes catalase, glutathione peroxidase, and superoxide dismutase seem not to play a protective role in this experimental model.
==== Body
Background
Reactive oxygen species (ROS) [1,2] and reactive nitrogen species such as peroxynitrite (ONOO-) [3] are involved in the damage induced by ischemia and reperfusion (IR). The damage by reactive nitrogen species has been made evident by the increase in protein tyrosine nitration [3-6]. The consequences of IR include alterations in DNA, lipids, and proteins (carbonyl formation and nitrosylation) [3-6]. Renal IR is associated with acute renal failure [3,4] as well as proximal tubular damage [1-3]. IR-induced damage is ameliorated by spin traps [2], inhibition of inducible nitric oxide synthase [3], lecithinized superoxide dismutase (SOD) [3], ebselen, a ONOO- scavenger [3], inhibitors of calpain activation [4], SOD and catalase (CAT) mimetic [6], antioxidants [7,8], or in the other circumstances such as hypothyroidism [9]. Paller [9] found that the renal damage and the increase in malondialdehyde (MDA) induced by IR were significantly lower in hypothyroid than in euthyroid rats. The specific mechanisms involved in the protective effect of hypothyroidism against renal IR remain to be fully elucidated.
The role of antioxidant enzymes in the oxidative damage to kidney has been studied. It has been found that the elevated expression of antioxidant enzymes including CAT, SOD, glutathione peroxidase (GPx) [10-15], and more recently heme oxygenase-1 [16,17], prior to renal oxidant insult, was able to ameliorate renal damage. Furthermore, the inhibition of CAT [18] or heme oxygenase-1 [16] aggravates renal damage induced by puromycin aminonucleoside [18] or IR [16], respectively. These data strongly suggest that the modulation of the antioxidant enzymes may alter the renal damage induced by oxidants. It is unknown if the antioxidant enzymes may be regulated differentially and involved in the protective effect of hypothyroidism against renal IR. Interestingly, the administration of some exogenous antioxidants is able to modulate antioxidants enzymes and renal damage induced by IR [19-21]. In addition, (-)-epicatechin 3-O-gallate [22] and Wen-Pi-Tang [23] induced renal antioxidant enzymes and protected against lipopolysaccharide- and IR-induced kidney damage and plasma 3-nitrotyrosine (3-NT) formation. Tyrosine nitration may be induced not only by ONOO-, but also by another reactive nitrogen species including nitrogen dioxide radical (NO2•) and dinitrogen trioxide (N2O3). ONOO- is a potent oxidation species that have been found to cause also lipid peroxidation and cytotoxicity [24-26].
Based on the above information, in the present paper we evaluated if hypothyroidism is able to prevent against the nitrosative stress induced by IR. Nitrosative stress was evaluated by measuring nitrated proteins by immunohistochemistry using antibodies against 3-NT [3-6,27,28]. Oxidative stress was evaluated using immunohistochemical techniques to evaluate the protein carbonyl content [29] and 4-hydroxy-2-nonenal (4-HNE) protein adducts [30,31]. The protein carbonyl content was measured using antibodies against dinitrophenol (DNP)-modified proteins [29]. In addition the activity of the antioxidant enzymes CAT, GPx, and SOD was studied before and after renal IR in control and hypothyroid rats.
Methods
Reagents
Xanthine, nitroblue tetrazolium (NBT), 3,3-diaminobenzidine, bovine serum albumin, xanthine oxidase, NADPH, glutathione reductase (GR), 2,4-dinitrophenylhydrazine, and reduced glutathione (GSH) were purchased from Sigma (St. Louis, MO, USA). Ethylenediaminetetraacetic acid disodium salt (EDTA Na2), ammonium sulfate, and copper chloride were purchased from JT Baker (Mexico City, México). Hydrogen peroxide (H2O2), formaldehyde, and sodium carbonate were obtained from Mallinckrodt (Paris, KY, USA). Sodium azide was obtained from Merck (Mexico City, México). Rabbit anti-3-NT polyclonal antibodies were from Upstate (Catalogue # 06-284, Lake Placid, NY, USA). Goat anti-DNP polyclonal antibodies (Catalogue # J06) were from Biomeda Corporation (Foster City, CA, USA). Mouse anti-4-HNE monoclonal antibodies (Catalogue #24325) were from Oxis International Inc. (Portland, OR, USA). Anti-rabbit Ig horseradish peroxidase antibody (Catalogue # NA-934) and anti-mouse Ig horseradish peroxidase antibody (Catalogue # NIF-825) were purchased from Amersham Life Sciences (Buckinghamshire, England). Donkey anti-goat horseradish peroxidase antibodies (Catalogue # SC2020) were from Santa Cruz Biotechnology, Inc. (Santa Cruz, CA, USA). All other chemicals were reagent grade and commercially available.
Induction of hypothyroidism
All animal procedures were approved by the Animal Care Committee of the Instituto Nacional de Cardiología "Ignacio Chavez" and followed the guidelines of Norma Oficial Mexicana (NOM-ECOL-087-1995). Male Wistar rats weighing 380 ± 17 g underwent surgical thyroidectomy with parathyroid reimplant (HTX), as previously described [32-35]. Briefly, the trachea was exposed under ether anesthesia, and under a stereoscopic microscope (Wild M5, Wild Heerbrugg, Switzerland), the parathyroid glands were visualized, dissected from the thyroid gland, and reimplanted into the surrounding neck muscles. The thyroid gland was then carefully dissected, to avoid injury to the laryngeal nerves, and completely excised. The effectiveness of this procedure was assessed by determining the concentration of calcium, phosphorous and thyroxine in 10 sham-operated control and 10 HTX rats, using standard techniques. The results obtained 15 days after surgery were: Ca2+ of 10.2 ± 3 in control vs. 10.3 ± 0.2 mM in HTX; phosphorous of 6.5 ± 0.3 in control vs. 6.3 ± 0.5 mM in HTX; and thyroxine of 6.4 ± 0.77 in control vs. 1.18 ± 0.19 μg dL-1 in HTX, P < 0.05. The sham group (375 ± 14 g) underwent a surgical procedure in which the animals were anesthetized, the trachea was exposed and then the incision was closed simulating the thyroidectomy surgery. The body weight for HTX and CT rats obtained 15 days after surgery was 389 ± 22 g and 417 ± 25 g, respectively.
Ischemia and reperfusion studies
Four groups of rats were studied: Control (CT), sham operated animals; hypothyroid (HTX), rats subjected to thyroidectomy; ischemia and reperfusion (IR), rats submitted to IR; and HTX+IR, rats subjected to HTX plus IR. The experimental protocol was performed 15 days after the thyroidectomy (HTX group) or the simulated surgery (CT). Under anesthesia and heparin administration, blood samples were obtained and the kidneys were reperfused and removed. Additional animals from CT and HTX groups were subjected to right nephrectomy and the left renal artery was occluded with a non-traumatic vascular clamp for 60 min. Then, the clamp was released allowing the reestablishment of renal blood flow or reperfusion and 24 h after the rats were anesthetized, blood samples were obtained and the kidney was washed with 0.9% saline solution and excised. These groups were named as IR and HTX+IR, respectively. Blood plasma was obtained and stored at -40°C. Kidney was used for histological and immunohistochemical studies and for determination of antioxidant enzymes activities. Areas of the kidney (renal cortex, outer medulla and inner medulla) were macroscopically dissected using a razor blade and frozen at -70°C for further measurement of enzymatic activities.
Determination of plasma creatinine and blood urea nitrogen (BUN)
Creatinine and BUN were measured using a creatinine analyzer model 2 and a BUN analyzer 2 (Beckman Instruments, Fullerton, CA, USA), respectively.
Histological studies
Kidney tissue slices were fixed in 10% neutral buffered formaldehyde solution, and embedded in paraffin [27,36,37]. Sections at 4 μm of thickness were obtained and stained with hematoxylin-eosin (H&E). Histologic assessment of tubular necrosis was determined semiquantitatively using a method described by Chatterjee et al. [38]. The score was graded from 0 to 3 where 0 = normal histology, 1 = tubular cells swellings, brush border loss, nuclear condensation, with up to one third of tubular profile showing nuclear loss; 2 = same as for score, but greater than one third and less than two thirds of tubular profile shows nuclear loss; 3 = grater of two thirds of tubular profile shows nuclear loss. In addition a quantitative histological damage was determined by using a Leica Qwin Image Analyzer (Leica Microsystems, Cambridge, UK). The following parameters were quantified: (a) the percentage of tubules with hyaline casts in the renal cortex, using the low power magnification objective, three random choice fields were studied counting the tubules without and with hyaline casts to determine the percentage of the latter, and (b) in those tubules with hyaline casts, the total lumen area and the area occupied by the cast were determined, then the percentage of the lumen area occupied by the hyaline cast was determined.
Immunohistochemical localization of 3-NT, DNP, and 4-HNE
For immunohistochemistry, 4 μm sections were deparaffinized with xylene and rehydrated with ethanol. Endogenous peroxidase was quenched/inhibited with 4.5% H2O2 in methanol by incubation for 1.5 h at room temperature. The sections used for DNP immunohistochemistry were incubated with 0.2% dinitrophenylhydrazine in 2 N HCl for 60 min at room temperature in absence of light and then were extensively washed. Nonspecific adsorption was minimized by leaving the sections in 3% bovine serum albumin in phosphate buffer saline for 30 min. Sections were incubated overnight with 1:700 dilution of anti-3-NT antibody [36] or with 1:500 dilution of anti-DNP antibody or with 1:100 dilution of anti-4-HNE antibody [36]. After extensive washing with phosphate buffer saline, the sections were incubated with 1:500 dilution of peroxidase conjugated anti-rabbit Ig antibody (for 3-NT) or with a 1:500 dilution of a peroxidase conjugated anti-goat Ig or anti-mouse IgG (for DNP and 4-HNE, respectively) for 1 h, and finally incubated with H2O2-diaminobenzidine for 1 min. Sections were counterstained with hematoxylin (for 3-NT and 4-HNE) or with methyl green (for DNP) and observed under light microscopy. All the sections from the four studied groups were incubated under the same conditions with the same antibodies concentration, and in the same running, so the immunostaining was comparable among the different experimental groups. Quantitative image analysis was performed with a Zeiss KS 300 Imaging System 3.0 (Carl Zeiss Vision GmbH, Hallbergmoos, Germany). This software determines densitometric means value of selected tissue regions. Thus, 10 fields/rat were randomly selected and the intensity of the 3-NT, 4-HNE, and DNP immunostaining was determined. We normalized the data (arbitrary units) to 1.0 using the control kidneys. We run negative controls omitting primary and/or secondary antibodies.
Tissue homogenization
Renal cortex, outer medulla and inner medulla were homogenized in a Polytron (Model PT 2000, Brinkmann, Westbury, NY, USA) for 10 seconds in cold 50 mM potassium phosphate, 0.1% Triton X-100, pH = 7.0 [28]. The homogenate was centrifuged at 19,000 × g and 4°C for 30 min and the supernatant was separated to measure total protein and the activities of CAT, GPx, and SOD. Total protein was measured by the method of Lowry et al. [39].
Catalase assay
Renal CAT activity was assayed at 25°C by a method based on the disappearance of H2O2 from a solution containing 30 mM H2O2 in 10 mM potassium phosphate buffer pH 7.0 at 240 nm [40]. The reaction was started by the addition of 25 μL of the sample to 725 μL of H2O2. Under the described conditions, the decomposition of H2O2 by CAT contained in the samples follows a first-order kinetic as given by the equation k = 2.3/t log Ao/A where k is the first-order reaction rate constant, t is the time over which the decrease of H2O2, due to CAT activity, was measured (15 s), and Ao/A is the optical density at times 0 and 15 s, respectively. The results were expressed in k/mg protein.
Glutathione peroxidase assay
Renal GPx activity was assayed by a method previously described [41]. Reaction mixture consisted of 50 mM potassium phosphate pH = 7.0, 1 mM EDTA, 1 mM sodium azide, 0.2 mM NADPH, 1 U/mL of glutathione reductase, and 1 mM GSH. One hundred μL of the appropriate dilution of tissue homogenates were added to 0.8 mL of mixture and allowed to incubate for 5 min at room temperature before initiation of the reaction by the addition of 0.1 mL 2.5 mM H2O2 solution. Absorbance at 340 nm was recorded for 3 min and the activity was calculated from the slope of these lines as μmoles of NADPH oxidized per min taking into account that the millimolar absorption coefficient for NADPH is 6.22 mM-1cm-1. Blank reactions with homogenates replaced by distilled water were subtracted from each assay. The results were expressed as U/mg protein.
Superoxide dismutase assay
SOD activity in kidney homogenates was assayed by using a previously reported method [41]. A competitive inhibition assay was performed using xanthine-xanthine oxidase system to reduce NBT. Mixture reaction contains in a final concentration: 0.122 mM EDTA, 30.6 μM NBT, 0.122 mM xanthine, 0.006% bovine serum albumin, and 49 mM sodium carbonate. Five hundred μL of tissue homogenate at the appropriate dilution, were added to 1.66 mL of the mixture described above, then 50 μL xanthine oxidase, in a final concentration of 2.8 U/L, were added and incubated in a water bath at 27°C for 30 min. The reaction was stopped with 066 mL of 0.8 mM cupric chloride and the optical density was read at 560 nm. One hundred percent of NBT reduction was obtained in a tube in which the sample was replaced by distilled water. The amount of protein that inhibited NBT reduction to 50% of maximum was defined as one unit of SOD activity. Results were expressed as U/mg protein.
Statistics
The data are expressed as the mean ± SD. Data were analyzed with a non-paired t-test or with ANOVA followed by multiple comparisons by Bonferroni t test, as appropriate. P value less than 0.05 was considered statistically significant.
Results
General and biochemical data
Creatinine and BUN were similar in CT and HTX groups (Table 1). Twenty four h after IR, both plasma creatinine and BUN increased significantly, however the increases were significantly lower in HTX+IR group (Table 1). These data confirm previous observations of Paller [9] who showed that the renal damage induced by IR was ameliorated in HTX rats.
Histological and immunohistochemical analysis
Representative histopathology and immunohistochemistry features in the kidney cortex of rats after 24 h of IR in control and HTX rats are presented in Figure 1. Surgical induced hypothyroidism does not produce any histological abnormality in the kidney. After IR, the kidney cortex shows extense ischemic tubular necrosis and hyaline cylinders are present in many tubular lumens (Figures 1 and 2).
The histological abnormalities, manifested by numerous necrotic tubules with a high percentage of tubules with hyaline casts, were significantly lower in the HTX+IR group (HTX+IR group vs. IR group P < 0.001, Figure 2) confirming that hypothyroidism prevented renal damage induced by IR (Figure 1). In a similar fashion than in CT rats, the immunostaining of 3-NT, 4-HNE, and DNP is negative in HTX rats, while it is strong in the IR group (Figure 3). The increase in 3-NT, 4-HNE, and DNP staining induced by IR was ameliorated in the HTX+IR group (Figures 1 and 3).
Representative histopathology and immunohistochemistry features in the medulla of the kidney rats in the four groups of animals is presented in Figure 4. There are no abnormalities in the renal medulla in HTX group. After IR, the kidney medulla shows extense ischemic tubular necrosis, and numerous tubular lumens have hyaline casts (Figure 4). Ischemia and reperfusion in HTX rats produced lesser histological damage, occasional medullar kidney tubules are revisted by necrotic epithelium (Figure 4) and a decrease in the area of tubular lumen occupied with hyaline casts (HTX+IR group vs. IR group, P < 0.001, Figure 5). Like in control rats, the immunostaining of 3-NT, 4-HNE and DNP is negative in HTX rats, whereas it is strong after IR (Figure 4). The staining of 3-NT, HNE and DNP in HTX was significantly lower in HTX+IR group compared to IR group (Figures 4 and 6).
Renal activity of antioxidant enzymes
The renal activity of antioxidant enzymes in CT and HTX rats before and after IR is shown in tables 2, 3, 4. Activities of CAT (Table 2) and GPx (Table 3) were measured in renal cortex and outer and inner medulla. Superoxide dismutase was measured in renal cortex and outer medulla (Table 4). There was no enough sample of inner medulla to measure SOD. With one exception (a marginal increase in GPx activity in outer medulla), the comparisons between CT and HTX rats were not significant. Ischemia and reperfusion induced a decrease in the antioxidant enzymes. Catalase activity was decreased in renal cortex of IR group compared to CT group and in renal cortex and outer medulla of HTX+IR group compared to HTX group (Table 2). Glutathione peroxidase activity decreased in renal cortex and outer medulla of IR group compared to CT group and in renal cortex and inner medulla of HTX+IR group compared to HTX group (Table 3). Superoxide dismutase activity decreased in outer medulla of HTX+IR group compared to HTX group (Table 4). With one exception (GPx in outer medulla) the IR-induced decrease in renal antioxidant enzymes was not prevented by HTX in HTX+IR group. In fact, the activities of antioxidant enzymes were significantly lower in HTX+IR group than in IR group. This may suggest that the antioxidant enzymes in outer medulla HTX group are more susceptible to inactivation by IR than those of CT rats.
Discussion
The data presented in this work show that hypothyroid rats were significantly more resistant to IR induced renal damage than euthyroid rats and are consistent with the protective effect of hypothyroidism against oxidative stress and tissue damage in several experimental models [42-44]. The protective effect of HTX was observed by histological (necrotic tubules, percentage of tubules with hyaline casts and area of tubular lumen occupied by hyaline casts) and biochemical (creatinine and BUN) analyses. In addition, this protective effect was associated with a decrease in oxidative damage which was evaluated by the immunohistochemical localization of protein carbonyls and 4-HNE modified proteins. The protective effect of HTX in ischemia and reperfusion associated with the amelioration of oxidative damage had been observed previously by Paller [9]. Although hypothyroidism protects the kidney during IR, the mechanism involved in this protective effect has not been completely elucidated. One of the major effects of thyroid hormones is to increase mitochondrial respiration [45] which results in increased generation of ROS, leading to oxidative damage to membrane lipids. There is a good deal of evidence to indicate that metabolic depression brought about by hypothyroidism is associated with a decrease in free radical production and a subsequent protection against lipid peroxidation [46,47]. This supports the notion that reduced demand for oxygen in hypothyroidism serves as a protective factor in tissue injury due to ROS. In fact, it has been shown that hypothyroidism is able to prevent the increase in lipid peroxidation and the diminution in GSH as well the tissue damage induced by intracolonic administration of trinitrobenzene sulfonic acid (experimental model of colitis) [42]. Hypothyroidism also prevents the increase in MDA and decreases the susceptibility to oxygen radical-induced lung damage in newborn rats exposed to prolonged hyperoxia [43]. The lower toxicity of arsenic in hypothyroid animals was associated with the prevention of arsenic-induced lipid peroxidation in liver and kidneys [44]. Furthermore, hypothyroidism was able to protect against acetaminophen hepatotoxicity [48] which has been associated to oxidative stress [49]. The majority of the studies in hypothyroid animals have found no change or a decrease in tissue markers of oxidative stress (thiobarbituric acid reactive substances, MDA or oxidized glutathione) (Table 5) supporting a decreased production of ROS in this experimental model.
In this study we also showed that IR damage was associated with an increase in tyrosine nitration which is consistent with previous studies [3-6]. Interestingly, it was observed that the protective effect of HTX was associated with a significant decrease in 3-NT immunostaining. Noiri et al. [3], Chatterjee et al. [4,6], and Patel et al. [5] found that the protective effect of ebselen [3], PD150606 and E-64 (inhibitors of calpain activation) [4], interleukin-6 deficiency [5], and EUK-134 (a SOD and CAT mimetic) [6] in renal ischemia and reperfusion damage was associated with attenuation of nitrosative stress evaluated by tyrosine nitration. More recently it was shown that the protective effect of soy feeding of renal damage induced by puromycin aminonucleoside was associated with a decrease in tyrosine nitration [58].
Tyrosine nitration is induced by reactive nitrogen species including ONOO- which is synthesized by the reaction between superoxide anion (O2•↓) and nitric oxide (NO•). The protective effect of ebselen, a ONOO- scavenger, in the IR-induced renal damage and tyrosine nitration suggests that ONOO- is enhanced and involved in tissue damage and nitrosative stress in this experimental model. There are evidences suggesting the increase of O2•↓ and NO• in IR-induced renal damage [reviewed in [59]] which may favor ONOO- formation. In hypothyroid rats, the decreased nitrosative stress may be explained, at least in part, by the diminution in oxygen consumption and O2•↓ production.
We wanted to know if the antioxidant enzymes may be involved in the protective effect of HTX in IR taking into account several reports of the literature showing that the enhanced expression of some antioxidant enzymes was able to attenuate renal damage induced by IR [11,16], H2O2 infusion [10,11], cisplatin [13], puromycin aminonucleoside [10], and cyclosporine A [15]. However, our data show that IR induced a decrease in the activity of CAT, GPx and SOD. Our data are consistent with previous data showing that renal ischemia and reperfusion damage is associated with a decrease in antioxidant enzymes [19,20,60]. Interestingly, the IR-induced decrease in antioxidant enzymes was not prevented by HTX in the HTX+IR group. In fact, in some cases, the IR-induced decrease in antioxidant enzymes was significantly higher in the HTX+IR group compared to IR group suggesting that these antioxidant enzymes are not involved in the protective effect of HTX on IR-induced renal damage and oxidative and nitrosative stress. Antioxidant enzymes activities in CT and HTX groups were essentially similar suggesting that hypothyroidism has no effect on the renal activity of these enzymes. There is no a consistent pattern on the effect of hypothyroidism on tissue antioxidant enzymes (see Table 6). Sawant et al. [67] found that SOD decreased, GPx increased and CAT remained unchanged in hypothyroid rats.
The most studied enzymes in hypothyroid animals are SOD, Cu, ZnSOD, MnSOD, CAT, GPx, and GR. The effect of hypothyroidism on the antioxidant enzymes in several tissues is not consistent (Table 6). In some cases the change of antioxidant enzyme activity seems to be tissue specific [47,68]. On the other hand, within a single tissue, the response of the antioxidant enzymes to hypothyroidism is not always similar [55,61-67].
Hypothyroidism attenuates not only renal but also cardiac damage induced by ischemia and reperfusion. Bobadilla et al. [69] have shown that hypothyroidism conferred protection against reperfusion arrhythmias and the cardiac release of creatine kinase and aspartate amino transferase and preserved the normal structure of myocardial tissue. In addition Chavez et al. [70] demonstrated that hypothyroidism renders mitochondria resistant to the opening of membrane permeability transition pore. This may be relevant to the protective effect of hypothyroidism in ischemia and reperfusion since it has been recognized that mitochondria play a key role in cell-death pathways by activating mitochondrial permeability transition pore and causing the release of cytochrome C and proapoptotic factors, as well as Ca2+ overload that promotes non-selective permeability of the inner membrane. The prolonged opening of the membrane permeability transition pore during the first few minutes of reperfusion is a critical determinant of cell death, and pharmacological inhibition of the pore at the time of reperfusion protects the cells [71].
Conclusion
It is concluded that HTX rats are more resistant to oxidative and nitrosative stress and renal damage induced by IR, which is not mediated by a differential regulation of the antioxidant enzymes CAT, GPx, and SOD.
List of abbreviations used
ANOVA Analysis of variance
BUN Blood urea nitrogen
CAT Catalase
CT Control rats
DNP Dinitrophenol
GPx Glutathione peroxidase
GR Glutathione reductase
GSH Glutathione, reduced form
GSSG Glutathione, oxidized form
4-HNE 4-hydroxy-2-nonenal
H&E Hematoxylin-eosin
HTX Hypothyroid rats
IR Ischemia and reperfusion
MDA Malondialdehyde
MMI Methimazole
3-NT 3-nitrotyrosine
NBT Nitroblue tetrazolium
ONOO-Peroxynitrite
PTU 6-n-propyl-2-thiouracil,
ROS Reactive oxygen species
SD Standard deviation
SOD Superoxide dismutase
Competing interests
The author(s) declare that they have no competing interests.
Authors' contributions
VMTV performed ischemia and reperfusion studies, collected samples and measured the activity of antioxidant enzymes. DB performed histological and immunohistochemical analyses and edited the manuscript. MF thyroidectomized rats and characterized the hypothyroid state, performed ischemia and reperfusion studies and edited the manuscript. ET performed ischemia and reperfusion studies and edited the manuscript. RHP advised about the histological and immunohistochemical analyses. ONMC performed the statistical analyses and edited the manuscript. JPCH conceived and coordinated the study and wrote and edited the manuscript.
Pre-publication history
The pre-publication history for this paper can be accessed here:
Acknowledgements
This work was supported by CONACYT 0359P-M to M Franco.
Figures and Tables
Figure 1 Representative images of histology (H&E) (first row) and immunohistochemical detection of 3-NT (second row), 4-HNE (third row) and DNP (fourth row) in the renal cortex of the four groups of rats studied: Control (CT), thyroidectomized (HTX), ischemia and reperfusion (IR) and HTX+IR (100× magnification).
Figure 2 Quantitative analysis of damage in the renal cortex of rats from IR and HTX+IR groups. (A) Percentage of tubules with hyaline casts. n = 4 rats and 60 tubuli/rat. (B). Severity score. Mean ± SD. n = 4 rats and 15 tubuli/rat. aP < 0.001 vs. IR.
Figure 3 Quantitative immunostaining of 3-NT, 4-HNE, and DNP-modified proteins in renal cortex of the four groups of rats studied: CT, HTX, IR, and HTX+IR. n = 4 rats and 10 determinations/rat. Mean ± SD. aP < 0.001 vs. control, bP < 0.001 vs. IR.
Figure 4 Representative images of histology (H&E) (first row) and immunohistochemical detection of 3-NT (second row), 4-HNE (third row) and DNP (fourth row) in renal medulla of the four groups of rats studied: CT, HTX, IR, and HTX+IR. (100× magnification).
Figure 5 Quantitative analysis of damage in the renal medulla of rats from IR and HTX+IR groups. Area of tubular lumen with hyaline casts is shown as percentage. n = 4 rats, and 60 tubular lumen/rat. Mean ± SD. aP < 0.001 vs. IR.
Figure 6 Quantitative immunostaining of 3-NT, 4-HNE, and DNP-modified proteins in renal medulla of the four groups of rats studied: CT, HTX, IR, and HTX+IR. n = 4 rats and 10 determinations/rat. Mean ± SD. aP < 0.001 vs. control, bP < 0.001 vs. IR.
Table 1 Plasma creatinine and BUN in the four groups of rats studied.
CT HTX IR HTX+IR
Creatinine, mg/dL 0.45 ± 0.05 0.40 ± 0.08 5.08 ± 0.55a 3.83 ± 0.41*
BUN, mg/dL 26 ± 2 23 ± 1 122 ± 11a 78 ± 17*
Data are mean ± SD. N = 16. aP < 0.001 vs. CT, *P < 0.001 vs. IR.
Table 2 Catalase activity (k/mg protein) in the four groups of rats studied.
CT HTX IR HTX+IR
Renal cortex 0.22 ± 0.04 (16) 0.22 ± 0.02 (16) 0.17 ± 0.05a (5) 0.14 ± 0.04b (5)
Outer medulla 0.07 ± 0.02 (16) 0.08 ± 0.02 (16) 0.08 ± 0.017 (7) 0.04 ± 0.004b,* (7)
Inner medulla 0.013 ± 0.002 (16) 0.014 ± 0.003 (16) 0.012 ± 0.003 (7) 0.012 ± 0.002 (7)
Data are mean ± SD. Number of determinations is in parentheses. aP < 0.05 vs. CT, bP < 0.001 vs. HTX, *P < 0.01 vs IR.
Table 3 Glutathione peroxidase activity (U/mg protein) in three sections of kidney from the four groups of rats studied.
CT HTX IR HTX+IR
Renal cortex 0.10 ± 0.01 (16) 0.11 ± 0.01 (16) 0.07 ± 0.03a (7) 0.04 ± 0.01b,** (7)
Outer medulla 0.07 ± 0.01 (16) 0.06 ± 0.01* (16) 0.03 ± 0.003a (7) 0.05 ± 0.007*** (7)
Inner medulla 0.06 ± 0.02 (16) 0.05 ± 0.01 (16) 0.05 ± 0.02 (6) 0.03 ± 0.003 (6)
Data are mean ± SD. Number of determinations is in parentheses. aP < 0.001 vs. CT, bP < 0.001 vs. HTX, *P < 0.05, vs. CT, **P < 0.01, ***P < 0.001 vs. IR.
Table 4 Superoxide dismutase activity (U/mg protein) in the four groups of rats studied.
CT HTX IR HTX+IR
Renal cortex 13.3 ± 1.0 (16) 14.4 ± 1.0 (16) 15.1 ± 2.1 (7) 13.4 ± 3.2 (7)
Outer medulla 10.9 ± 1.3 (16) 10.4 ± 1.1 (16) 8.2 ± 1.2a (8) 5.6 ± 1.5b,* (8)
Data are mean ± SD. Number of determinations is in parentheses. aP < 0.001 vs. CT, bP < 0.001 vs. HTX, *P < 0.001 vs. IR.
Table 5 Effect of experimental hypothyroidism on oxidative stress markers.
Ref. Specie MHI Change in oxidative stress markers
9 Rat HTX ↔ MDA and ↑ GSH in renal cortex.
50 Rat PTU ↑ Brain total antioxidant status.
51 Rat PTU ↓ MDA and GSH levels in cerebral, hepatic and cardiac tissues.
52 Rat PTU ↓ Advanced glycation end-products and MDA-lysine in liver.
53 Rat HTX ↑ GSH and ↓ MDA levels in liver.
54 Rat PTU ↑ MDA in plasma, erythrocytes, and liver tissue.
↔ MDA in kidney.
↔ GSH levels of kidney and liver.
55 Rat MMI ↔ Brain TBARS.
56 Mouse PTU ↔ LPx and GSH, GSSG and GSSG/GSH ratio in skeletal muscle.
47 Rat PTU ↓ TBARS in extensor digitorum longus muscle.
↔ TBARS in heart, liver, and soleus muscle.
57 Rat PTU ↓ MDA and GSH in renal and testicular tissue.
Ref. = reference, MHI = Method of hypothyroidism induction, HTX = Thyroidectomized, PTU = 6-n-propyl-2-thiouracil, MMI = Methimazole, ↑ = increase, ↓ = decrease, ↔ = without change, LPx = lipid peroxidation, MDA = malondialdehyde, GSH = reduced glutathione, GSSG = oxidized glutathione, TBARS = Thiobarbituric acid reactive substances.
Table 6 Effect of experimental hypothyroidism on antioxidant enzymes activities.
Ref. Specie MHI Change in antioxidant enzymes
61 Rat MMI ↑ MnSOD and CAT and ↔ Cu, ZnSOD in brown adipose tissue.
62 Rat PTU In liver mitochondria:
↑ Total and Cu, ZnSOD, ↔ MnSOD, ↓ CAT.
↑ Total and Se-independent and Se-dependent GPx
63 Rat PTU ↓ SOD and ↔ CAT in heart.
64 Rat MMI ↑ GPx, ↔ CAT, and total SOD in heart.
65 Rat PTU ↓ SOD and CAT and ↑ GPx in testis.
66 Rat PTU In brain:
↑ Total and MnSOD, CAT, total and Se-dependent GPx.
↓ Se-independent GPx, and GR.
55 Rat MMI In cerebral cortex:
↑ Total SOD, Cu, ZnSOD, and GPx, and ↓ MnSOD.
↔ CAT.
67 Rat Na131I ↓ SOD, ↔ CAT, and ↑ GPx in kidney.
↓ Plasma GPx.
47 Rat PTU Cu, ZnSOD ↔ in extensor digitorum longus and soleus muscles, ↑ heart, and ↓ liver.
GPx ↔ in soleus muscle and liver ↑ extensor digitorum longus muscle and heart.
68 Rat MMI GPx ↑ in gastrocnemius muscle and heart, and ↔ in liver.
GR in ↔ heart, gastrocnemius muscle, and liver.
Ref. = reference, MHI = Method of hypothyroidism induction, PTU = 6-n-propyl-2-thiouracil, MMI = methimazole, ↑ = increase, ↓ = decrease, ↔ = without change, SOD = superoxide dismutase, MnSOD = manganese-dependent superoxide dismutase, Cu, ZnSOD = copper,zinc-dependent superoxide dismutase, Se = selenium, GPx = Glutathione peroxidase, GR = glutathione reductase, CAT = catalase.
==== Refs
Nath KA Norby SM Reactive oxygen species and acute renal failure Am J Med 2000 190 665 678 10.1016/S0002-9343(00)00612-4
Pedraza-Chaverrí J Tapia E Bobadilla N Ischemia-reperfusion induced acute renal failure in the rat is ameliorated by the spin-trapping agent α-phenyl-N-tert-butyl nitrone (PBN) Ren Fail 1992 14 467 471 1461997
Noiri E Nakao A Uchida K Tsukahara H Ohno M Fujita T Brodsky S Goligorsky MS Oxidative and nitrosative stress in acute renal ischemia Am J Physiol Renal Physiol 2001 281 F948 F957 11592952
Chatterjee PK Todorovic Z Sivarajah A Mota-Filipe H Brown PA Stewart KN Mazzon E Cuzzocrea S Thiemermann C Inhibitors of calpain activation (PD150606 and E-64) and renal ischemia-reperfusion injury Biochem Pharmacol 2005 69 1121 1131 15763548 10.1016/j.bcp.2005.01.003
Patel NS Chatterjee PK Di Paola R Mazzon E Britti D De Sarro A Cuzzocrea S Thiemermann C Endogenous interleukin-6 enhances the renal injury, dysfunction, and inflammation caused by ischemia/reperfusion J Pharmacol Exp Ther 2005 312 1170 1178 15572648 10.1124/jpet.104.078659
Chatterjee PK Patel NS Kvale EO Brown PA Stewart KN Mota-Filipe H Sharpe MA Di Paola R Cuzzocrea S Thiemermann C EUK-134 reduces renal dysfunction and injury caused by oxidative and nitrosative stress of the kidney Am J Nephrol 2004 24 165 177 14752229 10.1159/000076547
Seth P Kumari R Madhavan S Singh AK Mani H Banaudha KK Sharma SC Kulshreshtha DK Maheshwari RK Prevention of renal ischemia-reperfusion-induced injury in rats by picroliv Biochem Pharmacol 2000 59 1315 1322 10736432 10.1016/S0006-2952(00)00268-9
Adin CA Croker BP Agarwal A Protective effects of exogenous bilirubin on ischemia-reperfusion injury in the isolated, perfused rat kidney Am J Physiol Renal Physiol 2005 288 F778 F784 15561977 10.1152/ajprenal.00215.2004
Paller MS Hypothyroidism protects against free radical damage in ischemia acute renal failure Kidney Int 1986 29 1162 1166 3747332
Kawamura T Yoshioka T Bills T Fogo A Ichikawa I Glucocorticoid activates glomerular antioxidant enzymes and protects glomeruli from oxidant injuries Kidney Int 1991 40 291 301 1942778
Yoshioka T Bills T Moore-Jarrett T Greene HL Burr IM Ichikawa I Role of intrinsic antioxidant enzymes in renal oxidant injury Kidney Int 1990 38 282 288 2402119
Yoshioka T Kawamura T Meyrick BO Beckman JK Hoover RL Yoshida H Ichikawa I Induction of manganese superoxide dismutase by glucocorticoids in glomerular cells Kidney Int 1994 45 211 219 8127010
Davis CA Nick HS Agarwal A Manganese superoxide dismutase attenuates cisplatin-induced renal injury: importance of superoxide J Am Soc Nephrol 2001 12 2683 2690 11729237
Yin M Wheeler MD Connor HD Zhong Z Bunzendahl H Dikalova A Samulski RJ Schoonhoven R Mason RP Swenberg JA Thurman RG Cu/Zn-superoxide dismutase gene attenuates ischemia-reperfusion injury in the rat kidney J Am Soc Nephrol 2001 12 2691 2700 11729238
Zhong Z Connor HD Yin M Wheeler MD Mason RP Thurman RG Viral delivery of superoxide dismutase gene reduces cyclosporine A-induced nephrotoxicity Kidney Int 2001 59 1397 1404 11260401 10.1046/j.1523-1755.2001.0590041397.x
Shimizu H Takahashi T Suzuki T Yamasaki A Fujiwara T Odaka Y Hirakawa M Fujita H Akagi R Protective effect of heme oxygenase induction in ischemic acute renal failure Crit Care Med 2000 28 809 817 10752834 10.1097/00003246-200003000-00033
Deshane J Wright M Agarwal A Heme oxygenase-1 expression in disease states Acta Biochim Pol 2005 52 273 284 15933765
Pedraza-Chaverrí J Granados-Silvestre MA Medina-Campos ON Hernández-Pando R Effect of the in vivo catalase inhibition on aminonucleoside nephrosis Free Radic Biol Med 1999 27 245 253 10468195 10.1016/S0891-5849(99)00023-4
Singh D Chander V Chopra K Protective effect of catechin on ischemia-reperfusion-induced renal injury in rats Pharmacol Rep 2005 57 70 76 15849379
Singh D Chander V Chopra K Carvedilol attenuates ischemia-reperfusion-induced oxidative renal injury in rats Fundam Clin Pharmacol 2004 18 627 634 15548233 10.1111/j.1472-8206.2004.00279.x
Aksoy Y Yapanoglu T Aksou H Yildirim AK The effect of dehydroepiandrosterone on renal ischemia-reperfusion-induced oxidative stress in rabbits Urol Res 2004 32 93 96 15250101 10.1007/s00240-003-0382-6
Yokozawa T Rhyu DY Cho EJ Aoyagi K Protective activity of (-)-epicatechin 3-O-gallate against peroxynitrite-mediated renal damage Free Radic Res 2003 37 561 571 12797478 10.1080/1071576031000083134
Rhyu DY Yokozawa T Choa EJ Park JC Prevention of peroxynitrite-induced renal injury through modulation of peroxynitrite production by the Chinese prescription Wen-Pi-Tang Free Radic Res 2002 36 1261 1269 12607816 10.1080/1071576021000016490
Tarpey MM Wink DA Grisham MB Methods for detection of reactive metabolites of oxygen and nitrogen: in vitro and in vivo considerations Am J Physiol Regul Integr Comp Physiol 2004 286 R431 R444 14761864
Szabo C Multiple pathways of peroxynitrite cytotoxicity Toxicol Lett 2003 140–141 105 112 10.1016/S0378-4274(02)00507-6
Radi R Nitric oxide, oxidants, and protein tyrosine nitration Proc Natl Acad Sci USA 2004 101 4003 4008 15020765 10.1073/pnas.0307446101
Barrera D Maldonado PD Medina-Campos ON Hernández-Pando R Ibarra-Rubio ME Pedraza-Chaverrí J HO-1 induction attenuates renal damage and oxidative stress induced by K2 Cr2O7 Free Radic Biol Med 2003 34 1390 1398 12757849 10.1016/S0891-5849(03)00068-6
Maldonado PD Barrera D Medina-Campos ON Hernández-Pando R Ibarra-Rubio ME Pedraza-Chaverrí J Aged garlic extract attenuates gentamicin induced renal damage and oxidative stress in rats Life Sci 2003 73 2543 2556 12967679 10.1016/S0024-3205(03)00609-X
Aksenov MY Aksenova MV Butterfield DA Geddes JW Markesbery WR Protein oxidation in the brain in Alzheimer's disease Neuroscience 2001 103 373 383 11246152 10.1016/S0306-4522(00)00580-7
Toyokuni S Uchida K Okamoto K Hattori-Nakakuki Y Hiai H Stadtman ER Formation of 4-hydroxy-2-nonenal-modified proteins in the renal proximal tubules of rats treated with a renal carcinogen, ferric nitrilotriacetate Proc Natl Acad Sci USA 1994 91 2616 2620 8146163
Zainal TA Weindruch R Szweda LI Oberley TD Localization of 4-hydroxy-2-nonenal-modified proteins in kidney following iron overload Free Radic Biol Med 1999 26 1181 1193 10381189 10.1016/S0891-5849(98)00312-8
Franco M Galicia O Quintana A Martinez F Experimental hypothyroidism modifies specific binding of A1 and A2A analogues to adenosine receptors in the rat kidney Br J Pharmacol 2004 142 461 468 15148254 10.1038/sj.bjp.0705799
Franco M Bobadilla NA Suárez J Tapia E Sánchez L Herrera-Acosta J Participation of adenosine in the renal hemodynamic abnormalities of hypothyroidism Am J Physiol 1996 270 F254 F262 8779885
Franco M Tapia E Martinez F Davila Ma E Grimaldo JI Medina K Herrera-Acosta J Adenosine regulates renal nitric oxide production in hypothyroid rats J Am Soc Nephrol 1999 10 1681 1688 10446935
Martinez F Franco M Quintana A Herrera-Acosta J Sodium-dependent adenosine transport is diminished in brush border membrane vesicles from hypothyroid rat kidney Pflugers Arch 1997 433 269 275 9064642
Pedraza-Chaverri J Barrera D Maldonado PD Chirino YI Macias-Ruvalcaba NA Medina-Campos ON Castro L Salcedo MI Hernandez-Pando R S-allylmercaptocysteine scavenges hydroxyl radical and singlet oxygen in vitro and attenuates gentamicin-induced oxidative and nitrosative stress and renal damage in vivo BMC Clin Pharmacol 2004 4 5 15119956 10.1186/1472-6904-4-5
Pedraza-Chaverri J Barrera D Medina-Campos ON Carvajal RC Hernandez-Pando R Macias-Ruvalcaba NA Maldonado PD Salcedo MI Tapia E Saldivar L Castilla ME Ibarra-Rubio ME Time course study of oxidative and nitrosative stress and antioxidant enzymes in K2Cr2O7-induced nephrotoxicity BMC Nephrol 2005 6 4 15854231 10.1186/1471-2369-6-4
Chatterjee PK Cuzzocrea S Brown PA Zacharowski K Stewart KN Mota-Filipe H Thiemermann C Tempol, a membrane-permeable radical scavenger, reduces oxidant stress-mediated renal dysfunction and injury in the rat Kidney Int 2000 58 658 673 10916089 10.1046/j.1523-1755.2000.00212.x
Lowry OH Rosenbrough NJ Farr AL Randall RJ Protein measurement with the Folin phenol reagent J Biol Chem 1951 193 265 275 14907713
Pedraza-Chaverrí J Granados-Silvestre MA Medina-Campos ON Maldonado PD Olivares-Corichi IM Ibarra-Rubio ME Post-transcriptional control of catalase expression in garlic treated rats Mol Cell Biochem 2001 216 9 19 11216869 10.1023/A:1011050619406
Pedraza-Chaverrí J Maldonado PD Medina-Campos ON Olivares-Corichi IM Granados-Silvestre MA Hernández-Pando R Ibarra-Rubio ME Garlic ameliorates gentamicin nephrotoxicity: relation to antioxidant enzymes Free Radic Biol Med 2000 29 602 611 11033412 10.1016/S0891-5849(00)00354-3
Isman CA Yegen BC Alican I Methimazole-induced hypothyroidism in rats ameliorates oxidative injury in experimental colitis J Endocrinol 2003 177 471 476 12773128 10.1677/joe.0.1770471
Rodriguez-Pierce M Sosenko IR Whitney P Frank L Propylthiouracil treatment decreases the susceptibility to oxygen radical-induced lung damage in newborn rats exposed to prolonged hyperoxia Pediatr Res 1994 35 530 535 8065833
Allen T Rana SV Oxidative stress by inorganic arsenic: modulation by thyroid hormones in rat Comp Biochem Physiol C Toxicol Pharmacol 2003 135 157 162 12860054 10.1016/S1532-0456(03)00086-3
Nishiki K Ericinska M Wilson DF Cooper S Evaluation of oxidative phosphorylation in hearts from euthyroid, hypothyroid and hyperthyroid rats Am J Physiol 1978 235 C212 C219 215035
Swaroo A Ramasarma T Heat exposure and hypothyroid conditions decrease hydrogen peroxide generation in liver mitochondria Biochem J 1985 226 403 408 3994666
Asayama K Dobashi K Hayashibe H Megata Y Kato K Lipid peroxidation and free radical scavengers in thyroid dysfunction in the rat: a possible mechanism of injury to heart and skeletal muscle in hyperthyroidism Endocrinology 1987 121 2112 2118 2824181
Bruck R Frenkel D Shirin H Aeed H Matas Z Papa M Zaidel L Avni Y Oren R Halpern Z Hypothyroidism protects rat liver from acetaminophen hepatotoxicity Dig Dis Sci 1999 44 1228 1235 10389702 10.1023/A:1026652913347
James LP Mayeux PR Hinson JA Acetaminophen-induced hepatotoxicity Drug Metab Dispos 2003 31 1499 1506 14625346 10.1124/dmd.31.12.1499
Carageorgiou H Pantos C Zarros A Mourouzis I Varonos D Cokkinos D Tsakiris S Changes in antioxidant status, protein concentration, acetylcholinesterase, (Na+, K+)-, and Mg2+ -ATPase activities in the brain of hyper- and hypothyroid adult rats Metab Brain Dis 2005 20 129 139 15938131 10.1007/s11011-005-4150-x
Mogulkoc R Baltaci AK Aydin L Oztekin E Sivrikaya A The effect of thyroxine administration on lipid peroxidation in different tissues of rats with hypothyroidism Acta Physiol Hung 2005 92 39 46 16003944 10.1556/APhysiol.92.2005.1.6
Pamplona R Portero-Otin M Ruiz C Bellmunt MJ Requena JR Thorpe SR Baynes JW Romero M Lopez-Torres M Barja G Thyroid status modulates glycoxidative and lipoxidative modification of tissue proteins Free Radic Biol Med 1999 27 901 910 10515595 10.1016/S0891-5849(99)00135-5
Teare JP Greenfield SM Marway JS Preedy VR Punchard NA Peters TJ Thompson RP Effect of thyroidectomy and adrenalectomy on changes in liver glutathione and malonaldehyde levels after acute ethanol injection Free Radic Biol Med 1993 14 655 660 8325537 10.1016/0891-5849(93)90147-M
Tas S Dirican M Sarandol E Serdar Z The effect of taurine supplementation on oxidative stress in experimental hypothyroidism Cell Biochem Funct 2004 15617030
Mano T Sinohara R Sawai Y Oda N Nishida Y Mokumo T Asano K Ito Y Kotake M Hamada M Nakai A Nagasaka A Changes in lipid peroxidation and free radical scavengers in the brain of hyper- and hypothyroid aged rats J Endocrinol 1995 147 361 365 7490566
Gredilla R Lopez Torres M Portero-Otin M Pamplona R Barja G Influence of hyper- and hypothyroidism on lipid peroxidation, unsaturation of phospholipids, glutathione system and oxidative damage to nuclear and mitochondrial DNA in mice skeletal muscle Mol Cell Biochem 2001 221 41 48 11506185 10.1023/A:1010930110382
Mogulkoc R Baltaci AK Oztekin E Ozturk A Sivrikaya A Short therm thyroxine administration leas to lipid peroxidation in renal and testicular tissues of rats with hypothyroidism Acta Biol Hung 2005 56 225 232 16196198 10.1556/ABiol.56.2005.3-4.6
Pedraza-Chaverri J Barrera D Hernandez-Pando R Medina-Campos ON Cruz C Murguia F Juarez-Nicolas C Correa-Rotter R Torres N Tovar AR Soy protein diet ameliorates renal nitrotyrosine formation and chronic nephropathy induced by puromycin aminonucleoside Life Sci 2004 74 987 999 14672755 10.1016/j.lfs.2003.07.045
Ueda N Mayeoux PR Baliga R Shah SV Molitoris BA, Finn WF Oxidant mechanisms in acute renal failure Acute renal failure A companion to Brenner & Rector's The Kidney 2001 Philadelphia: WB Saunders 60 77
Dobashi K Ghosh B Orak JK Singh I Singh AK Kidney ischemia-reperfusion: modulation of antioxidant defenses Mol Cell Biochem 2000 205 1 11 10821417 10.1023/A:1007047505107
Petrovic N Cvijic G Davidovic V The activity of antioxidant enzymes and the content of uncoupling protein-1 in the brown adipose tissue of hypothyroid rats: comparison with effects of iopanoic acid Physiol Res 2001 50 289 297 11521740
Das K Chainy GB Modulation of rat liver mitochondrial antioxidant defence system by thyroid hormone Biochim Biophys Acta 2001 1537 1 13 11476958
Chattopadhyay S Zaidi G Das K Chainy GB Effects of hypothyroidism induced by 6-n-propylthiouracil and its reversal by T3 on rat heart superoxide dismutase, catalase and lipid peroxidation Indian J Exp Biol 2003 41 846 849 15248482
Mano T Sinohara R Sawai Y Oda N Nishida Y Mokuno T Kotake M Hamada M Masunaga R Nakai A Nagasaka A Effects of thyroid hormone on coenzyme Q and other free radical scavengers in rat heart muscle J Endocrinol 1995 145 131 136 7798018
Choudhury S Chainy GB Mishro MM Experimentally induced hypo- and hyper-thyroidism influence on the antioxidant defence system in adult rat testis Andrologia 2003 35 131 140 12780529 10.1046/j.1439-0272.2003.00548.x
Das K Chainy GB Thyroid hormone influences antioxidant defense system in adult rat brain Neurochem Res 2004 29 1755 1766 15453272 10.1023/B:NERE.0000035812.58200.a9
Sawant BU Nadkarni GD Thakare UR Joseph LJ Rajan MG Changes in lipid peroxidation and free radical scavengers in kidney of hypothyroid and hyperthyroid rats Indian J Exp Biol 2003 41 1334 1337 15332508
Venditti P Balestrieri M Di Meo S De Leo T Effect of thyroid state on lipid peroxidation, antioxidant defences, and susceptibility to oxidative stress in rat tissues J Endocrinol 1997 155 151 157 9390017 10.1677/joe.0.1550151
Bobadilla I Franco M Cruz D Zamora J Robles SG Chávez E Hypothyroidism provides resistance to reperfusion injury following myocardium ischemia Int J Biochem Cell Biol 2001 33 499 506 11331205 10.1016/S1357-2725(01)00016-4
Chávez E Franco M Reyes-Vias H Zazueta C Ramírez J Carrillo R Hypothyroidism renders liver mitochondria resistant to the opening of membrane permeability transition pore Biochim Biophys Acta 1998 1407 243 248 9748606
Halestrap AP Clarke SJ Javadov SA Mitochondrial permeability transition pore opening during myocardial reperfusion – a target for cardioprotection Cardiovasc Res 2004 61 372 385 14962470 10.1016/S0008-6363(03)00533-9
|
16274486
|
PMC1291371
|
CC BY
|
2021-01-04 16:03:43
|
no
|
BMC Nephrol. 2005 Nov 7; 6:12
|
utf-8
|
BMC Nephrol
| 2,005 |
10.1186/1471-2369-6-12
|
oa_comm
|
==== Front
BMC NeurolBMC Neurology1471-2377BioMed Central London 1471-2377-5-191625577110.1186/1471-2377-5-19Research ArticleEarly B-cell Factor gene association with multiple sclerosis in the Spanish population Martínez Alfonso [email protected] Ana [email protected] las Heras Virginia [email protected] Rafael [email protected]ández-Arquero Miguel [email protected] la Concha Emilio G [email protected] Elena [email protected] Department of Clinical Immunology, Hospital Clinico San Carlos, Madrid, Spain2 Department of Neurology, Hospital Clinico San Carlos, Madrid, Spain2005 28 10 2005 5 19 19 9 8 2005 28 10 2005 Copyright © 2005 Martínez et al; licensee BioMed Central Ltd.2005Martínez et al; licensee BioMed Central Ltd.This is an Open Access article distributed under the terms of the Creative Commons Attribution License (), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
Background
The etiology of multiple sclerosis (MS) is at present not fully elucidated, although it is considered to result from the interaction of environmental and genetic susceptibility factors. In this work we aimed at testing the Early B-cell Factor (EBF1) gene as a functional and positional candidate risk factor for this neurological disease. Axonal damage is a hallmark for multiple sclerosis clinical disability and EBF plays an evolutionarily conserved role in the expression of proteins essential for axonal pathfinding. Failure of B-cell differentiation was found in EBF-deficient mice and involvement of B-lymphocytes in MS has been suggested from their presence in cerebrospinal fluid and lesions of patients.
Methods
The role of the EBF1 gene in multiple sclerosis susceptibility was analyzed by performing a case-control study with 356 multiple sclerosis patients and 540 ethnically matched controls comparing the EBF1 polymorphism rs1368297 and the microsatellite D5S2038.
Results
Significant association of an EBF1-intronic polymorphism (rs1368297, A vs. T: p = 0.02; OR = 1.26 and AA vs. [TA+TT]: p = 0.02; OR = 1.39) was discovered. This association was even stronger after stratification for the well-established risk factor of multiple sclerosis in the Major Histocompatibility Complex, DRB1*1501 (AA vs. [TA+TT]: p = 0.005; OR = 1.78). A trend for association in the case-control study of another EBF1 marker, the allele 5 of the very informative microsatellite D5S2038, was corroborated by Transmission Disequilibrium Test of 53 trios (p = 0.03).
Conclusion
Our data support EBF1 gene association with MS pathogenesis in the Spanish white population. Two genetic markers within the EBF1 gene have been found associated with this neurological disease, indicative either of their causative role or that of some other polymorphism in linkage disequilibrium with them.
==== Body
Background
Multiple sclerosis (MS) is one of the most common neurological diseases of young adults in Europe and North America [1]. Similarly to other common complex diseases, the interplay of genome and environment as MS susceptibility factors seems to determine the final outcome. Its precise etiology is at present unknown, even though the first genetic association with the MHC was published more than thirty years ago [2]. Genomic screens support the hypothesis that susceptibility to develop MS is determined by multiple genes with small individual contributions. To decipher those combinations of genes resulting in MS is a major goal of research. Association of MS with the HLA-DRB1*1501-DQB1*0602 haplotype has been unambiguously demonstrated [3]. The diversity of the predisposition genes is evident if we consider that the major risk allele HLA-DRB1*1501 is present only in 33% of our Spanish MS patients. New susceptibility genes are therefore actively sought worldwide. In the past, candidate gene approaches have successfully revealed associations with disease susceptibility, severity or disease course.
MS has been traditionally considered an autoimmune demyelinating disorder of the central nervous system (CNS) due to autoreactive T cells on myelin proteins. However, other cells and processes have also been involved in the MS immune attack. A role for B cells in MS pathogenesis has been suggested from their presence in the cerebrospinal fluid and lesions of MS patients [4,5]. Axonal degeneration [6] has been found in early stages of the disease [7]. Axonal loss is a reliable marker of MS clinical disability [8], although the mechanism underlying axonal damage in MS remains elusive [9,10].
B cells derive from a common lymphoid progenitor, itself derived from a multipotent bone marrow progenitor. The development of a B lymphocyte comprises multiple stages with sequential expression of genes participating in immunoglobulin gene rearrangements and signaling. B cell development depends on a number of transcription factors [11] including early B cell factor (EBF), as shown for the dramatic phenotype of EBF-deficient mice [12]. B cell differentiation to plasma cell in secondary lymphoid organs is an exquisitely regulated process requiring EBF inhibition [13] and a full recapitulation of this B cell final differentiation has been recently described in the CSF [14].
EBF [15] belongs to a family of proteins present in the animal kingdom, the Collier/Olf1/EBF proteins, and it is also expressed in neural cells of different origins. EBF plays an evolutionarily conserved role in the expression of proteins essential for axonal pathfinding and neuronal differentiation in both sensory and motor neurons [16]. In addition to the action of the EBF protein in embryonic neural development, it is expressed in the adult nervous system too [17]. Furthermore, EBF binding activates the Herpes Simplex Virus Type I ICP0 (Infected Cell Protein 0) gene promoter, important for productive infection and reactivation from latency [18]. This virus has been related to MS [19,20].
All this evidence prompted us to determine whether the EBF1 gene (coding for the first member of this family cloned in humans), located at chromosome 5q has any role in MS pathogenesis. Unfortunately, no description of functional EBF1 polymorphisms exists in the literature, and therefore two markers were selected based on strictly genetic parameters. The first is the highly polymorphic D5S2038 microsatellite mapping to the EBF1 gene and the second an intronic single nucleotide polymorphism (rs1368297). The present work shows association of these markers with MS in the Spanish cohorts tested. Most probably MS is consequence of the interaction of a limited number of genetic and environmental risk factors in a patient, which may vary from those present in other patients.
Methods
Patients and controls
Three hundred and fifty six consecutively recruited MS patients from a single center and 540 healthy controls, mainly blood donors and staff, were included in a case-control study approved by the Hospital Ethical committee. The MS diagnosis was established based on the Poser criteria [21] and most of these patients have been described in previous studies from our group [22].
Genotyping
D5S2038 microsatellite was amplified with annealing temperature of 56°C using the following set of primers:
Forward: 5' FAM-GTT CAA ATC TTG CCT TTG CC-3'
Reverse: 5'-GCC ATT GCT TTG TTT ATG CA-3'
Samples were subsequently denatured and run on an ABI Prism 3100 automatic sequencer (Applied Biosystems, Foster City, CA, USA). Each sample included an internal size standard in order to achieve a highly consistent measure and the results were analyzed using the GeneScan software (Applied Biosystems) and the Local Southern method.
The EBF1 polymorphism (rs1368297) was analyzed by TaqMan Assays-on-Demand (C___2085085_10) from Applied Biosystems, following manufacturer suggestions.
Both genetic markers conformed to Hardy-Weinberg equilibrium in the control population.
Statistical analysis
Allele and genotype frequencies in patients and controls were compared by the χ2 test; p values were considered significant at a level of < 0.05. Odds ratio (OR) and p values were calculated using a standard computer package (Epi Info v. 6.02, CDC, Atlanta, USA). The power of the study for the SNP analyzed is above 80% considering a relative risk of 1.4 and the observed allelic frequency of 0.5 at the standard significance level of 0.05.
Results
As microsatellites are very polymorphic and informative genetic markers, we decided to select one within the EBF1 gene, D5S2038, and to study its allele distribution in our cohorts. The case-control analysis of this microsatellite in 356 MS patients and 540 healthy controls yielded the results summarized in Table 1. As shown, the only allele displaying a trend for association was D5S2038*5, while the frequency of the other ten alleles did not change significantly when patients and controls were compared. No differences were observed when different clinical forms were compared (primary progressive vs relapsing-remitting/secondary progressive). Considering these case-control results, we performed a Transmission Disequilibrium Test (TDT) with 53 trios formed by the patients and their progenitors. Distorted transmission of the D5S2038*5 was observed too (Table 2).
Table 1 Allele frequencies of EBF1 microsatellite D5S2038 in MS patients and healthy controls.
EBF1 D5S2038 Controls (n = 540) MS patients (n = 356) p
1 2 2 0.99
2 27 15 0.58
3 55 36 0.97
4 272 176 0.78
5* 246 183 0.08
6 119 81 0.80
7 79 47 0.54
8 146 94 0.83
9 8 7 0.57
10 7 4 1.00
11 3 0 0.28
* p = 0.08; OR (95%CI) = 1.26 (0.96–1.67)
Table 2 Transmission Disequilibrium Test (TDT) of EBF1 microsatellite D5S2038 in trios (MS patient and progenitors).
EBF1 D5S2038 Transmitted Non transmitted p
3 5 9 0.91
4 23 11 0.44
5 28 15 0.03
6 3 8 0.96
7 3 7 0.94
8 11 10 0.50
9 0 2 1.00
10 1 1 0.75
We pursued to check another polymorphism within the EBF1 gene in order to provide more evidence in support of the association of the gene with MS. Data corresponding to the allele and genotype frequencies of the intronic SNP rs1368297 are found in Table 3. Allele A was significantly increased in the diseased population and, under a recessive inheritance model, the AA homozygous genotype also conferred predisposition to MS. Moreover, the ratio of risk: non-risk homozygous genotypes was significantly higher among multiple sclerosis patients (AA: TT: p = 0.03; OR [95%CI] = 1.56 [1.03–2.37]).
Table 3 Allele and genotype frequencies of the EBF1 polymorphism in multiple sclerosis patients and controls.
EBF rs1368297 AA AT TT A T
MS patients 125 168 58 418 284
(n = 351) 35.6% 47.8% 16.5% 59.5% 40.5%
Controls 149 267 108 565 483
(n = 524) 28.4% 50.9% 20.6% 53.9% 46.1%
AA vs. (AT+TT): p = 0.02; OR (95% CI) = 1.39 (1.03–1.88).
A vs. T: p = 0.02; OR (95% CI) = 1.26 (1.03–1.53).
The MS cohort was stratified for the well-known MHC susceptibility factor DRB1*1501 and the difference between DRB1*1501+ patients and healthy controls was even stronger than in the unconditioned analysis (Table 4).
Table 4 Genotype frequencies of the EBF1 polymorphism in HLA-DRB1*1501 positive and negative multiple sclerosis patients.
EBF rs1368297 AA AT TT A T
DRB1*1501+ MS patients* 51 53 19 155 91
(n = 123) 41.5% 43.1% 15.4% 63% 37%
DRB1*1501- MS patients 74 115 39 263 193
(n = 228) 32.5% 50.4% 17.1% 57.7% 42.3%
Controls 149 267 108 565 483
(n = 524) 28.4% 50.9% 20.6% 53.9% 46.1%
* AA vs. (AT+TT): p = 0.005; OR (95% CI) = 1.78 (1.16–2.73)
A vs. T: p = 0.01; OR (95% CI) = 1.46 (1.08–1.96)
Finally, when simultaneous carriage of both susceptibility alleles, D5S2038*5 and EBF SNP*A, was compared between MS patients and controls an increment was observed within the diseased cohort (p = 0.03; OR [95%CI] = 1.38 [1.03–1.84]).
Discussion
Information from genomic screens proposed the 5q chromosomal region as linked to MS [23,24]. Additionally, a recent report compared chromosomal regions, quantitative trait loci (QTLs), of MS patients and of EAE animal models and, by analysis of sequence similarities, defined consensus genes potentially conferring susceptibility to MS [25]. Among them, the EBF1 gene in chromosome 5q34 was cited, providing positional evidence of the role of this gene in MS predisposition.
Moreover, the simultaneous measurement of thousands of genes upregulated by EBF through microarray technology allowed the detection of 3.5-fold increase in the expression of interleukin 6 (IL-6) and of the microtubule associated protein tau listed among the top twelve most abundant transcripts [26]. IL-6 has been detected in MS brain and its expression elevated in cerebrospinal fluid of patients [27,28]. IL-6 knockout mice showed resistance to induced EAE, too [29]. The physiological function of tau is to bind to and stabilize microtubules [30] and it is involved in regulation of axonal transport [31]. Tau protein concentration has been found repeatedly increased in cerebrospinal fluid of MS patients [32]. Also morphological examination demonstrated accumulation of amorphous deposits of abnormally phosphorylated tau in the cell body and axons of neurons within demyelinating plaques in EAE [33]. Axonopathy has been involved recently in early stages of the pathogenesis of another neurological disease, Alzheimer's disease [34]. Aberrant accumulation of proteins may be crucial to the impairment of axonal transport.
Our results evidence association of the EBF1 gene with MS. There are no functional studies of these gene polymorphisms, although it was cloned more than a decade ago. However, several reports showed transcriptional regulatory elements located in intronic regions of different genes [35-38]. In fact, the susceptibility allele of this EBF-intronic polymorphism allows the putative binding of an AP-1 transcription factor and this binding site is disrupted in the presence of the T allele (as predicted by TFSEARCH ver 1.3). Functional studies of EBF1 will aid in clarifying the role of this gene in MS pathogenesis. Nonetheless, the polymorphisms studied in this work act as genetic markers, which could potentially be the etiologic variants or be in linkage disequilibrium with them.
The EBF prototypical regulatory activity in B lymphocyte differentiation alone justifies the functional involvement of the EBF1 gene in an autoimmune disease as MS. Increasing evidence supports the role in MS disease course of IgM antibodies [39] produced by CD5+ B-lymphocytes, that are elevated in CSF of patients with aggressive forms of MS [40]. These natural IgM antibodies recognize myelin antigens and are strong complement activators [41]. Both, antibodies and complement, have been shown to contribute to MS disability through demyelination and axonal damage [42]. IgG antibodies with hypermutated V regions have been also described [43]. Moreover, EBF1 is a potent modulator of adipogenesis [44] and the IgM bands in cerebrospinal fluid of MS patients were directed against myelin lipids [40].
Conclusion
Our data suggest that the EBF1 gene involved in B-cell development, adipogenesis and axonal damage play a causative role in MS. Many mechanistic ties between axonal damage, tau pathology, intrathecal B1 subpopulation responsible for IgM secretion, conventional B cells, and the EBF1 gene role in MS susceptibility could be thought up. Confirmation in an independent cohort would substantiate our hypothesis about the implications of this gene in MS. Further understanding of the MS pathogenesis will help in the selection of therapeutic targets and characterization of the specific susceptibility genetic pattern in an individual will aid in a better diagnosis and ultimately in achievement of a personalized therapy.
List of abbreviations used
Early B-cell Factor gene (EBF1).
Multiple sclerosis (MS).
Major Histocompatibility Complex (MHC).
Human leukocyte antigen (HLA).
Central nervous system (CNS).
Infected Cell Protein 0 (ICP0).
Transmission Disequilibrium Test (TDT).
Single nucleotide polymorphism (SNP).
Quantitative trait loci (QTLs).
Experimental autoimmune encephalitis (EAE).
Odds ratio (OR).
Competing interests
The author(s) declare that they have no competing interests.
Authors' contributions
AMartínez carried out the genotyping of some samples and participated in the statistical analysis and writing of the manuscript.
AMas carried out the genotyping of most of the patients and a great part of the controls and participated in the statistical analysis.
VdlH made the diagnosis, participated in the recollection of samples and collaborated in the statistical analysis.
RA made the diagnosis, participated in the recollection of samples and collaborated in the statistical analysis.
MFA participated in the coordination of the study and helped to collect the DNA samples and to interpret the data.
EgdlC participated in the design and coordination of the study and critically revised the article.
EU conceived of the study and participated in the statistical analysis and drafted the major part of the manuscript.
All authors have read and approved the final manuscript.
Pre-publication history
The pre-publication history for this paper can be accessed here:
Acknowledgements
The authors thank Carmen Martínez for her skilful technical assistance. Elena Urcelay is recipient of a Ramón y Cajal contract of the Spanish Government. Alfonso Martínez is recipient of a research contract of the Spanish Health Ministry (CP04/00175). Ana Mas is a fellow of the Alfonso Martín Escudero Fundation. The Spanish FIS 04/0991 and the Rodriguez-Pascual Fundation supported this work.
==== Refs
Noseworthy JH Lucchinetti C Rodriguez M Weinshenker BG Multiple sclerosis N Engl J Med 2000 343 938 952 11006371 10.1056/NEJM200009283431307
Jersild C Svejgaard A Fog T HL-A antigens and multiple sclerosis Lancet 1972 1 1240 1241 4113225 10.1016/S0140-6736(72)90962-2
Olerup O Hillert J HLA class II-associated genetic susceptibility in multiple sclerosis: a critical evaluation Tissue Antigens 1991 38 1 15 1926129
Owens GP Kraus H Burgoon MP Smith-Jensen T Devlin ME Gilden DH Restricted use of VH4 germline segments in an acute multiple sclerosis brain Ann Neurol 1998 43 236 243 9485065 10.1002/ana.410430214
Qin Y Duquette P Zhang Y Talbot P Poole R Antel J Clonal expansion and somatic hypermutation of V(H) genes of B cells from cerebrospinal fluid in multiple sclerosis J Clin Invest 1998 102 1045 1050 9727074
Trapp BD Peterson J Ransohoff RM Rudick R Mork S Bo L Axonal transection in the lesions of multiple sclerosis N Engl J Med 1998 338 278 285 9445407 10.1056/NEJM199801293380502
De Stefano N Narayanan S Francis GS Arnaoutelis R Tartaglia MC Antel JP Matthews PM Arnold DL Evidence of axonal damage in the early stages of multiple sclerosis and its relevance to disability Arch Neurol 2001 58 65 70 11176938 10.1001/archneur.58.1.65
De Stefano N Matthews PM Fu L Narayanan S Stanley J Francis GS Antel JP Arnold DL Axonal damage correlates with disability in patients with relapsing-remitting multiple sclerosis. Results of a longitudinal magnetic resonance spectroscopy study Brain 1998 121 1469 1477 9712009 10.1093/brain/121.8.1469
Pitt D Werner P Raine CS Glutamate excitotoxicity in a model of multiple sclerosis Nat Med 2000 6 67 70 10613826 10.1038/71555
Lappe-Siefke C Goebbels S Gravel M Nicksch E Lee J Braun PE Griffiths IR Nave KA Disruption of Cnp1 uncouples oligodendroglial functions in axonal support and myelination Nat Genet 2003 33 366 374 12590258 10.1038/ng1095
Busslinger M Transcriptional control of early B cell development Annu Rev Immunol 2004 22 55 79 15032574 10.1146/annurev.immunol.22.012703.104807
Lin H Grosschedl R Failure of B-cell differentiation in mice lacking the transcription factor EBF Nature 1995 376 263 267 7542362 10.1038/376263a0
Lin KI Tunyaplin C Calame K Transcriptional regulatory cascades controlling plasma cell differentiation Immunol Rev 2003 194 19 28 12846804 10.1034/j.1600-065X.2003.00040.x
Corcione A Casazza S Ferretti E Giunti D Zappia E Pistorio A Gambini C Mancardi GL Uccelli A Pistoia V Recapitulation of B cell differentiation in the central nervous system of patients with multiple sclerosis Proc Natl Acad Sci U S A 2004 101 11064 11069 15263096 10.1073/pnas.0402455101
Wang MM Reed RR Molecular cloning of the olfactory neuronal transcription factor Olf-1 by genetic selection in yeast Nature 1993 364 121 126 8321284 10.1038/364121a0
Prasad BC Ye B Zackhary R Schrader K Seydoux G Reed RR unc-3, a gene required for axonal guidance in Caenorhabditis elegans, encodes a member of the O/E family of transcription factors Development 1998 125 1561 1568 9502737
Kudrycki KE Buiakova O Tarozzo G Grillo M Walters E Margolis FL Effects of mutation of the Olf-1 motif on transgene expression in olfactory receptor neurons J Neurosci Res 1998 52 159 172 9579406 10.1002/(SICI)1097-4547(19980415)52:2<159::AID-JNR4>3.0.CO;2-9
Devireddy LR Jones CJ Olf-1, a neuron-specific transcription factor, can activate the herpes simplex virus type 1-infected cell protein 0 promoter J Biol Chem 2000 275 77 81 10617588 10.1074/jbc.275.1.77
Baig S Olsson O Olsson T Love A Jeansson S Link H Cells producing antibody to measles and herpes simplex virus in cerebrospinal fluid and blood of patients with multiple sclerosis and controls Clin Exp Immunol 1989 78 390 395 2558823
Ferrante P Mancuso R Pagani E Guerini FR Calvo MG Saresella M Speciale L Caputo D Molecular evidences for a role of HSV-1 in multiple sclerosis clinical acute attack J Neurovirol 2000 6 S109 114 10871797
Poser CM Paty DW Scheinberg L McDonald WI Davis FA Ebers GC Johnson KP Sibley WA Silberberg DH Tourtellotte WW New diagnostic criteria for multiple sclerosis: guidelines for research protocols Ann Neurol 1983 13 227 231 6847134 10.1002/ana.410130302
Martinez Doncel A Rubio A Arroyo R de las Heras V Martin C Fernandez-Arquero M de la Concha EG Interleukin-10 polymorphisms in Spanish multiple sclerosis patients J Neuroimmunol 2002 131 168 172 12458048 10.1016/S0165-5728(02)00248-5
Sawcer S Ban M Maranian M Yeo TW Compston A Kirby A Daly MJ De Jager PL Walsh E Lander ES A high-density screen for linkage in multiple sclerosis Am J Hum Genet 2005 77 454 467 16080120 10.1086/444547
Kenealy SJ Babron MC Bradford Y Schnetz-Boutaud N Haines JL Rimmler JB Schmidt S Pericak-Vance MA Barcellos LF Lincoln RR A second-generation genomic screen for multiple sclerosis Am J Hum Genet 2004 75 1070 1078 15494893 10.1086/426459
Serrano-Fernandez P Ibrahim SM Zettl UK Thiesen HJ Godde R Epplen JT Moller S Intergenomic consensus in multifactorial inheritance loci: the case of multiple sclerosis Genes Immun 2004 5 615 620 15573086 10.1038/sj.gene.6364134
Mansson R Tsapogas P Akerlund M Lagergren A Gisler R Sigvardsson M Pearson correlation analysis of microarray data allows for the identification of genetic targets for early B-cell factor J Biol Chem 2004 279 17905 17913 14960572 10.1074/jbc.M400589200
Maimone D Guazzi GC Annunziata P IL-6 detection in multiple sclerosis brain J Neurol Sci 1997 146 59 65 9077497 10.1016/S0022-510X(96)00283-3
Navikas V Matusevicius D Soderstrom M Fredrikson S Kivisakk P Ljungdahl A Hojeberg B Link H Increased interleukin-6 mRNA expression in blood and cerebrospinal fluid mononuclear cells in multiple sclerosis J Neuroimmunol 1996 64 63 69 8598391 10.1016/0165-5728(95)00155-7
Okuda Y Sakoda S Saeki Y Kishimoto T Yanagihara T Enhancement of Th2 response in IL-6-deficient mice immunized with myelin oligodendrocyte glycoprotein J Neuroimmunol 2000 105 120 123 10742553 10.1016/S0165-5728(00)00192-2
Lee VM Goedert M Trojanowski JQ Neurodegenerative tauopathies Annu Rev Neurosci 2001 24 1121 1159 11520930 10.1146/annurev.neuro.24.1.1121
Ebneth A Godemann R Stamer K Illenberger S Trinczek B Mandelkow E Overexpression of tau protein inhibits kinesin-dependent trafficking of vesicles, mitochondria, and endoplasmic reticulum: implications for Alzheimer's disease J Cell Biol 1998 143 777 794 9813097 10.1083/jcb.143.3.777
Kapaki E Paraskevas GP Michalopoulou M Kilidireas K Increased cerebrospinal fluid tau protein in multiple sclerosis Eur Neurol 2000 43 228 232 10828654 10.1159/000008181
Schneider A Araujo GW Trajkovic K Herrmann MM Merkler D Mandelkow EM Weissert R Simons M Hyperphosphorylation and aggregation of tau in experimental autoimmune encephalomyelitis J Biol Chem 2004 279 55833 55839 15494405 10.1074/jbc.M409954200
Stokin GB Lillo C Falzone TL Brusch RG Rockenstein E Mount SL Raman R Davies P Masliah E Williams DS Axonopathy and transport deficits early in the pathogenesis of Alzheimer's disease Science 2005 307 1282 1288 15731448 10.1126/science.1105681
Surinya KH Cox TC May BK Identification and characterization of a conserved erythroid-specific enhancer located in intron 8 of the human 5-aminolevulinate synthase 2 gene J Biol Chem 1998 273 16798 16809 9642238 10.1074/jbc.273.27.16798
Ghayor C Herrouin JF Chadjichristos C Ala-Kokko L Takigawa M Pujol JP Galera P Regulation of human COL2A1 gene expression in chondrocytes. Identification of C-Krox-responsive elements and modulation by phenotype alteration J Biol Chem 2000 275 27421 27438 10856292
Ozaki K Ohnishi Y Iida A Sekine A Yamada R Tsunoda T Sato H Hori M Nakamura Y Tanaka T Functional SNPs in the lymphotoxin-alpha gene that are associated with susceptibility to myocardial infarction Nat Genet 2002 32 650 654 12426569 10.1038/ng1047
Tokuhiro S Yamada R Chang X Suzuki A Kochi Y Sawada T Suzuki M Nagasaki M Ohtsuki M Ono M An intronic SNP in a RUNX1 binding site of SLC22A4, encoding an organic cation transporter, is associated with rheumatoid arthritis Nat Genet 2003 35 341 348 14608356 10.1038/ng1267
Villar LM Masjuan J Gonzalez-Porque P Plaza J Sadaba MC Roldan E Bootello A Alvarez-Cermeno JC Intrathecal IgM synthesis is a prognostic factor in multiple sclerosis Ann Neurol 2003 53 222 226 12557289 10.1002/ana.10441
Villar LM Sadaba MC Roldan E Masjuan J Gonzalez-Porque P Villarrubia N Espino M Garcia-Trujillo JA Bootello A Alvarez-Cermeno JC Intrathecal synthesis of oligoclonal IgM against myelin lipids predicts an aggressive disease course in MS J Clin Invest 2005 115 187 194 15630459 10.1172/JCI200522833
Sellebjerg F Christiansen M Garred P MBP, anti-MBP and anti-PLP antibodies, and intrathecal complement activation in multiple sclerosis Mult Scler 1998 4 127 131 9762660 10.1191/135245898678909475
Mead RJ Singhrao SK Neal JW Lassmann H Morgan BP The membrane attack complex of complement causes severe demyelination associated with acute axonal injury J Immunol 2002 168 458 465 11751993
Uccelli A Aloisi F Pistoia V Unveiling the enigma of the CNS as a B-cell fostering environment Trends Immunol 2005 26 254 259 15866238 10.1016/j.it.2005.02.009
Akerblad P Lind U Liberg D Bamberg K Sigvardsson M Early B-cell factor (O/E-1) is a promoter of adipogenesis and involved in control of genes important for terminal adipocyte differentiation Mol Cell Biol 2002 22 8015 8025 12391167 10.1128/MCB.22.22.8015-8025.2002
|
16255771
|
PMC1291372
|
CC BY
|
2021-01-04 16:28:53
|
no
|
BMC Neurol. 2005 Oct 28; 5:19
|
utf-8
|
BMC Neurol
| 2,005 |
10.1186/1471-2377-5-19
|
oa_comm
|
==== Front
BMC OphthalmolBMC Ophthalmology1471-2415BioMed Central London 1471-2415-5-251626290710.1186/1471-2415-5-25Research ArticleA functional profile of gene expression in ARPE-19 cells Sharma Rajesh K [email protected] William E [email protected] Allyson D [email protected] Dianna A [email protected] Department of Ophthalmology and Hamilton Eye Institute, University of Tennessee Health Science Center, 930 Madison Ave, Memphis, TN 38163, USA2005 1 11 2005 5 25 25 21 1 2005 1 11 2005 Copyright © 2005 Sharma et al; licensee BioMed Central Ltd.This is an Open Access article distributed under the terms of the Creative Commons Attribution License (), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
Background
Retinal pigment epithelium cells play an important role in the pathogenesis of age related macular degeneration. Their morphological, molecular and functional phenotype changes in response to various stresses. Functional profiling of genes can provide useful information about the physiological state of cells and how this state changes in response to disease or treatment. In this study, we have constructed a functional profile of the genes expressed by the ARPE-19 cell line of retinal pigment epithelium.
Methods
Using Affymetrix MAS 5.0 microarray analysis, genes expressed by ARPE-19 cells were identified. Using GeneChip® annotations, these genes were classified according to their known functions to generate a functional gene expression profile.
Results
We have determined that of approximately 19,044 unique gene sequences represented on the HG-U133A GeneChip® , 6,438 were expressed in ARPE-19 cells irrespective of the substrate on which they were grown (plastic, fibronectin, collagen, or Matrigel). Rather than focus our subsequent analysis on the identity or level of expression of each individual gene in this large data set, we examined the number of genes expressed within 130 functional categories. These categories were selected from a library of HG-U133A GeneChip® annotations linked to the Affymetrix MAS 5.0 data sets. Using this functional classification scheme, we were able to categorize about 70% of the expressed genes and condense the original data set of over 6,000 data points into a format with 130 data points. The resulting ARPE-19 Functional Gene Expression Profile is displayed as a percentage of ARPE-19-expressed genes.
Conclusion
The Profile can readily be compared with equivalent microarray data from other appropriate samples in order to highlight cell-specific attributes or treatment-induced changes in gene expression. The usefulness of these analyses is based on the assumption that the numbers of genes expressed within a functional category provide an indicator of the overall level of activity within that particular functional pathway.
==== Body
Background
The retinal pigment epithelium (RPE) is a monolayer of hexagonal cells separating the neural retina from the underlying choroidal vascular bed. RPE cells are essential for development, survival, and physiological activity of photoreceptor cells [1,2]. RPE cells provide the molecular machinery for recycling the inactive form of the photoisomerized visual pigment back to the active isomer for subsequent formation of rhodopsin [3]. RPE phagocytizes spent photoreceptor outer segments; provides nutrients to, and removes metabolic waste from, the photoreceptors; contributes to retinal adhesion and maintenance of the blood-retinal barrier; and absorbs light and dissipates heat energy derived from incident light [4,5]. Recent evidence shows that RPE cells also participate in the immunologic functions in the retina. RPE cells can express major histocompatibility complex (MHC) class I and II antigens and the intercellular adhesion molecule-1 (ICAM-1). These cells process and present the antigen to helper T cells [6-9]. RPE responds to proinflammatory cytokines and secretes IL-6, IL-8, and monocyte chemotactic protein [10-14]. Through these mechanisms RPE cells play a key role in inflammatory, infectious, and degenerative diseases of the retina. Impairment of RPE functions have been implicated in a number of hereditary retinal degenerations [15-18], and more importantly in the pathogenesis of age-related macular degeneration (AMD), one of the most prevalent causes of visual impairment in elderly [19]. Given the importance of RPE cells in the normal physiology and disease of retina, RPE has become the subject of intense investigation especially those elucidating the role of RPE cells in the molecular mechanisms of AMD. Transplantation of normal as well as genetically modified RPE cells is being envisaged as a possible treatment of retinal degenerations [20,21]. Given the pivotal role of RPE in retinal development, physiology and diseases it is important to investigate the gene expression profile of these cells, which will than lay the foundation for further molecular characterization of RPE cells in both normal and diseased states.
DNA microarray technology provides a view of the expression profiles of a cell sample that encompasses virtually the entire genome. Microarray technology has a number of distinct applications including DNA sequencing, mutation analysis, gene discovery, and gene expression analysis [22-26]. Microarray technology allows a rapid quantitative measurement of gene expression within a tissue sample, as defined by messenger RNA (mRNA) abundance. The opportunity to quantitate gene expression on a genome-wide scale has added a new dimension to our understanding of many biologic and disease processes. However, analysis of large data sets derived from microarray analysis can be problematic. It is an overwhelming task to consider the expression levels of each of the twenty or so thousand known genes individually. An alternative strategy is to group individual genes into functional categories in order to generate what has been termed a "functional gene profile". Different types of analyses can then be applied to gene profiles. For example, functional categories of genes displaying the highest levels of expression can be identified and thus provide a means for focusing on groups of functionally related genes that may be highly expressed by a specific cell type or physiological state. "Cluster analysis" is another more complex type of gene profile analysis in which significant changes in gene expression due to some experimental variable are mapped with respect to functional categories.
We propose a novel approach to gene profile analysis based on simply the total number of genes within a functional category that are expressed above some pre-determined level. In order to meet the goal of generating a data set that will be comprehensive but not too large to be readily useful, we have generated a functionally classified list of genes whose expression level met the specific criteria established by Affymetrix analysis instead of considering the absolute level of expression. Below we describe 130 functional categories that account for 68% of the genes represented on the Affymetrix microarray chip, and 70% of the genes expressed by the ARPE-19 cell line. Key words for the functional categories were chosen from the HG-U133A GeneChip® Library, allowing us to use Affymetrix annotation terms (i.e., key words) to sort genes into categories. We used this classification scheme to calculate the number of functionally related genes expressed within a category and to produce a ARPE-19 Gene Expression Profile. Since data for each category consists of a single number, the entire database for RPE gene expression can be represented in a Profile with 130 data points. We suggest that comparing the ARPE-19 Profile with the profile of genes represented on the Affymetrix HG-U133A GeneChip® may provide a measure of the cell-specific pattern of gene expression unique for RPE cells. A RPE-specific expression profile data base would have a number of potential uses, such as selecting specific genes and functionally related groups of genes for further analysis with microarray and validation by RT-PCR. Data already available in the literature demonstrates that many of the genes included in our expression profile are known to be present in ARPE-19 cells as validated by quantitative RT-PCR (for example, see Chowers I 2004 IOVS [27]). We further suggest that functional categories with large numbers of expressed genes may reflect high relative importance of those specific functions to RPE. Another unique aspect of our approach was that it excluded from analysis any genes whose expression was dependent on the substrate upon which the ARPE-19 cells were grown. There is no uniformly accepted substrate utilized researchers in this field, yet expression of certain classes of genes is known to be substrate dependent. (See Discussion.) We grew cells on four commonly used substrates (fibronectin, Matrigel, collagen, and uncoated plastic culture dishes) and included in our analysis only those genes that were uniformly expressed by cells on all four substrates. Our intent was to focus on "substrate-independent" genes that would be likely to be expressed under most experimental conditions. The advantages and disadvantages related to this overall analysis strategy are reviewed in the Discussion section.
Methods
RPE culture
ARPE-19 cells were used in the experiments. These are diploid non-transformed human RPE cells that display many properties typical of differentiated RPE in vivo [28]. ARPE-19 cells were obtained from a commercial source (ATCC, Manassas, VA). The cells were plated on 75-cm2 flasks at a density of 10,000 cells/cm2 and maintained in culture until the plates became >95% confluent. Cultures were fed three times a week with Dulbecco's modified Eagle's medium-nutrient mixture F-12 (DMEM-F-12; GIBCO, Invitrogen Corporation, Grand Island, NY) supplemented with 10% fetal bovine serum, 100 U/mL penicillin, and 100 μg/mL streptomycin. Cultures were passaged by dissociation in 0.05% (wt/vol) trypsin. For the microarray experiments, cells belonging to passage 4 were used 24 hrs. after removing serum from the medium in order to further synchronize the metabolic and physiological state of cells.
Substrate-dependent genes
Our overall goal in these experiments was to generate a functional catalogue of "core" ARPE-19 genes. To achieve that goal, we sought to avoid genes that might be highly sensitive to exact culturing conditions, such as the choice of substrate. While substrate-sensitive genes are likely to be very important to the cell, as a group they may confound our results. Since there is no uniformly accepted substrate for APRE-19 culture, we reasoned that a catalogue of genes that were expressed regardless of substrate would provide the best standard. Cells were grown on four commonly used substrates: fibronectin, Matrigel, collagen, and uncoated plastic. Each sample was run on a separate chip and analyzed as described below. Any genes not expressed on all four substrates were identified as "substrate-specific genes" and were removed from further analysis. Data from the remaining substrate-independent genes were considered as n = 4 (In other words, since we only included genes uniformly expression in all four samples, there is no variation in the gene expression profile among the n = 4). Variability in expression levels of substrate-independent genes among these four chips was relatively low (see Results) suggesting that overall variability due to technical factors was low.
Microarray methods
The RNA isolation procedures for the Affymetrix analysis were conducted using TRIzol Reagent (GIBCO, Carlsbad, CA) according to the manufacturer's instructions. Initially, the quality of total RNA was assessed by electrophoresis through a 1% agarose gel, then the Agilent Bioanalyzer System (Agilent Technologies, Palo Alto, CA) immediately prior to cRNA synthesis. The procedures for the Affymetrix gene chips, beginning with first strand cDNA synthesis, were conducted by Genome Explorations (Memphis, Tennessee). The Human Genome U133A GeneChip® contains 22,283 probe sets together with expressed sequence tag (EST) sequences. The RNA (isolated using TRIzol) was run over of a G50 spin column. First and second strand cDNA were synthesized from 15 μg of total RNA using the SuperScript Double-Stranded cDNA Synthesis Kit (GIBCO, Carlsbad, CA) and oligo-dT24-T7 (5'-GGC CAG TGA ATT GTA ATA CGA CTC ACT ATA GGG AGG CGG-3') primer according to the manufacturer's instructions. cRNA was synthesized and labeled with biotinylated UTP and CTP by in vitro transcription using the T7 promoter coupled double stranded cDNA as template and the T7 RNA Transcript Labeling Kit (ENZO Diagnostics Inc. Farmingdale NY). The fragmented cRNA was hybridized to the olygonucleotide array, washed, stained with phycoerythrein conjugated streptavidin (Molecular Probes, Eugene OR), and scanned. Intensities were determined using a laser confocal scanner (Hewlett-Packard; Palo Alto, CA).
The scanned images were analyzed using Microarray Suite Version 5.0 (MAS 5.0, Affymetrix, Inc., Santa Clara, CA). The MAS 5.0 statistical algorithms calculate signal intensity, probe set detection, probe set (gene expression) change, and signal log ratio. The signal intensity for each gene was calculated as the average intensity difference, represented by [S(PM – MM)/(number of probe pairs)], where PM and MM denote perfect-match and mismatch probes.
The analysis applied a decision matrix based on the hybridization behavior of all 11 probe pairs per probe set. These matrices are used to determine if the gene is expressed above a threshold level (i.e., called Present by Absolute Call decision matrix). Investigators can change the specificity and sensitivity criteria for the present call by changing the alpha-1 value in the MAS 5.0 software. We used a value of 0.05 as a standard and 0.07 or 0.18 for less restrictive present calls.
Many genes on the chip are represented by more than one probe set. Since we were interested in the number of expressed genes, not the number of probe sets, a method was devised for selecting a single probe set to represent each gene. First, we performed a search for multiple probe sets using the designated Unigene ID or title for each gene. Once the redundant probe sets were identified, we determined the coefficient of variation of the expression levels of each probe set for the n of 4 chips. The probe set with the lowest coefficient of variation was chosen. Data from replicate probe sets for the same gene were omitted from further analysis. This choice of method was somewhat arbitrary, but it had the advantage of being based strictly on a statistical criterion and it should not introduce any bias in the data since all probe sets, all genes, and all chips were treated the same. The coefficient of variation data was not used for any subsequent steps in the analysis. The resulting database, free of redundant probes for any given gene, was used for determining the number of genes expressed within functional categories.
Functional categories were established for classification of the majority of genes represented on the HG-U133A GeneChip® , including those expressed by ARPE-19 cells. (See Results for additional details.) HG-U133A GeneChip® annotations were downloaded from Affymetrix. Using FileMaker Pro 5.5 (FileMaker, Inc., Santa Clara, CA), the following annotations were used to create a HG-U133A GeneChip® Library: Probe Set ID, Title, Unigene ID, Sequence Derived from, Sequence Description, Archival Reference Group, Gene Symbol, GO Biological Process, GO Molecular Function, Proteome Biochemical Function, Proteome Cellular Role, Interpro ID and Classification, Ortholog-Homolog, and Pathways. Relational databases were created by linking the HG-U133A GeneChip® Library to the four data sets provided by the Affymetrix MAS 5.0 analysis. Search terms (i.e. keywords) were used to extract the probe sets for each functional category. The resulting classification scheme consisted of 130 functional categories that could be grouped under 15 major subheadings. Many of the genes could be classified in more than one category. Genes were included in all the categories in which they were classifiable. Additional literature searches were used to clarify ambiguities. In order to determine whether the relative number of genes expressed by ARPE-19 cells within a particular functional category was significantly different from that predicted based on numbers represented on the chip, we performed a z-test
Results
Numbers of genes expressed
The Affymetrix HG-U133A GeneChip® contains 22,283 probe sets representing 19,044 distinct genes, of which 6,438 genes were called as present in all four samples of ARPE-19 cells with an alpha-1 value of 0.05. When present calls were made with a less restrictive sensitivity and specificity criterion (alpha-1 value was changed to 0.18), 8,671 genes were called present.
Establishment of functional categories
In order to establish a functional profile of genes, each of the 19,044 distinct genes represented on the HG-U133A GeneChip® was assigned a functional classification using Affymetrix software search programs. By a process of trial and error, these individual functional classifications were organized into a limited number of function-related categories. Our goal was to establish a classification scheme that would include as many genes as possible within a "reasonable" number of functional categories. By this process, we were able to select 130 functional categories that accounted for 66.48% of the genes printed on the HG-U133A GeneChip® and 70.94% of the genes expressed by ARPE-19 cells. These categories are listed on the abscissa of Figure 1 and 2, arranged under 15 major functional subheadings. Most of the genes could be classified in more than one category. Genes were counted in each functional category they were associated with, which resulted in multiple listings of most genes. Roughly 33% of the genes printed on the HG-U133A GeneChip® and 29% of the genes expressed by ARPE-19 cells could not be classified in functional categories because the probes represented genes with unknown products or unknown functions. In addition, certain functional categories represented by very small numbers of genes were also placed under the category of unclassified for practical reasons.
A functional profile of gene expression
Rather than display the data as absolute numbers of genes in each functional category, we chose to represent each number as a percentage. Results are shown in Figure 1 and 2, using percentages calculated from the formulae listed below.
The blue bars = the number of genes in a given category that are represented on the HG-U133A GeneChip®/the total number of genes represented on the HG-U133A GeneChip® .
The red bars = the number of genes in a given category that are expressed by ARPE-19/the total number of genes represented on the HG-U133A GeneChip® .
Results show that the functional profile of the genes expressed by ARPE-19 was significantly different from the functional profile of genes represented on the HG-U133A GeneChip® . Overall, ARPE-19 expressed 33.8% of the represented genes. However, the percentage of genes expressed in each functional category varied considerably from that norm. Based on a z-test analysis, a total of 60 of the 130 functional categories contained significantly more or significantly less than the predicted number of expressed genes (See Fig. 1 and 2).
An additional indication of the uniqueness of the ARPE-19 functional gene profile can be seen in a comparison of functional categories containing the largest percentages of both represented and expressed genes. The ten categories containing the largest numbers of genes are shown in Table 1. These ten categories represent the same four basic functional groups in both lists: Receptors, Signal Transduction, Metabolism, and Transport. In spite of this overall similarity, these two lists differ in several important aspects. 1) Genes related to Metabolism and to Signal Transduction-associated Metabolism were among the most numerously expressed by ARPE-19, whereas these genes were not among the most numerously represented on the chip. The reverse was true for genes related to Transcription Factors. 2) Of particular interest were several cases in which ARPE-19 expressed approximately 50% or more of the genes represented on the HG-U133A GeneChip® . (See Fig. 1 and 2.) This occurred in functional categories that were large (i.e., Binding Protein, Hydrolase, and Metabolism), as well as those that were small (i.e., Ribosomal Proteins and Protein Synthesis). Based on these data, we conclude that the ARPE-19 Functional Gene Expression Profile does reflect characteristics of the ARPE-19 cell type and that it is not simply a reflection of the functional groups of probe sets represented on the HG-U133A GeneChip® .
We completed an additional calculation in which the Functional Gene Expression Profile was expressed as a different type of percentage also shown in Fig. 1 and 2, using the formula below:
Yellow bars = the number of genes within a given functional category that are expressed by ARPE-19/Total number of genes expressed.
This calculation is comparable to that used in constructing a standard "pie chart," which has been routinely used by other investigators to display the function of genes expressed in a given cell type. Inclusion of 130 functional categories used in our analysis adds considerably more detail than could be contained on a normal pie chart format and thus cannot be presented as such. However, the bar graph presentation does allow an expanded overview of virtually all known genes within a reasonably simple format. Results show that several functional categories of genes account for a large portion of the total number of genes expressed.
In the three largest functional categories (Binding Protein, Cell Surface Receptors, and Receptors), the number of genes expressed accounted for approximately 32% of the total number of genes expressed. It must be kept in mind that our classification scheme includes most genes in more than one category and that there is likely to be considerable overlap among these three categories. Thus, the actual number of distinct genes in this functional grouping could be as low as 10%. Even so, this represents a major functional class of genes expressed by ARPE-19.
In each of eight large categories, the number of genes expressed accounted for approximately 5% of the total number expressed. These latter categories represent genes associated with cell signaling, transport, gene/protein expression, and energy metabolism.
Effect of specificity/sensitivity parameters and substrate on gene classification
In order to determine the degree to which specificity and sensitivity parameters influence the analysis and functional categorization of the expressed genes, we altered the alpha-1 value in the Microarray Suit 5.0 so that the present calls were less specific but more sensitive, i.e., less restrictive. When the alpha-1 value was changed from 0.05 to 0.075 only 570 additional genes were called present, suggesting that our data were not overly influenced by the sensitivity parameters originally selected. To increase the number of present calls by approximately 30% required setting the alpha-1 value as high as 0.18. With this alpha-1 value, 8671 genes were called present in all four samples.
To determine the degree to which our results were influenced by variations in genes expressed by samples from the four different culture substrates used, we compared the number of genes expressed in all four samples with the number of genes expressed in at least one of the samples. Results show that 6438 genes were called present in all samples and that 9749 genes were called present in at least one sample. This suggests that substrate may have significant effects on gene expression.
Discussion
Mapping of chromosomal positions and genomic organization of human genes has elucidated the chemical background of the genome [29], linking specific genes to various human diseases. However, to understand the pathophysiological mechanisms, it is prudent to resort to functional genomics approaches [30,31]. Identifying the genes expressed in a particular tissue and profiling their function, as we have done in this study for a widely used human retinal pigment epithelium cell line, lays the foundations for such understanding.
An additional objective was to focus primarily on the genes likely to be consistently expressed even under varying culture conditions, specifically when different substrates were used. By including only substrate-independent genes, the Profile may be more widely applicable to labs using different culture conditions and substrates. There is no standard substrate that is uniformly accepted by ARPE-19 researchers; several are in common usage (plastic, collagen, Matrigel, and fibronectin). Our aim was to include genes that are expressed by the ARPE-19 cells irrespective of the substrate. We cultured four samples for microarray analysis, each on one of these four substrates and then eliminated from our analysis any genes that were not uniformly expressed by all four samples. Theoretically then, our analysis should be independent of specific substrate effects. It should be kept in mind, however, that substrate-specific effects may play an important role in RPE cells. In vivo, RPE cells grow in a penta-lamellar structure called Bruch's membrane that is thought to play an important role in the health and disease of RPE cells [1]. Changes in the Bruch's membrane have been implicated in the pathogenesis of age-related macular degeneration where RPE also plays an important role. It has been shown that attachment of RPE cells to the basement membrane is essential for its survival. All these facts imply that the nature of basement membrane affects the gene expression in RPE cells. These genes may not be included in our database of 6,438 genes called present in all four samples. For comparison, we calculated that 9,749 genes were expressed by at least one of the four samples. This suggests that up to 3,311 genes could be substrate specific. Additional experiments would be required in order to confirm this suggestion and to the identity genes that are specifically expressed in response to a given substrate.
It is estimated that a typical mammalian cell expresses about 10,000 to 20,000 mRNA species and in diseased conditions between 0.2–10% of this may be differentially expressed. Considering that approximately half of the human genome is represented on the HG-U133A GeneChip® , the detection of 6,438 genes falls within the expected range. It is also estimated that approximately 10–20% of the entire genome is expressed in any cell type. Our study gave a slightly higher value, with 33.8% of the genes expressed in the ARPE-19 cells. This might reflect the fact that we examined only genes that encode for proteins whose identity and function are known. This group includes many of the common housekeeping genes that are expected to be expressed in most cells, and thus might have a higher probability for detection in our analysis. The group of unidentified genes not included in our analysis may more likely include rare genes that would not be expected to be as widely expressed.
The Functional Gene Expression Profile is not a strict quantitative indicator of any given function because, firstly, it takes into account only the expression of a gene and not its level of expression. Secondly, individual genes might have excitatory or inhibitory influence on a particular function. Lastly, the level of gene expression may not be quantitatively related to the function. Other parameters such as rate of translation, RNA turnover, post-transitional modification and degradation rate of proteins can all affect the degree to which a given gene and its protein product contribute to the functional state of cells. Nevertheless, it is reasonable to assume that if a cell is actively involved in a given function it will likely express many of the genes involved in the corresponding functional pathways. Likewise, if the cell is not involved in a specific function it will express fewer of the genes related to those functionally related pathways. Even though some genes activate and others inhibit the function, a cell must maintain homeostasis and thus is likely to regulate any ongoing activity – for example cell division – by balancing the expression of excitatory and inhibitory factors. If this is the case, all genes within appropriate cell division pathways would have a greater probability of being expressed in an actively dividing cell than in one that is quiescent. The actively dividing cell would have a higher level of "gene chatter" within cell division pathways, which should be reflected in a shift in the percentage of genes expressed within that functional category. An additional aim in developing a Functional Gene Expression Profile was to facilitate analysis of a large microarray data set. The major contribution that our work provides in this regard is the development of a classification scheme of limited size, which includes virtually all ARPE-19 expressed genes whose functions are known. The classification scheme was constructed using Affymetrix search terms (i.e., key words) that appear in the HG-U133A GeneChip® annotations, which provide a readily accessible, standard vocabulary for the uniform classification of gene expression data sets by other investigators. These GeneChip® annotations are updated quarterly by Affymetrix. They can be easily downloaded and used to create or update GeneChip® libraries and searchable relational databases.
The Profile is essentially an expanded pie chart, that contains more information than can feasibly be presented in a standard pie chart format. Nevertheless, it can be displayed in a reasonably sized bar graph with 130 data points. By representing expression results as a percentage of the total number of expressed genes, direct comparisons of expression information (albeit in compressed form) can be made for virtually all functionally identified genes across cell types, treatments, physiological states, etc. Recently, somewhat similar approaches have been used to make data mining SoftWear tools (EASE) that allow comparisons of gene lists and search for gene categories over represented in a sample.
If the Function Gene Expression Profile is to be a useful as a genetic blueprint for cell types or functional states, it must be sensitive enough to reflect substantive differences in gene expression that are unique for those specific cell types or physiological sates. Experiments are underway to prepare Profiles of appropriate data sets from other cell types and to carry out comparative analyses. From these comparisons, we will determine the degree to which Profiles differ, and more importantly, if these differences can provide the basis for identification of genes and functional pathways that are of particular relevance to the cell or physiological state in question. Our current results do show a significant difference in the profile of genes expressed by ARPE-19 compared to the profile of the genes represented on the HG-U133A GeneChip® . Thus, we have one comparison that shows unique aspects of ARPE-19 gene expression compared to all genes expressed by all cells. Even in the absence of further comparative data, the Profile provides a useful gene expression snapshot of a confluent monolayer of ARPE-19 cells. The highest percentages of genes expressed were in categories that could be related to specialized RPE functions (receptors and binding proteins) and those that may be related to housekeeping genes (energy metabolism, transport, and gene/protein expression). The quiescent state of the culture is consistent with low percentages of genes expressed in functional categories that include cell division, cell growth, and cell structure/mobility.
Conclusion
We present a system of profiling the expressed genes based on their functions. The Profile can be compared with equivalent microarray data from other appropriate samples in order to highlight cell-specific attributes or treatment-induced changes in gene expression. The usefulness of these analyses is based on the assumption that the numbers of genes expressed within a functional category provide an indicator of the overall level of activity within that particular functional pathway.
Competing interests
The author(s) declare that they have no competing interests
Pre-publication history
The pre-publication history for this paper can be accessed here:
Acknowledgements
This work was supported by grant from NEI (EY13080 to D.J.), Research to Prevent Blindness, International Retina Research Foundation, and the UT Memphis Neuroscience Center. The authors wish to thank Dr. Peter A. Netland for support, and Danny Morse for his assistance in preparing Figures.
Figures and Tables
Figure 1 Functional categories of genes represented on the HG-U133A GeneChip174 and expressed by ARPE-19 cells. A total 130 functional categories are displayed on the abscissa. These are arranged alphabetically in groups under 15 large functional subheadings indicated on the right and separated by dashed lines. The numbers of genes represented and expressed were converted into percentages indicated on the ordinate. The formulae used for calculating percentages were different for each color-coded bar. Yellow bar = total number of genes expressed by ARPE-19 within the functional category/total number of genes expressed in all categories. The heights of the yellow bars may reflect the relative importance of a given function in ARPE-19 cells. Red bar = total number of genes expressed within the functional category/total number of genes represented on the HG-U133A GeneChip174 in all categories. Blue bar = total number of genes represented within the functional category/total number of genes represented in all categories. Comparisons of the heights of the red versus blue bars in each category show considerable variation from the overall average of approximately 33% expressed. Based on a z-test analysis, categories that have statistically higher-than-predicted percentages expressed are noted by "+;" those with statistically lower-than-predicted percentages expressed are noted by "-." These differences indicate that the functional profile of genes expressed are not a simple reflection of the functional profile of genes represented on the chip. Abbreviations: Ca – calcium; BDNF – brain derived neurotrophic factor ; IGF – insulin growth factor; TRH – thyroid releasing hormone; GABA – gamma amino butyric acid; NADPH – nicotinamide adenine dinucleotide phosphate (reduced form); 5HT – 5 hydroxytryptamine; AMPA – alpha amino-3-hydroxy-5-methylisoxazole-4-proprionic acid; NMDA – n-methyl-d-aspartate.
Figure 2 Functional categories of genes represented on the HG-U133A GeneChip174 and expressed by ARPE-19 cells. A total 130 functional categories are displayed on the abscissa. These are arranged alphabetically in groups under 15 large functional subheadings indicated on the right and separated by dashed lines. The numbers of genes represented and expressed were converted into percentages indicated on the ordinate. The formulae used for calculating percentages were different for each color-coded bar. Yellow bar = total number of genes expressed by ARPE-19 within the functional category/total number of genes expressed in all categories. The heights of the yellow bars may reflect the relative importance of a given function in ARPE-19 cells. Red bar = total number of genes expressed within the functional category/total number of genes represented on the HG-U133A GeneChip174 in all categories. Blue bar = total number of genes represented within the functional category/total number of genes represented in all categories. Comparisons of the heights of the red versus blue bars in each category show considerable variation from the overall average of approximately 33% expressed. Based on a z-test analysis, categories that have statistically higher-than-predicted percentages expressed are noted by "+;" those with statistically lower-than-predicted percentages expressed are noted by "-." These differences indicate that the functional profile of genes expressed are not a simple reflection of the functional profile of genes represented on the chip. Abbreviations: Ca – calcium; BDNF – brain derived neurotrophic factor ; IGF – insulin growth factor; TRH – thyroid releasing hormone; GABA – gamma amino butyric acid; NADPH – nicotinamide adenine dinucleotide phosphate (reduced form); 5HT – 5 hydroxytryptamine; AMPA – alpha amino-3-hydroxy-5-methylisoxazole-4-proprionic acid; NMDA – n-methyl-d-aspartate.
Table 1 Functional Categories that Contain the Largest Percentages of Genes
Categories with ≥ 4% of genes represented on the chip Categories with ≥ 0.5% of genes expressed by ARPE
Functional Category Percentage* Functional Category Percentage 1 Percentage 2
Cell Sur face Receptors 9.6 Binding Proteins 4.2 12.3
Receptors 9.6 Cell Surface Receptors 2.3 6.8
Binding Proteins 9.6 Receptors 2.3 6.8
Kinase 5.4 Kinase 2.0 5.9
Signal transduction 5.4 Hydrolase 1.9 5.7
Transport (Metabolism) 5.1 Transport 1.9 5.6
Transport 5.1 Transport (Metabolism) 1.8 5.6
Transcription Factor 4.8 Signal transduction 1.7 4.9
Channels/Transport Proteins 4.2 Metabolism (signal transduction) 1.6 4.9
Hydrolase 4.3 Metabolism 1.6 4.7
* number of genes in a given category that are represented on the chip/ total number of genes represented on the chip
1 number of genes in a given category that are expressed by ARPE/ total number of genes represented on the chip
2 number of genes in a given functional category that are expressed by ARPE/ total number of genes expressed
==== Refs
Sharma RK Ehinger B Kaufman PL, Alm A Development and Structure of the Retina Adler's Physiology of the Eye: Clinical Application 2003 St. Louis: Mosby 319 347
Sharma RK Johnson DA Molecular signals for development of neuronal circuitry in the retina Neurochem Res 2000 25 1257 1263 11059800 10.1023/A:1007696112956
Wald G Molecular basis of visual excitation Science 1968 162 230 239 4877437
Bok D The retinal pigment epithelium: a versatile partner in vision J Cell Sci Suppl 1993 17 189 195 8144697
Bok D Retinal photoreceptor-pigment epithelium interactions. Friedenwald lecture Invest Ophthalmol Vis Sci 1985 26 1659 1694 2933359
Dhawan J Lichtler AC Rowe DW Farmer SR Cell adhesion regulates pro-alpha 1(I) collagen mRNA stability and transcription in mouse fibroblasts J Biol Chem 1991 266 8470 8475 2022661
Schwogler S Odenthal M Meyer zum Buschenfelde KH Ramadori G Alternative splicing products of the tenascin gene distinguish rat liver fat storing cells from arterial smooth muscle cells and skin fibroblasts Biochem Biophys Res Commun 1992 185 768 775 1376992 10.1016/0006-291X(92)91692-J
Tucker RP Hammarback JA Jenrath DA Mackie EJ Xu Y Tenascin expression in the mouse: in situ localization and induction in vitro by bFGF J Cell Sci 1993 104 69 76 7680659
LaFleur DW Fagin JA Forrester JS Rubin SA Sharifi BG Cloning and characterization of alternatively spliced isoforms of rat tenascin. Platelet-derived growth factor-BB markedly stimulates expression of spliced variants of tenascin mRNA in arterial smooth muscle cells J Biol Chem 1994 269 20757 20763 7519614
Elner VM Scales W Elner SG Danforth J Kunkel SL Strieter RM Interleukin-6 (IL-6) gene expression and secretion by cytokine-stimulated human retinal pigment epithelial cells Exp Eye Res 1992 54 361 368 1381679 10.1016/0014-4835(92)90048-W
Elner VM Strieter RM Elner SG Baggiolini M Lindley I Kunkel SL Neutrophil chemotactic factor (IL-8) gene expression by cytokine- treated retinal pigment epithelial cells Am J Pathol 1990 136 745 750 2183623
Elner SG Strieter RM Elner VM Rollins BJ Del Monte MA Kunkel SL Monocyte chemotactic protein gene expression by cytokine-treated human retinal pigment epithelial cells Lab Invest 1991 64 819 825 2046333
Jaffe GJ Roberts WL Wong HL Yurochko AD Cianciolo GJ Monocyte-induced cytokine expression in cultured human retinal pigment epithelial cells Exp Eye Res 1995 60 533 543 7615019 10.1016/S0014-4835(05)80068-5
Nagineni CN Detrick B Hooks JJ Synergistic effects of gamma interferon on inflammatory mediators that induce interleukin-6 gene expression and secretion by human retinal pigment epithelial cells Clin Diagn Lab Immunol 1994 1 569 577 8556503
Bok D Hall MO The role of the pigment epithelium in the etiology of inherited retinal dystrophy in the rat J Cell Biol 1971 49 664 682 5092207 10.1083/jcb.49.3.664
Mullen RJ La Vail MM Inherited retinal dystrophy: primary defect in pigment epithelium determined with experimental rat chimeras Science 1976 192 799 801 1265483
Marlhens F Bareil C Griffoin JM Zrenner E Amalric P Eliaou C Mutations in RPE65 cause Leber's congenital amaurosis Nat Genet 1997 17 139 141 9326927 10.1038/ng1097-139
Gu SM Thompson DA Srikumari CR Lorenz B Finckh U Nicoletti A Murthy KR Rathmann M Kumaramanickavel G Denton MJ Gal A Mutations in RPE65 cause autosomal recessive childhood-onset severe retinal dystrophy Nat Genet 1997 17 194 197 9326941 10.1038/ng1097-194
Sarks SH Ageing and degeneration in the macular region: a clinico-pathological study Br J Ophthalmol 1976 60 324 341 952802
Sharma RK Ehinger B Management of hereditary retinal degenerations: present status and future directions Surv Ophthalmol 1999 43 427 444 10340561 10.1016/S0039-6257(99)00006-5
Sharma RK Bergström A Ehinger B Retinal cell transplants Prog Retinal Eye Res 1995 15 197 230 10.1016/1350-9462(95)00010-0
Lander ES Array of hope Nat Genet 1999 21 3 4 9915492 10.1038/4427
Brown PO Botstein D Exploring the new world of the genome with DNA microarrays Nat Genet 1999 21 33 37 9915498 10.1038/4462
Hacia JG Resequencing and mutational analysis using oligonucleotide microarrays Nat Genet 1999 21 42 47 9915500 10.1038/4469
Lipshutz RJ Fodor SP Gingeras TR Lockhart DJ High density synthetic oligonucleotide arrays Nat Genet 1999 21 20 24 9915496 10.1038/4447
Shoemaker DD Schadt EE Armour CD He YD Garrett-Engele P McDonagh PD Loerch PM Leonardson A Lum PY Cavet G Wu LF Altschuler SJ Edwards S King J Tsang JS Schimmack G Schelter JM Koch J Ziman M Marton MJ Li B Cundiff P Ward T Castle J Krolewski M Meyer MR Mao M Burchard J Kidd MJ Dai H Phillips JW Linsley PS Stoughton R Scherer S Boguski MS Experimental annotation of the human genome using microarray technology Nature 2001 409 922 927 11237012 10.1038/35057141
Chowers I Kim Y Farkas RH Gunatilaka TL Hackam AS Campochiaro PA Finnemann SC Zack DJ Changes in retinal pigment epithelial gene expression induced by rod outer segment uptake Invest Ophthalmol Vis Sci 2004 45 2098 2106 15223782 10.1167/iovs.03-0863
Dunn KC Aotaki-Keen AE Putkey FR Hjelmeland LM ARPE-19, a human retinal pigment epithelial cell line with differentiated properties Exp Eye Res 1996 62 155 169 8698076 10.1006/exer.1996.0020
Venter JC Adams MD Myers EW Li PW Mural RJ Sutton GG Smith HO Yandell M Evans CA Holt RA Gocayne JD Amanatides P Ballew RM Huson DH Wortman JR Zhang Q Kodira CD Zheng XH Chen L Skupski M Subramanian G Thomas PD Zhang J Gabor Miklos GL Nelson C Broder S Clark AG Nadeau J McKusick VA Zinder N Levine AJ Roberts RJ Simon M Slayman C Hunkapiller M Bolanos R Delcher A Dew I Fasulo D Flanigan M Florea L Halpern A Hannenhalli S Kravitz S Levy S Mobarry C Reinert K Remington K Abu-Threideh J Beasley E Biddick K Bonazzi V Brandon R Cargill M Chandramouliswaran I Charlab R Chaturvedi K Deng Z Di FV Dunn P Eilbeck K Evangelista C Gabrielian AE Gan W Ge W Gong F Gu Z Guan P Heiman TJ Higgins ME Ji RR Ke Z Ketchum KA Lai Z Lei Y Li Z Li J Liang Y Lin X Lu F Merkulov GV Milshina N Moore HM Naik AK Narayan VA Neelam B Nusskern D Rusch DB Salzberg S Shao W Shue B Sun J Wang Z Wang A Wang X Wang J Wei M Wides R Xiao C Yan C Yao A Ye J Zhan M Zhang W Zhang H Zhao Q Zheng L Zhong F Zhong Zhu S Zhao S Gilbert D Baumhueter S Spier G Carter C Cravchik A Woodage T Ali F An H Awe A Baldwin D Baden H Barnstead M Barrow I Beeson K Busam D Carver A Center A Cheng ML Curry L Danaher S Davenport L Desilets R Dietz S Dodson K Doup L Ferriera S Garg N Gluecksmann A Hart B Haynes J Haynes C Heiner C Hladun S Hostin D Houck J Howland T Ibegwam C Johnson J Kalush F Kline L Koduru S Love A Mann F May D McCawley S McIntosh T McMullen I Moy M Moy L Murphy B Nelson K Pfannkoch C Pratts E Puri V Qureshi H Reardon M Rodriguez R Rogers YH Romblad D Ruhfel B Scott R Sitter C Smallwood M Stewart E Strong R Suh E Thomas R Tint NN Tse S Vech C Wang G Wetter J Williams S Williams M Windsor S Winn-Deen E Wolfe K Zaveri J Zaveri K Abril JF Guigo R Campbell MJ Sjolander KV Karlak B Kejariwal A Mi H Lazareva B Hatton T Narechania A Diemer K Muruganujan A Guo N Sato S Bafna V Istrail S Lippert R Schwartz R Walenz B Yooseph S Allen D Basu A Baxendale J Blick L Caminha M Carnes-Stine J Caulk P Chiang YH Coyne M Dahlke C Mays A Dombroski M Donnelly M Ely D Esparham S Fosler C Gire H Glanowski S Glasser K Glodek A Gorokhov M Graham K Gropman B Harris M Heil J Henderson S Hoover J Jennings D Jordan C Jordan J Kasha J Kagan L Kraft C Levitsky A Lewis M Liu X Lopez J Ma D Majoros W McDaniel J Murphy S Newman M Nguyen T Nguyen N Nodell M The sequence of the human genome Science 2001 291 1304 1351 11181995 10.1126/science.1058040
Fields S The future is function Nat Genet 1997 15 325 327 9090371 10.1038/ng0497-325
Lander ES The new genomics: global views of biology Science 1996 274 536 539 8928008 10.1126/science.274.5287.536
|
16262907
|
PMC1291373
|
CC BY
|
2021-01-04 16:03:50
|
no
|
BMC Ophthalmol. 2005 Nov 1; 5:25
|
utf-8
|
BMC Ophthalmol
| 2,005 |
10.1186/1471-2415-5-25
|
oa_comm
|
==== Front
BMC OphthalmolBMC Ophthalmology1471-2415BioMed Central London 1471-2415-5-261626291210.1186/1471-2415-5-26Research ArticleIndications and outcome of repeat penetrating keratoplasty in India Vanathi M [email protected] Namrata [email protected] Rajesh [email protected] Radhika [email protected] Jeewan S [email protected] Rasik B [email protected] Rajendra Prasad Centre For Ophthalmic Sciences, All India Institute of Medical Sciences, New Delhi, India2005 2 11 2005 5 26 26 27 12 2004 2 11 2005 Copyright © 2005 Vanathi et al; licensee BioMed Central Ltd.This is an Open Access article distributed under the terms of the Creative Commons Attribution License (), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
Background
Repeat penetrating keratoplasty is quite often required as there is high chance of failure of the primary graft particularly in the developing world. We planned a study to analyze the indications and outcome of repeat penetrating keratoplasty in a tertiary care centre in India.
Methods
A retrospective analysis of all the patients who underwent repeat penetrating keratoplasty, between January 1999 and December 2001 was performed. The parameters evaluated were indication for the primary penetrating keratoplasty, causes of failure of the previous graft, and final visual outcome and clarity of the repeat corneal grafts.
Results
Of fifty-three eyes of 50 patients with repeat penetrating keratoplasty (three patients underwent bilateral corneal regrafts), 37 eyes had undergone one regraft each, 14 eyes two regrafts and two eyes had three regrafts. The follow-up of the patients ranged from one to three years. The most common primary etiologic diagnosis was vascularized corneal scars (66%), of which the scars related to infection were most common (68.5%). Twenty-eight regrafts (52.8%) remained clear at a mean follow-up of 1.54 ± 0.68 years, of which 25 were single regrafts (89.3%). The commonest cause of failure of regraft was infection to the corneal graft (recurrence of herpetic infection in 9 eyes and perforated graft ulcers in 3 eyes). Three (18.6%) of the 16 eyes with multiple corneal regrafts achieved a BCVA of 6/60. Overall, only five eyes (all with single regraft) achieved a BCVA of 6/18 or better at the end of follow-up.
Conclusion
Graft infection is the leading cause of failure of repeat keratoplasty in this part of the world. Prognosis for visual recovery and graft survival is worse in eyes undergoing multiple regrafts.
==== Body
Background
Corneal graft failure constitutes a common indication for penetrating keratoplasty [1-20]. There are a few studies in literature which have reported the indications and outcome of corneal regrafts [21-27]. The primary indications for repeat corneal transplantation show a changing trend. Aphakic and pseudophakic bullous keratopathy were the common primary indications for regrafts in developed countries in previous studies [22-26], others being, infectious keratitis [23] and corneal dystrophies [22,25]. However, a more recent study on the profile of repeat keratoplasty [27] identified vascularized corneal scar as the most common primary indication for corneal regrafting. The paucity of studies on repeat keratoplasty from developing countries prompted us to evaluate the indications and outcome of repeat keratoplasty at a tertiary eye care referral centre.
Methods
We retrospectively reviewed the records of 50 patients of corneal regrafts performed at the Cornea services of Rajendra Prasad Centre for Ophthalmic Sciences, All India Institute of Medical Sciences, New Delhi, during the period from January 1999 to December 2001. Of all these 50 patients, 32 patients (64%) were from rural and peri-urban areas and 18 (36%) were from urban areas.
The parameters evaluated were the patient's age, gender, indication for the primary corneal transplan tation, duration of follow-up, number of regrafts, associated procedures performed and outcome of corneal regrafts. More than one corneal regraft was considered as a multiple regraft. Details of previous grafts in cases of multiple regrafts were collected from old records wherever available. Graft outcome was defined in terms of the clarity over a period of time till last follow-up or graft failure, whichever was earlier. Graft was considered to be clear if the clarity was of grade 3 or grade 4. By grade 4 clarity, we mean an absolutely clear graft with good visualization of the iris details behind it. By grade 3 clarity, we mean a graft with minimal haze but still with reasonably good visualization of the iris details.
Allograft rejection was diagnosed by the presence of endothelial or epithelial rejection line or both and corneal edema with anterior chamber reaction. Corneal graft failure was diagnosed when irreversible graft edema was present with or without vascularization or scarring of the graft. Intraocular pressure greater than 21 mm Hg on two separate occasions was taken as secondary glaucoma.
Postoperatively patients were prescribed topical betamethasone sodium phosphate 0.1% and ciprofloxacin 0.3% QID each and ocular lubricants (polyvinyl alcohol) QID. Patients with healed herpetic keratitis were given oral acyclovir 400 mg twice a day for 1 year after keratoplasty. The follow up schedule after the surgery was daily from day 1 till epithelial healing, 1 week, 1 month, 3 months, 6 months, 1 year and yearly thereafter. The duration of follow-up was taken as the time between the last regraft and the final follow-up. The outcome of the surgery was analyzed statistically using paired 't' test and p-value smaller than 0.05 was considered as statistically significant.
Results
Sixty eight patients had undergone corneal regrafts at our centre during the study period. Of these, 50 patients (73.5 %) (41 males and 9 females) had a regular follow-up with us of which, 3 patients had undergone bilateral corneal regrafts and these were included in the study (N = 53 eyes). The primary keratoplasty was performed at our centre in 39 eyes of 36 patients and 14 patients were referred from other centres after failure of the primary graft.
The mean age of the patients at the time of repeat penetrating keratoplasty was 45.2 ± 16.5 years. The mean follow-up after regraft was 1.54 ± 0.68 years. Of a total of 53 eyes, 37 eyes had one corneal regraft, 14 had two corneal regrafts and two eyes had undergone three regrafts each (i.e. multiple regrafts in 16 eyes). Forty-eight eyes had undergone other associated intraocular procedures such as goniosynechiolysis (32 eyes), iridectomy with pupilloplasty (7 eyes), cataract extraction (7 eyes) and anterior vitrectomy (11 eyes) at the time of regraft.
The most common indication for primary penetrating keratoplasty in these eyes was vascularized corneal scars (35 of the 53 eyes; 66%) followed by perforations secondary to microbial keratitis (6/ 53; 11.3%) (Table 1). The most common cause of vascularized corneal scars was healed microbial keratitis (non-herpetic) (37.1%), followed by herpetic keratitis (31.4%) (Table 1). The amount of vascularization was variable. Fifteen eyes showed a single quadrant deep vascularization along with one quadrant of superficial vascularization; eleven eyes showed one quadrant deep vascularization and 2 quadrants of superficial vascularization; 8 eyes had 2 quadrants of deep vascularization and 2 quadrants of superficial vascularization, and 1 eye had 3 quadrants of deep and 3 quadrants of superficial corneal vascularization.
The failure of primary graft was attributable to poor ocular surface in 18 eyes (33.9%), recurrence of herpetic keratitis in 8 eyes (15%), perforated graft ulcers in 4 (7.5%), scarring due to graft infection in 4 (7.5%), allograft rejection in 7 eyes (13.2%), endothelial decompensation in 8 eyes (15 %) and raised intraocular pressure in 4 eyes (7.5%). Since a high proportion of the grafts failed due to poor ocular surface, all these eyes were put on intensive (1 hourly) preservative free lubricants and repeat grafting was performed only after these eyes attained a reasonably good ocular surface. Two eyes with failed primary graft and ocular surface problems required entropion surgery and in two eyes permanent punctual plugs were inserted before regraft.
After repeat graft, these eyes were prescribed 1 hourly preservative free lubricant for 3 months and QID later along with topical chloramphenicol and topical dexamethasone QID each for 3 months and BD each later.
Of the 53 eyes with regrafts, 28 eyes (52.8%) had clear grafts at the end of follow up of which, 25 eyes (89.3%) had undergone single regraft and 3 eyes had multiple regrafts (Table 2). The reasons for failure in the remaining 25 eyes (47.2 %) was recurrence of herpetic infection in 9 eyes (36%), uncontrolled glaucoma in 5 eyes (20%), allograft rejection in 4 eyes (16%), perforated graft ulcers in 3 eyes (12%), endothelial decompensation in 2 eyes (8%), poor ocular surface in 2 eyes (8%). The survival of the regrafts has been depicted in Figure 1.
The pre-operative visual acuity ranged from light perception to 1/60. The best corrected visual acuity of the 28 clear grafts ranged from 4/60 to 6/9. Only five eyes (9.4%) achieved a BCVA of 6/18 or better at the end of 1 year after re graft, of which all had single regrafts (Table 3). In twenty-five eyes with failed repeat grafts, visual acuity ranged from light perception to 1/60.
Causes of suboptimal visual outcome (post-refraction) in regrafts that remained clear (grade 3/4) was poor ocular surface (9 eyes), post penetrating keratoplasty astigmatism (6 eyes), macular scarring (5 eyes), treated graft rejection (4 eyes) and post-penetrating keratoplasty glaucoma (4 eyes).
Discussion
Repeat corneal grafts remain a drain on existing resources for corneal transplantation and the rise in the number of regrafts parallels the rise in number of primary keratoplasties. Several studies [1-20] on the indications for penetrating keratoplasty have cited variable figures comprising the proportion of regrafts varying from 6.6% to 41%. On the basis of these studies it is apparent that corneal regrafts show an increasing trend among indications for penetrating keratoplasty. Dandona et al [19] studied the indications for penetrating keratoplasty in India and reported regrafts as the second most common indication for penetrating keratoplasty in India (17.1%), next only to corneal scarring of varied etiology.
Different success rates and visual outcome in corneal regrafts have been reported in literature [21-27]. Reported percentages of clear regrafts vary from 51% to 74% in the earlier studies [21,25-27]. In the present study the outcome of repeat penetrating keratoplasty in terms of graft clarity was comparable (52.8%).
Eyes in which repeat keratoplasty was performed necessitated multiple intraocular manipulations like goniosynechiolysis, iridectomy with pupilloplasty, cataract extraction and anterior vitrectomy. Post-keratoplasty glaucoma resulting due to these manipulations is an important factor that can result in graft failure. In the present study, a higher proportion of eyes developed secondary glaucoma (5 out of 25) resulting in graft failure in regrafts in comparison to primary keratoplasty (4 out of 53). This perhaps can be attributed to the fact that repeated surgeries result in greater development of anterior chamber reaction and multiple synechiae, and multiple intraocular manipulations required in regrafts can further be the compounding factor for the same.
Patients who undergo repeat corneal grafts carry the risk of developing variable amount of corneal neovascularisation as was seen in our cases. Corneal neovascularisation is an independent risk factor that can jeopardize the outcome of a successfully performed keratoplasty by causing episodes of graft rejection. However, in contrast to the studies from the developed countries which report graft rejection and recurrence of dystrophies as the main causes for failure of regrafts [22-25], our study highlights that higher prevalence of graft infection including recurrence of previous infection and secondary glaucoma are the leading causes of failure of repeat grafts. Ocular surface problem was the leading cause of failure of primary graft in this study. These eyes were intensively treated with preservative free lubricants to improve the ocular surface before performing regraft. This improved graft survival in these eyes and very few regrafts failed due to ocular surface problems. However, some amount of ocular surface problem persisted in some eyes resulting in mild haze and hence remained the leading cause for suboptimal best corrected visual acuity in these regrafts in spite of reasonably good graft clarity.
Recent studies [25-27] report a visual acuity of 20/40 or better in only 15% to 41% of clear regrafts. Similarly, the present study reports that among the 28 eyes with clear regrafts, a BCVA of 6/18 or better was seen in five eyes only (17.9%) and less than 6/18 in 23 eyes (82.1%). Contrary to few studies [22,25,26] reporting that visual prognosis in multiple regrafts to be comparable to that in single regrafts (these regrafts had been performed mainly for pseudophakic bullous keratopathy), Bersudsky et al [27] had concluded that graft survival and visual outcome is inversely proportional to the number of corneal regrafts. Similarly our study also demonstrates a poorer visual outcome in eyes undergoing multiple regrafts (Table 3). Though the number of eyes with multiple regrafts was small in our study, 13 eyes (81.3%) out of 16 eyes with multiple regrafts failed in our study suggesting that prognosis for graft survival in multiple regrafts seems to be poorer. Owing to the poor graft survival, permanent keratoprosthesis may be a suitable alternative in eyes with multiple regrafts.
Conclusion
The study demonstrates that visual prognosis for multiple corneal regrafting is suboptimal, owing to high chance of graft infection, secondary glaucoma and allograft rejection. However, repeat penetrating keratoplasty should be considered when required depending upon patient's need and motivation and in the absence of any contraindication.
Competing interests
The author(s) declare that they have no competing interests.
Authors' contributions
MV performed data collection, NS designed the study, RS wrote the manuscript, RT followed up and evaluated the patients, JST and RBV performed the surgeries. All authors read and approved the manuscript.
Pre-publication history
The pre-publication history for this paper can be accessed here:
Figures and Tables
Figure 1 Corneal graft survival.
Table 1 Indications for primary corneal transplantation in corneal regrafts
Indications No of eyes (%) (n = 53) Single regraft (n = 37) Multiple Regraft (n = 16)
1 Vascularised corneal scars 35 (66) 22 13
Healed infectious keratitis (non-herpetic) 13 (37.1) 9 4
Herpetic scarring 11 (31.4) 3 8
Trachomatous 5 (14.3) 5 0
Trauma 4 (11.4) 4 0
Steven Johnson's syndrome 1 (2.9) 1 0
Aniridia 1 (2.9) 0 1
2 Infectious keratitis 6 (11.3) 5 1
3 Aphakic Bullous Keratopathy 4 (7.5) 3 1
4 Pseudophakic Bullous Keratopathy 3 (5.7) 3 0
5 Corneal dystrophy 4 (7.5) 3 1
6 Congenital glaucoma 1 (1.9) 1 0
Table 2 Outcome of corneal regrafting surgery
Multiple regrafts
No of eyes (n = 53) 1 regraft (n = 37) 2 regraft (n = 14) 3 regraft (n = 2)
Clear regrafts (n = 28) 25 3 0
Failed regrafts (n = 25) 12 11 2
p value (Paired 't' test) 0.001 (statistically significant) 0.006 (not significant)
Table 3 Visual Acuity of single Vs multiple regrafts at last follow up
Clear regrafts (n = 28) BCVA 6/18 or better (n = 5) BCVA Less than 6/18 (n = 23)
Single regrafts (n = 25) 5 20
Multiple regrafts (n = 3) 0 3
==== Refs
Arentsen J Morgan B Green WR Changing indications for keratoplasty Am J Ophthalmol 1976 81 313 318 769559
Robin JB Gindi JJ Koh K Schanzlin DJ Rao NA York KK Smith RE An update for indications of keratoplasty. 1979 through 1983 Arch Ophthalmol 1986 104 87 89 3510613
Morris RJ Bates AK Changing indications for keratoplasty Eye 1989 3 455 459 2606220
Mohamadi P McDonnell JM Irvine JA McDonnel PJ Rao N Smith RE Changing indications for penetrating keratolasty, 1984–88 Am J Ophthalmol 1989 107 550 552 2653047
Brady SE Rapuano CJ Arentsen JJ Cohen EJ Laibson PR Clinical indications for the procedures associated with penetrating keratoplasty, 1983–88 Am J Ophthalmol 1989 108 118 122 2667369
Damji KF Rootman J White VA Richards JS Changing indications for penetrating keratoplasty in Vancouver, 1978–87 Can J Ophthalmol 1990 25 243 248 2207870
Lindquist TD McGlothan JS Rotkis WM Chandler JW Indications for penetrating keratoplasty: 1980–1988 Cornea 1991 10 210 216 2055026
Hyman L Wittpen J Yang C Indications and techniques for penetrating keratoplasty, 1985–1988 Cornea 1992 11 573 576 1468221
Mamalis N Anderson CW Kreisler KR Lundergan MK Olson RJ Changing trends in the indications for penetrating keratoplasty Arch Ophthalmol 1992 110 1409 1411 1417539
Vail A Gore SM Bradley BA Easty Dl Rogers CA Corneal transplantation in the United Kingdon and Republic of Ireland Br J Ophthalmol 1993 77 650 656 8218035
Williams KA Muehlberg SM Wing SJ Coster DJ The Australian Corneal Graft Registry: 1990 to 1992 Report Aust N Z J Ophthalmol 1993 21 1 48 8333942
Price FW Whitson WE Collins KS Marks RG Five-year corneal graft survival: large, single-center patient cohort Arch Ophthalmol 1993 111 799 805 8512481
Haamann P Jensen OM Schmidt P Changing indications for penetrating keratoplasty Acta Ophthalmol 1994 72 443 446 7825409
Flowers CW Chang KY McLeod SD Irvine JA McDonnell PJ Rao N Smith RE Changing indications for penetrating keratoplasty 1989–1993 Cornea 1995 14 583 588 8575177
Liu E Slomovic AR Indications for penetrating keratoplasty in Canada, 1986–1995 Cornea 1997 16 414 419 9220238
Lois N Kowal VO Cohen EJ Rapuano CJ Gault JA Raber IM Laibson PR Indications for penetrating keratoplasty and associated procedures, 1989–1995 Cornea 1997 6 623 629
Frucht-Pery J Shtibel H Solomon A Siganos CS Yassur Y Pe'er J Thirty years of penetrating keratoplasty in Israel Cornea 1997 16 16 20 8985628
Wong TY Chan C Lim L Lim TH Tan DT Changing indications for penetrating keratoplasty: a newly developed country's experience Aust N Z J Ophthalmol 1997 25 145 150 9267601
Dandona L Ragu K Janarthanan M Naduvilath TJ Shenoy R Rao GN Indications for penetrating keratoplasty in India Indian J Ophthalmol 1997 45 163 168 9475018
Chen WL Hu FR Wang IJ Changing indications for penetrating keratoplasty in Taiwan from 1987 to 1999 Cornea 2001 20 141 144 11248815 10.1097/00003226-200103000-00004
Cowden KJ Kaufman HE Polack FM The prognosis of keratoplasty after previous graft failures Am J Ophthalmol 1974 78 523 555 4607225
Robinson CH Indications, complications and prognosis for repeat penetrating keratoplasty Ophthalmic Surg 1979 10 27 34 384314
Insler MS Pechous B Visual results in repeat penetrating keratoplasty Am J Ophthalmol 1986 102 371 375 3529972
MacEwen Cj Khan Zuh Anderson E Macewen CG Corneal re-graft; indications and outcome Ophthalmic Surg 1988 19 706 712 3057403
Rapuano CJ Cohen EJ Brady SE Arentsen JJ Laibson PR Indications for and outcomes of repeat penetrating keratoplasty Am J Ophthalmol 1990 109 689 695 2346198
Patel NP Kim T Rapuano CJ Cohen EJ Laibson PR Indications for and outcomes of repeat penetrating keratoplasty, 1989–1995 Ophthalmology 2000 107 719 724 10768334 10.1016/S0161-6420(00)00003-8
Bersudsky V Blum-Hareuveni T Rehany U Rumelt S The profile of Repeated Corneal Transplantation Ophthalmology 2001 108 461 469 11237899 10.1016/S0161-6420(00)00544-3
|
16262912
|
PMC1291374
|
CC BY
|
2021-01-04 16:03:50
|
no
|
BMC Ophthalmol. 2005 Nov 2; 5:26
|
utf-8
|
BMC Ophthalmol
| 2,005 |
10.1186/1471-2415-5-26
|
oa_comm
|
==== Front
BMC PhysiolBMC Physiology1472-6793BioMed Central London 1472-6793-5-171627115410.1186/1472-6793-5-17Research ArticleShort-term administration of growth hormone (GH) lowers blood pressure by activating eNOS/nitric oxide (NO)-pathway in male hypophysectomized (Hx) rats Nyström Henrik C [email protected] Natalia [email protected] Kenneth [email protected]öm Göran [email protected] Anna [email protected] Department of Physiology, Institute of Physiology & Pharmacology, The Sahlgrenska Academy, Göteborg University, P.O. Box 432, SE-405 30 Göteborg, Sweden2 Department of Clinical Physiology, Sahlgrenska University Hospital/SU, SE-413 45 Göteborg, Sweden2005 7 11 2005 5 17 17 18 5 2005 7 11 2005 Copyright © 2005 Nyström et al; licensee BioMed Central Ltd.This is an Open Access article distributed under the terms of the Creative Commons Attribution License (), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
Background
The aim of the study was to evaluate the acute and continuous (up to 14 days of treatment) effect of growth hormone (GH) on blood pressure (BP) regulation and to investigate the interplay between GH, nitric oxide (NO) and BP.
In un-supplemented and GH supplemented hypophysectomized (Hx) male rats as well as intact rats, continuous resting mean arterial blood pressure (MAP) was measured using telemetry. Baroreceptor activity and the influences of NO on BP control were assessed during telemetric measurement. Furthermore, basal plasma and urine nitrate levels and aortic endothelial nitric oxide synthase (eNOS) expression were analysed. Endothelial function as well as vascular structure in the hindquarter vascular bed was estimated using an in vivo constant-flow preparation.
Results
Hypophysectomy was associated with decreased MAP (Hx: 83 ± 3 vs Intact: 98 ± 6 mmHg, p < 0.05) and heart rate (HR) (Hx: 291 ± 4 vs Intact: 351 ± 7 beat/min, p < 0.05). Endothelial dysfunction and reduced vasculature mass in the hindquarter vascular bed was found in Hx rats. GH substitution caused a further transient decrease in MAP and a transient increase in HR (14% and 16% respectively, p < 0.05). The reduction in MAP appeared to be NO dependent. Aortic eNOS expression was unchanged. GH substitution resulted in an impaired baroreceptor function. Two weeks of GH treatment did not normalise the BP, vascular structure and the endothelial function in the resistance vessels.
Conclusion
GH substitution seems to have a short lasting effect on lowering blood pressure via activation of the NO-system. An interaction between GH, NO-system and BP regulation can be demonstrated.
==== Body
Background
Adult hypopituitarism and untreated growth hormone (GH) deficiency (GHD) is associated with endothelial dysfunction [1], decreased systemic formation of nitric oxide (NO) [2] and increased risk for cardiovascular morbidity and mortality [3] as compared to the normal population [4]. There are disparate reports on the effect on blood pressure (BP) in this condition; different studies have found normotension, hypotension or hypertension in GHD-patients (for review see [5]). Growth hormone replacement has been shown to reverse many of the adverse cardiovascular risk factors associated with GHD, including decreased systolic and diastolic BP, decreased vascular resistance [6], increased cardiac output (CO), increased heart rate (HR) [2,6], improved endothelial function [1,7] and increased NO formation [2].
Hypophysectomized (Hx) rats express decreased systolic BP and mean arterial blood pressure (MAP), CO and HR compared with intact rats [8,9]. As in human adult GHD, these rats show endothelial dysfunction and impaired vascular reactivity [10,11], while aortic endothelial nitric oxide synthase (eNOS) expression is not affected [12]. Furthermore, these rats also display decreased heart weights [12,13], and reduced vasculature mass in the muscle vascular bed suggesting an hypotropic remodelling of the cardiovascular system [13]. One week of GH substitution in Hx rats does not result in a change in systolic BP measured by the tail-cuff technique, whereas HR, heart weight and aortic eNOS expression are increased [12]. The resistance vessels (mesenteric artery) still demonstrate endothelial dysfunction [10]. Previous studies on BP regulation in Hx rats [10,12,13] have used less precise techniques to measure BP i.e. tail-cuff technique and therefore showed imprecise GH effect on acute and continuous BP regulation.
In the present study we aimed to investigate the acute and continuous (up to 14 days of treatment) effect of growth hormone substitution on blood pressure regulation. In order to measure the true, unstressed BP, we used a telemetry technique with implantable transmitters. We were also interested in investigating the interplay between GH, NO and BP. In order to do this we used a specific NOS blocker and measured the aortic eNOS expression, basal plasma and urine nitrate levels. In addition, baroreceptor activity was examined both early and late after onset of GH treatment. At the end of the study, endothelial function and the vascular structural properties in the hindquarter were studied using an in vivo constant-flow preparation.
Results
Growth hormone causes a transient decrease in mean arterial blood pressure and a transient increase in heart rate, Fig 1 and 2
Hypophysectomy per se, caused a 17% decrease in MAP and HR before onset of hormonal treatment (MAP; Intact: 98 ± 6, Hx: 83 ± 3, GH: 81 ± 5, p < 0.05 Intact vs GH or Hx, Fig 1, HR; Intact: 351 ± 7, Hx: 291 ± 4, GH: 295 ± 7, p < 0.05 Intact vs GH or Hx, Fig 2). Neither MAP nor HR changed when the Hx rats were treated with [T4+GC] compared with pre-treatment levels or with intact rats (day -2, Fig 1 and 2). Growth hormone caused an immediate drop in MAP by approx 14% (to 70 ± 4 mmHg, p < 0.05 vs Hx, Fig 1) with a concomitant increase in HR by approx. 16% (to 351 ± 6 beat/min, p < 0.05 vs Hx, Fig 2). After 14 days of GH therapy, the MAP returned to pre-treatment levels (day 14 = 81 ± 4 mmHg, Fig 1), indicating that GH had a transient effect on BP regulation. Similarly, HR returned to pre-treatment levels (pre-treatment = 319 ± 7 vs day 14 = 319 ± 8 beat/min, Fig 2). Mean arterial pressure was unchanged in both Hx and intact rats during the treatment period (Hx: 81 ± 3 to 83 ± 6 mmHg, Intact: 98 ± 6 to 95 ± 6 mmHg). In addition, no change in HR was detected in Hx and intact rats during the treatment period (Hx:291 ± 4 to 290 ± 6 beat/min, Intact: 351 ± 7 to 368 ± 7 beat/min). At day 14, the HR, but not MAP, was significantly changed between the groups (p < 0.05 Hx vs GH, p < 0.05 GH vs Intact and p < 0.05 Hx vs Intact).
Growth hormone decreasing effect on blood pressure was NO dependent, Fig 3 and 4
Early L-NAME treatment
Both intact and Hx rats had a similar BP response to L-NAME with an increase of approx 45% (Hx from 82 ± 4 to 120 ± 7 mmHg, Intact from 99 ± 6 to 142 ± 8 mmHg, Fig 3). Growth hormone treatment caused a significantly greater BP response of approx 56% (GH from 71 ± 4 to 110 ± 6 mmHg, p < 0.05 vs. Intactand Hx, Fig 3).
Late L-NAME treatment
There was no difference in the response of MAP to L-NAME between the groups (Intact from 91 ± 6 to 144 ± 9 mmHg, Hx from 76 ± 5 to 126 ± 5 mmHg, GH from 74 ± 4 to 120 ± 9 mmHg, Fig 4).
Aortic eNOS expression was unaffected after 14 days of GH substitution
Hypophysectomy per se, had no effect on aortic eNOS protein levels (Intact: 121 ± 31 vs Hx:91 ± 26%). There was no significant change in aortic eNOS expression after GH treatment compared with both Hx and Intact rats (GH: 110 ± 16 %).
Growth hormone effects on basal urine and plasma nitrate and nitrate clearance
The urine nitrate was lower, but not significantly decreased in the Hx group compared to GH and intact (Hx: 216.16 ± 72.23, Intact: 400.15 ± 94.65, GH: 641.1 ± 208.7 μM/ml/100 g BW n.s.). Both Hx and GH rats had significantly higher levels of plasma nitrate compared with intact rats (Hx: 79.4 ± 5.9, GH: 78.2 ± 5.7, Intact: 46.4 ± 2.3 μM, p < 0.05 between Hx and intact and between GH and intact). The nitrate clearance was calculated in all groups, demonstrating that GH had an increased clearance compared with Hx rats, but similar with intact rats (Hx:1.8 ± 0.6, GH: 6.2 ± 2.0, Intact: 6.2 ± 1.5 μl/min/100 g, p < 0.05 between Hx and intact and between GH and intact).
Growth hormone caused a blunted baroreceptor activity, Fig 5
The baroreceptor activity, tested by phenylephrine, was similar in intact and Hx rats (Intact: -0.94 ± 2.29, Hx: -2.35 ± 0.69 bpm/mmHg, Fig 5). In contrast, GH resulted in a blunted baroreceptor activity (0.12 ± 0.82, p < 0.05 vs Hx, Fig 5). When the rats were pre-treated with L-NAME, none of the groups differed in baroreceptor activity (Intact: -2.41 ± 2.4, Hx:-2.23 ± 0.68, GH: -0.51 ± 0.40 bpm/mmHg).
Growth hormone increases body weight
The BW of the Hx group did not change over time, indicating a complete hypophysectomy (from arrival to the onset of treatment: Intact: 295 ± 11 to 391 ± 24 g, Hx: 251 ± 19 to 225 ± 13 g, GH: 256 ± 14 to 232 ± 15 g, p < 0.05 between Intact and Hx, Intact and GH at both time points). After 14 days of treatment, the [T4+GC] group did not change in BW compared to pre-treated weight (Hx: 232 ± 8 to 232 ± 7 g). The Intact group increased by approx. 2% in BW during this period (391 ± 24 to 399 ± 28 g, N.S.). The GH group increased in BW by 5% during the treatment period (from 256 ± 14 to 268 ± 19 g, p < 0.05).
Growth hormone caused a increased plasma IGF-I
Plasma IGF-I decreased in Hx rats by approx. 85% compared with Intact rats (Intact:1098 ± 46 vs Hx 164 ± 27 ng/ml, p < 0.05). There was a substantial increase in plasma IGF-I after 14 days of GH therapy by approx. 8-fold increase (Hx: 164 ± 27 vs GH: 736 ± 56 ng/ml, p < 0.05). Growth hormone treated rats displayed a significant decrease in plasma IGF-I compared with intact rats (p < 0.05).
Growth hormone effect on heart weight and structural parameters of the resistance vessels, Table 1 and Fig 6
The relative LV weight was not changed between Hx and intact animals. Surprisingly, there was a significant loss of weight in the LV after GH therapy by approx. 2% (p < 0.05 vs Hx rats, Table 1). The RV weight was not changed between the groups (Table 1).
The skeletal muscle vascular bed in Hx rats showed an increased sensitivity (left shift) to administered NA by approx. 22% compared with intact rats (Table 1), indicating an increased sensitivity of the vascular á-adrenoceptors. Treatment with GH resulted in a normalisation of the NA sensitivity (right shift) (Table 1). The perfusion resistance (PR) obtained at maximal dilation was decreased by approx. 19% in both GH and Hx group compared with intact rats (Intact: 1.94 ± 0.06, Hx: 1.16 ± 0.04, GH: 1.12 ± 0.08 mmHg/ml/min/100 g dry HQ, p < 0.05 vs intact and GH and Hx, Fig 6A). When vasoconstriction was induced by means of NA or with angiotensin II+phenylephrine, both Hx and GH group showed approx. 42% and 40%, respectively, decreased resistance response compared with intact rats (Table 1). Similar resistance changes were obtained in GH, Hx and intact rats after maximal constriction was induced by BaCl2 (Intact: 61 ± 7, Hx: 26 ± 5, GH: 31 ± 4 mmHg/ml/min/100g dry HQ, p < 0.05 Intact vs GH and Hx, Fig 6B). The average slope calculated from individual pressure-flow curves, also obtained during full vascular relaxation, showed similar magnitude of structural changes between Hx, GH and intact rats (Intact: 0.84 ± 0.04, Hx: 0.47 ± 0.02, GH: 0.44 ± 0.03 mmHg/ml/min/100 g dry HQ, p < 0.05 Intact vs GH and Hx, Fig 6C). Both the GH treated and Hx rats showed endothelial dysfunction compared with Intact rats, demonstrated by decreased dilation after a single dose of acetylcholine (Intact: 73 ± 1%, Hx: 65 ± 3%, GH: 61 ± 4%, Fig 6D).
Discussion
The major findings in this study are; (i) Hypophysectomy per se, caused a decrease in MAP and HR, endothelial dysfunction and reduced vasculature mass in the hindquarter vascular bed. (ii) Supplementation with [T4+GC] did not change any of the studied parameters. (iii) GH substitution resulted in a transient decrease in MAP and a transient increase in HR. (iv) GH substituted rats had an increased MAP response after administration of early L-NAME compared with intact and Hx rats, indicating a transient activation of NO by GH. (v) Long-term GH treatment resulted in an impaired baroreceptor activity. (vi) 14 days of GH therapy to Hx rats neither improved the endothelial function nor restored the vascular structure. Taken together, GH substitution seems to have a short lasting effect on lowering blood pressure via activation of the NO-system. The transient increase in HR is most likely mediated by a baroreceptor activation. These data also suggest that a longer period of GH therapy is required to improve endothelial function and to restore the morphology of the vasculature. Finally, there is interplay between GH, NO-system and blood pressure regulation.
The effect of GH on blood pressure
Growth hormone therapy to Hx rats caused an immediate and transient decrease in MAP. This reduction in MAP appears to be NO-dependent since L-NAME treatment caused a greater MAP response than it did in both Hx and intact rats. We speculate that this finding may reflect an increase in eNOS activity and/or expression, resulting in enhanced NO bioavailability. This may in turn lead to decreased total peripheral resistance [6,14], resulting in reduced MAP. Growth hormone and IGF-I has been shown to increase eNOS in vitro [15] as well as to increase eNOS expression [12] and NO formation [2]in vivo.
It has been suggested that GH treatment can be associated with an increase in extra cellular volume (ECV) [16-18]. An increased volume load may theoretically lead to an increase in BP. However, a previous published study showed that increased body sodium concentration and increased ECV after GH substitution to GHD patients were not associated with an increase in BP [18]. Taken together, this suggests that the load of ECV seems to be of minor importance for the BP regulation at least in GH treated GHD patients.
The GH activated NO-dependent decrease in MAP appears to be transient since 12 days of GH therapy, results in a similar increase in MAP after a single dose of L-NAME as it did in both intact and Hx animals. This result is supported by unchanged aortic eNOS expression in all groups and the return of MAP to pre-treatment levels in GH treated rats. The present study showed increases of urine nitrate after 12 days of GH treatment, which may suggest that there is still some GH stimulated NO production left, although this increase appears to be of minor importance for the BP regulation. The level of MAP in GH rats did not return to the level in intact rats. It is possible that the observed reduction of the resistance vessels mass in the hindquarters vascular bed could be the explanation. Folkow et al [13] have shown that 6 weeks of treatment with both GH and T4 results in normalisation of BP as well as an enhanced vascular mass in the hindquarter vascular bed. In GH over-expressed mice, increased vascular mass has also been described [19,20]. These mice displayed either hypertension [19] or normotension [20]. Thus, it seems that a longer supplementation period is required to give an effect of vascular morphology on the resistance vessels as well as to normalise BP. However, hGH treatment to Hx rats for longer periods than two to three weeks have been reported to result in antibodies against GH [21]. Guided by this information, we decided to limit the study to two weeks.
Growth hormone effect on heart rate and the baroreceptor activity
In the present study, GH substitution results in an immediate and transient increase in HR, which is partly in accordance with other studies [2,6]. This direct effect in HR can probably be explained by an increased sympathetic activity that could originate from increased firing of baroreceptors in response to decreased MAP. Two weeks of GH therapy resulted in a normalisation of HR. This transient change in HR might be explained by deactivation of baroreceptors due to increased BP. After twelve days of GH therapy, baroreceptor activity was blunted. The blunted effect of the baroreceptors can most likely be explained in two ways; by loss of structure or by "resetting". Both GH and Hx rats showed reduced vasculature mass in the hindquarter vascular bed, but only GH treated rats exhibit impaired function of the baroreceptors, suggesting that the blunted function of the baroreceptors could not be due to loss of structure in the baroreceptors in GH rats. Therefore, our data suggest that the impaired function of the baroreceptors seems to be related to "resetting". It has been suggested that NO appears to act as a sympatholytic agent to modulate the central sympathetic outflow [22]. The impaired baroreceptor activity appears not to be NO-dependent, since the blunted effect of the baroreceptors was unchanged after L-NAME treatment.
Growth hormone effect on adreno-receptor sensitivity and endothelial function
Growth hormone treated Hx rats showed a right shift and/or normalisation of the noradrenalin ED50 value in the hindquarters vascular bed, which is in accordance with other studies [11,13,19]. Long-term GH replacement in GHD patients has been shown to decrease sympathetic activity to the muscular vascular bed [23], whereas heart rate variability displays increased ratio of sympato-vagal tone, decreased vagal tone and increased sympathetic tone [24]. This might suggest that the rightward shift and/or normalisation of the noradrenalin ED50 value in the hindquarter vascular bed is caused by increased sympathetic tone and changed ratio between the sympato-vagal and the vagal tone.
It is well known that GHD patients exhibit endothelial dysfunction [1,25]. This phenomenon has been detected in both conduit [11] and resistance vessels in Hx rats [10]. Accordingly, in this study the resistance vessels in the hindquarter vascular bed also exhibit endothelial dysfunction in Hx rats. Neither one week [10] nor two weeks of GH therapy resulted in improved endothelial function in the resistance vessels. In contrast, both GH supplemented GHD patients [7] as well as GH treated Hx rats [11] show improved endothelial function in the conduit vessels. This suggests that longer treatment is required to enhance the endothelial function in the resistance vessels.
Conclusion
Hypophysectomy per se, is associated with decreased MAP, HR and loss of vascular structure. Growth hormone substitution to Hx rats results in a transient decrease in MAP and a transient increase in HR. This transient decrease of MAP was NO dependent. Growth hormone treatment caused an impaired baroreceptor activity. Two weeks of GH therapy neither improved the endothelial function nor restored vascular parameters. To summarize, GH substitution seems to have a short lasting effect on lowering blood pressure via activation of the NO-system. These data also suggest that a longer period of GH therapy is required to improve endothelial function and to restore the morphology of the vasculature. Finally, there is an interplay between GH, blood pressure and NO-system.
Methods
Animals
Male Wistar rats were obtained from M&B (Ejby, Denmark). The rats underwent hypophysectomy at approximately eight weeks of age (approx. 280–300 g) at M&B one week prior arrival to our facility. The protocol conformed to guidelines on the conduct of animal experiments issued by the Swedish National Board for Laboratory Animals and was approved by the Ethics Committee for Animal Experiments at Göteborg University. The animals were housed at constant temperature (20°C) at a relative humidity of 50–60%. A 12 h dark/light cycle was maintained with lights on/off at 07.00 AM to 7.00 PM. The rats had free access to standard pellet chow and tap water throughout the study. All rats were acclimatized for one week before the onset of the experiment. Body weight (BW) was measured throughout the study once a week.
Experimental protocol
Two different experimental protocols were used.
Substitution therapy
In both protocol 1 and 2 (see below), all Hx rats received thyroxine (T4, 20 μg/kg/day) and glucocorticoid (GC, 400 μg/kg/day) ([T4+GC]) treatment by mini-osmotic pumps s.c. (model 2004, Alza Pharmaceuticals, Palo Alto, CA, USA). After two days of [T4+GC] treatment, the Hx rats were divided into two groups of which one was treated with GH (1 mg/kg/day) by a second mini-osmotic pump s.c. (2ML2) for 14 days.
Protocol 1
Hypophysectomized rats and intact controls were equipped with telemetry transmitters for measurement of conscious unrestrained MAP and HR for 18 days (Intact n = 5, Hx n = 8, and GH n = 5). MAP and HR were measured two days before onset of treatment (Day -4 to -2) and continued throughout the start of substitution with [T4+GC] (Day -1 to 0) and treatment with GH (Day 1–14). In addition to these baseline measurements, two different experiments were performed; 1) the activity of the NO system was assessed by acute administration of L-NAME (early, Day 3–4) and (late, Day 11–12), 2) baroreceptor activity was assessed by acute administration of phenylephrine (Day 11–12). Two days of baseline measurements were recorded before onset of these two experiments. At the end of the 14 day protocol, rats were anesthetized, sacrificed and blood and tissues samples were taken for further analyses.
Protocol 2
Rats in the second protocol were treated identically as in protocol 1, except that we did not implant telemetry transmitters (Intact n = 14, Hx n = 12, and GH n = 12). In this protocol, two different experiments were performed; 1) plasma and urine nitrate concentrations were measured (Day 10–12), 2) at the end of the 14 day protocol, vascular structure as well as endothelial function in the hind-quarter vascular bed was assessed using an in vivo constant-flow preparation.
Experimental procedures
Implantation of radio-telemetric implants
During the high dose substitution therapy (see below), a radio telemetric transducer catheter (o.d. 0.76 mm, Data Science International, Inc, St. Paul, MN, USA) was implanted into the lower aorta and glued into position (3 M Vetbond™, 3 M Animal Care Products, St Paul, MN, USA) in Hx (n = 13) and control (n = 5) rats. Rats were anaesthetized using Ketalar:Rompun (39:5 mg/kg) and isoflurane (Baxter Healthcare, Chicago, MI, USA). The catheter tip was placed at least 1 cm below the renal arteries. The transmitter (TA11PA-C40) was secured to the abdominal wall and the abdomen closed with sutures. After four weeks of recovery and without hormonal supplementation, the animal in its home cage was placed on a receiver plate and the signal collected using the Dataquest LabPRO Acquisition System (Ver. 3.0, Data Sciences international, Inc, St. Paul, MN, USA). The following sampling parameters were used; sampling frequency 500 Hz, sample duration 15 sec., save period 5 min. The mean arterial signal was corrected for electronic offset, the average of one measurement outside the animal before and after implantation.
High dose substitution therapy to improve surgical survival
To be able to perform surgery on the vulnerable Hx rats, high-dose steroid therapy and salt (NaCl) enriched diet was given before and after surgery according to the following protocol. For five days all rats (Hx and intact rats) had free access to normal tap water and dexamethasone in tap water (orally, 1.5 mg/l). Rats also received standard pellet- and salt pellet chow (248 mmol/100 g, Lactamin, Vadstena, Sweden). The rats were also treated with glucocorticoid at day 1: 2 mg/kg × 2, day 2: 2 mg/kg × 3, day 3: (day of surgery): 2 mg/kg × 4, day 4: 2 mg/kg × 3 and day 5: 2 mg/kg × 2 (s.c.). This high dose regime was washed out during the following four weeks before onset of the experiment. During this period the rats received no hormonal treatment, but had free access to salt pellet and standard pellet chow. Using this regime 75% of Hx animals survived the surgical procedure.
NO dependency protocol
To test if the NO system was involved in the GH-dependent BP regulation, a NOS antagonist (L-NAME 15 mg/kg, s.c.) was given during two occasions, early (day 3–4) and late (day 11–12). Saline was used as a control and given in an equal volume (0.1 ml/kg, s.c.). L-NAME and saline were given in randomized order.
Test of baroreceptor activity
The baroreceptor activity was tested by using phenylephrine (300 μg/kg, s.c.) after pre-treatment with NaCl or L-NAME at day 11–12. Three hours after late L-NAME or saline were given; all rats received a single dose of phenylephrine to generate a slow gradual increase in MAP and a corresponding baroreceptor elicited decrease in HR within 60 seconds. L-NAME and NaCl were given in randomized order. To determine the optimal dose of phenylephrine, a dose response curve was established in separate experiments (data not shown).
End of experiment
On day 13–14 the rats were allowed to recover and baseline measurements were performed. At day 14, the rats were anesthetized using isoflurane and decapitated. Blood samples were taken for IGF-I analysis. The heart and the aorta were quickly excised. The heart was separated into left (LV) (including septum), and right ventricles (RV) and weighed. The aorta was trimmed free of fat and adherent tissues, frozen in liquid nitrogen and stored at -80°C for further analysis.
Measurements of IGF-I
In randomly selected plasma samples from protocol 1, the plasma IGF-I content was analyzed by a commercial RIA kit (Mediagnost, Reutlingen, Germany) in intact rats (n = 3), Hx rats (n = 5) and in GH (n = 5) treated rats [12].
Immunoblotting
Immunoblotting and protein extraction techniques have previously been described [12]. Briefly, protein was extracted from the aorta from intact rats (n = 5), Hx rats (n = 8) and GH (n = 5) treated Hx rats. The total protein concentration was determined by a commercial protein assay (Bio-Rad, Hercules, CA, USA). 25 μg of aorta and liver of total proteins were loaded in each lane on gels (10% NuPAGE® Bis-Tris gels; Novex, San Diego, CA, USA). The gels were run for 90 minutes at constant voltage (150 V). Molecular weight standards (See Blue®; Novex, San Diego, CA, USA) were used on each gel. The proteins were transferred to a polyvinyldifluoride (PVDF) membrane (Amersham, Buckinghamshire, UK). The membranes were incubated with a mouse monoclonal antibody against eNOS (dilution 1:1,000; Transduction Laboratories, Lexington, KY, USA). Immunoreactive protein was visualized by chemiluminescence using an alkaline phosphatase-conjugated secondary antibody (dilution 1:30,000; Sigma, St. Louis, MO, USA) and CDP-Star® (Tropix, Bredford, MA, USA) as a substrate. The membranes were exposed to ECL film (Amersham, Buckinghamshire, UK) at room temperature for 1–5 minutes and the films were subsequently developed. Semi quantitative measurements of proteins from the immunoblots were made by densitometry (Fluor-S™ Multimager, Quantity One ver. 4.1.0, Bio-Rad, Hercules, CA, USA). The optical density (OD) of each band was measured. Lane containing extract of liver was used as a reference on each gel.
Nitrate measurements
The animals were housed individually in metabolic cages. During the first day of measurement, the rats had free access to tap water and standard pellet chow. During the second and the third day in the metabolic cages, fasting was induced in the animals over night with free access to distillate water. This was in order to avoid interference of food and water on the nitrate measurement [26]. Water intake, urine volume and BW were measured. During the last day of measurements, the urine was collected for 24 hours, weighed and frozen for further nitrate analysis. At the end of the protocol, a blood sample of approximately 200 μl was drawn from the tail vein. The whole blood was centrifuged and plasma was collected and frozen for further analysis.
Plasma and urine nitrate analyses
Plasma and urine samples were analyzed for total nitrate concentration using a gas chromatography/mass spectrometric method as previously described [27]. Briefly, after the samples had been prepared, the samples were injected into a Varian 3400 gas chromatograph equipped with a 30 m BPX-5 capillary column operated with a temperature program (60–240°C). A Varian Saturn II mass spectrometer served as detector operating in the positive ion/chemical ionization mode using methane as the reactant gas and selective monitoring of mass equivalent m/z 124 for endogenous nitrate and mass equivalent m/z 125 for the 15N-labelled internal standard. Basal urine and plasma nitrate concentrations and nitrate clearance were calculated.
Hemodynamic analysis of resistance-vessel design
The structural characteristics of the skeletal muscle vasculature were analysed hemodynamically using a method which has been described in detail previously [13]. In brief, the isolated hind limbs of randomly treated and untreated rats were perfused in pairs via an aortic cannula. The perfusate consisted of an oxygenated 2% dextran-Tyrode solution, to which 0.5% bovine serum albumin (Sigma Chemicals, St Louis, MISS, USA) had been added. Perfusion pressure was measured via the cannulated tail artery. The hind limb vascular bed was initially maximally dilated by repeated injections of papaverine (3 mg totally), using constant flow of 10 ml/100 g hind limbs. Pressure-flow curves during maximal dilatation were constructed by altering the speed of the perfusion pump. During this constant flow condition, noradrenalin (NA) in increasing concentrations was added to the perfusate (20–64 μg). At maximal NA constriction, an injection of acetylcholine (0.01 mg, the correct dose was tested out by dose response curves, data not shown) was given. Finally, angiotensin II (200 ng), phenylephrine (4 mg) and BaCl2 (150 mg) were injected. Noradrenalin dose-resistance curves were constructed of hind limb vascular beds, and the ED50 value was calculated. After the experiment, the hindquarters were dried at 70°C for 48 h and weighed. The dry weight of the HQ was used to standardize the calculation of the hemodynamic parameters.
Hormones, solutions and drugs
The following hormones were used; Thyroxine (T4, L-thyroxine; Nycomed, Oslo, Norway); Glucocorticoid (GC, cortisol phosphate; Solu-Cortef, Upjohn, Puurs, Belgium); human Growth hormone (hGH, Pharmacia, Stockholm, Sweden). Thyroxine, glucocorticoids and hGH were dissolved in 0.9% NaCl.
The composition (in mM) of the 2% dextran Tyrode solution used in the infusion experiments was Na+ 148.0, Cl- 133.4, K+ 4.3, Ca2+ 2.5, Mg2+ 0.8, HCO3 25, H2PO4 0.5, D-glucose 5.6.
The following drugs were used: acetylcholine (Sigma); Angiotensin II (Sigma); Phenylephrine (Sigma); Noradrenalin (Sigma); Nω-nitro-L-arginine methyl ester (L-NAME, Sigma); Dexamethasone (Sigma); Papaverine (Sigma).
Statistical analysis
Values are expressed as MEAN ± SE. Statistical analyses of BW, plasma IGF-I, eNOS expression, basal urine and plasma nitrate and nitrate clearance, all HQ-data, BP and HR comparisons between the groups and percentages changes after L-NAME treatment were performed by one-way ANOVA followed by Tukey's as a post-hoc test. Hormonal effects on the BP and the HR was calculated by using paired t-test of mean of 24 h before onset of treatment compared to mean of 24 h after onset of treatment. L-NAME percentage effect on BP was calculated by comparing 10 values pre-treatment and 10 values post-treatment. Individual k-values was calculated when barorecteptor activity was studied.
Authors' contributions
HCN: participated in the design of the study, performed some statistical analysis, acquisition of data, analysis and interpretation of data, been involved in the drafting of the manuscript
NK: carried out the nitrate measurements
KC: involved in the drafting of the manuscript
GB: participated in the design of the study, involved in the drafting of the manuscript
AW: made substantial contribution to concept and design, analysis and interpretation of data, performed statistical analysis, acquisition of data, drafting the manuscript
All authors read and approved the final manuscript
Acknowledgements
The critical review of the manuscript by Dr. Yrsa Bergmann Sverrisdottir is highly appreciated. The authors thank Mrs Gunnel Andersson for excellent help with the telemetric probe implantation. This study was supported by the Swedish Research council (GB; 12 580, KC; 14 231), the Swedish National Heart and Lung foundation (GB, KC), the memory of Lars Hierta (AW), King Gustaf V:s and Queen Victoria foundation (AW), the Emelle foundation (AW), the Swedish Hypertension Society (AW, HN), the Swedish Society of Medicine (AW), the foundation of Tore Nilsson for Medical Research (AW), the foundation of Västra Götalands Regionen (KC) and the foundation of Eva and Oscar Ahrén (HN). The Swedish Heart and Lung foundation and SWEGENE contributed to the post-doctoral position of Anna Wickman.
Figures and Tables
Figure 1 Mean arterial blood pressure (MAP) in hypophysectomized (Hx, solid line), growth hormone supplemented Hx (GH, solid squares) and in intact rats (solid triangles). The first two days (-4 to -2) represent non-supplemented basal measurements of MAP in all groups. At day -2, all Hx rats received thyroxine and glucocorticoid [T4+GC] treatment with mini-osmotic pumps, whereas the intact rats underwent sham operation. At day zero, the onset of GH treatment (mini-osmotic pumps implantation) was performed in the GH animals, whereas both intact and Hx animals underwent sham operation. * indicate p<0.05 between pre- and post-GH treatment in GH group.
Figure 2 Heart rate (HR) in hypophysectomized (Hx, solid line), growth hormone supplemented Hx (GH, solid squares) and in intact rats (solid triangles). The first two days (-4 to -2) represent non-supplemented basal measurements of HR in all groups. At day -2, all Hx rats received thyroxine and glucocorticoid [T4+GC] treatment with mini-osmotic pumps, whereas the intact rats underwent sham operation. At day zero, the onset of GH treatment (mini-osmotic pumps implantation) was performed in the GH animals, whereas both intact and Hx animals underwent sham operation. * indicate p<0.05 between pre- and post-GH treatment in GH group.
Figure 3 Effect of a single dos of L-NAME (15 mg/kg, s.c.) on mean arterial blood pressure (MAP) in hypophysectomized (Hx, solid line), growth hormone supplemented Hx (GH, solid squares) and in intact rats (solid triangles). The L NAME injection was given early, at day three or four (see in method section for further information) after onset of GH treatment. *indicates p<0.05 between GH and intact or Hx animals.
Figure 4 Effect of a single dos of L-NAME (15 mg/kg, s.c.) on mean arterial blood pressure (MAP) in hypophysectomized (Hx, solid line), growth hormone supplemented Hx (GH, solid squares) and in intact rats (solid triangles). The L NAME injection was given late, at day eleven to twelve (see in method section for further information) after onset of GH treatment.
Figure 5 Baroreceptor activity was tested in hypophysectomized (Hx, open squares), growth hormone treated Hx rats (GH, solid squares) and in intact rats (solid triangles). Baroreceptor activity test was performed with a single dose of phenylephrine (4 mg/kg, s.c.). Mean arterial blood pressure (MAP) and heart rate (HR) were correlated to each other and the k-values were calculated in all groups. *indicate a significant difference of the slope between Hx and GH rats.
Figure 6 Graphs A-C showing the effect of hypophysectomy (Hx) and 14 days of growth hormone substitution to Hx rats (GH) on structural properties and acetylcholine response (D) in the skeletal vascular bed. (A) Graphs showing the perfusion resistance (PR) at maximal dilation (max dil), reflecting the average internal radius of the skeletal vasculature. (B) Graph showing the PR at maximal constriction (max con) obtained by barium chloride, reflecting average medial bulk of contractile tissue. (C) Graph showing slope, calculated by linear regression from individual pressure-flow curves during perfusion at maximal dilation. (D) Linear graph showing the resistance pre (during maximal noradrenaline constriction) and post (after a single dose of acetylcholine) in the skeletal vascular bed, reflecting the endothelial function. *indicate p < 0.05 between intact and GH or Hx rats.
Table 1 Effects of 14 days of administration of thyroxine (T4), glucocorticoids (GC) and growth hormone (GH) were studied in male hypophysectomized (Hx) rats. Intact rats were also used. The parameters that were studied were: body weight (BW), wet hindquarter (HQ weight), dry HQ, oedema, left (LV) and right (RV) ventricular and heart weight (HW) as well on structural properties in the skeletal muscle bed. The perfusion resistance (PR) was calculated as PR/flow (ml/min)/100g dry HQ weight for noradrenalin and angiotensin II (Ang II)/phenylephrine (Phe) responses. The ED-50 value demonstrates the half-maximal effect of the effective noradrenalin dose, demonstrating the adreno-receptor sensitivity. Maximal noradrenaline and Ang and Phe responses reflect the average medial bulk of contractile tissues. Values are expressed as MEAN±SE. §denotes that both protocol 1 and 2 are included. *denotes that the number in the groups are changed, intact group n=8, Hx-group n=6 and GH group n=6. adenotes p<0.05 vs. intact, bdenotes p<0.05 vs. Hx-[T4+GC].
Intact Hx [T4+GC] Hx GH+ [T4+GC]
N 14 12 12
BW (g) 402 ± 5 222 ± 3a 265 ± 5a, b
Wet HQ weight (g/100 g BW) 44.4 ± 1.0 43.3 ± 3.1 41.7 ± 1.7a
Dry HQ weight (g/100 g BW) 21.6 ± 0.3 18.4 ± 0.6a 15.6 ± 1.7a, b
Oedema (%) 51 ± 1 57 ± 2 62 ± 2a, b
LV weight (mg/100 g BW)§ 161.6 ± 12.6 164.4 ± 16.2 158.5 ± 14.7b
RV weight (mg/100 g BW)§ 41.9 ± 9.9 42.1 ± 9.8 39.9 ± 9.9
HW (mg/100 g BW)§ 203.3 ± 15.0 206.5 ± 20.3 186.4 ± 29.3a, b
ED 50 value of noradrenalin 0.593 ± 0.066 0.463 ± 0.05a 0.584 ± 0.074b
PR Noradrenalin response (mmHg/ml/min/100 g dry HQ) 23.64 ± 2.4 13.8 ± 0.8a 13.8 ± 1.4a
PR Ang II + Phe response (mmHg/ml/min/100 g dry HQ)* 33 ± 7 17 ± 2 22 ± 5
Values are expressed as MEAN ± SE. §denotes that animals from both protocol 1 and 2 are included. *denotes that the number in the groups are changed, intact group n = 8, Hx-group n = 6 and GH group n = 6. adenotes p < 0.05 vs. intact, bdenotes p < 0.05 vs. Hx- [T4+GC]
==== Refs
Capaldo B Guardasole V Pardo F Matarazzo M Di Rella F Numis F Merola B Longobardi S Sacca L Abnormal vascular reactivity in growth hormone deficiency Circulation 2001 103 520 524 11157716
Böger RH Skamira C Bode-Böger SM Brabant G von zur Muhlen A Nitric oxide may mediate the hemodynamic effects of recombinant growth hormone in patients with acquired growth hormone deficiency J Clin Invest 1996 98 2706 2713 8981915
Khan AS Sane DS Wannenburg T Sonntag WE Growth hormone, insulin-like growth factor-1 and the aging cardiovascular system Cardiovasc Res 2002 54 25 35 12062358 10.1016/S0008-6363(01)00533-8
McCallum RW Petrie JR Dominiczak A Connell MC Growth hormone deficiency and vascular risk Clin Endocrinol 2002 57 11 24 10.1046/j.1365-2265.2002.01559.x
Saccá L Cittadini A Fazio S Growth hormone and the heart. Endocrine Reviews 1994 15 555 573 7843068 10.1210/er.15.5.555
Caidahl K Edén S Bengtsson B Cardiovascular and renal effects of growth hormone Clin Endocrinol (Oxf) 1994 40 393 400 8187304
Evans LM Davies JS Anderson RA Ellis GR Jackson SK Lewis MJ Frenneaux MP Rees A Scanlon MF The effect of GH replacement therapy on endothelial function and oxidative stress in adult growth hormone deficiency Eur J Endocrinol 2000 142 254 262 10700719 10.1530/eje.0.1420254
László FA Csernay L Kocsis J Effects of hypophysectomy on haemodynamic changes induced by administration of oestrone + vasopressin Res exp Med 1974 163 219 227 10.1007/BF01851669
Shen YT Wiedmann RT Lynch JJ Grossman W Johnson RG GH replacement fails to improve ventricular function in hypophysectomized rats with myocardial infarction Am J Physiol 1996 271 H1721 H1727 8945883
Gustafsson H Wickman Tordby A Brandin L Hedin L Jonsdottir IH The interactive effects of growth hormone and oestrogen on vascular responses in hypophysectomized female rats Eur J Endo 2002 146 267 274 10.1530/eje.0.1460267
Rossoni G Locatelli V De GCV Torsello A Schweiger F Boghen M Nilsson M Bernareggi M Muller EE Berti F Growth hormone and hexarelin prevent endothelial vasodilator dysfunction in aortic rings of the hypophysectomized rat J Cardiovasc Pharmacol 1999 34 454 460 10471007 10.1097/00005344-199909000-00021
Wickman A Jonsdottir IH Bergstrom G Hedin L GH and IGF-I regulate the expression of endothelial nitric oxide synthase (eNOS) in cardiovascular tissues of hypophysectomized female rats Eur J Endocrinol 2002 147 523 533 12370116 10.1530/eje.0.1470523
Folkow B Isaksson OGP Karlström G Lever AF Nordlander M Trophic effects of hypophyseal hormones on resistance vessels and the heart in normotensive and renal hypertensive rats Acta Physiologica Scandinavica 1992 144 291 306 1585813
Napoli R Guardasole V Angelini V D'Amico F Zarra E Matarazzo M Sacca L Acute effects of growth hormone on vascular function in human subjects J Clin Endocrinol Metab 2003 88 2817 2820 12788893 10.1210/jc.2003-030144
Thum T Tsikas D Frölich JC Borlak J Growth hormone induces eNOS expression and nitric oxide release in a cultured human endothelial cell line FEBS Letters 2003 555 567 571 14675775 10.1016/S0014-5793(03)01356-5
Honeyman TW Goodman HM Fray JCS The effects of growth hormone on blood pressure and renin secretion in hypophysectomized rats Endocrinology 1983 112 1613 1617 6339210
Moller J Jorgensen JO Moller N Hansen KW Pedersen EB Christiansen JS Expansion of extracellular volume and suppression of atrial natriuretic peptide after growth hormone administration in normal man J Clin Endocrinol Metab 1991 72 768 772 1826008
Hoffman DM Crampton L Sernia C Nguyen TV Ho KKY Short-term growth hormone (GH) treatment of GH-deficient adults increases body sodium and extracellular water, but not blood pressure J Clin Endocrinol Metab 1996 81 1123 1128 8772586 10.1210/jc.81.3.1123
Bohlooy-Y M Carlsson L Olsson B Gustafsson H Andersson IJL Törnell J Bergström G Vascular function and blood pressure in growth hormone transgenenic mice Endocrinology 2001 142 3317 3323 11459773 10.1210/en.142.8.3317
Dilley RJ Schwartz SM Vascular remodeling in the growth hormone transgenic mouse. Circulation Research 1989 65 1233 1240 2805241
Fielder PJ Mortensen DL Mallet P Carlsson B Baxter RC Clark R Differential long-term effects of insulin-like growth factor-I (IGF-I), growth hormone (GH), and IGF-I plus GH on body growth and IGF-I binding proteins in hypophysectomized rats Endocrinology 1996 137 1913 1920 8612531 10.1210/en.137.5.1913
Sakuma I Togashi H Yoshioka M Saito H Yanagida M Tamura M Kobayashi T Yasuda H Gross SS Levi R NG-methyl-L-arginine, an inhibitor of L-arginine-derived nitric oxide synthesis, stimulates renal sympathetic nerve activity in vivo. A role for nitric oxide in the central regulation of sympathetic tone? Circ Res 1992 70 607 611 1537096
Sverrisdottir YB Elam M Herlitz H Bengtsson BA Johannsson G Intense sympathetic nerve activity in adults with hypopituitarism and untreated growth hormone deficiency J Clin Endocrinol Metab 1998 83 1881 1885 9626113 10.1210/jc.83.6.1881
Leong KS Mann P Wallymahmed M MacFarlane IA Wilding JP Influence of growth hormone replacement on heart rate variability in adults with growth hormone deficiency Clin Endocrinol 2001 54 819 826 10.1046/j.1365-2265.2001.01267.x
Evans LM Davies JS Goodfellow J Rees JA Scanlon MF Endothelial dysfunction in hypopituitary adults with growth hormone deficiency Clin Endocrinol (Oxf) 1999 50 457 464 10468904 10.1046/j.1365-2265.1999.00671.x
Jayachandran M Hayashi T Sumi D Thakur NK Kano H Ignarro LJ Iguchi A Up-regulation of endothelial nitric oxide synthase through beta(2)- adrenergic receptor--the role of a beta-blocker with NO-releasing action Biochem Biophys Res Commun 2001 280 589 594 11162560 10.1006/bbrc.2000.4177
Wennmalm A Benthin G Edlund A Jungersten L Kieler-Jensen N Lundin S Westfelt UN Petersson AS Waagstein F Metabolism and excretion of nitric oxide in humans. An experimental and clinical study Circ Res 1993 73 1121 1127 8222083
|
16271154
|
PMC1291375
|
CC BY
|
2021-01-04 16:03:51
|
no
|
BMC Physiol. 2005 Nov 7; 5:17
|
utf-8
|
BMC Physiol
| 2,005 |
10.1186/1472-6793-5-17
|
oa_comm
|
==== Front
BMC PsychiatryBMC Psychiatry1471-244XBioMed Central London 1471-244X-5-401626290310.1186/1471-244X-5-40Research ArticleTesting assumptions for endophenotype studies in ADHD: Reliability and validity of tasks in a general population sample Kuntsi Jonna [email protected] Penny [email protected] Jonathan [email protected]örger Norbert A [email protected] der Meere Jaap J [email protected] MRC Social, Genetic and Developmental Psychiatry Centre, Institute of Psychiatry, King's College London, De Crespigny Park, London SE5 8AF, UK2 Department of Developmental and Experimental Clinical Psychology, University of Groningen, Grote Kruisstraat 2/1, 9712 TS Groningen, The Netherlands2005 1 11 2005 5 40 40 22 5 2005 1 11 2005 Copyright © 2005 Kuntsi et al; licensee BioMed Central Ltd.2005Kuntsi et al; licensee BioMed Central Ltd.This is an Open Access article distributed under the terms of the Creative Commons Attribution License (), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
Background
Advances in both genetic and cognitive-experimental studies on attention deficit hyperactivity disorder (ADHD) have opened new opportunities for cognitive endophenotype research. In such genetic designs the focus is on individual differences in characteristics, associated with ADHD, that can be measured reliably over time. Genetic studies that take a 'quantitative trait loci' approach hypothesise that multiple susceptibility genes contribute to a continuous dimension of ADHD symptoms. As an important initial step, we aimed to investigate the underlying assumptions that (1) key cognitive-experimental tasks indicate adequate test-retest reliability and (2) ADHD symptom scores in a general population sample are associated with performance on these tasks.
Methods
Forty-nine children were assessed on a go/no-go task and a reaction time task (the 'fast task') that included manipulations with event rate and incentives. The children were assessed twice, with a test-retest interval of two weeks.
Results
The majority of the task variables demonstrated moderate-to-good test-retest reliability. The correlations between teacher ratings of ADHD symptoms and key task variables were .4–.6: ADHD symptoms were associated with poor performance (especially high reaction time variability) in a slow baseline condition, whereas there was low or no association in conditions with a faster event rate or incentives. In contrast, no clear pattern of findings emerged based on parent ratings of ADHD symptoms.
Conclusion
The data support the usefulness of the go/no-go and fast tasks for genetic studies, which require reliable and valid indices of individual differences. The overall pattern of associations between teacher ratings of ADHD symptoms and task variables is consistent with effects of event rate and incentives on performance, as predicted by the model of activation and arousal regulation. The lack of a clear pattern of findings with parent ratings of ADHD symptoms warrants further study.
==== Body
Background
Attention deficit hyperactivity disorder (ADHD) has been the focus of numerous both genetic and cognitive-experimental studies in recent years. The symptom cluster of impulsivity, overactivity and inattentiveness has been shown to be highly heritable and molecular genetic studies have obtained initial evidence of genes that contribute to the disorder (reviewed in [1-3]). Cognitive-experimental studies have indicated tasks that are sensitive to ADHD, pointing to possible underlying psychological processes [4,5]. These advances have led to an increased interest in planning investigations on cognitive endophenotypes in ADHD [4]. Endophenotypes refer to quantifiable intermediate constructs that index an underlying liability to a disorder or behavioural trait. By studying candidate endophenotypes we can start to unravel the causal pathways from etiological factors through to psychological processes and behaviour.
Certain key assumptions underlie such an approach that combines genetic and cognitive-experimental methods. As the focus is on differences between individuals in etiological factors that contribute to the disorder, it is critical that the candidate endophenotypes reflect characteristics on which individuals differ reliably over time. Demonstrating adequate test-retest reliability has been acknowledged as an important initial step [6], but few studies have as yet obtained such data. Although an unusually high variability in performance is one of the strongest findings yet to emerge from cognitive-experimental research on ADHD [4,5,7], this needs not pose a problem for test-retest reliability: if variability in performance indicates a true underlying process, we can aim to obtain reliable indicators of this characteristic.
Whereas some genetic studies on ADHD follow a strict diagnostic (categorical) approach, others are based on an underlying QTL (quantitative trait loci) assumption of multiple susceptibility genes contributing to a continuous dimension of ADHD symptoms. Twin study data are consistent with the hypothesis that ADHD represents the extreme of a behaviour that varies continuously throughout the entire population [8,9]. From this hypothesis it follows that performance on tasks that are sensitive to ADHD is expected to correlate with continuous ADHD symptom scores in the general population. This has received scant attention in research to date, in contrast to a wealth of studies comparing task performance between children with ADHD and control children.
Amongst the most commonly used tasks in ADHD research are inhibition measures, such as the go/no-go and stop tasks. Whereas findings have been inconsistent with regard to a hypothesised inhibition deficit (see below), other data give further clues to the possible psychological processes implicated in ADHD. Two studies indicate an association between ADHD and poor inhibition on the stop task only in a standard, non-incentive condition, with the children with ADHD performing as well as the control children in a high incentive condition [10,11]. A third study did not find effects of reward and response cost on inhibition in ADHD [12], but the lack of comparison to a non-contingency condition limits the conclusions that can be made. Van der Meere and colleagues [13] found that inhibition on the go/no-go task was dependent on the presentation rate of stimuli: in contrast to both the slow and fast conditions, the performance of children with ADHD was comparable to that of control children in the medium event rate condition. In two further studies no differences were observed between ADHD and control groups on the inhibition indices of the stop task under two [14] or three [15] different event rates. Consistently across the studies, however, the slow condition produced disproportionately slow and variable reaction times in the children with ADHD. Event rate manipulation within the go/no-go task has also distinguished children with ADHD from control children with regard to cardiac [14] and evoked potential [16] measures: group differences, consistent with poor effort allocation in children with ADHD, emerged only in the slow and not in the fast condition.
Other recent studies similarly report the highly characteristic pattern on reaction time data, while not finding evidence for an association between ADHD and an inhibition deficit [7,17]. As Castellanos and Tannock [4] point out, "response variability is the one ubiquitous finding in ADHD research across a variety of speeded-reaction-time tasks, laboratories and cultures" (p. 624). A detailed statistical analysis of response times on a four-choice reaction time task demonstrated explicitly how a greater proportion of abnormally slow responses, mixed with fast responses on some trials, leads to the inconsistent pattern of responding in ADHD [18].
One approach to explaining the findings of highly variable cognitive performance and the interaction between cognitive performance and factors affecting the energetic state (incentives, presentation rate and predictability of stimuli) proposes an underlying regulation problem in ADHD. For example, the optimal stimulation theory placed the emphasis on the regulation of arousal [19] and, more recently, the state regulation theory proposes a critical role for the regulation of activation and effort (reviewed in [5]; see also [20]). Within this approach, the focus has also shifted from attentional deficits to the regulation of attention in ADHD [5,21,22]. An alternative account suggests that multiple psychological processes, both cognitive and reward-related processes, are affected [4].
Studying the psychological processes in ADHD within a genetic design is only at its beginning: candidate endophenotypes are being speculated upon and frameworks developed [4], but evidence is as yet limited. Initial studies have provided inconsistent findings, which has been attributed to methodological limitations, such as small sample sizes (reviewed in [1].
In this study we addressed two underlying assumptions that are highly relevant for large-scale genetic investigations on ADHD that include cognitive testing: (1) adequate test-retest reliability for key experimental tasks and (2) an association between performance on the tasks and ADHD symptom scores in a general population sample. We have previously reported test-retest reliability data for selected executive function measures and a 'delay aversion' task [23], and used these data to select tasks for a subsequent, initial twin study on hyperactivity [7,24]. The initial twin data supported reaction time variability as a potential endophenotype [24], with further supportive evidence emerging from a recent family study on ADHD [25]. In the current study we focused on two different tasks that appear particularly promising based on recent data – the go/no-go task and a reaction time task – and included within-task manipulations (some novel) with event rate and incentives. As working memory deficits are also proposed to characterise ADHD in some models [4], we additionally examine the association between ADHD symptom scores and performance on a working memory measure.
Methods
Ethical approval was obtained from the Institute of Psychiatry Ethical Committee (Research).
Sample
The children were recruited from local primary (ages 8–9 years) and secondary (ages 12–13 years) schools in London, UK. We asked head teachers to randomly select pupils who fulfilled the inclusion criteria of being fluent in English, within the appropriate age range and without special educational needs (severe sensory, physical or cognitive impairments). We specified that 'random selection' means a procedure such as choosing every fifth child in alphabetical order and that by using such random selection we aim to obtain a sample that is representative of children living in the UK. Consent forms and information packs were given to parents, who returned them to the school. Of the 66 parents contacted, 53 (80%) gave consent for their child to take part in the study. Each school that participated was given a £40 voucher to a book/stationery shop and the children earned small prizes (vouchers and stationery) on some tasks. Parents did not receive any financial reward for participation.
Children with prorated IQs below 70 (n = 3) were excluded. A fourth child was excluded due to a refusal to follow task instructions. The sample consists of 49 children: 19 boys and 30 girls. The children had a mean age of 11.03 years (SD = 2.08). The ethnic origin of the children was classified as follows: 51% white Caucasian, 35% African/Caribbean, 6% Indian/Pakistani/Bangladeshi, 2% Chinese, 4% 'other' and 2% unknown. Three children were not at school during re-test days and therefore time 2 data are missing for them. In addition, data are missing for a few children on specific tasks, due to technical problems with the portable computers during task administration (the numbers of children in each analysis are reported in the tables of the Results section).
Procedure
Two trained research workers visited the children at school, assessing the children individually in separate, quiet rooms. The same research worker tested the same children at both time points. The test-retest interval was two weeks.
The tasks were administered in the following fixed order: fast task, WISC block design, WISC vocabulary, go/no-go task and WISC digit span. The tests, as well as conditions within each test, were presented in the same order at both time points. Testing sessions were arranged around the school's timetable and the children were given short breaks as required. The total length of the testing session, including breaks, was approximately 1.5–2 h.
Measures
The Revised Conners' Parent [26] and Teacher [27] Rating Scales
Ratings on the Conners' scales were obtained from parents and teachers, with response rates of 98% and 83%, respectively. Adding up the scores on the inattentive and hyperactive-impulsive DSM-IV symptoms subscales forms a total ADHD DSM-IV symptoms subscale. A prorating procedure was applied in the few instances where there was missing data, as recommended in the manual [28].
Wechsler Intelligence Scales for Children (WISC-IIIUK [29]
The vocabulary and block design subtests from the WISC were used to obtain an estimate of the child's IQ (prorated following procedures described by Sattler [30]). The children's IQs ranged from 72 to 145, normally distributed (M = 97.88, SD = 14.13). In addition, the digit span subtest was administered to obtain an estimate of working memory (digit span backward score).
The go/no-go task
We used a version of the task developed by van der Meere and colleagues [13,14]. On each trial, one of two possible stimuli appeared for 300 ms in the middle of the computer screen. The child was instructed to respond only to the 'go' stimuli and to react as quickly as possible, but to maintain a high level of accuracy. The proportion of 'go' stimuli to 'no-go' stimuli was 4:1. The response variables are commission errors, mean reaction time to 'go' stimuli and standard deviation of the reaction times. (Omission errors were rare – mean in each condition 1–5% – and are therefore not included in analyses.)
The children performed the task under three different conditions, matched for length of time on task. The fast condition consisted of 462 trials and had an inter-stimulus-interval (ISI) of 1 s. The ISI dropped to 8 s in the slow presentation condition, which consisted of 72 trials. The order of presentation of the slow and fast conditions was varied randomly across children.
The incentive condition was always administered last. This new condition, designed specifically for this study, is a modification of the incentive condition used in the study on the stop task by Slusarek et al. [11]. Each correct response to the letter X and each correct non-response to the letter O earned the child one point. The child lost one point for each omission error (failure to respond to X) and for each failure to respond within 2 s. Each commission error (incorrect response to O) led to the loss of five points. The points were shown in a box, immediately right of the screen centre, and were updated continuously throughout. The child started with 40 points, to avoid the possibility of a negative tally. The child was asked to try to win as many points as possible, and was told that the points will be exchanged for a real prize when the game ends. This condition consisted of 72 trials and had an ISI of 8 s.
A practice session preceded each experimental condition.
The fast task
The baseline condition followed a standard warned four-choice reaction time task, as outlined in Leth-Steensen et al. [18]. A warning signal (four empty circles, arranged side by side) first appeared on the screen. At the end of the foreperiod (presentation interval for the warning signal), the circle designated as the target signal for that trial was filled (coloured) in. The child was asked to make a compatible choice response by pressing the response key that directly corresponded in position to the location of the target stimulus. Following a response, the stimuli disappeared from the screen and a fixed inter-trial interval of 2500 ms followed. Speed and accuracy were emphasised equally. If the child did not respond within 10 s, the trial terminated.
First a practice session was administered, during which the child had to respond correctly to five consecutive trials. The baseline condition, with a foreperiod of 8 s and consisting of 72 trials, then followed.
To investigate the extent to which a response style characterised by slow and variable speed of responding can be maximally reduced, we developed a novel comparison condition that used a fast event rate (1 s) and incentives. This condition started immediately after the baseline condition and consisted of 80 trials (following the faster event rate conditions in Leth-Steensen et al. [18]). The child was told that if she will respond really quickly one after another, she will win smiley faces and will get real prizes in the end. The child won a smiley face each time she responded faster than her own mean reaction time during the baseline (first) condition consecutively for three trials. The baseline mean reaction time was calculated here based on the middle 94% of responses, therefore excluding extremely fast and extremely slow responses. The smiley faces appeared below the circles in the middle of the screen and were updated continuously.
Following the procedure recommended by Leth-Steensen et al. [18], observations on the fast task that were more than four standard deviations from a participant's mean reaction time for a specific condition were excluded. This conservative criterion aims to minimise the risk of removing any 'real' data in ADHD research on standard reaction time tasks, while still controlling for very extreme observations. Only 0.5–0.8% of the observations from each condition were excluded.
The response variables are mean reaction time and standard deviation of the reaction times, calculated for each condition based on correct responses only. (Trials that were therefore excluded consisted of omission errors (mean in each condition below 0.5%) and incorrect responses (mean in each condition 4–9%). We report data from the baseline condition both including all trials and including the first 30 trials only; the latter provides a match on length of time on task with the fast-incentive condition. Only the first 30 trials of the baseline condition were used to calculate difference scores, which indicate improvement between the baseline and fast-incentive conditions.
Results
Test-retest reliability
To establish test-retest reliability, we calculated inter-class Pearson product moment correlations and partial correlations, controlling for age, for each of the response variables (Table 1). Following the practice recommended by Rousson, Gasser and Seifert [31], we focus on inter-class rather than intra-class correlations (although report both to enable a comparison with previous research). Rousson et al. [31] recommend only using intra-class correlations, which take into account systematic error (learning effects), for intra-rater and inter-rater reliability, and not for test-retest reliability. Learning effects are a natural phenomenon and not a shortcoming associated with measures; in test-retest reliability the focus is on the consistency of the performance of the participants in relation to one another. To indicate the extent to which performance improved consistently across participants, we also report t-test results for the comparisons between mean scores at time 1 and time 2 for each measure (Table 1). For the test-retest correlations, where significance is not a sufficient criterion for adequate reliability, we focus on the size of the correlations. In further correlational analyses, we additionally indicate those correlations that reached significance at alpha .01 or .05 level (ie, if significance level is not given, this indicates a non-significant p-value).
Table 1 Test-retest reliability results
Measure inter-class r partial ra intra-class r time 1 time 2 t-value df p
mean SD mean SD
Fast Task (n = 41)
baseline (all trials)
mean RT .85 .75 .77 730.31 177.76 659.24 156.71 4.89 40 <.001
SD of RTs .60 .53 .58 193.77 103.55 171.86 118.49 1.40 40 .17
baseline (30 trials)
mean RT .76 .53 .68 670.64 155.09 603.32 138.74 3.75 40 .001
SD of RTs .47 .37 .43 155.78 68.90 141.59 97.80 0.93 40 .36
fast-incentive
mean RT .89 .79 .82 537.69 121.55 499.78 100.97 4.29 40 <.001
SD of RTs .76 .65 .70 122.80 50.63 105.34 47.43 3.25 40 .002
Go/No-Go (n = 44–47)
slow condition
mean RT .76 .63 .74 512.00 105.13 532.07 112.06 -1.77 44 .08
SD of RTs .58 .53 .58 177.51 115.86 172.33 123.30 .315 44 .75
commission errors .56 .56 .53 45.19 23.18 36.59 22.46 2.69 44 .01
fast condition
mean RT .88 .85 .87 387.20 74.69 385.24 64.58 0.37 44 .71
SD of RTs .83 .82 .83 147.42 69.48 147.02 65.80 0.07 44 .95
commission errors .70 .67 .64 42.54 16.99 38.62 17.89 1.93 44 .06
incentive condition
mean RT .70 .63 .69 508.64 91.44 526.48 79.26 -1.77 43 .08
SD of RTs .26b .09b .26 119.52 47.43 122.91 40.47 -0.42 43 .68
commission errors .54 .47 .52 25.95 16.30 21.82 16.97 1.72 43 .09
a controlling for age b inter-class r increases to .61 and partial r to .44, if exclude six potential outliers (see text for further information)
The test-retest correlations for both tasks were in the moderate to high range (.5–.9), with two exceptions (see below). The results were overall similar whether or not age was controlled for, with only slight decreases in the size of the correlation for the partial correlations. Significant t-test results, indicating learning effects, emerged for all other fast task variables, except the standard deviation of reaction times in the baseline condition, but for only one go/no-go task variable (commission errors in the slow condition).
For the baseline fast task data there was a small decrease in the size of the test-retest correlations (from .5–.9 to .4–.8), when considering the first 30 trials separately. However, the fast-incentive condition data indicated good reliability, despite the relatively short administration time.
The low inter-class correlation of .26 for the standard deviation of reaction times in the incentive condition of the go/no-go task increased to .61, and the partial correlation to .44, if six children with highly discrepant values between time 1 and time 2 sessions (differences of 92–158 ms) were excluded. To examine the extent to which the lower reliability of this variable may affect its validity, insofar as its association with theoretically related variables is concerned, we calculated a Pearson product moment correlation between this variable (including the potential outliers) and the standard deviation of reaction times in the fast-incentive condition of the fast task. For time 1 data the correlation was .56 (p < .01); for time 2 data the correlation dropped to .29.
Effects of task manipulations on the total sample
Paired t-tests between each pair of comparison conditions within a task indicated improved performance (p < .01) following task manipulations (fast event rate and/or incentives) for all but three go/no-go task comparisons: (1) SD of RTs between slow and fast conditions, (2) commission errors between slow and fast conditions and (3) mean RT between slow and incentive conditions. To minimise possible practice effects (given that the incentive conditions were always administered last), we performed these analyses on time 2 data.
To explore the extent to which the two types of manipulations in the go/no-go task – fast event rate and incentives – had a similar effect on performance within individual children, we calculated Pearson product moment correlations, and partial correlations controlling for age, between the two difference scores for each variable. In these as well as other correlational analyses we present both types of correlations to allow for an easy visualisation of the effects of age on the results. However, as the main interest is on associations between performance across conditions (or between task performance and ADHD ratings, below) independent of age effects, we focus mainly on the partial correlations. The slow-fast condition difference score and the slow-incentive condition difference score correlated strongly for each of the three variables (all p < .01): mean RT (r = .63/partial r = .62), SD of RTs (r = .78/partial r = .77) and commission errors (r = .73/partial r = .74).
Association between task performance and ratings of ADHD symptoms
The next question we addressed was whether individual differences between children on task performance and, in particular, the extent of improvement from the slow baseline condition to a condition with a faster event rate or incentives (or both) is associated with parents' and teachers' ratings of ADHD symptoms (T-scores). The model of arousal and activation regulation predicts an association between ADHD symptoms and poor performance in a baseline condition, but lack of (or reduced) association with performance in conditions with such task manipulations. Here we focus on time 1 data only, which reflect the usual situation of a one-off assessment session. We calculated Pearson product moment correlations, and partial correlations controlling for age, between these variables and report also the associated effect sizes (Tables 2, 3, 4).
Table 2 Association between difference scores on the go/no-go task and parent and teacher ratings on the Conners' scales (T-scores): correlations (r) and partial correlations, controlling for age (partial r), with associated effect sizes (d)
Difference score Teacher (n = 38–39): total ADHD symptom score Parent (n = 45–46): total ADHD symptom score
r d partial r d r d partial r d
slow – fast condition
mean RT .19 0.4 .24 0.5 -.11 -0.2 -.06 -0.1
SD of RTs .44** 1.0 .44** 1.0 .08 0.2 .13 0.3
commission errors .39* 0.8 .40* 0.9 .11 0.2 .12 0.2
slow – incentive condition
mean RT .36* 0.8 .38* 0.8 .15 0.3 .19 0.4
SD of RTs .44** 1.0 .46** 1.0 .13 0.3 .17 0.3
commission errors .21 0.4 .20 0.4 -.03 -0.1 -.06 -.01
* p < .05 (2-tailed); ** p < .01 (2-tailed)
Table 3 Association between go/no-go task variables and parent and teacher ratings on the Conners' scales (T-scores): correlations (r) and partial correlations, controlling for age (partial r), with associated effect sizes (d)
Measure Teacher (n = 38–39): total ADHD symptom score Parent (n = 45–46): total ADHD symptom score
r d partial r d r d partial r d
slow condition
mean RT .10 0.2 .22 0.5 -.19 -0.4 -.09 -0.2
SD of RTs .39* 0.8 .48** 1.1 .05 0.1 .14 0.3
commission errors .38* 0.8 .42** 0.9 .02 0.0 .06 0.1
fast condition
mean RT -.08 -0.2 -.01 -0.0 -.15 -0.3 -.04 -0.1
SD of RTs -.06 -0.1 -.03 -0.1 -.05 -0.1 -.01 -0.0
commission errors .10 0.2 .14 0.3 -.10 -0.2 -.05 -0.1
incentive condition
mean RT -.26 -0.6 -.23 -0.5 -.35* -0.7 -.29 -0.6
SD of RTs -.09 -0.2 -.03 -0.1 -.21 -0.4 -.12 -0.2
commission errors .21 0.4 .31 0.7 .02 0.0 .13 0.3
* p < .05 (2-tailed); ** p < .01 (2-tailed)
Table 4 Association between fast task variables and parent and teacher ratings on the Conners' scales (T-scores): correlations (r) and partial correlations, controlling for age (partial r), with associated effect sizes (d)
Measure Teacher (n = 36): total ADHD symptom score Parent (n = 43): total ADHD symptom score
r d partial r d r d partial r d
baseline (all trials)
mean RT .19 0.4 .42** 0.9 -.06 -0.1 .13 0.3
SD of RTs .45** 1.0 .58** 1.4 .16 0.3 .29 0.6
baseline (30 trials)
mean RT .14 0.3 .35* 0.7 -.11 -0.2 .06 0.1
SD of RTs .44** 1.0 .55** 1.3 .17 0.4 .28 0.6
fast-incentive
mean RT -.03 -0.1 .10 0.2 -.08 -0.2 .10 0.2
SD of RTs .09 0.2 .22 0.5 .08 0.2 .26 0.5
difference score (baseline 30 trials – fast-incentive)
mean RT .20 0.4 .24 0.5 .07 0.1 -.03 0.0
SD of RTs .47** 1.1 .47** 1.1 .14 0.3 .14 0.3
* p < .05 (2-tailed); ** p < .01 (2-tailed)
For the go/no-go task the teacher-rated ADHD symptoms correlated moderately strongly (.4–.5) and significantly with four out of the six difference scores (Table 2), indicating a greater improvement in performance between the slow and the other two conditions in children with more ADHD symptoms. An additional examination of the correlations from the individual conditions confirmed this pattern of findings (Table 3): moderately strong correlations (.4–.5) were obtained for performance in the slow condition, and low correlations – indeed in some cases negative correlations, indicating better performance – were obtained for performance in the fast and incentive conditions. The correlation between teacher-rated ADHD symptoms and the difference score for commission errors between the slow and incentive conditions was less strong (.2) and non-significant. In contrast, no clear pattern emerged for correlations with parent-rated ADHD symptoms.
For the fast task the teacher ratings were moderately strongly (.5) associated with the standard deviation of reaction times difference score, indicating an association between teacher-rated ADHD symptoms and improvement in the variability in the speed of responding from the baseline to the fast-incentive condition (Table 4). This was also reflected in the highly significant correlations with baseline data, and the lower and non-significant correlations with fast-incentive condition data. Correlations with parent ratings were all non-significant and do not indicate a clear pattern.
We calculated additional correlations, controlling for age, separately for the inattentiveness and hyperactivity-impulsivity subscales for the key measures on both tasks (the difference scores), to examine whether a similar pattern of results emerges for both subscales. These suggested few differences for either teacher or parent ratings. The magnitude of the difference between the correlation of a difference score with inattentiveness vs hyperactivity-impulsivity was less than .12 in each case, with neither dimension consistently associated with higher correlations.
The digit span backward score, a measure of working memory, was not associated with either teacher (r = .09) or parent (r = .04) ratings of ADHD symptoms. The correlations with digit span forward scores were similarly near zero (r = -.03 and r = .01, respectively).
As a greater proportion of teacher ratings were missing for boys in the older age range, than for the other subgroups of children, we additionally examined correlations between task performance and parent ratings, excluding children for whom teacher ratings were missing. The correlations indicated a similar pattern of results as with the total sample.
The teacher and parent inattentiveness ratings correlated .28 with one another, and the teacher and parent hyperactivity-impulsivity ratings .33. For the total ADHD symptoms, the correlation between teacher and parent ratings was .33 (p < .05).
The association between IQ and task performance is not a specific focus of this paper, but we note that the pattern of associations between the key task variables (difference scores) and parent and teacher ratings of ADHD symptoms did not change, when controlling for IQ, as well as for age.
Discussion
With a general population sample of children, we demonstrated moderate to good test-retest reliability for the majority of variables from the go/no-go and fast tasks, and an association between performance on these tasks and teacher, but not parent, ratings of ADHD symptoms.
An acceptable level of test-retest reliability depends on the nature of the measures. For tests to be useful in clinical practice, high reliability coefficients are commonly required (around .8 or above [33]). With experimental tasks, it may be unrealistic to expect reliability coefficients of such magnitude. We previously reported test-retest inter-class correlation coefficients of between .2 and .7 (with a median value of .66) for a range of task variables [23].
Considering correlations of .7 or higher as indicating good test-retest reliability, and correlations of .5 and .6 as indicating moderate test-retest reliability, the variables from the fast and go/no-go tasks were within such a moderate-to-good reliability range, with two exceptions. Controlling for age effects did not have a noticeable effect on the reliability coefficients, with only slight decreases in the size of the correlation for some variables.
Reaction time variability in the baseline condition of the fast task indicated adequate test-retest reliability when including all trials, but the inter-class and partial correlation coefficients were slightly lower (.5 and .4) when focusing on the first 30 trials only. This illustrates how task length can affect reliability. The lower reliability coefficient for the reaction time variability in the incentive condition of the go/no-go task seems due to a few children having highly discrepant values between the test and retest sessions. The reason for this is not clear, but we note that the incentive condition was administered late in the testing session and this may have contributed to fatigue in some children. It seems relevant that the reaction time variability in the go/no-go task incentive condition was strongly associated with the reaction time variability in the fast-incentive condition of the fast task at time 1, as predicted, but that the association was less strong (and non-significant) at time 2. Given that these tasks are aimed to challenge children's ability to concentrate over time, fatigue may have affected some children's performance particularly at time 2 (when novelty of the tasks had worn off). Note that in the other two conditions the reaction time variability demonstrated adequate test-retest reliability.
When using these tasks within genetic designs, we aim to use multivariate genetic model-fitting approaches. From a test-retest reliability viewpoint such analyses may have an additional advantage: a priori one would expect a better reliability for summed components compared to single variables [31].
The second main research question addressed the extent of association between task performance and ratings of ADHD symptoms in a general population sample. The within-task manipulations with event rate and incentives test the prediction of the arousal and activation regulation model of improved performance following such manipulations; therefore associations are predicted with difference scores, which indicate improvement from 'baseline' to a comparison condition. When considering the data separately from each condition, this would be reflected in associations with baseline performance and lack of (or reduced) associations with performance in the conditions with incentive or event rate manipulations.
The associations with teacher ratings of ADHD symptoms were overall in line with these predictions. The strongest associations were obtained for reaction time variability. In both tasks the decrease in reaction time variability from baseline condition to a condition with a faster event rate or incentives (or both) was significantly associated with teacher ratings of ADHD symptoms. The data from each condition separately indicate how the difference score results reflect the significant associations of teacher ADHD ratings with reaction time variability in the baseline conditions, and low or no association with this variable in the conditions with event rate and incentive manipulations. The magnitude of the correlations (.4 – .6) is greater than, for example, that between ADHD symptoms and IQ scores, which is typically -.2 – -.4 [32,34-36]. (Note that a correlation of -.3 translates to a 9 point difference in mean IQ scores between children with ADHD and control children [32]). In the fast task the association with teacher ADHD ratings emerged in the baseline condition whether focusing on the first 30 trials only or the full 72 trials. Similar results, though somewhat less strong, emerged also for mean reaction times. The association between teacher-rated ADHD symptoms and disproportionately slow and variable reaction times in the slow condition is consistent with previous studies on children with ADHD [13-15]. In addition to our focus on individual differences in relation to ADHD symptoms, we also demonstrated that the fast task manipulations improved the speed and the variability of speed in the total sample of children.
On the go/no-go task, the rate of commission errors is considered indicative of response inhibition. Teacher ratings of ADHD symptoms were associated with commission errors in the slow but not in the fast condition. Findings on the effects of event rate on inhibition in children with ADHD have been somewhat inconsistent, with uncertainty surrounding an optimal event rate and some studies reporting no association between commission errors and ADHD in any event rate condition [14,15]. The present data on associations with teacher ratings of ADHD symptoms are consistent with the suggestion that a fast event rate optimises the activation or arousal state of the child [15]. The findings were less strong regarding improvement in the rate of commission errors from the slow to the incentive condition: although in the predicted direction (r = .2), the correlation between teacher ratings of ADHD symptoms and the difference score was not significant. In our future research we will investigate the effects of the incentive manipulation in the go/no-go task in children with diagnosed ADHD.
The data on the go/no-go task and teacher (or parent) ratings of ADHD symptoms are not consistent with an inhibition deficit hypothesis that predicts an association between ADHD symptoms and commission errors across all conditions. We similarly obtained no evidence for an association between ADHD symptom scores and working memory, as measured by the digit span backward score. The present data demonstrated effects of event rate and incentives and further indicated that the two types of manipulations in the go/no-go task had a similar effect on performance within individual children: improvement in inhibition and reaction time performance following a faster event rate correlated strongly with improvement in performance following the introduction of incentives. Whereas effects of individual manipulations may be open to multiple interpretations, the overall pattern of findings obtained in the present study, including the association between the effects of the two types of manipulations in the go/no-go task, seems most parsimonious with the model of arousal and activation regulation. Yet the possible relationships between different current models need to be investigated in more detail.
In contrast to the encouraging findings with teacher ratings, no clear pattern of findings emerged for parent ratings of ADHD symptoms. Oosterlaan and colleagues [37] similarly reported recently that only teacher ratings of ADHD symptoms predicted performance on cognitive tasks that were sensitive to ADHD, with parent ratings not contributing to the association. The sample in their study included children with research diagnoses of pervasive ADHD and control children.
Teacher and parent ratings of ADHD symptoms reflect only partially overlapping phenotypes (as exemplified by the correlation of .3 reported here) and genotypes [38]. Potential strengths of teacher ratings of ADHD symptoms include a better awareness of population norms, observing children in situations that are challenging for children with ADHD symptomatology and greater objectivity. Teacher ratings show higher internal consistency and stability [39], are free of the rater bias typically found in parent ratings [40] and show also greater genetic stability [41]. However, the validity of teacher and parent ratings is a complex issue and a detailed discussion is beyond the scope of this paper.
An exploration of the degree of association of inattentiveness and hyperactivity-impulsivity subscales separately with key task variables suggested few differences. Data from studies on diagnosed ADHD on the strength of the association between each subtype and task performance have not yielded a consistent pattern [10,37,42,43].
A limitation of the current study is that a greater proportion of teacher ratings were missing for boys in the older age range, than for the other subgroups of children. However, the response rate from teachers was high overall (83%) and we controlled for age effects in the analyses. Additional analyses also indicated that the missing teacher ratings are unlikely to have led to any systematic bias. Another limitation, though not a specific focus of the study, is that the sample was not large enough to analyse data separately for girls and boys to study possible sex effects. Although ADHD diagnoses are more common among boys than girls [44], from a QTL perspective the focus is not only on diagnosed ADHD, but also on ADHD symptoms in an unselected population; hence an interest in studying the reliability and validity of the tasks in a mixed-sex sample.
Conclusion
The demonstration of moderate-to-good test-retest reliability for the majority of the variables from the go/no-go and fast tasks supports their usefulness for genetic studies, which require reliably measured indices of individual differences. The association of task performance (including variables previously associated with clinically diagnosed ADHD) with teacher ratings of ADHD symptoms in a general population sample is consistent with the assumption of a continuously distributed ADHD trait. The lack of a clear pattern of findings with parent ratings of ADHD symptoms warrants further study. The next step will involve an examination of the QTL hypothesis of a quantitative dimension of ADHD symptoms in endophenotype research, by using the same tasks in genetic designs with a general population sample with continuously measured ADHD symptoms, as well as with children with diagnosed ADHD. Combining quantitative genetic (family or twin designs) and molecular genetic methods will enable an investigation of both the overall familial or genetic influences on task performance and associations between specific genes and task performance.
Competing interests
The author(s) declare that they have no competing interests.
Authors' contributions
JK conceived of the study, supervised task development, recruitment and data collection, performed statistical analyses and drafted the manuscript. PA participated in recruitment, task development and data collection, and performed statistical analyses on data on the go/no-go task. JM participated in recruitment, task development and data collection. NAB and JJvdM developed the go/no-go task and advised on statistical analyses on this task. All authors read and approved the manuscript.
Pre-publication history
The pre-publication history for this paper can be accessed here:
Acknowledgements
We are grateful to the children, parents and teachers for their participation. Our thanks to Stuart Newman and Kayley O'Flynn. We also thank Philip Asherson for helpful comments on the manuscript. The study was supported by a project grant from the Wellcome Trust (Principal Investigator Jonna Kuntsi).
==== Refs
Asherson PJ IMAGE consortium Attention deficit hyperactivity disorder in the post-genomic era Eur J Child Adolesc Psychiatry 2004 150 70
DiMaio S Grizenko N Joober R Dopamine genes and attention-deficit hyperactivity disorder: a review J Psychiatry Neurosci 2003 28 27 38 12587848
Faraone SV Doyle AE The nature and heritability of attention-deficit/hyperactivity disorder Child Adolesc Psychiatr Clin N Am 2001 10 299 316 11351800
Castellanos FX Tannock R Neuroscience of attention-deficit/hyperactivity disorder: the search for endophenotypes Nat Rev Neurosci 2002 3 617 628 12154363
van der Meere JJ Sandberg S The role of attention Hyperactivity disorders of childhood 2002 2 Cambridge: Cambridge University Press 162 213
De Geus EJC Boomsma DI A genetic neuroscience approach to human cognition Eur Psychol 2001 6 241 253 10.1027//1016-9040.6.4.241
Kuntsi J Oosterlaan J Stevenson J Psychological mechanisms in hyperactivity: I Response inhibition deficit, working memory impairment, delay aversion, or something else? J Child Psychol Psychiatr 2001 42 199 210 10.1017/S0021963001006709
Levy F Hay D McStephen M Wood C Waldman I Attention-deficit hyperactivity disorder: A category or a continuum? Genetic analysis of a large-scale twin study J Am Acad Child Adolesc Psychiatry 1997 36 737 744 9183127 10.1097/00004583-199706000-00009
Willcutt EG Pennington BF DeFries JC Etiology of inattention and hyperactivity/impulsivity in a community sample of twins with learning difficulties J Abnorm Child Psychol 2000 28 149 159 10834767 10.1023/A:1005170730653
Konrad K Gauggel S Manz A Schöll M Lack of inhibition: a motivational deficit in children with attention deficit/hyperactivity disorder and children with traumatic brain injury Child Neuropsychol 2001 6 286 296 11992192
Slusarek M Velling S Bunk D Eggers C Motivational effects on inhibitory control in children with ADHD J Am Acad Child Adolesc Psychiatry 2001 40 355 363 11288778 10.1097/00004583-200103000-00016
Oosterlaan J Sergeant JA Effects of reward and response cost on response inhibition in AD/HD, disruptive, anxious, and normal children J Abnorm Child Psychol 1998 26 161 174 9650623 10.1023/A:1022650216978
van der Meere JJ Stemerdink N Gunning B Effects of presentation rate of stimuli on response inhibition in ADHD children with and without tics Percept Mot Skills 1995 81 259 262 8532467
Börger N van der Meere JJ Motor control and state regulation in children with ADHD: a cardiac response study Biol Psychiatry 2000 5 247 267
Scheres A Oosterlaan J Sergeant JA Response execution and inhibition in children with AD/HD and other disruptive disorders: the role of behavioural activation J Child Psychol Psychiatry 2001 42 347 357 11321204 10.1017/S0021963001006898
Wiersema RJ van der Meere JJ Roeyers H State regulation and response inhibition in children with ADHD and children with early and continuously treated phenylketonuria: an event-related potential comparison J Inherit Metab Dis
Banaschewski T Brandeis D Heinrich H Albrecht B Brunner E Rothenberger A Questioning inhibitory control as the specific deficit of ADHD – evidence from brain electrical activity J Neural Trans 2004 111 841 864 10.1007/s00702-003-0040-8
Leth-Steensen C King Elbaz Z Douglas VI Mean response times, variability, and skew in the responding of ADHD children: a response time distributional approach Acta Psychologica 2000 104 167 190 10900704 10.1016/S0001-6918(00)00019-6
Zentall SS Zentall TR Optimal stimulation: a model of disordered activity and performance in normal and deviant children Psychol Bull 1983 94 446 471 6657825 10.1037//0033-2909.94.3.446
Sergeant JA Oosterlaan J van der Meere JJ Quay HC, Hogan A Information processing and energetic factors in attention-deficit/hyperactivity disorder Handbook of disruptive behaviour disorders 1999 New York: Plenum Press 75 104
Mason DJ Humphreys GW Kent LS Exploring selective attention in ADHD: visual search through space and time J Child Psychol Psychiatry 2003 44 1158 1176 14626457 10.1111/1469-7610.00204
van der Meere J Gozal D, Molfese DL ADHD and state regulation Attention deficit hyperactivity disorder: from genes to patients 2005 NJ: Humana Press Inc 413 422
Kuntsi J Stevenson J Oosterlaan J Sonuga-Barke EJS Test-retest reliability of a new delay aversion task and executive function measures B J Dev Psychol 2001 19 339 338 10.1348/026151001166137
Kuntsi J Stevenson J Psychological mechanisms in hyperactivity: II The role of genetic factors J Child Psychol Psychiatr 2001 42 211 219 10.1017/S0021963001006710
Nigg J Blaskey L Stawicki JA Sachek J Evaluating the endophenotype model of ADHD neuropsychological deficit: results for parents and siblings with DSM-IV ADHD combined and inattentive subtypes J Abnorm Psychol 2004 113 614 625 15535793 10.1037/0021-843X.113.4.614
Conners CK Sitarenios G Parker JDA Epstein JN The revised Conners' Parent Rating Scale (CPRS-R): factor structure, reliability, and criterion validity J Abnorm Child Psychol 1998 26 257 268 9700518 10.1023/A:1022602400621
Conners CK Sitarenios G Parker JDA Epstein JN Revision and restandardization of the Conners Teacher Rating Scale (CTRS-R): factor structure, reliability, and criterion validity J Abnorm Child Psychol 1998 26 279 291 9700520 10.1023/A:1022606501530
Conners CK Conners' Rating Scales-Revised: technical manual 1997 North Tonawanda: Multi-Health Systems Inc
Wechsler D Wechsler Intelligence Scale for Children 1991 Third UK. London: The Psychological Corporation
Sattler JM Assessment of children: WISC-III and WPPSI-R supplement San Diego: CA 1992
Rousson V Gasser T Seifert B Assessing intrarater, interrater and test-retest reliability of continuous measurements Stat Med 2002 21 3431 3446 12407682 10.1002/sim.1253
Kuntsi J Eley TC Taylor A Hughes C Asherson P Caspi A Moffitt TE Co-occurrence of ADHD and low IQ has genetic origins Am J Med Genet (Neuropsychiatric Genetics) 2004 124B 41 47 10.1002/ajmg.b.20076
Ryan JJ Lopez SJ Sumerall SW Dorman WI, Hersen M Understanding test construction Understanding psychological assessment 2001 New York: Kluwer Academic/Plenum Publishers
Fergusson DM Horwood LJ Lynskey MT The effects of conduct disorder and attention deficit in middle childhood on offending and scholastic ability at age 13 J Child Psychol Psychiatr 1993 34 899 916
Goodman R Simonoff E Stevenson J The impact of child IQ, parent IQ and sibling IQ on child behavioural deviance scores J Child Psychol Psychiatr 1995 36 409 425
Rapport MD Scanlan SW Denney CB Attention-deficit/hyperactivity disorder and scholastic achievement: A model of dual developmental pathways J Child Psychol Psychiatry 1999 40 1169 1183 10604396 10.1017/S0021963099004618
Oosterlaan J Scheres A Sergeant JA Which executive functioning deficits are associated with AD/HD, ODD/CD and combordid AD/HD + ODD/CD? J Abnorm Child Psychol 2005 33 69 85 15759592 10.1007/s10802-005-0935-y
Martin N Scourfield J McGuffin P Observer effects and heritability of childhood attention-deficit hyperactivity disorder symptoms B J Psychiatry 2002 180 260 265 10.1192/bjp.180.3.260
Gomez R Harvey J Quick C Scharer I Harris G DSM-IV AD/HD: confirmatory factor models, prevalence, and gender and age differences based on parent and teacher ratings of Australian primary school children J Child Psychol Psychiatr 1999 40 265 274 10.1017/S0021963098003321
Nadder TS Silberg JL Rutter M Maes HH Eaves LE Comparisons of multiple measures of ADHD symptomatology: a multivariate genetic analysis J Child Psychol Psychiatry 2001 42 475 486 11383963 10.1017/S0021963001007119
Nadder TS Rutter M Silberg JL Maes HH Eaves LJ Genetic effects on the variation and covariation of attention deficit-hyperactivity disorder (ADHD) and oppositional-defiant disorder/conduct disorder (ODD/CD) symptomatologies across informant and occasion of measurement Psychol Med 2002 32 39 53 11883729
Chhabildas NA Pennington BF Willcutt EG A comparison of the neuropsychological profiles of the DSM-IV subtypes of ADHD J Abnorm Child Psychol 2001 29 529 540 11761286 10.1023/A:1012281226028
Todd RD Sitdhiraksa N Reich W Ji TH Joyner CA Heath AC Neuman RJ Discrimination of DSM-IV and latent class attention-deficit/hyperactivity disorder subtypes by educational and cognitive performance in a population-based sample of child and adolescent twins J Am Acad Child Adolesc Psychiatry 2002 41 820 828 12108807 10.1097/00004583-200207000-00014
Swanson JM Sergeant JA Taylor E Sonuga-Barke EJS Jensen PS Cantwell DP Attention-deficit hyperactivity disorder and hyperkinetic disorder Lancet 1998 351 429 433 9482319 10.1016/S0140-6736(97)11450-7
|
16262903
|
PMC1291376
|
CC BY
|
2021-01-04 16:33:02
|
no
|
BMC Psychiatry. 2005 Nov 1; 5:40
|
utf-8
|
BMC Psychiatry
| 2,005 |
10.1186/1471-244X-5-40
|
oa_comm
|
==== Front
BMC Vet ResBMC Veterinary Research1746-6148BioMed Central London 1746-6148-1-71625963110.1186/1746-6148-1-7Research ArticlePorcine Circovirus type 2 (PCV2) causes apoptosis in experimentally inoculated BALB/c mice Kiupel Matti [email protected] Gregory W [email protected] Elizabeth J [email protected] Adam [email protected] Harm [email protected] Suresh K [email protected] Animal Disease Diagnostic Laboratory, Purdue University, West Lafayette, Indiana, 47906-1175, USA2 Department of Veterinary Pathobiology, Purdue University, West Lafayette, Indiana, 47906, USA3 Lilly Research Laboratories, Indianapolis, Indiana 46285, USA2005 31 10 2005 1 7 7 8 6 2005 31 10 2005 Copyright © 2005 Kiupel et al; licensee BioMed Central Ltd.2005Kiupel et al; licensee BioMed Central Ltd.This is an Open Access article distributed under the terms of the Creative Commons Attribution License (), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
Background
We have previously described microscopic and electron microscopic alterations in lymphoid organs of PCV2 inoculated mice as apoptosis. In this study we wanted to investigate the molecular pathogenetic mechanism of PCV2-induced apoptosis. Eight-week old BALB/c mice were either sham inoculated (control mice) or inoculated intraperitoneally (ip) and intranasally (in) with a single (sPCV mice) or multiple (mPCV mice) doses of PCV2. Four control mice and 4 sPCV mice were sacrificed 7, 14, 28 and 42 days post inoculation (PI). All 4 mPCV mice were sacrificed 42 days PI. Following necropsy, immunohistochemistry for caspase 3 and in-situ TUNEL assay were performed on sections of spleen, lymph nodes, thymus and ileum from control, sPCV and mPCV mice. In addition, total RNA was extracted from spleens of control, sPCV and mPCV mice for simultaneous detection and semiquantitation of bcl-2 homologues and various caspase mRNAs using a multiprobe RNase protection assay system.
Results
PCV2 replicated and was associated with apoptosis in spleens, lymph nodes and Peyer's patches of infected BALB/c mice. Upregulation of caspase 1, 2, 3, 6, 7, 8, 11 and 12 and upregulation for the transcripts of apoptosis inhibitors bcl-2, bcl-w and bcl-X and apoptosis promoters' bax, bak and bad was detected in spleens of sPCV and mPCV mice, but not control mice. Apoptosis was further confirmed by light and electron microscopic morphology as well as by positive TUNEL assay and detection of activated caspase 3. PCV2 nucleic acid was detected by in-situ hybridization in the nuclei and cytoplasm of such apoptotic cells.
Conclusion
The data presented here support the hypothesis that PCV2 induces apoptosis mediated through the activation of caspases 8 and 3 in the spleens of infected mice.
==== Body
Background
Circoviruses, the smallest animal DNA viruses known so far, have a single copy of circular single-stranded ambisense DNA genome that varies in size between 1.7 and 2.3 kb. Animal circoviruses have been demonstrated in chickens (chicken anemia virus, ChAV, [49]), pigs (porcine circovirus, PCV, [45]), pigeons (pigeon circovirus, [47]) and psittacines (psittacine beak and feather disease virus, PBFDV, [36]). Porcine circovirus (PCV), an approximately 17 nm in diameter, non-enveloped virus with icosahedral symmetry, was originally identified as a noncytopathic contaminant of the PK-15 porcine kidney cell line [44]. The genome of PK-15 derived virus has been sequenced [28] and isolates of PCV that are genetically like PK-15 cell PCV are referred to as PCV1 [29]. Inoculation studies in pigs using PK-15 derived PCV1 did not result in clinical disease [1,46]. In 1990's, field strains of PCV have been found in lesions of pigs with postweaning multisystemic wasting syndrome (PMWS) [2,5,6,10,17,31,33,42]. Isolates of PMWS-associated PCV are genetically and antigenically different from the PK-15 cell PCV and are referred to as PCV2 [29].
PMWS is clinically characterized by progressive weight loss, dyspnea, tachypnea and less frequent diarrhea, pallor and icterus in pigs [5]. Gross lesions in pigs with PMWS consist of generalized lymphadenopathy in combination with less frequent lesions in the lungs, liver, kidneys and stomach [5,16]. The most consistent microscopic lesions in affected pigs are in lymphoid organs and include lymphoid cell depletion and granulomatous inflammation with inconsistently occurring intracytoplasmic viral inclusion bodies in macrophages [5,10,17,31,40]. PCV nucleic acid and antigen have been demonstrated within lesions in multiple organs of naturally diseased pigs with PMWS [5,6,10,18,31,40]. So far, isolates of PCV from pigs with PMWS have been identified nearly exclusively as PCV2 [2,14,15,29,31]. However, the role of PCV2 in PMWS remains unclear. PCV2 infection alone produces asymptomatic infection in germ-free pigs without evidence of overt PMWS [21]. In contrast, coinfection of PCV2 with porcine parvovirus (PPV) or concurrent injection with keyhole limpet hemocyanin in incomplete Freund's adjuvant enhanced replication of PCV, and caused PMWS [11,19,21,22].
On the basis of histopathological changes in naturally and experimentally infected pigs, it appears that PCV2 induces apoptosis in pigs in vivo. Hepatic disease has been implicated as the major cause of icterus, wasting and death in naturally occurring and experimentally reproduced cases of PMWS [21,22,39]. The predominant hepatic lesion has been described as single cell necrosis [3,11,21,22] or apoptosis [39] of hepatocytes. Only recently, ORF3 of PCV2 has been shown to play a major role in the induction of virus-induced apoptosis through activation of caspase-8 and caspase-3 pathways, but not caspase-9 [24]. However, ORF3 is not essential for viral replication and recent studies indicate that apoptosis is not a remarkable feature in PMWS lymphoid lesion development [38]. On the contrary, when assessing the proliferation/apoptosis ratio to determine cell turnover, decreased cell proliferation and not increased apoptosis was concluded to be the most important variable leading to cell depletion in PMWS lymphoid tissues [25].
We demonstrated the replication of PCV2 in 8-week old BALB/c mice following experimental inoculation [20]. In mice PCV2 caused only mild microscopic lesions in lymphoid organs characterized by proliferation and morphologic evidence of apoptosis of cells in germinal centers and mild lymphoid depletion of the paracortex. PCV2 nucleic acid was detected in the nuclei and cytoplasm of apoptotic cells in germinal centers.
The aim of this study was to confirm the described microscopic and electron microscopic alterations in lymphoid organs of mice as apoptosis and to study the molecular pathogenetic mechanism of PCV2-induced apoptosis.
Results
PCV2 induces apoptosis in lymphoid organs of BALB/c mice
PCV2-inoculated mice, but not control mice, had enlarged germinal centers in lymphoid tissues that were composed of cells morphologically typical of lymphoblasts and histiocytes. Numerous apoptotic cells were detected in spleen, lymph nodes and Peyer's patches of PCV2-inoculated mice but not control mice by concurrent positive TUNEL assay and immunostaining for activated caspase 3. Apoptosis was further confirmed by electron microscopy as previously described [20]. Apoptotic cells were detected in enlarged germinal centers of spleens and lymph nodes (Figures 1a–c and 2a) of sPCV mice on days 7, 14, 28 and 42 PI and mPCV mice on day 42 PI. In addition, many apoptotic cells were found in the sinusoids of the spleen (Figure 2b) of the sPCV mice at 28 days PI. There were few apoptotic cells in spleens of sPCV mice on day 7 PI. The number of apoptotic cells was significantly increased on day 14 PI and remained the same on days 28 and 42 PI in sPCV mice and on day 42 PI in mPCV mice. Apoptotic cells, identified by activated caspase-3 and TUNEL corresponded to cells with microscopic changes typical of apoptosis (Figure 1b and 2a). The cytoplasm and less frequently the nuclei of many apoptotic cells in germinal centers were positive for PCV nucleic acid by in-situ hybridization (Figure 1c). However, no PCV nucleic acid was identified in apoptotic cells in the sinusoids of the spleen of sPCV mice at 28 days PI.
Figure 1 Spleen from a PCV2-inoculated sPCV mouse 28 days PI. Large numbers of apoptotic cells (arrows) in germinal centers. H & E staining, Bar = 40 μm. Same germinal center in spleen as in figure 1. Apoptotic cells (arrows) stain dark brown with TUNEL. In-situ hybridization, hematoxylin counterstain, Bar = 20 μm. Same germinal center in spleen as in figure 1. The majority of apoptotic cells (arrows) stain dark blue for PCV2 nucleic acid. In-situ hybridization, hematoxylin counterstain, Bar = 60 μm
Figure 2 Same germinal center in spleen as in figure 1. Apoptotic cells (arrows) stain dark brown for active caspase 3. Immunohistochemistry, hematoxylin counterstain, Bar = 80 μm. Spleen from a PCV2-inoculated sPCV mouse 28 days PI. Large numbers of apoptotic cells (arrows) in sinusoids and germinal centers stain dark brown with TUNEL. In-situ hybridization, hematoxylin counterstain, Bar = 100 μm
PCV2 increases mRNA of caspases in spleens of BALB/c mice
RPAs from spleens of control mice demonstrated barely detectable levels for caspase 2, 3 and 8 mRNA, but caspase 1, 6, 7, 11 and 12 mRNA were not detected. The responses were essentially unchanged over the 42-day period. In contrast, transcripts for caspase 1, 2, 3, 6, 7, 8, 11 and 12 increased at 7 days PI and were even stronger at 14, 28 and 42 days PI in sPCV mice and at 42 days PI in mPCV mice. The enhanced mRNA expression was strongest for caspase 1, 2, 3 and 8 (Figure 3). Transcripts of caspase 14 were not upregulated.
Figure 3 RNAse protection assay (RPA) for various caspase mRNAs from spleens of individual BALB/c mice. The probe and a representative RPA of a control mouse 28 days post sham inoculation are shown on the left. RPAs for representative BALB/c mice that were inoculated with a single (sPCV2) or multiple (mPCV2) doses of PCV2 and were euthanized at 14, 28 or 42 days PI, are shown on the right. Transcripts for caspases 1, 2, 3, 6, 7, 8, 11 and 12 are increased in sPCV and mPCV mice when compared to control mice. Upregulation of caspases 1, 2, 3, 6 and 8 are strongest and bands for caspase 11 and 12 are only faint. Quantitation of mRNA is identical for all samples.
PCV2 increases mRNA of promoters and inhibitors of apoptosis in spleens of BALB/c mice
Spleens of control mice had detectable levels of apoptosis inhibitors bcl-2, bcl-w and bcl-X mRNA and apoptosis promoters bax and bak mRNA. Bands for bcl-2 and bak were only faint. The responses were essentially unchanged over the 42-day period. In contrast, the transcripts of bcl-2, bcl-w, bcl-X, bax, bak and bad increased at 7 days PI and remained elevated at 14, 28 and 42 days PI in sPCV mice and at 42 days PI in mPCV mice. The expression of these transcripts was strongest in sPCV mice at 28 days PI. In addition, there was increased expression of bfl-1 mRNA in sPCV mice at 7, 14, 28 and 42 days PI and in mPCV mice at 42 days PI (Figure 4).
Figure 4 RNAse protection assay (RPA) for various bcl-2 homologue mRNAs from spleens of individual BALB/c mice. A representative RPA of a control mouse 28 days post sham inoculation is shown on the left. RPAs for representative BALB/c mice that were inoculated with a single (sPCV2) or multiple (mPCV2) doses of PCV2 and were euthanized at 14, 28 or 42 days PI, are shown on the right. Transcripts of apoptotic inhibitors bcl-2, bcl-w and bcl-X and apoptotic promoters bax, bak and bad are increased in sPCV and mPCV mice when compared to control mice. In addition, there was also upregulation of transcripts for bfl-1 in sPCV and mPCV mice. Quantitation of mRNA is identical for all samples.
Discussion
PCV2 caused apoptosis in spleens, lymph nodes and Peyer's patches of infected BALB/c mice, as confirmed by light and electron microscopic morphology [20] as well as by positive TUNEL assay and detection of activated caspase 3. Previously, PCV2 nucleic acid had been detected in the cytoplasm and rarely in the nucleus of putative histiocytes in follicular centers [20]. In serial sections of lymphoid tissue, putative histiocytes in follicular centers that were positive for activated caspase 3 were also positive for PCV2 nucleic acid. In contrast, no PCV2 nucleic acid was identified in apoptotic cells in the sinusoids of spleens from PCV2 infected mice.
Morphologic evidence of apoptosis was detected in PCV2 infected mice, but not in control mice that had been inoculated with tissue culture fluid. However, RPAs from spleens of control mice demonstrated barely detectable levels for caspase 3 mRNA. These data most likely represent some "basal" level of apoptosis involved in morphogenesis and homeostasis of lymphoid tissues that were not appreciated microscopically.
The mechanism of PCV2-induced apoptosis in mice is unknown. A possible mechanism could be the direct effect of a PCV2-encoded protein on virus infected and uninfected bystander cells. Viruses use various strategies to induce or inhibit apoptosis as a mechanism to enhance replication. Chicken anemia virus (ChAV), a member of the family of Circoviridae, encodes a 14-kDa proline-rich protein referred to as apoptin that causes apoptosis in thymic lymphoblasts, intra- and extra-sinusoidal hemocytoblasts, and reticular cells. Expression of apoptin induces apoptosis in cultured transformed chicken mononuclear cells and in chicken thymocytes, both cells that are susceptible to infection with ChAV, but not in chicken embryo fibroblasts, which are not susceptible to infection with ChAV [34,35]. Apoptin also induces apoptosis in transformed and tumorigenic human cells in culture [34,35], but does not induce apoptosis in normal diploid human cells [9,50]. Apoptosis induced by apoptin is p53-independent, but requires activation of caspase 3 [7,8,37,51,52]. In normal cells, apoptin is localized mainly in the cytoplasm, whereas in transformed cells it is found in the nucleus [9]. A specific protein encoded by PCV2, similar to apoptin, could cause the detected apoptosis in lymphoid organs of PCV2 infected mice. So far, 11 potential open reading frames (ORF) have been identified in the genome of PCV2 [14] with six ORFs encoding putative proteins larger than 6 kDa. The function of the putative proteins encoded by the other ORFs 4 to 6 of PCV2 is unknown. Only ORF1 and ORF2 have been shown to encode the origin of replication and a replication-associated protein [26,27] and a structural protein [32], respectively. ORF2 of PCV2 shares a highly conserved basic N-terminal region that is similar to the major structural protein found in ChAV [30]. ORF3 has recently been characterized as a non-structural protein that is not essential for PCV2 replication in cultured PK15 cells and plays a major role in virus-induced apoptosis in cultured cells by activating initiator caspase-8 and effector caspase-3 pathways [24].
In an attempt to understand the mechanism of induction of apoptosis in the spleen of PCV2-infected BALB/c mice we investigated the regulation of mRNAs of various caspases and bcl-2 homologues using a multiprobe RNAse protection assay (RPA). Upregulation of caspase 1, 2, 3, 6, 7, 8, 11 and 12, but not 14 was detected. Among the 14 caspases that have been identified so far, caspases 3, 6 and 7 play a central role in driving the apoptotic effector pathway [23,43]. The expression of all three was increased in spleens of PCV2 infected mice. Caspases are produced as zymogens and have to be activated to initiate their function. In this study we confirmed activation of only caspase 3 in apoptotic macrophages using immunohistochemistry. Activated caspase 3 cleaves and inactivates the inhibitor for caspase-activated DNase (CAD), allowing CAD to enter the nucleus and degrade chromosomal DNA [12]. Activation of caspase 3 has been observed in various types of cells undergoing apoptosis induced by a variety of stimuli. In immune system-responsive cells, such as macrophages and lymphocytes, activation of caspase 3 has been shown to be required for apoptosis induced by Fas-FasL or TNF-α-TNFR interactions [43]. Because caspase 3 was the most significantly upregulated effector caspase and because it was detected in activated form in the cytoplasm of apoptotic macrophages, apoptosis in the spleens of PCV2 infected mice is likely mediated through the activation of caspase 3. Caspase 3 following initiation by caspase 8 has been shown to play a major role in the induction of PCV2-induced apoptosis in porcine cells [24]. In our studies transcripts for caspases 3 and 8 were both strongly elevated possibly suggesting a similar mechanism. Apoptosis in PCV2 infected porcine cells was not associated with initiator caspase 9. Unfortunately, caspase 9 was not included in the commercially available RPA, but should be tested in the future. The increased expression of the transcripts of apoptosis inhibitors bcl-2, bcl-w and bcl-X and apoptosis promoters' bax, bak and bad in the spleens of PCV2 infected mice, may indicate activation of an apoptotic pathway and attempts by the cells to prevent apoptosis. Based on samples of RNA extracted from whole tissue, it is impossible to conclude the specific pathway that led to apoptosis. In vitro experiments are required to further elucidate this mechanism.
In contrast to the other caspases, the primary functions of caspase 1 and 11 are generally believed to be proinflammatory [13,43]. Caspase 1 processes IFN-γ-inducing factor and regulates LPS-induced IFN-γ production. Activation of caspase 1, also referred to as IL-1β converting enzyme, in macrophages leads to cleavage of the precursor IL-1β into active IL-1 that may subsequently initiate an intense host inflammatory response. Therefore, it has been proposed that activation of caspase 1 converts a proapoptotic event into a proinflammatory one [53]. The strong upregulation of caspase 1 in this study correlated with upregulation of IL-1β mRNA in spleens of PCV2 infected mice (data not shown). We speculate that caspase 1 might play a role in inducing inflammation in lymphoid tissues of PCV2 infected mice.
Apoptosis has been confirmed in lymphoid tissues of PCV2 infected pigs [25,38]. However, it has been speculated, that lymphoid tissue depletion in PCV2-infected pigs is mainly related to decreased proliferative activity of lymphoid cells, and is caused by a long-standing absence of lymph node positive growth factors (mainly cytokines) produced by lymphocyte activation [41] rather than apoptosis [25]. In view of the knowledge on apoptosis in lymphoid tissues in pigs with PCV2 infection, the murine model presented here appears different from PMWS in pigs. PCV2 infected mice not only don't develop clinical signs and lesions typical of PMWS, but also exhibit an increase of apoptosis in lymphoid tissues, which does not seem to be the cause of lymphoid depletion in pigs.
Methods
Origin of tissues
Sixteen BALB/c mice (control mice) were sham inoculated and sixteen BALB/c mice were inoculated with a single (sPCV) dose of PCV2. In addition 4 BALB/c mice were inoculated with multiple (mPCV) doses of PCV2 6 times on days 0, 7, 14, 21, 28 and 35. Mice were inoculated intraperitoneally and intranasally with 0.2 ml of tissue culture fluid (control mice) or with 0.2 ml of PCV2 inoculum with a titer of 105 TCID50/1.0 ml (sPCV and mPCV mice). Four control and 4 sPCV mice were sacrificed 7, 14, 28 and 42 days post-inoculation (PI) and all 4 mPCV mice were sacrificed on day 42 PI. PCV2 infection was confirmed by PCR and in-situ hybridization and microscopic lesions as described [20]. Sections of tissue, including spleen, mediastinal and mesenteric lymph nodes, thymus and ileum were fixed in 10% buffered formalin for microscopic and immunohistochemical evaluation. Samples of spleen from each mouse were collected during necropsy and were frozen immediately in liquid nitrogen at -80°C for RNA extraction. This study met the standards of the Guide for the Care and Use of Laboratory Animals and the study protocol was approved by the Purdue Animal Care and Use Committee (PACUC #98–112).
PCV2 detection
In-situ hybridization for demonstration of PCV2 nucleic acid was performed as previously described using a PCV2-specific oligoprobe [20]. Briefly, tissue sections were deparaffinized, digested with 0.25% pepsin and prehybridized. Hybridization was performed for 5 minutes at 105°C and 60 minutes at 37°C with a specific 3'-end digoxigenin labeled oligoprobe (5'-CCAACAAAATCTCTATACCC-3') at a concentration of 5 μl/1 ml using a commercial workstation (Fischer Scientific, Pittsburgh, PA). The detection system consisted of an anti-digoxigenin antibody (Boehringer Mannheim Biochemica, Indianapolis, IN) conjugated with alkaline phosphatase (dilution 1:500) and the substrates "NBT/X-Phos"(Nitro-blue tetrazolium/5-Bromo-4-chloro-3-indolylphosphate, Boehringer Mannheim Biochemica, Indianapolis, IN). Controls included lymphoid tissue from PCV2 infected pigs and sections of spleen and liver from PCV2 negative pigs and mice [19].
Transmission electron microscopy
Samples of spleen from control and inoculated mice were fixed in 4% glutaraldehyde and postfixed by osmium tetroxide as previously described [20]. Tissues were embedded in epon and ultrathin sections were cut and stained with lead citrate and uranyl acetate. Sections were examined and photographed using a Joel JEM-1000CX transmission electron microscope.
Cell death analysis
Following deparaffinization, selected sections of spleen, lymph nodes, thymus and ileum from control, sPCV and mPCV mice were incubated in target retrieval solution (Dako Corporation, Santa Barbara, CA) for 15 min at 95–99°C. Immunostaining was performed using the Dako autostainer (Dako Corporation, Santa Barbara, CA) by incubation with a polyclonal anti-mouse anti-activated caspase 3 antibody (R&D Systems, Minneapolis, MN) at a dilution of 1:100 at -4°C overnight, followed by a 1:500 dilution of a biotinylated goat anti-rabbit secondary antibody (Dako Corporation, Santa Barbara, CA). The antibody binding was localized using a peroxidase labeled streptavidin-biotin complex (Dako Corporation, Santa Barbara, CA) followed by diaminobenzidine as a chromogen substrate. After a final wash in automation buffer, the sections were counterstained with Lerner's hematoxylin. As negative controls, sections were incubated with isotype control antibodies.
For selected sections of spleen from control, sPCV and mPCV mice, DNA fragmentation was detected in situ by terminal deoxynucleotidyl transferase-mediated dUTP-DIG nick end labelening (TUNEL) using a commercial system (R&D Systems, 2000).
RNAse Protection Assay (RPA)
Prior to analysis, total RNA was extracted from spleens of control, sPCV and mPCV mice that had been frozen at -80°C as described elsewhere [4]. Simultaneous detection and semiquantitation of bcl-2 homologues (mAPO-2) and caspase (mAPO-1) mRNAs were accomplished with the multiprobe RNase protection assay system from Pharmingen (San Diego, Calif). Briefly, a mixture of [32P]CTP-labelled antisense riboprobes was generated from bcl-2 homologues and caspase templates. These panels included templates for the murine housekeeping genes encoding L32 (a murine ribosomal protein) and glyceraldehyde-3-phosphate dehydrogenase (GAPDH, Young & Trowsdale, 1985), to ensure equal loading of total RNA onto the gels. A predetermined amount of total spleen RNA was hybridized overnight at 56°C with 300 pg of the 32P-antisense riboprobe mixture. After hybridization, the samples were digested with 2,500U of T1 nuclease (Gibco-BRL, Gaithersburg, Md.). Nuclease-protected RNA fragments were purified by ethanol precipitation. After purification, the samples were resolved on a 4.5% polyacrylamide sequencing gel. Protected bands were observed after exposure of the gels to Fuji X-ray film (Fisher, Itasca, Ill.). The specific bcl-2 homologues and caspase bands were identified on the basis of their individual migration patterns in comparison with the undigested probes. The intensities of the bands from the mRNA of spleens of sPCV and mPCV mice were compared to those of uninfected control spleens.
RPAs of each panel of bcl-2 homologues and caspase mRNAs were performed at least 3 times with similar results for samples of spleen from control, sPCV and mPCV mice.
Authors' contributions
MK designed and coordinated the study, carried out TEM and PCV detection and drafted the manuscript, GWS participated in the study design, coordination and interpretation, SKM participated in study design and interpretation, AN carried out the RPAs, EJG carried out the cell death analysis, HH participated in study design and interpretation. All authors read and approved the final manuscript
Acknowledgements
We would like to thank Carmen Zambrano for her laboratory work and Sam Royer for his photographic work.
==== Refs
Allan GM McNeilly F Cassidy JP Reilly GAC Adair B Ellis JA McNulty MS Pathogenesis of porcine circovirus; experimental infections of colostrum deprived piglets and examination of pig foetal material Vet Micro 1995 44 49 64 10.1016/0378-1135(94)00136-K
Allan GM McNeilly F Kennedy S Daft B Clarke EG Ellis JA Haines DM Meehan BM Adair B Isolation of porcine circovirus-like viruses from pigs with a wasting disease in the USA and Europe J Vet Diagn Invest 1998 10 3 10 9526853
Bolin SR Stoffregen WC Nayar GPS Hamel AL Postweaning multisystemic wasting syndrome induced after experimental inoculation of ceasarian-derived, colostrums-deprived piglets with type 2 porcine circovirus J Vet Diagn Invest 2001 13 185 194 11482594
Chomczynski P Sacchi M Single-step method of RNA isolation by acid guanidinium thiocyanate-phenol-chloroform extraction Anal Biochem 1987 162 156 159 2440339 10.1016/0003-2697(87)90021-2
Clark EG Post-weaning multisystemic wasting syndrome Proc 28th Ann Meet Am Assoc Swine Pract, Quebec City, Quebec 499 501 March 1–4 1997
Daft B Nordhausen RW Latimer KS Niagro FD Interstitial pneumonia and lymphadenopathy associated with circoviral infection in a six-week-old pig Proc 39th Ann Meet Am Assoc Vet Lab Diagn, Little Rock, AR 32 October 12–18 1996
Danen-van Oorschot AA van der Eb AJ MH Noteborn BCL-2 stimulates apoptin-induced apoptosis Adv Exp Med Biol 1999 457 245 249 10500799
Danen-van Oorschot AA van der Eb AJ Noteborn MH The chicken anemia virus-derived protein apoptin requires activation of caspases for induction of apoptosis in human tumor cells J Virol 2000 74 7072 7028 10888647 10.1128/JVI.74.15.7072-7078.2000
Danen-van Oorschot AA Fischer DF Grimbergen JM Klein B Zhuang S Falkenburg JH Backendorf C Quax PH van der Eb AJ Noteborn MH Apoptin induces apoptosis in human transformed and malignant cells but not in normal cells Proc Natl Acad Sci USA 1997 94 5843 5847 9159162 10.1073/pnas.94.11.5843
Ellis JA Hassard L Clark EG Harding JCS Allan GM Willson P Strokappe J Martin K McNeilly F Meehan BM Todd D Haines DM Isolation of circovirus from lesions of pigs with postweaning multisystemic wasting syndrome Can Vet J 1998 39 44 51 9442952
Ellis JA Krakowka SJ Lairmore M Haines DM Bratanich AE Clark G Allan GM Konoby C Hassard L Meehan BM Martin K Harding SJC Kennedy S McNeilly F Reproduction of lesions of postweaning multisystemic wasting syndrome in gnotobiotic pigs J Vet Diagn Invest 1999 11 3 14 9925205
Enari M Sakahira H Yokoyama H Okawa K Iwamatsu A Nagata S A caspase-activated DNase that degrades DNA during apoptosis, and its inhibitor ICAD Nature 1998 391 43 50 9422506 10.1038/34112
Ghayur T Banerjee S Hugunin M Butler D Herzog L Carter A Quintal L Sekut L Talanian R Paskind M Caspase-1 processes IFN-gamma-inducing factor and regulates LPS-induced IFN-gamma production Nature 1997 386 619 623 9121587 10.1038/386619a0
Hamel AL Lin LL Nayar GPS Nucleotide sequence of porcine circovirus associated with postweaning multisystemic wasting syndrome in pigs J Virol 1998 72 5262 5267 9573301
Hamel AL Lin LL Sachvie C Grudeski E Nayar GPS PCR detection and characterization of type 2 porcine circovirus Can J Vet Res 2000 64 44 52 10680656
Harding JCS Clark EG Recognizing and diagnosing postweaning multisystemic wasting syndrome (PMWS) Swine Health Prod 1997 5 201 203
Kiupel M Stevenson GW Mittal SK Clark EG Haines DM Circovirus-like viral associated disease in weaned pigs in Indiana Vet Pathol 1998 35 303 307 9684976
Kiupel M Stevenson GW Kanitz CL Anothayanontha L Latimer KS Mittal SK Cellular localization of porcine circovirus in postweaning pigs with chronic wasting disease Eur J Vet Pathol 1999 5 45 50
Kiupel M Stevenson GW Choi J Latimer KS Kanitz CL Mittal SK Production of clinical disease and lesions typical of Postweaning Multisystemic Wasting Syndrome (PMWS) in experimentally inoculated gnotobiotic pigs Proc 43rd Ann Meet Am Assoc Vet Lab Diagn, Birmingham, AL 2 October 21–23 2000
Kiupel M Stevenson GW Latimer KS Choi J Kanitz CL Mittal SK Viral replication and lesions in BALB/c mice experimentally inoculated with Porcine Circovirus (PCV) isolated from a pig with Postweaning Multisystemic Wasting Disease (PMWS) Vet Path 2001 38 74 82 11199167 10.1354/vp.38-1-74
Krakowka S Ellis JA Meehan B Kennedy S McNeilly F Allan G Viral wasting syndrome of swine: Experimental reproduction of Postweaning multisystemic wasting syndrome in gnotobiotic swine by coinfection with porcine circovirus 2 and porcine parvovirus Vet Path 2000 37 254 263 10810990 10.1354/vp.37-3-254
Krakowka S Ellis JA McNeilly F Ringler S Rings DM Allan G Activation of the immune system is the pivotal event in the reproduction of wasting disease in pigs infected with porcine circovirus-2 (PCV-2) Vet Path 2001 38 31 42 11199162 10.1354/vp.38-1-31
Li P Nijhawan D Budihardjo I Srinivasula SM Ahmad M Alnemri E Wang X Cytochrome c and dATP-dependent formation of Apaf-1/caspase-9 complex initiates an apoptotic protease cascade Cell 1997 91 479 489 9390557 10.1016/S0092-8674(00)80434-1
Liu J Chen I Kwang J Characterization of a previously unidentified viral protein in porcine circovirus type 2-infected cells and its role in virus-induced apoptosis J Virol 2005 79 8262 8274 15956572 10.1128/JVI.79.13.8262-8274.2005
Mandrioli L Sarli G Panarese S Baldoni S Marcato PS Apoptosis and proliferative activity in lymph node reaction in postweaning multisystemic wasting syndrome (PMWS) Vet Immunol Immunopathol 2004 97 25 37 14700535 10.1016/j.vetimm.2003.08.017
Mankertz A Hillenbrand B Replication of porcine circovirus type 1 requires two proteins encoded by the viral rep gene Virol 2001 279 429 438 10.1006/viro.2000.0730
Mankertz A Persson F Mankertz J Blaess G Buhk HJ Mapping and characterization of the origin of DNA replication of porcine circovirus J Gen Virol 1997 71 2562 2566
Meehan BM Creelan JL McNulty MS Todd D Sequence of porcine circovirus DNA: affinities with plant circoviruses J Gen Virol 1997 78 221 227 9010307
Meehan BM McNeillly F Todd D Kennedy S Jewhurst VA Ellis JA Hassard LE Clark EG Haines DM Allan GM Characterization of novel circovirus DNAs associated with wasting syndromes in pigs J Gen Virol 1998 79 2171 2179 9747726
Meehan BM Todd D Creelan JL Earle JAP Hoey EM McNulty MS Characterization of viral DNAs from cells infected with CAA: sequence analysis of the cloned replicative form and transfection capabilities of cloned genome fragments Arch Virol 1992 124 301 309 1605740 10.1007/BF01309811
Morozov I Sirinarumitr T Sorden SD Halbur PG Morgan MK Yoon KJ Paul PS Detection of a novel strain of porcine circovirus in pigs with postweaning multisystemic wasting syndrome J Clin Microbiol 1998 36 2535 2541 9705388
Nawagitgul P Morozow I Bolin SR Harms PA Sorden SD Paul PS Open reading frame 2 of porcine circovirus type 2 encodes a major capsid protein J Gen Virol 2000 81 2281 2287 10950986
Nayar GP Hamel SA Lin L Detection and characterization of porcine circovirus associated with post-weaning multisystemic wasting syndrome in pigs Can Vet J 1997 38 385 386 9187809
Noteborn MH GF DeBoer Van Roozelaar DJ Karreman C Kranenburg O Vos JG Jeurissen SHM Hoeben RC Zantema A Koch G Van Ormondt H Van der Eb AJ Characterization of cloned chicken anemia virus DNA that contains all elements for the infectious replication cycle J Virol 1991 65 3131 3139 1851873
Noteborn MH Todd D Verschueren CAJ De Gauw HWFM Curran WL Veldkamp S Douglas AJ McNulty MS Van der Eb AJ Koch G A single chicken anemia virus protein induces apoptosis J Virol 1994 68 346 351 8254747
Pass DA Perry RA The pathology of psittacine beak and feather disease Austral Vet J 1984 61 69 74 6743145
Pietersen A Noteborn HM Apoptin Adv Exp Med Biol 2000 465 153 161 10810623
Resendes AR Majo Nl Segales J Mateu E Calsamiglia M Domingo M Apoptosis in lymphoid organs of pigs naturally infected by porcine circovirus type 2 J Gen Virol 2004 85 2837 2844 15448345 10.1099/vir.0.80221-0
Rosell C Segales J Domingo M Hepatitis and staging of hepatic damage in pigs naturally infected with porcine circovirus type 2 Vet Path 2000 37 687 692 11105965 10.1354/vp.37-6-687
Rosell C Segales JM Plana-Duran J Balasch M Rodriguez-Arrioja GM Kennedy S Allan GM McNeilly F Latimer KS Domingo M Pathological, immunohistochemical, and in-situ hybridization studies of natural cases of Postweaning Multisystemic Wasting Syndrome (PMWS) in pigs J Comp Path 1999 120 59 78 10098016 10.1053/jcpa.1998.0258
Sarli G Mandrioli L Laurenti M Sidoli L Cerati C Rolla G Marcato PS Immunohistochemical characterization of the lymph node reaction in pig post-weaning multisystemic wasting syndrome (PMWS) Vet Immunol Immunopathol 2001 83 53 67 11604161 10.1016/S0165-2427(01)00363-4
Segales JM Sitjar M Domingo S First report of post-weaning syndrome in piglets in Spain Vet Rec 1997 141 600 601 9429277
Thornberry NA Lazebnik Y Caspases: Enemies within Science 1998 281 1312 1316 9721091 10.1126/science.281.5381.1312
Tischer I Rasch R Tochtermann G Characterization of papovavirus and picornavirus-like particles in permanent pig kidney cell lines Zentralblatt Bakt Hyg A 1974 226 153 167
Tischer I Gelderblom H Vettermann W Koch MA A very small porcine virus with circular single-stranded DNA Nature 1982 295 64 66 7057875 10.1038/295064a0
Tischer I Mields W Wolff D Vagt M Griem W Studies on epidemiology and pathogenity of porcine circovirus Arch Virol 1986 91 271 276 3778212 10.1007/BF01314286
Woods LW Latimer KS Barr BC Niagro FD Campagnoli RP Nordhausen RW Castro AE Circovirus-like infection in a pigeon J Vet Diagn Invest 1993 5 609 612 8286462
Young JA Trowdale J A processed pseudogene in an intron of the HLA-DP beta 1 chain is a member of the ribosomal protein L31 gene family Nucleic Acid Res 1985 13 8883 8891 3866218
Yuasa N Taniguchi T Yoshida I Isolation and some characteristics of an agent inducing anemia in chicks Avian Dis 1979 23 366
Zhang YH Abrahams PJ van der Eb AJ Noteborn MH The viral protein apoptin induces apoptosis in UV-C-irradiated cells from individuals with various hereditary cancer-prone syndromes Cancer Res 1999 59 3010 3015 10383168
Zhuang SM Landegent JE Verschueren CA Falkenburg JH van Ormondt H van der Eb AJ Noteborn MH Apoptin, a protein encoded by chicken anemia virus, induces cell death in various human hematologic malignant cells in vitro Leukemia Suppl 1995 1 S118 120
Zhuang SM Shvarts A van Ormondt H Jochemsen AG van der Eb AJ Noteborn MH Apoptin, a protein derived from chicken anemia virus, induces p53-independent apoptosis in human osteosarcoma cells Cancer Res 1995 55 486 489 7834613
Zychlinsky A Sansonetti PJ Apoptosis as a proinflammatory event: what we can learn from bacteria-induced cell death Trends Microbiol 1997 5 201 204 9160509 10.1016/S0966-842X(97)01044-5
|
16259631
|
PMC1291377
|
CC BY
|
2021-01-04 16:29:46
|
no
|
BMC Vet Res. 2005 Oct 31; 1:7
|
utf-8
|
BMC Vet Res
| 2,005 |
10.1186/1746-6148-1-7
|
oa_comm
|
==== Front
BMC Vet ResBMC Veterinary Research1746-6148BioMed Central London 1746-6148-1-81626643710.1186/1746-6148-1-8Research ArticleATP-sensitive potassium channel (KATP channel) expression in the normal canine pancreas and in canine insulinomas Donley Vicky R [email protected] Erin K [email protected] Aimee C [email protected] Thomas [email protected] Kansas State University, Department of Clinical Sciences, 1800 Denison Ave, Manhattan, KS 66506-5606, USA2005 2 11 2005 1 8 8 9 6 2005 2 11 2005 Copyright © 2005 Donley et al; licensee BioMed Central Ltd.2005Donley et al; licensee BioMed Central Ltd.This is an Open Access article distributed under the terms of the Creative Commons Attribution License (), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
Background
Pancreatic beta cells express ATP-sensitive potassium (KATP) channels that are needed for normal insulin secretion and are targets for drugs that modulate insulin secretion. The KATP channel is composed of two subunits: a sulfonylurea receptor (SUR 1) and an inward rectifying potassium channel (Kir6.2). KATP channel activity is influenced by the metabolic state of the cell and initiates the ionic events that precede insulin exocytosis. Although drugs that target the KATP channel have the expected effects on insulin secretion in dogs, little is known about molecular aspects of this potassium channel. To learn more about canine beta cell KATP channels, we studied KATP channel expression by the normal canine pancreas and by insulin-secreting tumors of dogs.
Results
Pancreatic tissue from normal dogs and tumor tissue from three dogs with histologically-confirmed insulinomas was examined for expression of KATP channel subunits (SUR1 and Kir6.2) using RT-PCR. Normal canine pancreas expressed SUR1 and Kir6.2 subunits of the KATP channel. The partial nucleotide sequences for SUR1 and Kir6.2 obtained from the normal pancreas showed a high degree of homology to published sequences for other mammalian species. SUR1 and Kir6.2 expression was observed in each of the three canine insulinomas examined. Comparison of short sequences from insulinomas with those obtained from normal pancreas did not reveal any mutations in either SUR1 or Kir6.2 in any of the insulinomas.
Conclusion
Canine pancreatic KATP channels have the same subunit composition as those found in the endocrine pancreases of humans, rats, and mice, suggesting that the canine channel is regulated in a similar fashion as in other species. SUR1 and Kir6.2 expression was found in the three insulinomas examined indicating that unregulated insulin secretion by these tumors does not result from failure to express one or both KATP channel subunits.
==== Body
Background
The ATP-sensitive potassium (KATP) channel is a crucial component of the pancreatic beta cell insulin secretion pathway and a target for drugs that modulate insulin secretion. In the endocrine cells of the pancreas, KATP channel is composed of two subunits: a sulfonylurea receptor (SUR 1), which contains two adenine nucleotide binding domains [1], and an inward rectifying potassium channel (Kir6.2), which conducts potassium ions [2]. The KATP channel is important for sensing the metabolic state of the pancreatic beta cell and initiating the changes in membrane potential that precede insulin exocytosis. In the presence of high concentrations of glucose, beta cell metabolism of the sugar increases ATP production. The resulting rise in the cellular ATP/ADP ratio causes closure of KATP channels, which bind ATP and ADP at different sites [3]. The decrease in potassium conductance caused by channel closure depolarizes the cell membrane and initiates a sequence of ionic events that ultimately trigger insulin secretion [3].
Drugs that target pancreatic KATP channels are used to modulate insulin secretion. Diazoxide and the sulfonylurea class of drugs are the most important examples of drugs that alter KATP channel activity as their principle mechanism of action. These drugs have opposite actions on KATP channel activity which is reflected by their effects on insulin secretion. Diazoxide, a potent KATP channel opener, inhibits insulin secretion, even in the presence of glucose, by "locking" KATP channels in the open configuration, which prevents cellular depolarization and prevents the ionic events that normally precede insulin exocytosis [4]. In contrast, the sulfonylurea drugs have the opposite effect on KATP channel activity. These drugs cause KATP channel closure, even in the absence of glucose, which leads to membrane depolarization and enhanced insulin secretion [5]. SUR 1 was discovered when research into the mechanism of action of the sulfonylurea drugs indicated that these drugs displayed high-affinity binding to a protein that modulated insulin secretion [6]. Diazoxide also binds to SUR 1, although at a different site than the sulfonylureas, since diazoxide does not prevent sulfonylurea binding [7,8]. Endogenous SUR1 ligands, such as alpha-endosulfine, may also regulate insulin secretion [9].
Although sulfonylurea drugs and diazoxide have the expected effects on insulin secretion from normal dogs [10-13], little information is known about the canine beta cell KATP channel that is the molecular target for these drugs. Some but not all canine insulinomas respond to diazoxide [14], suggesting the possibility that mutations in the KATP channel could be responsible for drug failure in some cases. Unregulated insulin secretion by the insulinoma could, in theory, result from any mutation of either SUR 1 or Kir6.2 that promotes KATP channel closure or impairs normal metabolic regulation. Unfortunately, little has been reported about KATP channel mutations in human insulin producing tumors and the authors are not aware of any reports examining the KATP channel in canine insulinoma. However, naturally occurring mutations of SUR1 and Kir6.2 are known to cause a group of congenital hyperinsulinemic disorders of humans collectively known as persistent hypoglycemic hyperinsulinism of infancy (PHHI) [15]. About 40 separate mutations in SUR1 and three mutations in Kir6.2 have been described thus far [15,16]. The development of methods to study the canine KATP channel would permit more detailed study of canine beta cell function and insulin secretion. Likewise, these methods could also be used to learn more about canine insulinoma, in particular the contribution of abnormal KATP channel function to the unregulated insulin secretion that characterizes these tumors.
The purpose of this study was to learn more about KATP channel expression in normal canine pancreas and in insulin-secreting endocrine tumors of the canine pancreas. The study goals were twofold: to develop RT-PCR methods and reagents to study KATP channel expression in the normal canine pancreas and to screen a series of canine insulinomas to determine whether these tumors have normal KATP channel expression.
Results
SUR 1 expression in normal canine pancreas
The results of RT-PCR for SUR1 expression in normal canine pancreas are shown in Figure 1. The primers, which were designed from the rat SUR1 nucleotide sequence, amplified a 248 bp product from rat and dog pancreases. Sequence analysis confirmed that the amplicon from rat pancreas was rat SUR1, indicating that the primers specifically amplified the desired product. The amplicon from normal dog pancreas was also determined to be SUR1 based on shared base identity with SUR1 from other species (Figure 2). The canine SUR1 sequence was verified by repeat sequencing and placed in GenBank [GenBank:AY927669]. The partial sequence of canine SUR 1 was found to have a high degree of nucleotide identity with corresponding SUR 1 sequences from humans (93%), rat (93%), and hamster (92%).
Figure 1 PCR primers designed against rat SUR1 amplified a single band of expected size (248 bp) from rat (R) and dog (D) pancreas. Sequence analysis confirmed the identity of each band as SUR 1. The size markers (M) are shown on the left. The arrow indicates the approximate location of the 300 bp marker.
Figure 2 Canine SUR1 shares a high degree of nucleotide identity with SUR1 from other mammalian species. The alignment compares the partial sequence of canine SUR1 [GenBank:AY927669] with SUR1 from human [GenBank:AF087138], rat [GenBank:L40624], and hamster [GenBank:L40623]. Areas of sequence identity are shown in black and areas of non-identity are shown in white.
Kir6.2 expression in normal canine pancreas
Expression of Kir6.2 was detected in normal canine pancreas using primers designed against rat Kir6.2 [GenBank:D50581]. These primers amplified a product of expected size (approximately 167 bp) from rat and dog pancreases. Sequence analysis revealed that the PCR product from rat pancreas was 99% identical to the published sequence for Kir6.2, indicating that RT-PCR yielded a specific product. However, these primers did not produce clean, defined bands from canine pancreas (data not shown). To further clarify canine Kir6.2 expression, a second set of primers that were specific for canine Kir6.2 was designed using sequence information contained in the canine genome database [18]. These primers amplified a well-defined DNA band of expected size (371 bp) from normal canine pancreas (Figure 3). Sequence analysis showed that the canine product had 100% identity with a corresponding region of canine chromosome 21 [GenBank:AAEX01017445] contained in the canine genome database [18]. The canine pancreatic Kir6.2 sequence was verified by repeat sequencing and placed in GenBank [GenBank:AY929390]. The partial sequence obtained for canine Kir6.2 sequence shared extensive identity with Kir6.2 from other species (Figure 4). The canine Kir6.2 shares 95% identity with the predicted chimpanzee Kir6.2 but also has substantial identities with Kir6.2 from human (93%), mouse (90%), and rat (87%). The canine Kir6.2 was also compared to a predicted sequence for canine Kir6.2 [GenBank:XM-542519] that is available in GenBank. The 271 bp Kir6.2sequence from normal canine pancreas has 100% identity with the predicted nucleotide sequence of canine Kir6.2.
Figure 3 PCR primers designed against canine Kir6.2 amplified a single band of expected size (371 bp) from dog pancreas (P). Sequence analysis confirmed the identity of the band as Kir6.2. The size markers (M) are shown on the left. The arrow indicates the approximate location of the 400 bp marker.
Figure 4 Canine Kir6.2 shares a high degree of nucleotide identity with Kir6.2 from other mammalian species. The alignment compares the partial sequence of canine Kir6.2 [GenBank:AY929390] with Kir6.2from chimpanzee [GenBank:XM521849], human [GenBank:BC040617], mouse [GenBank:NM-010602], and rat [GenBank:NM-031358] Areas of sequence identity are shown in black and areas of non-identity are shown in white.
KATP channel subunit expression by canine insulinomas
Histologically-confirmed insulinomas from three dogs were analyzed for SUR1 and Kir6.2 expression. For each insulinoma, subunit expression was determined using the previously described rat SUR 1 and canine Kir6.2 primers. All three insulinomas were found to express the genes for SUR 1 and Kir6.2 (Figure 5). The identities of the PCR products obtained from each insulinoma were confirmed by nucleotide sequencing. Analysis of SUR1 and Kir6.2 sequences obtained from insulinomas did not reveal any mutations and were identical to the respective sequences obtained from normal canine pancreas. The intensity of SUR1 bands was noted to differ substantially among the insulinomas (Figure 5, panel A). Semiquantitative analysis comparing the OD of the SUR1 band with the OD of GAPDH band (the reference gene) from each insulinoma confirmed that SUR1 expression varied. Insulinoma #3 expressed SUR1 at a level that was 3.5-fold greater than insulinoma #1 and 2-fold greater than insulinoma #2. In contrast, expression levels of Kir6.2 were approximately equal in all three insulinomas (data not shown).
Figure 5 Panel A – PCR primers designed against rat SUR1 amplified a single band of the expected size (248 bp) from each of the three canine insulinomas examined (lanes labeled SUR1 1–3). Sequence analysis confirmed the identity of each band as SUR 1. GAPDH expression was used as a psoitive control (lanes labeled GAPDH 1–3). Semiquantitative analysis showed that SUR1 expression by insulinoma #3 was 3.5 and 2 fold greater than insulinomas #1 and #2, respectively (see text for explanation). Size markers (M) are shown on the left. The arrows indicate the approximate location of the 300 bp marker. Panel B – PCR primers designed against canine Kir6.2 amplified a single band of expected size (371 bp) from normal pancreas (P) and from each of the three canine insulinomas (labeled 1–3) examined. Sequence analysis confirmed the identity of each band as Kir6.2. Size markers (M) are shown on the left. The arrows indicate the approximate location of the 400 bp marker
Discussion
The results demonstrate that the SUR 1 and Kir6.2 subunits of the KATP channel are expressed in the pancreas of normal dogs. The combination of SUR1 and Kir6.2 comprises the classical beta cell KATP channel and expression of this subunit combination in the normal dog pancreas is consistent with previous reports showing that this same combination of subunits is expressed in the pancreatic beta cells by other species [2,19] and by insulin-secreting beta cell lines derived from the rat and mouse [2]. It is likely that the SUR1 and Kir6.2 detected in whole pancreatic samples represent islet specific expression since KATP channel subunit expression has not been reported in the exocrine pancreas. SUR 1 and Kir6.2 expression is not specific to beta cells, however, and are also expressed in islet alpha cells, where they are involved in glucagon secretion [20,21]. As the KATP channel subunits have similar function in all islet cells where they are expressed (e.g. sulfonylureas also stimulate glucagon secretion [21]) and since beta cells are the predominant endocrine cell in the canine islet, it may be concluded that the SUR1 and Kir6.2 detected in DNA isolated from whole pancreas reflects beta cell expression in the dog, although it is likely that other islet cells express the KATP channel also.
Because the subunit composition is the same, we predict that the canine beta cell KATP channel is subject to the same manner of metabolic regulation as the KATP channel in other species, such as the rat, mouse, and human, which have an identical KATP channel composition. The mechanisms of sulfonyluea and diazoxide on the pancreatic beta cell have not been studied at the molecular level in the dog but the results of this study predict that these pharmacologic agents will act via the KATP channel to alter insulin secretion. Indeed, in vivo studies have shown the expected changes in serum insulin and glucose levels in dogs treated with sulfonylureas or diazoxide [10-13].
The RT-PCR results also showed that all three insulinomas examined expressed SUR 1 and Kir6.2. Thus, failure to express one or both subunits cannot account for the abnormal insulin secretion by these tumors. However, variation in SUR1 expression was noted among the insulinomas. In particular, expression in one tumor (insulinoma #3) was several fold greater than the other tumors. It is not known whether the expression level of one or both subunits plays a role in the abnormal insulin secretion by canine insulinomas. The expression of the KATP channel by insulinomas is consistent with the hypothesized origin of these tumors from pancreatic beta cells or other pancreatic endocrine tissue. The islet origin of canine insulinomas is further supported by our findings that these tumors also express glucokinase and insulin but not amylase (V. Donley and T. Schermerhorn, unpublished observations).
The current results do not eliminate the possibility that the canine tumors express truncated versions of SUR1 or Kir6.2, such as those described for some humans with PHHI and unregulated insulin secretion [16,22,23]. Simple expression experiments do not provide information about functional aspects of the KATP channel, however, and our experiments would not have been able to detect functional abnormalities of KATP channels that might be present in the insulinomas. Many mutations do not affect gene expression but instead affect one or more of the steps beyond expression that are needed for proper channel function. Among the reported subunit mutations associated with KATP channel dysfunction are mutations that cause impaired translation into functional proteins [16,22], abnormal protein trafficking [23], impaired protein folding [24], defective channel assembly [22,24], and altered nucleotide sensitivity and substrate binding [16]. Many of the known KATP channel mutations leading to unregulated insulin secretion are associated with human PHHI. Most of these mutations affect SUR1 (there are over 40 reported human SUR1 mutations [15], while relatively few are Kir6.2 mutations [16,22,23]. Many of the mutations causing PHHI result in the production of KATP channels that lack activity. In the absence of KATP channel activity to couple metabolic activity and membrane potential, persistent calcium entry via voltage-sensitive calcium channels promotes insulin secretion that continues despite profound hypoglycemia. It is possible that mutations occurring in canine insulinomas might produce similar functional consequences. We screened canine insulinomas for mutations in SUR1 and Kir6.2 by comparing sequences obtained from insulinomas with those obtained from normal dogs. Using direct sequence comparison, no mutations were detected in the relatively short SUR1 or Kir6.2 sequences obtained from the insulinomas.
However, as the complete sequences of SUR1 and Kir6.2 from each insulinoma were not exhaustively screened for mutations, it is not possible to determine with any certainty whether or not KATP channel mutations were responsible for the abnormal secretory function exhibited by these tumors. The partial DNA sequence of canine SUR1 obtained from these insulinomas is homologous to a region that spans exons 7 and 8 of human SUR1 which codes for the c-terminal end of the linker region between the 7th and 8th transmembrane regions as well as a portion of the 8th transmembrane domain of the SUR1 protein. Several mutations in exon 8 have been found in humans with PHHI [16], suggesting that this region should be included when screening canine insulinomas for SUR1 mutations. However, the human SUR1 gene in humans is large, containing 39 exons [16]. Mutations causing hyperinsulinemia have been reported along the entire gene, including introns [15,16]. Assuming the canine gene is arranged similarly, mutation screening of SUR1 will not be readily performed until the complete nucleotide sequence of the normal canine gene is available. The same holds true for Kir6.2, although this gene is considerably smaller than SUR1 in the human and is likely to be similar in the dog.
Conclusion
In conclusion, this study documents the expression of the KATP channel subunits, SUR 1 and Kir6.2, in the normal canine pancreas. The partial sequences of canine SUR1 and Kir6.2 obtained are highly similar to those from other mammals, including humans, non-human primates, and rodents. KATP channel subunit expression was detected in all three canine insulinomas examined, indicating that absence of subunit expression cannot account for the abnormal insulin secretion observed with these tumors. SUR 1 and Kir6.2 sequence comparisons with normal canine pancreas did not reveal any mutations within the sequences obtained from the canine insulinomas.
Methods
Animals
All normal pancreatic and insulinoma tissues were obtained from dogs submitted necropsy examination or from dogs that underwent surgery to remove an insulinoma. All pancreases used to determine KATP channel expression in "normal" tissue was from dogs with no known pathology or that had non-pancreatic pathology. All dogs from which insulinoma tissue was obtained exhibited clinical signs suggestive of insulinoma and had a pancreatic mass at surgery. The diagnosis of insulinoma was confirmed by histological examination in each case. Rat pancreas served as control tissue for experiments unless otherwise indicated. Rat pancreatic tissue was obtained from normal Wistar rats that had been sacrificed for reasons unrelated to this study. Samples were immediately snap frozen in liquid nitrogen and maintained at -80 C until used for RNA extraction.
PCR primers
Primers used in PCR reactions were designed against a rat sequence for SUR 1 [GenBank:L40624] and a mouse sequence for Kir6.2 [GenBank:D50581] [17]. A second set of Kir6.2 primers was designed in our laboratory using canine-specific sequence information recently made available through the canine genome project [18]. GAPDH expression was used as an internal control for each experiment. Primer pairs and their expected product size were SUR1 (sense 5'-GGAGCAATCCAGACCAAG AT-3'; antisense 5'-AGCCAGCAGAATGATGACAG-3', expected product 248 bp), rat Kir6.2 (sense 5'-TCCAACAGCCCGCTCTAC-3'; antisense 5'-GATGGGGACAAAACGCTG-3', expected product 167 bp); canine Kir6.2 (sense 5'-CCTACACCAGGTGGACATCC-3', antisense 5'-CAGGCTGCGGTCCTCATCAA-3', expected product 371 bp); GAPDH (sense 5'-ATCTTCCAGGAGCGAGAT-3'; antisense 5'-TGGTCATGAGTCCTTCCACGATA-3'; expected product 300 bp).
RNA isolation and reverse transcription-polymerase chain reaction (RT-PCR) RT-PCR was performed to determine the expression of SUR1 and Kir6.2 in canine pancreas and in a series of three canine insulinomas. GAPDH gene expression was used as an internal PCR control. To prepare RNA, frozen tissue samples were homogenized using a QIAshredder and RNA extracted in Trizol Reagent according to manufacturer's instructions (Qiagen RNeasy Mini Kit). cDNA synthesis was carried out using a mixture (19 μl) that included total RNA (5 μg), dNTP 10 mmol/L, random hexamers 100 ng/μl, RNasin 20 IU and reverse transcriptase (Superscript II RT, Invitrogen) 25 IU. The mixture was incubated for 50 min at 42 C according to manufacturer's instructions, after which the reaction was terminated by heating the reaction to 70 C for 15 min.
PCR amplification was carried out in 25 μL reactions that contained: 0.5 μL template cDNA, 1 μL of each specific oligonucleotide primer, 0.1 μL Invitrogen Platinum Taq DNA Polymerase, 2.5 μL 10× buffer, 0.75 μl MgCl2, 1 μL dNTPs and 18.95 μL distilled water. PCR cycles for amplification of SUR1 and Kir6.2 were as follows. SUR1: 94 C for 5 min; followed by 35 cycles of 94 C for 0.5 min, 59 C for 0.5 min, 72 C for 1.5 min and then 72 C for 30 min. Kir6.2: 94 C for 5 min; followed by 35 cycles of 94 C for 0.5 min, 58.6 C for 0.5 min, 72 C for 1.5 min; then 72 C for 30 min. Products from the PCR reaction were resolved by electrophoresis on a 1.2% agarose gel and detected with ethidium bromide staining and UV light. The identity of the PCR products was confirmed by sequence analysis performed using an automated ABI 3700 DNA Analyzer (Kansas State University, Department of Plant Pathology).
Semiquantitative analysis of KATP subunit expression in insulinomas
Optical densities (OD) of PCR bands were determined from gel images using ImageJ software (Rasband, W.S., ImageJ, U. S. National Institutes of Health, Bethesda, Maryland, USA, , 1997–2005). Semiquantitative analysis was performed by comparing the OD of the SUR1 and Kir6.2 bands to the OD of a reference gene (GAPDH) from the same insulinoma. The ratio of SUR1 (or Kir6.2) to GAPDH provides a semiquantitative basis for comparison of expression levels between different tumors.
List of abbreviations
KATP – ATP-sensitive potassium channel
Kir6.2 – inward rectifying potassium channel 6.2
PHHI – persistent hypoglycemic hyperinsulinism of infancy
OD – optical density
SUR 1 – sulfonylurea receptor 1
Authors' contributions
VRD processed the tissues, performed the RT-PCR experiments, and prepared samples for sequencing. EKH participated in primer design, performed some of the RT-PCR experiments and participated in drafting the manuscript. ACK contributed important RT-PCR data. TS conceived and designed the study, drafted the manuscript, and performed sequence comparisons. All authors have read and approved the final manuscript.
Acknowledgements
The authors wish to thank the DNA Sequencing Facility, K-State Dept. Plant Pathology and Katie Gleason for excellent technical assistance.
Supported by a KSU Dept Clinical Sciences Research Grant (TS) and KSU Dept Anatomy and Physiology Clinical Resident Research Grant (AK).
The funding sources had no roles in study design, data acquisition or management, manuscript writing, or submission decision.
==== Refs
Mikhailov MV Mikhailova EA Ashcroft SJ Investigation of the molecular assembly of beta-cell K(ATP) channels FEBS Lett 2000 482 59 64 11018523 10.1016/S0014-5793(00)02035-4
Sakura H Ammala C Smith PA Gribble FM Ashcroft FM Cloning and functional expression of the cDNA encoding a novel ATP-sensitive potassium channel subunit expressed in pancreatic beta-cells, brain, heart and skeletal muscle FEBS Lett 1995 377 338 44 8549751 10.1016/0014-5793(95)01369-5
Bratanova-Tochkova TK Cheng H Daniel S Gunawardana S Liu YJ Mulvaney-Musa J Schermerhorn T Straub SG Yajima H Sharp GW Triggering and augmentation mechanisms, granule pools, and biphasic insulin secretion Diabetes 2002 51 S83 90 11815463
Gribble FM Tucker SJ Ashcroft FM The essential role of the Walker A motifs of SUR1 in K-ATP channel activation by Mg-ADP and diazoxide EMBO J 1997 16 1145 52 9135131 10.1093/emboj/16.6.1145
Proks P Reimann F Green N Gribble F Ashcroft F Sulfonylurea stimulation of insulin secretion Diabetes 2002 51 S368 76 12475777
Kramer W Oekonomopulos R Punter J Summ HD Direct photoaffinity labeling of the putative sulfonylurea receptor in rat beta-cell tumor membranes by [3H]glibenclamide FEBS Lett 1988 229 355 9 2831099 10.1016/0014-5793(88)81155-4
Schwanstecher M Brandt C Behrends S Schaupp U Panten U Effect of MgATP on pinacidil-induced displacement of glibenclamide from the sulphonylurea receptor in a pancreatic beta-cell line and rat cerebral cortex Br J Pharmacol 1992 106 295 301 1393263
Schwanstecher M Sieverding C Dorschner H Gross I Aguilar-Bryan L Schwanstecher C Bryan J Potassium channel openers require ATP to bind to and act through sulfonylurea receptors EMBO J 1998 17 5529 35 9755153 10.1093/emboj/17.19.5529
Heron L Virsolvy A Peyrollier K Gribble FM Le Cam A Ashcroft FM Bataille D Human alpha-endosulfine, a possible regulator of sulfonylurea-sensitive KATP channel: molecular cloning, expression and biological properties Proc Natl Acad Sci U S A 1998 95 8387 91 9653196 10.1073/pnas.95.14.8387
Hramiak IM Dupre J McDonald TJ Effects of galanin on insulin responses to hormonal, neuropeptidal, and pharmacological stimuli in conscious dogs Endocrinology 1988 122 2486 91 2453342
Porksen NK Munn SR Steers JL Schmitz O Veldhuis JD Butler PC Mechanisms of sulfonylurea's stimulation of insulin secretion in vivo: selective amplification of insulin secretory burst mass Diabetes 1996 45 1792 7 8922367
Lorrain J Angel I Eon MT Oblin A Langer SZ Inhibition of insulin release induced by galanin in the dog is not exclusively mediated by activation of ATP-sensitive K+ channels Pharmacology 1993 46 109 14 7680135
Hillaire-Buys D Ribes G Blayac JP Loubatieres-Mariani MM A new benzothiadiazine derivative, LN 5330: effects on pancreatic hormones Diabetologia 1984 27 583 6 6152237 10.1007/BF00276972
Leifer CE Peterson ME Matus RE Insulin-secreting tumor: diagnosis and medical and surgical management in 55 dogs J Am Vet Med Assoc 1986 188 60 4 3003016
Sharma N Crane A Gonzalez G Bryan J Aguilar-Bryan L Familial hyperinsulinism and pancreatic beta-cell ATP-sensitive potassium channels Kidney Int 2000 57 803 8 10720932 10.1046/j.1523-1755.2000.00918.x
Aguilar-Bryan L Bryan J Molecular biology of adenosine triphosphate-sensitive potassium channels Endocr Rev 1999 20 101 35 10204114 10.1210/er.20.2.101
Liu Y He HR Ding JH Gu B Wang H Hu G Iptkalim inhibits cocaine challenge-induced enhancement of dopamine levels in nucleus accumbens and striatum of rats by up-regulating Kir6.1 and Kir6.2 mRNA expression Acta Pharmacol Sin 2003 24 527 33 12791178
National center for biotechnical information (NCBI). Dog Genome Resources
Seino S Miki T Physiological and pathophysiological roles of ATP-sensitive K+ channels Prog Biophys Mol Biol 2003 81 133 76 12565699 10.1016/S0079-6107(02)00053-6
Bokvist K Olsen HL Hoy M Gotfredsen CF Holmes WF Buschard K Rorsman P Gromada J Characterisation of sulphonylurea and ATP-regulated K+ channels in rat pancreatic A-cells Pflugers Arch 1999 438 428 36 10519134 10.1007/s004240051058
Hoy M Olsen HL Bokvist K Buschard K Barg S Rorsman P Gromada J Tolbutamide stimulates exocytosis of glucagon by inhibition of a mitochondrial-like ATP-sensitive K+ (KATP) conductance in rat pancreatic A-cells J Physiol 2000 527 109 20 10944174 10.1111/j.1469-7793.2000.00109.x
Nestorowicz A Inagaki N Gonoi T Schoor KP Wilson BA Glaser B Landau H Stanley CA Thornton PS Seino S Permutt MA A nonsense mutation in the inward rectifier potassium channel gene, Kir6.2, is associated with familial hyperinsulinism Diabetes 1997 46 1743 8 9356020
Cartier EA Conti LR Vandenberg CA Shyng SL Defective trafficking and function of KATP channels caused by a sulfonylurea receptor 1 mutation associated with persistent hyperinsulinemic hypoglycemia of infancy Proc Natl Acad Sci U S A 2001 98 2882 7 11226335 10.1073/pnas.051499698
Thomas P Ye Y Lightner E Mutation of the pancreatic islet inward rectifier Kir6.2 also leads to familial persistent hyperinsulinemic hypoglycemia of infancy Hum Mol Genet 1996 5 1809 12 8923010 10.1093/hmg/5.11.1809
|
16266437
|
PMC1291378
|
CC BY
|
2021-01-04 16:29:46
|
no
|
BMC Vet Res. 2005 Nov 2; 1:8
|
utf-8
|
BMC Vet Res
| 2,005 |
10.1186/1746-6148-1-8
|
oa_comm
|
==== Front
Biomed Eng OnlineBioMedical Engineering OnLine1475-925XBioMed Central London 1475-925X-4-621625963910.1186/1475-925X-4-62ResearchAssessment of pulse rate variability by the method of pulse frequency demodulation Hayano Junichiro [email protected] Allan Kardec [email protected] Atsunori [email protected] Nobuyuki [email protected] Fumihiko [email protected] Core Laboratory, Nagoya City University Graduate School of Medical Sciences, Nagoya 467-8601, Japan2 Universidade Federal do Maranhao, Sao Luis-Ma, Brazil3 Department of Cardiovascular Dynamics, National Cardiovascular Center Research Institute, Suita, Osaka 565-8565, Japan4 Department of Internal Medicine and Pathophysiology, Nagoya City University Graduate School of Medical Sciences, Nagoya 467-8601, Japan5 Department of Internal Medicine, Suzuka National Hospital, Suzuka 513-8501, Japan2005 1 11 2005 4 62 62 9 5 2005 1 11 2005 Copyright © 2005 Hayano et al; licensee BioMed Central Ltd.2005Hayano et al; licensee BioMed Central Ltd.This is an Open Access article distributed under the terms of the Creative Commons Attribution License (), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
Background
Due to its easy applicability, pulse wave has been proposed as a surrogate of electrocardiogram (ECG) for the analysis of heart rate variability (HRV). However, its smoother waveform precludes accurate measurement of pulse-to-pulse interval by fiducial-point algorithms. Here we report a pulse frequency demodulation (PFDM) technique as a method for extracting instantaneous pulse rate function directly from pulse wave signal and its usefulness for assessing pulse rate variability (PRV).
Methods
Simulated pulse wave signals with known pulse interval functions and actual pulse wave signals obtained from 30 subjects with a trans-dermal pulse wave device were analyzed by PFDM. The results were compared with heart rate and HRV assessed from simultaneously recorded ECG.
Results
Analysis of simulated data revealed that the PFDM faithfully demodulates source interval function with preserving the frequency characteristics of the function, even when the intervals fluctuate rapidly over a wide range and when the signals include fluctuations in pulse height and baseline. Analysis of actual data revealed that individual means of low and high frequency components of PRV showed good agreement with those of HRV (intraclass correlation coefficient, 0.997 and 0.981, respectively).
Conclusion
The PFDM of pulse wave signal provides a reliable assessment of PRV. Given the popularity of pulse wave equipments, PFDM may open new ways to the studies of long-term assessment of cardiovascular variability and dynamics.
==== Body
Background
Analysis of heart rate variability (HRV) has been standardized with using R-R intervals of electrocardiogram (ECG) as the source signal [1]. The measurement of ECG, however, requires multiple electrode attachments and cable connections, which precludes frequent assessments of HRV in general populations. Beyond the uses as an autonomic functional index [2,3] or as a prognostic marker of long-term survival [4-6], the applications of HRV have extended into many areas of health sciences and technologies, such as those as markers for assessing the levels of physical and mental demands [7,8], the degree of fatigue [9], the depth of relaxation and resting [10,11] and the comfortableness of living and occupational environments [12,13]. For these applications, self-applicable devices that allow frequent, preferably everyday measurement in general population may be more useful.
From this aspect, analysis of pulse rate variability (PRV) from pulse wave signal has been studied as a potential surrogate of HRV analysis [14]. In contrast to ECG, pulse wave can be recorded with a single sensor without electrode and indeed, pulse wave equipments are popular and widely used not only in hospital cares but also in health sciences and clinical homecare practices. However, there is also a problem, i.e., smoother waveform of pulse wave precludes accurate measurement of pulse-to-pulse intervals by fiducial-point algorithms such as those used for measuring R-R intervals of ECG.
Here, we report a technique of pulse frequency demodulation (PFDM) as a method for estimating the function of instantaneous pulse rate. Taking advantage of smoother waveform of pulse wave, the PFDM extracts instantaneous pulse rate directly from pulse wave signal as a function of time. The continuous function of pulse rate can be converted into instantaneous pulse interval function, which can be directly used for spectral analysis. In this study, we tested the performance of the PFDM using both simulated data and actual pulse wave signals that were recorded with a wireless trans-dermal photoelectric device. We also compared PRV estimated by the PFDM with HRV measured by R-R intervals of simultaneously recorded ECG.
Methods
Principle of PFDM
The core process of PFDM is frequency demodulation based on the method of complex demodulation (CDM) [15-17]. CDM is a non-linear time-domain method for time series analysis, which provides amplitude and frequency of non-stationary/unstable oscillatory signal as a continuous function of time. Principle and computer algorithm of CDM have been published previously [16]. Briefly, CDM extracts the time dependent functions of instantaneous frequency through the following four steps: (1) the spectral region of interest (the frequency range of target oscillation) is shifted to zero frequency by forming a product, throughout the record, of the original signal and a complex sinusoid at a reference frequency (Fr, the center frequency of the spectral region of interest), (2) the resultant complex signal is low-pass filtered so that only frequency components around zero remain, (3) the real and imaginary parts of the low-pass filtered signal are converted to a polar form, yielding the instantaneous phase, as a function of time, of the component identified at or near the Fr, and (4) the time series of frequency of the target oscillation is calculated through adding the first-order derivative of the phase to the Fr, since the slope of the phase versus time curve indicates deviation of instantaneous operative frequency from the Fr.
In PFDM, the CDM is customized for analyzing pulse wave signal. Pulse frequency (instantaneous pulse rate) could change widely from less than 0.5 Hz (30 bpm) to more than 3 Hz (180 bpm). Because CDM extracts frequency of oscillatory components within an assigned spectral region (range of CDM filter), an appropriate frequency range needs to be selected so that it covers the possible range of pulse frequency. This is, however, a trade-off with the requirement that the frequency range of CDM filter needs to be narrow enough to avoid influence of harmonics and subharmonics of the fundamental oscillation (pulse wave).
When frequency of pulse wave decreases to a frequency as low as f Hz, the upper limit of the frequency range should be less than 2f Hz (frequency of the 2nd harmonic of the pulse wave). Likewise, when frequency of pulse wave increases to a frequency as high as f Hz, the lower limit of the frequency range should be greater than f/2 Hz (frequency of the 2nd subharmonic of the pulse wave). If the range for CDM filter is expressed as Fr ± Fc, where Fr is the reference frequency (the center frequency of the spectral region of interest) and Fc is the corner frequency of the low-pass filter, then the necessary condition for avoiding the influence of harmonics and subharmonics is
Fc < Fr/3.
For example, when mean pulse rate is 60 bpm, the widest possible range for CDM filter is 60 ± 20 (40 to 80) bpm; if pulse frequency deviates from this range during analyzing period, CDM would not provide an adequate estimation of pulse rate.
To overcome this problem, we devised the algorithm of PFDM incorporating the following three features: (1) overlapping short-segmentation of data, (2) adaptive determination of the Fr, (3) iterative algorithm for optimizing Fr, and (4) stepwise convergence of Fc toward Fr/3 during the iterative process. Briefly, data are first divided into short segments so that pulse frequency within each segment can be expected to remain within the range that can be covered by a feasible CDM filter. In each segment, the mean pulse frequency of the previous segment is used as the initial Fr value (adaptation). The Fr is further optimized through iteration processes, i.e., mean pulse frequency calculated is iteratively used as the Fr in the next iteration until the mean pulse frequency calculated agrees with the Fr used. Finally, to further guard against the possibility that pulse frequency deviates from the range of CDM filter, an Fc of Fr/2 is used for the initial process and it is converged to Fr/3 in the following iteration processes. The effectiveness of each of these processes is demonstrated in the simulation studies.
Simulation data
Simulated pulse wave signal was generated as a sequence of a unit pulse wave that was sampled from actual pulse wave of a healthy subject. The pulse interval was modulated by various source interval functions (modulator functions), which were used as the reference signals for evaluating the performance of PFDM (Fig. 1). The width of each unit pulse wave was also modulated according to the changes in pulse interval as 18.97 × (PI)1/2 ms, where PI is the pulse interval.
Figure 1 Method for generating simulated pulse wave signal. Simulated pulse wave (panel B) was generated as a sequence of a unit pulse wave. The pulse interval was modulated by source interval functions (panel A). The width of each unit pulse wave was also modulated as 18.97 × (PI)1/2 ms, where PI is the pulse interval. a.u. = arbitrary unit.
Measurement of actual data
We studied 33 subjects (23 males and 10 females); the mean age (standard deviation [SD]) was 34 (7) yr and age range was 22–47 yr. They were recruited from the staffs in the work places of the authors. All subjects gave their written informed consent and the procedure was performed according to the Ethical Guidelines of Nagoya City University Graduate School of Medical Sciences. Continuous pulse wave was recorded from the dorsal side of the wrist with a wireless trans-dermal photoelectric pulse wave sensor system (Prototype C, DENSO, Japan, Fig. 2 and 3), by which the pulse wave data were digitized (14 bit). ECG was recorded simultaneously with a digital Holter recorder (RAC-2102, Nihon Koden, Tokyo, Japan). In order to obtain exact matching of time between the two recordings, several event markers were also recorded simultaneously. Both continuous pulse wave and ECG data were recorded during all-night sleep in a laboratory chamber, because the pulse wave device allowed stable measure of pulse wave signal only during rest at present. To our knowledge, no noninvasive ambulatory devices are currently available for measuring stable pulse wave signal during activities.
Figure 2 Block diagram of wireless trans-dermal photoelectric pulse wave monitoring system. This system detects pulse wave as the amount of light absorption by erythrocyte hemoglobin in pulsating epidermal capillary bed. Green light is emitted from a light-emitting diode (LED) to the skin and the reflected light is detected by a photodiode (PD) embedded in a sensor tip. The detected signal is amplified, digitized by an analog-to-digital (A/D) converter and processed with a serial signal conversion interface (SCI). The signal is then transmitted from a wireless micro-power module (303.825MHz). The signal is received with a receiving wireless module and processed by a microcomputer for data communications and error detection. The pulse wave signals are sent to board PC on which time series analyses including PFDM are performed and the results are presented on a liquid crystal display (LCD). AMP = amplifier, ANT = antenna, ASK = amplitude shift keying, OOK = on-off keying.
Figure 3 A prototype of the wireless trans-dermal photoelectric pulse wave monitoring system (DENSO, Japan). Panels A, B and C show a pulse wave sensor unit that applies green light to the skin and detects the reflected light through a small window (panel B) on the reverse side of the unit facing to the skin. The sensor unit transmits pulse wave signal to a receiver unit (panel D). The receiver unit has a small personal computer with LCD. The unit also has a slot used for charging the battery of the sensor unit.
Data analysis
For both simulated and actual data analysis, pulse wave data were sampled at 20 Hz. The PFDM was performed with custom-made software written with FORTRAN 95 (Salford Software Ltd, Old Trafford, Manchester, UK). For the PFDM, the data segments had a length of 30 s with overlapping for 10 s at both ends. For the iteration for optimizing Fr, the tolerance for the difference between the mean pulse frequency and Fr was set at 0.001 bpm. Although the instantaneous pulse frequencies were calculated for every sampling point (at 20 Hz), they were averaged over every 500 ms and converted into pulse intervals in order to compare with the source interval functions (for the simulation data) or R-R intervals of ECG (for the actual data).
HRV, by definition, is the beat-to-beat variability of sinus rhythm [1,18]. Thus, all R-R interval data involving ectopic beat(s) or heart block(s) were excluded from the analysis. The pulse wave data, however, have less information about arrhythmias. An ectopic beat, either supra-ventricular or ventricular, could result in consecutive short pulse intervals, a long pulse interval or even a normal interval depending on the timing and whether it generates a detectable pulse wave or not. Heart blocks may also result in long intervals. To avoid the effects of rhythm disturbances and noises on the analysis of PRV, we excluded all abnormal pulse intervals, which were defined as those deviating 12% or more from the local moving average over the preceding 20 s. The definition of abnormal pulse interval was determined tentatively considering expected range of physiological PRV for the subject population.
The ECG data were analyzed with a Holter scanner (DSC-3100, Nihon Koden, Tokyo, Japan), on which the results of automatic labeling of QRS complexes were reviewed and manually edited for all errors. The ECG were analyzed with a sampling frequency of 125 Hz and, thus, the R-R intervals were measured at a time resolution of 8 ms. The time series data of R-R interval were interpolated along the time axis with a horizontal-step function (R-R interval was considered as constant during each R-R interval) and resampled at 2 Hz.
Spectral analysis
Fast Fourier transformation with a Hanning window was performed for sequential 5-minute segments of both pulse interval and R-R interval data. A segment was excluded from the analysis, if the ratio of valid data points in the segment was <80%. In the segments analyzed, defected parts of data, if any, were interpolated by the horizontal-step function. After correcting for the loss of variance resulting from the window process, power spectral density was integrated over 0.04–0.15 Hz and 0.15–0.45 Hz for assessing the power of low frequency (LF) and high frequency (HF) components, respectively. The power of these components was converted into mean amplitude ([2 × power]1/2) to reduce the skewness from the normal distribution.
Statistical analysis
The agreement between PRV and HRV measures was evaluated from two aspects; (1) agreement when these measures are used for assessing intra-individual variations and (2) agreement when they are used for inter-individual comparisons.
To examine the former agreement, minute-to-minute pulse rate and heart rate and spectral components of PRV and HRV for 5-min segments were compared within each subject. The agreement between corresponding values were evaluated with (a) the difference-against-mean plot and the 'limits of agreement' of Bland and Altman method [19] and (b) intraclass correlation coefficients for 2-way mixed effects analysis of variance with defining segments as the random factor and methods as the fixed factor [20]. The statistical adequacy of the segments as random factor is partly supported by the fact that long-term heart rate fluctuation has the characteristics of random fractal noise [1,21].
To examine the latter agreement for inter-individual comparisons, the minute-to-minute and 5-min segment values were averaged over the entire recording length for individual subjects. The agreement between the corresponding mean values were evaluated with (a) the Bland and Altman method same as above [19] and (b) intraclass correlation coefficients for 2-way mixed effects analysis of variance with defining subjects as the random factor and methods as the fixed factor [20].
Data were presented as mean (SD) or median (range). Type 1 error level was set at probability (p value) of < 0.05.
Results
Simulation studies
Simulation studies were performed for evaluating the performance of the PFDM, particularly the effects of (1) Fr adaptation with data segmentation, (2) Fr optimization with iterative algorithm, (3) Fc convergence, and (4) robustness against the fluctuation in pulse height and baseline trend. Also, the frequency characteristics of the PFDM were examined to test if the output from the PFDM is appropriate for frequency domain analyses.
Effects of Fr adaptation and iterative optimizations
Simulated pulse wave signal was generated with modulating the pulse interval by a sinusoidal source interval function (Fig. 4A) that fluctuated between 600 and 1400 ms (corresponding to 43 to 100 bpm). The signal was analyzed with PFDM algorithms with different levels of Fr adjustment. Fig. 4B is the result with an algorithm with fixed Fr at 60 bpm and Fc at 40 bpm (no adaptation). Although the range of CDM filter covered pulse rate from 20 to 100 bpm, the output is affected by the 2nd harmonic and the 2nd subharmonic when the pulse interval reached the upper and lower ends. This problem was partly resolved with an algorithm furnishing Fr adaptation with data segmentation (length 30 s with overlapping 10 s at both ends) and an appropriate Fc (Fr/3; Fig. 4C); however, estimation errors still occurred at the portions where the pulse interval changed fast, due to transient deviation of the pulse frequency outside the range of CDM filter (spectral leakage). Finally, the pulse interval was satisfactorily estimated with adding iterative Fr optimization for each segment (Fig. 4D). The estimation error (differences between the estimated pulse intervals and the source interval function) was within ± 2 ms. The median (range) of the number of iteration with a tolerance of 0.001 bpm was 3 (3–5).
Figure 4 Effects of adaptation and optimization algorithms. Simulated pulse wave signal was generated by a sinusoidal source interval function (panel A) that fluctuated between 600 and 1400 ms (corresponding pulse rate, from 43 to 100 bpm). Panel B: the demodulated pulse interval with fixed Fr (60 bpm) and Fc (40 bpm). Panel C: the demodulated pulse interval with data segmentation (length, 30 s with 10-sec overlapping at both end) and Fr adaptation (Fc was set at Fr/3). Panel D: the demodulated pulse interval with data segmentation, Fr adaptation and Fr optimization with iteration (Fc was set at Fr/3).
Effects of Fc convergence
The PFDM algorithm even furnishing the Fr adaptation and optimization failed to follow this rapid change due to spectral leakage and erroneous adaptation to the harmonics or subharmonics of pulse wave (Fig. 5B). This possibility can be reduced with widening the range of CDM filter, which subsequently converged to an appropriate width after the fundamental pulse wave is detected. In the PFDM algorithm, the Fc was set at Fr/2 at the first iteration and then converged to Fr/3 in the following iterations, by which the rapid change in pulse interval was detected appropriately (Fig. 5C).
Figure 5 Effects of convergence algorithm. Simulated pulse wave signal was generated by a source interval function with an abrupt increase from 700 to 1200 ms during 10 s (panel A). Panel B: demodulated pulse interval with data segmentation, Fr adaptation and Fr optimization (iteration) without Fc convergence (Fc was set at Fr/3). Panel C: demodulated pulse interval with data segmentation, Fr adaptation, Fr optimization (iteration) and Fc convergence (Fc was set at Fr/2 initially and then converged to Fr/3).
Robustness against pulse height and baseline fluctuations
Fluctuations of pulse height and baseline, such as those due to respiration and body movements, are quite common. Simulation study, however, revealed that such fluctuations have only small influence on the estimation of pulse interval by the PFDM (Fig. 6).
Figure 6 Effect of pulse height and baseline fluctuations. Simulated pulse wave signals were generated by a source interval function oscillating at 0.25 Hz (panel A) with stable pulse height and baseline (panel B), with fluctuating pulse height (panel C), and with fluctuating baseline (panel D). These fluctuations had no substantial influence (the estimation error, within ± 2 ms) on the results of the PFDM (panel E).
Frequency characteristics
To test the transfer function of PFDM, the simulated pulse wave was generated using an oscillatory source interval function with linearly increasing frequency from 0 to 0.5 Hz at 0.001 Hz/sec (Fig. 7A). The fluctuation of demodulated pulse interval by the PFDM appeared to reflect faithfully both frequency and amplitude between 0 and 0.43 Hz (Fig. 7B). The transfer magnitude and phase between the source and demodulated pulse interval from 0.00 to 0.43 Hz were between 0.97 and 1.02 and between -0.1 and 0 π, respectively (Fig. 7C).
Figure 7 Frequency characteristics of PFDM. Simulated pulse wave signal was generated by a source interval function oscillating at linearly increasing frequency form 0 to 0.5 Hz at 0.001 Hz/sec (panel A). Panel B: demodulated pulse interval obtained from the simulated pulse wave with the PFDM. Panel C: transfer magnitude (solid line) and transfer phase (dashed line) calculated between the source interval and the demodulated pulse interval with the PFDM. Insets indicate the frequency ranges corresponding low frequency (LF) and high frequency (HF) bands.
Analysis of actual data
Actual data of overnight recordings of pulse wave together with simultaneous ECG were obtained from 30 subjects out of 33. Data were lost due to technical problems in a subject and were excluded due to atrial fibrillation during the entire recording periods in two subjects. The mean ± SD length of data in the 30 subjects was 6.0 ± 0.8 hr. In these data, median ratio (range) of 5-min segments that met the criteria (valid data ≥80%) for spectral analysis was 87 (63–95) %.
The presences of atrial fibrillation in the two excluded subjects were detected by the PFDM of pulse wave. In both subjects, more than 90% of the pulse intervals deviated 12% or more from the local moving average and none of the 5-min segments met the inclusion criteria (valid data ≥80%).
Agreement between pulse rate and heart rate
Minute-to-minute pulse rate assessed by the PFDM showed overnight trends almost identical to those of minute-to-minute heart rate assessed by ECG (Fig. 8). Good agreement was observed between pulse rate and heart rate within each subject (Fig. 8C and 8D). For 30 subjects, the median (range) of intraclass correlation coefficients was 0.999 (0.998–0.999), the mean (SD) of the differences between them was -0.00 (0.03) bpm and the upper and lower limits of agreement were 0.21 (SD, 0.06) and -0.22 (SD, 0.09) bpm, respectively.
Figure 8 Agreement of demodulated pulse rate by PFDM and ECG heart rate. Panels A and B: minute-to-minute pulse rate (PR) demodulated by the PFDM (panel A) and minute-to-minute heart rate (HR) from simultaneously recorded ECG (panel B) in a representative subject during all-night sleep. Panel C: correlation between PR and HR with the intraclass correlation coefficient (ICC). Panel D: the Bland and Altman plot with the upper and lower limits of agreement (+2 SD and -2 SD, respectively). SD = standard deviation.
The analysis of agreement for inter-individual comparisons showed almost perfect agreement between the mean pulse rate and heart rate of individual subjects. The intraclass correlation coefficient was 0.999, the mean difference was 0.00 bpm, and the upper and lower limits of agreement were 0.03 and -0.03 bpm, respectively.
Agreement of spectral components of PRV and HRV
The amplitude of LF and HF components of PRV and HRV were calculated for sequential 5-min segments of demodulated pulse interval by PFDM and ECG R-R interval, respectively. The overnight trends of amplitude of both components were similar between PRV and HRV (Fig. 9). The analysis of agreement between them showed unexpectedly good agreements (Fig. 9). For 30 subjects, the median (range) of intraclass correlation coefficients for the LF and HF amplitudes were 0.989 (0.911–0.998) and 0.940 (0.676–0.990), respectively. The absolute means (SD) of the differences were -0.1 (0.6) and 0.1 (0.5) ms for both LF and HF amplitudes and the upper and lower limits of agreement were 2.1 (SD, 1.3) and -2.4 (SD, 1.3) ms for LF amplitude and 4.1 (SD, 2.2) and -3.9 (SD, 1.6) ms for HF amplitude.
Figure 9 Agreement of spectral components of pulse rate variability (PRV) and heart rate variability (HRV). The amplitudes of spectral components were calculated for sequential 5-min segments over the entire length of data in a representative subject. Left panels: trendgram (panels A and B), scatter plot (panel C) and the Bland and Altman plot (panel D) for the amplitudes of low-frequency components of pulse interval variability (LFPRV) and R-R interval (LFHRV). Right panels: trendgram, scatter plot and the Bland and Altman plot of the amplitudes of high-frequency components of pulse interval variability (HFPRV) and R-R interval (HFHRV). ICC = intraclass correlation coefficient. Broken lines in the Bland and Altman plots represent the upper and lower limits of agreement (+2 SD and -2 SD, respectively). SD = standard deviation.
Table 1 Agreement between PRV and HRV for inter-individual comparisons
N PRV Mean (SD) ms HRV Mean (SD) ms Mean difference ms Upper limit of agreement ms Lower limit of agreement ms ICC
LF 30 27.4 (7.5) 27.6 (7.3) 0.2 1.2 -0.7 0.997
HF 30 31.6 (10.7) 31.4 (10.5) -0.2 3.9 -4.4 0.981
Analyses were performed for mean values of individual subjects.
HF = high frequency component, HRV = heart rate variability, ICC = intraclass correlation coefficient, LF = low frequency component, PRV = pulse rate variability, SD = standard deviation.
To evaluate the agreement for inter-individual comparisons, mean values of LF and HF amplitude of individual subjects were compared between PRV and HRV. The results showed good agreement for the mean LF amplitude and acceptable agreement for the mean HF amplitude (Table 1).
Influence of ventricular ectopies (VE)
In this study, abnormal pulse intervals caused by ectopies and heart blocks were excluded whenever they deviated 12% or more from the local average. To examine the influence of this process on the assessments of pulse rate and spectral components, the effects of the frequency of VE on the agreement between values calculated from the pulse interval by PFDM and ECG R-R interval were analyzed.
Of the 30 subjects, only 3 subjects had relatively frequent VE (187, 676, and 699 beats, respectively, during the overnight recordings). In these 3 subjects, when the frequency of VE in 5-min segments was <~10%, VE showed no substantial effects on the agreement between the values calculated from two methods (Fig. 10). When the frequency of VE was >~10%, however, the HF amplitude of PRV was greater than that of HRV (Fig. 10C).
Figure 10 Influence of ventricular ectopies on assessment by PFDM. The amplitudes of spectral components were calculated for 5-min sequential segments of demodulated pulse interval by PFDM and ECG R-R interval in three subjects with frequent ventricular ectopies. Panel A: the difference in pulse rate (PR) and heart rate (HR). Panel B: the difference in the amplitude of low-frequency component of pulse interval (LFPRV) and that of R-R interval (LFHRV). Panel C: the difference in the amplitude of high-frequency component of pulse interval (HFPRV) and that of R-R interval (HFHRV). All data are plotted against percentage of ventricular ectopies within each 5-min segment.
Discussion
In this study we proposed the PFDM technique as a method for demodulating instantaneous pulse rate from pulse wave signal and demonstrated its usefulness for assessing spectral components of PRV. The simulation studies revealed that (1) the PFDM provides reliable measurement of instantaneous pulse rate even if it fluctuates rapidly and over a wide range, (2) it is robust to the variations of pulse height and baseline trend, and (3) it preserves the frequency characteristics of the source modulator function, a feature necessary for spectral analysis. The analysis of actual pulse wave revealed that (1) minute-to-minute pulse rate assessed by the PFDM agreed perfectly with minute-to-minute heart rate measured by ECG, (2) the amplitudes of LF and HF components of PRV show overnight trends quite similar to those of HRV, and (3) the mean values of LF and HF amplitudes during night show good agreement between PRV and HRV. These observations indicate that the PFDM provides a reliable assessment of pulse rate and PRV and suggest that this technology makes pulse wave a potentially useful source signal for assessing cardiovascular variability and dynamics.
It is noteworthy that the pulse wave was sampled at a frequency of 20 Hz, while the ECG was sampled at 125 Hz. Nevertheless, we observed surprisingly good agreement not only between pulse rate and heart rate but also between the spectral components of PRV and HRV. This indicates that the accuracy of the PFDM analysis of pulse interval is not directly dependent on the time resolution of data. Indeed, the simulation studies revealed that the estimation errors of pulse interval by the PFDM are within ± 2 ms despite the fact that the sampling interval is 50 ms (20 Hz). This seems attributable to the fact that only PFDM, but not R-R interval measurement, utilizes the periodicity of signals; i.e., the PFDM analyzed pulse wave with assuming it as a cosine function with slowly changing amplitude and phase. Even for ECG, a periodicity analysis method can estimate instantaneous heart rate from signals sampled at a low frequency (5 Hz) [22].
Although measurement of R-R intervals of ECG is the standard for HRV analysis, this method has practical limitations [1]. Recordings of ECG require attachment of multiple electrodes and cables, which limit applications of HRV analysis in public or home health cares. Also, R-R intervals are intervals of event series. An appropriate interpolation is necessary to estimate the underlying modulator function that is the putative subject of HRV analysis. For this purpose, several interpolation methods with differing mathematic features have been proposed; but convincing physiological reasoning for selecting a method is lacking.
In contrast to ECG, recording of pulse wave requires only a single sensor, which allowed development of devices that can be used in daily life even everyday. Also, smooth waveform of pulse wave requires a lower sampling frequency (10–20 Hz) and its sinusoidal feature is advantageous to the PFDM as mentioned above. Furthermore, by the use of the PFDM, the signal putatively modulating the pulse interval is directly extracted as a continuous function. The signal can be used directly for spectral analysis without interpolation processes.
An apparent limitation of the PFDM of pulse wave is inability to assess the type of arrhythmias. A substantial part of such arrhythmia, however, can be detected by using appropriate criteria such as those using deviation of intervals from the local mean. In fact, the PFDM was able to detect persistent atrial fibrillation in this study. Also, even for the segments including frequent VE, the agreement for the pulse rate and LF amplitude are unaffected, although the HF amplitude was overestimated for the segments including VE>~10% compared with the HF amplitude obtained by R-R interval analysis. Interestingly, however, the exclusion of arrhythmic data has been reported to result in an underestimation of the HF amplitude for HRV analysis with R-R interval [18]. It should be also noted that the optimal criteria for excluding abnormal beats are not uniform but subject to change depending on age and other conditions that could affect the magnitude of physiological PRV. This issue, however, is not specific to PFDM but common to PRV and HRV assessment at least in detection of atrial ectopic beats and heart blocks.
Another limitation of the PFDM may be caused by the physiological differences between pulse interval and R-R interval. Theoretically, the variability of pulse rate is the sum of the variability existing in R-R interval, pre-ejection period and pulse wave velocity. Constant et al [14] suggested that, in the standing position, respiratory fluctuation of pulse wave velocity might be important cause of respiratory pulse rate variation. Although the present study indicates usefulness of the PFDM for assessing PRV and good agreement between PRV and HRV during night sleep, its usefulness as a surrogate of HRV for assessing autonomic functions and mortality risk need to be examined de novo.
Conclusion
The PFDM of pulse wave signal provides a reliable assessment of PRV. Given the popularity of pulse wave equipments, this technology may open new ways to studies of long-term assessment of cardiovascular variability and dynamics among general populations.
Authors' contributions
JH conceived and designed the study, collected the data and drafted manuscript. AKB participated in the conception of the study and revised the manuscript critically. AK participated in the design of the study and was involved in acquisition of data and revising the manuscript critically. NO was involved in analysis of data and revising the manuscript. FY participated in the conception of the study and was involved in acquisition of data and revising the manuscript. All authors read and approved the final manuscript.
Acknowledgements
This work was supported in part by the Research Grant from Suzuken Memorial Foundation (2003), by the Grant-in-Aid for Scientific Research (C) from the Japanese Ministry of Education, Culture, Sports, Science and Technology (15590765) and by the Research Grant (14A-9-09) for Nervous and Mental Disorders from the Japanese Ministry of Health, Labor and Welfare (2002–2004).
==== Refs
Malik M Camm AJ Eds Dynamic Electrocardiography 2004 New York: Blackwell Futura
Pomeranz B Macaulay RJ Caudill MA Kutz I Adam D Gordon D Assessment of autonomic function in humans by heart rate spectral analysis Am J Physiol 1985 248 H151 H153 3970172
Pagani M Lombardi F Guzzetti S Rimoldi O Furlan R Pizzinelli P Power spectral analysis of heart rate and arterial pressure variabilities as a marker of sympatho-vagal interaction in man and conscious dog Circ Res 1986 59 178 193 2874900
La Rovere MT Pinna GD Hohnloser SH Marcus FI Mortara A Nohara R Baroreflex sensitivity and heart rate variability in the identification of patients at risk for life-threatening arrhythmias: implications for clinical trials Circulation 2001 103 2072 2077 11319197
Kleiger RE Miller JP Bigger JT JrMoss AJ Decreased heart rate variability and its association with increased mortality after acute myocardial infarction Am J Cardiol 1987 59 256 262 3812275 10.1016/0002-9149(87)90795-8
Tsuji H Venditti FJ JrManders ES Evans JC Larson MG Feldman CL Reduced heart rate variability and mortality risk in an elderly cohort. The Framingham Heart Study Circulation 1994 90 878 883 8044959
Garde AH Laursen B Jorgensen AH Jensen BR Effects of mental and physical demands on heart rate variability during computer work Eur J Appl Physiol 2002 87 456 461 12172887 10.1007/s00421-002-0656-7
Kristal-Boneh E Raifel M Froom P Ribak J Heart rate variability in health and disease Scand J Work Environ Health 1995 21 85 95 7618063
Pichot V Bourin E Roche F Garet M Gaspoz JM Duverney D Quantification of cumulated physical fatigue at the workplace Pflugers Arch 2002 445 267 272 12457247 10.1007/s00424-002-0917-7
Bonnet MH Arand DL Heart rate variability in insomniacs and matched normal sleepers Psychosom Med 1998 60 610 615 9773766
Hayano J Yasuma F Hypothesis: respiratory sinus arrhythmia is an intrinsic resting function of cardiopulmonary system Cardiovasc Res 2003 58 1 9 12667941 10.1016/S0008-6363(02)00851-9
Dayawansa S Umeno K Takakura H Hori E Tabuchi E Nagashima Y Autonomic responses during inhalation of natural fragrance of Cedrol in humans Auton Neurosci 2003 108 79 86 14614968 10.1016/j.autneu.2003.08.002
Umemura M Honda K Influence of music on heart rate variability and comfort – a consideration through comparison of music and noise J Hum Ergol (Tokyo) 1998 27 30 38 11579697
Constant I Laude D Murat I Elghozi JL Pulse rate variability is not a surrogate for heart rate variability Clin Sci (Lond) 1999 97 391 397 10491338
Bloomfield P Bloomfield P Complex demodulation Fourier Analysis of Time Series: An Introduction 1976 New York: Wiley 118 150
Hayano J Taylor JA Mukai S Okada A Watanabe Y Takata K Assessment of frequency shifts in R-R interval variability and respiration with complex demodulation J Appl Physiol 1994 77 2879 2888 7896636
Hayano J Taylor JA Yamada A Mukai S Hori R Asakawa T Continuous assessment of hemodynamic control by complex demodulation of cardiovascular variability Am J Physiol 1993 264 H1229 H1238 8476100
Lippman N Stein KM Lerman BB Comparison of methods for removal of ectopy in measurement of heart rate variability Am J Physiol 1994 267 H411 H418 7519408
Bland JM Altman DG Statistical methods for assessing agreement between two methods of clinical measurement Lancet 1986 1 307 310 2868172
Shrout PE Fleiss JL Intraclass correlations: uses in assessing rater reliability Psychological Bulletin 1979 86 420 428 10.1037//0033-2909.86.2.420
Saul JP Albrecht P Berger RD Cohen RJ Analysis of long term heart rate variability: method, 1/f scaling and implications Comp Cardiol 1987 14 419 422
Barros AK Ohnishi N Heart instantaneous frequency (HIF): an alternative approach to extract heart rate variability IEEE Trans Biomed Eng 2001 48 850 855 11499522 10.1109/10.936361
|
16259639
|
PMC1291379
|
CC BY
|
2021-01-04 16:37:36
|
no
|
Biomed Eng Online. 2005 Nov 1; 4:62
|
utf-8
|
Biomed Eng Online
| 2,005 |
10.1186/1475-925X-4-62
|
oa_comm
|
==== Front
Cell Commun SignalCell communication and signaling : CCS1478-811XBioMed Central London 1478-811X-3-121623616810.1186/1478-811X-3-12ResearchXenopus frizzled-4S, a splicing variant of Xfz4 is a context-dependent activator and inhibitor of Wnt/β-catenin signaling Swain Rajeeb Kumar [email protected] Masaru [email protected] Araceli [email protected] Herbert [email protected] Institute of Human Genetics, University of Heidelberg, Im Neuenheimer Feld 366, 69120 Heidelberg, Germany2 Deptartment of Cell Biology, Max Planck Institute for Developmental Biology, Spemann Str. 35, 72076 Tübingen, Germany3 Genetics and Cell Biology Section, Genetics Division, National Cancer Center Research Institute, Tsukiji 5-chome, Chuo-ku, Tokyo, 104-0045, Japan2005 19 10 2005 3 12 12 17 6 2005 19 10 2005 Copyright © 2005 Swain et al; licensee BioMed Central Ltd.2005Swain et al; licensee BioMed Central Ltd.This is an Open Access article distributed under the terms of the Creative Commons Attribution License (), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
Background
Secreted Frizzled related proteins (SFRPs) are extracellular regulators of Wnt signaling. These proteins contain an N-terminal cysteine rich domain (CRD) highly similar to the CRDs of the Frizzled family of seven-transmembrane proteins that act as Wnt receptors. SFRPs can bind to Wnts and prevent their interaction with the Frizzled receptor. Recently it has been reported that a splice variant of human Frizzled-4 (FZD4S) lacking the transmembrane and the cytoplasmic domains of Frizzled-4 can activate rather than inhibit Wnt-8 activity in Xenopus embryos. This indicates that secreted CRD containing proteins such as Frizzled ecto-domains and SFRPs may not always act as Wnt inhibitors. It is not known how FZD4S can activate Wnt/β-catenin signaling and what biological role this molecule plays in vivo.
Results
Here we report that the Xenopus frizzled-4 is alternatively spliced to give rise to a putative secreted protein that lacks the seven-transmembrane and the cytoplasmic domains. We performed functional experiments in Xenopus embryos to investigate how this novel splicing variant, Xfz4S, can modulate the Wnt/β-catenin pathway. We show that Xfz4S as well as the extracellular domain of Xfz8 (ECD8) can act as both activators and inhibitors of Wnt/β-catenin signaling dependent on the Wnt ligand presented. The positive regulation of Wnt/β-catenin signaling by the extracellular domains of Frizzled receptors is mediated by the members of low density lipoprotein receptor-related protein (LRP-5/6) that act as Wnt coreceptors.
Conclusion
This work provides evidence that the secreted extracellular domains of Frizzled receptors may act as both inhibitors and activators of Wnt signaling dependent on the Wnt ligand presented.
==== Body
Background
Wnts are secreted glycoproteins that control an array of signaling processes in embryos and adult tissues [1-4]. These proteins act through the members of the Frizzled family of seven-transmembrane receptors [5,6]. Wnt and Frizzled interaction leads to the stabilization of cytoplasmic β-catenin, its nuclear translocation and subsequent transcriptional activation of Wnt/β-catenin target genes [1,7]. Two members of low-density lipoprotein receptor-related protein, LRP-5 and -6, act as coreceptors in the Wnt/β-catenin signaling [8-10]. These transmembrane proteins can interact with Wnts and form a ternary complex with Frizzled receptors [9]. This leads to the binding of axin to the cytoplasmic domain of LRP and its recruitment to the membrane [11]. Axin is a scaffolding protein necessary in the cytoplasm for assembly of the protein complex that phosphorylates β-catenin and promotes its degradation by ubiquitin proteasome dependent pathway [12,13]. Recruitment of axin to the membrane by LRP leads to the reduced phosphorylation of β-catenin and subsequent activation of Wnt/β-catenin pathway. In the extracellular space, various secreted molecules negatively regulate Wnt/β-catenin signaling [14]. Prominent among them are the members of the secreted Frizzled related protein family (SFRP) that inhibit Wnt/β-catenin signaling primarily by binding to the Wnts and preventing Wnt/Frizzled interaction. Dickkopf family of extracellular proteins can bind to the Frizzled coreceptor LRP-5/6 and inhibit Wnt/β-catenin signaling [15,16]. SFRPs contain a cystein rich domain (CRD) that is also found in the Frizzled receptors [14]. The CRD of Frizzleds and SFRPs is required for their interaction with Wnts [5,14]. In this paper, we present evidence that Xenopus frizzled-4 is alternatively spliced to generate a putative secreted protein (Xfz4S), containing a part of the extracellular domain but lacking the seven-transmembrane and cytoplasmic domains. Xfz4S can activate or inhibit the Wnt/β-catenin signaling dependent on the Wnt ligand presented. We show that the extracellular domain of Xenopus frizzled-8 (ECD8) not only inhibits Wnt signaling induced by a variety of Wnt ligands, but can also act synergistically with Wnt-5a in inducing Wnt/β-catenin signaling. We further show that the activation of Wnt/β-catenin pathway by the extracellular domains of Frizzled receptors is dependent on LRP.
Results
Xenopus frizzled- 4 is alternatively spliced
It has been reported that the human Frizzled 4 (FZD4) is alternatively spliced to give rise to a transcript (FZD4S) in which the intron-1 of FZD4 is retained [17]. In order to investigate if Xenopus frizzled-4 (Xfz4) could also be alternatively spliced, the genomic organization of this gene was studied. The Frizzled-4 gene in humans, Xenopus laevis and Xenopus tropicalis (), consists of 2 exons and one intron. In X. laevis, the intron has a length of 6 kb (Fig. 1A, 1B and unpublished observations of MK). Thus, the exon-intron structure of Frizzled-4 genes is conserved in human, mouse, X. laevis and X. tropicalis. The conserved exon-intron structure of the vertebrate Frizzled-4 genes prompted us to investigate if Xenopus frizzled-4 is alternatively spliced like its human ortholog. This putative splicing variant of Xfz4 (Xfz4S) in which the intron is retained should generate a protein of 128 amino acids. The first 81 amino acids of which are identical to the seven-transmembrane type Xfz4 and the other 47 amino acids are unique to Xfz4S (Fig. 1A and 1B). To determine if Xfz4S is expressed during the Xenopus development, RT-PCR was performed using PCR primers which recognize only the splicing variant of Xfz4 (Fig. 1A). The forward primers were selected from the exon-1 and the reverse primers were selected from intron-1. All the RNA samples were treated with DNaseI to eliminate any genomic DNA from the RNA preparation. The results show that Xfz4 is alternatively spliced and that the intron is retained in this splicing variant (Fig. 1C and 1B). Developmental RT-PCR showed that Xfz4S is expressed only after mid blastula transition (MBT) and expression persists during all stages of development studied (Fig. 1C). In contrast, Xfz4 message is maternally supplied and is present during all examined stages of Xenopus development (Fig. 1C and Ref. [18]). In order to study the spatial distribution of Xfz4S mRNA, a part of the intron1 of Xfz4 (Xfz4-intron1) was used as an in situ probe as this will specifically recognize the Xfz4S transcripts and not that of Xfz4. The in situ pattern of Xfz4S shows that this mRNA is ubiquitously expressed after MBT (data not shown). At the tail bud stage (stage 34), the strongest staining was observed in the head (Fig. 1D). The expression pattern of Xfz4S is not identical with Xfz4 but there is overlapping expression in the eye.
Figure 1 Molecular structure and expression pattern of Xfz4S. (A) Nucleotide and amino acid sequences of Xfz4S. The nucleotide sequence of Exon1 of Xfz4 is in capital and that of Intron1 is in small letters. Two sets of primers (P1F/P1R and P2F/P2R) were used for detection of Xfz4S by RT-PCR. (B) Schematic diagram showing the structure of Xfz4 gene containing two exons (boxes) and one intron. The splicing variants, Xfz4S retaining the intron and the Xfz4 are shown. The coding regions of these splicing variants are indicated by closed boxes and the UTRs by open boxes. (C) Developmental RT-PCR of Xenopus embryos with indicated Nieuwkoop and Faber (NF) stages. Xfz4S mRNA is first detected after mid blastula transition and the expression persist into tadpole stages. Xfz4 mRNA is maternally supplied and is expressed in all stages of development studied. ODC is a loading control. (D) Spatial expression pattern of Xfz4 and Xfz4S in tailbud stage embryos (stage 34). The embryos were hybridized with digoxigenin labelled RNA probes for antisense Xfz4 (Xfz4-AS), antisense Xfz4-intronI (Xfz4S-AS) or sense probe for Xfz4-intronI (Xfz4S-S).
Xfz4S acts synergistically with a specific group of Wnt ligands
Next, we tested the ability of Xfz4S to modulate the Wnt/β-catenin signaling. It has been shown, that FZD4S, a splice variant of human Frizzled-4 can enhance the activity of Wnt-8 in inducing secondary body axis when injected into the ventral marginal zone of Xenopus embryos [17]. This prompted us to ask if Xfz4S has similar activity in positively regulating Wnt/β-catenin signaling. Xfz4S, which is a putative secreted protein, may also inhibit Wnt signaling by sequestering the Wnts in the extracellular space. To test these possibilities, synthetic mRNA encoding Xfz4S was injected either alone or in combination with Wnt ligands into the animal blastomeres at 4-cell stage and the activation of Wnt/β-catenin target gene Xnr3 was monitored by RT-PCR at stage 10.5 (Fig. 2A). mRNAs for Wnt-1 type ligands (Wnt-3a, -8 or -8b) which can induce Wnt/β-catenin signaling were titrated to such low doses that they did not induce Xnr3 expression. When these Wnts were coinjected with Xfz4S, Xnr3 expression was induced (Fig. 2B). This suggests that Xfz4S acts synergistically with Wnt-3a, -8 and -8b in activating the Wnt/β-catenin pathway.
Figure 2 Xfz4S acts synergistically with canonical Wnt ligands in activating the Wnt/β-catenin target gene Xnr3 and inhibits non-canonical Wnt ligands. (A) Experimental scheme. Embryos were injected at 4 cell stage into the animal blastomeres with synthetic mRNA for Xwnt-3a (0.5 pg/embryo), Xwnt-4 (150 pg/embryo), Xwnt-5a (50 pg/embryo), Xwnt-8 (0.5 pg/embryo), Xwnt-8b (15 pg/embryo) or Xwnt-11 (50 pg/embryo), either alone or in combination with 500 pg Xfz4S. Animal caps were dissected out at stage 8 – 9, grown until stage 10.5 at which expression of Xnr3 was analyzed by RT-PCR. (B) Xwnt-3a, -8, -8b induce Xnr3 expression only in combination with Xfz4S. Xwnt-5a, -4 or -11 do not synergize with Xfz4S in inducing Xnr3. (C) Xfz4S inhibits Xnr3 expression induced by coinjection of Hfz5 and non-canonical Wnt-5a class ligands. Wnt-5a class ligands such as Wnt-4 (150 pg/embryo), -5a (50 pg/embryo) or -11 (50 pg/embryo) when injected in combination with 250 pg Hfz5, Xnr3 expression was induced. Induction of Xnr3 expression by these Wnt ligands in combination with Hfz5 was inhibited by coinjection of 500 pg Xfz4S mRNA. (D) Coimmunoprecipitation of Xfz4S and Wnt-5a. Myc-tagged Xfz4S (500 pg/embryo) and flag-tagged Wnt-5a (500 pg/embryo) were injected into Xenopus embryos at 2–4 cell stage. Myc-tagged Xfz4S coimmunoprecipitates with flag-tagged Wnt-5a indicating that they interact. A part of the embryo extract was incubated with mouse IgG, which serves as a control against non-specific binding of proteins. Total embryo extract (EE) shows the expression of Xfz4S-myc and Wnt5a-flag constructs.
Xfz4S can inhibit the Wnt activity
In our experiments, Xfz4S was not able to synergize with non-canonical Wnts such as Wnt-4, -5a or -11 in activating Wnt/β-catenin pathway (Fig. 2B). This could be due to the inability of these ligands to interact with Xfz4S. To test the interaction of Xfz4S and non-canonical Wnts we took advantage of the fact that the non-canonical Wnts that do not activate Wnt/β-catenin pathway when expressed alone, can do so in combination with Hfz5 [19]. We injected Wnt-4, -5a or -11 in combinations with Hfz5 into the animal blastomeres at 4-cell stage and monitored the expression of Wnt/β-catenin target gene Xnr3 by RT-PCR at stage 10.5. As expected, Xnr3 expression was induced in these animal caps. Coinjection of Xfz4S inhibited the activation Xnr3 by Hfz5 and Wnt-4, -5a or -11 (Fig. 2C). This suggests that Xfz4S can interact with non-canonical Wnts and can act as an inhibitor of the Wnt/β-catenin signaling.
Consistent with the functional interaction between Xfz4S and Wnt ligands in modulating the Wnt/β-catenin signaling, we found that myc-tagged Xfz4S coimmunoprecipitates with flag-tagged Wnt-5a (Fig. 2D). A flag-epitope tagged Xfz4S also coimmunoprecipitated with myc-tagged Wnt-11 (data not shown) indicating that Xfz4S forms a complex with these Wnt ligands.
The extracellular domain of Xfz8 can activate Wnt/β-catenin pathway
Based on our observation that Xfz4S, which resembles the extracellular domain of a Frizzled receptor can act as a positive regulator of Wnt signaling, we tested if the extracellular domains of other Frizzled receptors could exhibit the same function. Extracellular domains of Xfz7, Xfz8 and Hfz5 were tested in combination with both canonical and non-canonical Wnt ligands for their ability to modulate Wnt/β-catenin signaling. The Frizzled ecto-domains were injected into animal blastomeres of Xenopus embryos, either alone or in combination with Wnt-3a, -4, -5a, -8, -8b or -11 mRNA. Neither low dose of canonical Wnts alone, nor in combination with ECD-7/-8/-5 were able to induce Xnr3 expression in explanted animal cap tissues. Similar results were obtained when these ecto-domains of Frizzled receptors were injected in combination with Wnt-4 or Wnt-11. When ECD8 was coexpressed with Wnt-5a, however, Xnr3 expression was induced (Fig. 3A). Coexpression of ECD8 and Wnt-5a in the ventral marginal zone of the Xenopus embryos resulted in ectopic expression of Xnr3 (Fig. 3B). In later stages, such embryos developed incomplete secondary body axes without head structures (Fig. 3C). The activation of the Wnt/β-catenin target gene Xnr3 and the induction of secondary axes were not observed after injection of either ECD8 or Wnt-5a alone. It has been reported that expression of high amounts of ECD8 mRNA in the ventral marginal zone can induce secondary body axis including head structures [20]. The induction of such type of axis by ECD8 is achieved by inhibition of Wnt and BMP signaling and is not accompanied by induction of Xnr3 expression. In contrast to the induction of secondary axis structures by high doses of ECD8, the secondary axis induced by low doses of ECD8 and Wnt-5a was accompanied by induction of Wnt/β-catenin target gene Xnr3 (Fig. 3B). This provides further evidence that ECD8 and Wnt-5a act cooperatively in activating the Wnt/β-catening pathway.
Figure 3 Extracellular domain of Xfz8 (ECD8) can activate Wnt/β-catenin pathway in combination with Xwnt-5a. (A) 500 pg/embryo ECD8 mRNA when injected in conjunction with Xwnt-5a (50 pg/embryo), Xnr3 expression was induced in animal cap tissues. Extracellular domains of Xfz7 (ECD7, 300 pg/embryo) or Hfz5 (ECD5, 500 pg/embryo) did not synergize with Xwnt-5a in inducing Xnr3 expression. (B) ECD8 (200 pg/embryo) and Wnt-5a (50 pg/embryo) when injected into the ventral marginal zone, Xnr3 expression was induced. (C) At later stages these embryos developed partial secondary axis without head structures. No secondary axes were observed when ECD8 or Wnt-5a was injected alone.
Activation of Wnt/β-catenin pathway by Xfz4S and ECD8 is mediated by LRP
We next asked how Frizzled ecto-domains that lack the transmembrane and the cytoplasmic domains could positively regulate Wnt/β-catenin signaling. Our working hypothesis is that the activation of Wnt/β-catenin signaling by the Frizzled ecto-domain is mediated by the Wnt coreceptor LRP5/6. We postulate that the Wnt-1/Xfz4S and Wnt5a/ECD8 complexes interact with the coreceptor LRP5/6 and activate the Wnt/β-catenin signaling in LRP5/6 and axin dependent manner. To test this hypothesis, we interfered with LRP mediated signaling in several ways. Xdkk1, a Wnt inhibitor, can bind directly to LRP5/6 and prevent LRP-Wnt-Frizzled ternary complex formation [15,16]. A LRP mutant lacking the cytoplasmic carboxy terminal domain (△CLRP) can not interact with axin but can sequester the Wnt-Frizzled complexes and prevent their interaction with endogenous LRP5/6 [9,11]. We also interfered with LRP-axin interaction by overexpression of the DIX domain of Xdsh. The DIX domain of Axin is required for its interaction with both disheveled and LRP [11,21]. Hence, overexpression of the Xdsh-DIX will sequester axin in the cytoplasm and will prevent its interaction with LRP. Low doses of Wnt-3a mRNA were injected into the animal caps in combination with Xfz4S to induce Wnt signaling as monitored by the activation of Wnt/β-catenin target gene Xnr3. The activation of Xnr3 expression was blocked by coexpression of Xdkk1 or ΔCLRP6 and was substantially reduced by coexpression of Xdsh-DIX (Fig. 4A). The Xdsh-dd1 mutant containing the DIX domain completely blocked Wnt signaling induced by Wnt-3a and Xfz4S (data not shown).
Figure 4 Activation of Wnt/β-catenin pathway by Xfz4S and ECD8 is LRP dependent. (A) Xfz4S (500 pg/embryo) and Xwnt-3a (0.5 pg/embryo) when coinjected into the animal caps, expression of Xnr3 was induced. This activation of Xnr3 was blocked by coinjection of 300 pg Xdkk1 or 1 ng ΔC-LRP6 and was greatly reduced by 250 pg Xdsh-DIX. (B) Induction of Xnr3 expression by injection of ECD8 (500 pg/embryo) and Xwnt-5a (50 pg/embryo) was blocked by coinjection of 300 pg Xdkk1, 1 ng ΔC-LRP6 or by 250 pg Dsh-dd2. Coinjection of a dishevelled mutant lacking the DIX domain (Dsh-ΔDIX 250 pg/embryo) had no effect on ECD8 plus Wnt-5a induced activation of Xnr3.
We employed the same strategy in animal cap assays to investigate if the activation of Wnt/β-catenin pathway by ECD8 and Wnt-5a is LRP dependent. When ECD8 was expressed together with Wnt-5a in animal caps, Xnr3 expression was induced. The activation of Xnr3 expression by ECD8 plus Wnt-5a was blocked by coinjection of either Xdkk1 or ΔCLRP6. A mutant Xdsh molecule lacking the carboxy-terminus DEP domain but containing the DIX domain (Xdsh-dd2) was also able to block Xnr3 expression induced by ECD8 and Wnt-5a. The Xdsh-dd1 and the Xdsh-DIX mutants both blocked Wnt-5a/ECD8 induced Wnt siganling (data not shown). Consistent with our argument, Xdsh mutant lacking the DIX domain (Xdsh-ΔDIX) was not able to interfere with ECD8 and Wnt-5a induced activation of Xnr3 (Fig. 4B). These results suggest that Wnt-3a/Xfz4S and Wnt-5a/ECD8 complexes can interact with LRP and activate the Wnt/β-catenin pathway in LRP-axin dependent manner.
Discussion
Regulation of Wnt signaling by a novel splice variant of Frizzled-4
In this study we report that Xenopus frizzled-4 is alternatively spliced to give rise to a transcript (Xfz4S) that is predicted to generate a secreted protein lacking the transmembrane and cytoplasmic domains. Xfz4S mRNA is expressed as a zygotic transcript and is present during all stages of Xenopus development (Fig. 1C).
Xfz4S when overexpressed alone in Xenopus embryos, does not activate Wnt/β-catenin signaling. When coexpressed with Wnt-1 type ligands such as Wnt-3, -8 and -8b, it acts synergistically with these ligands in activating Wnt/β-catenin target gene Xnr3 (Fig. 2B). This is in agreement with the observation that human FZD4S acts synergistically with Wnt-8 in activation Wnt/β-catenin signaling [17]. Our results show that the ability of Xfz4S to synergize with Wnt ligands in activating Wnt/β-catenin signaling is dependent on Wnt coreceptor LRP (Fig. 4A). When Xfz4S was coexpressed with the non-canonical Wnt ligands such as Wnt-4, -5a and -11, it inhibited the ability of these ligands to activate Wnt/β-catenin signaling in conjunction with the receptor Hfz5 (Fig. 2C). This shows that Xfz4S can interact with both the canonical Wnt-1 and non-canonical Wnt-5a class ligands, but has opposite effects on Wnt/β-catenin signaling. This can be explained by postulating that only the complex between Xfz4S and the Wnt-1 type ligands is recognized by LRP and the Wnt/β-catenin pathway is activated in a LRP dependent manner whereas Wnt-5a/Xfz4S complex is not recognized by LRP. In this situation, Xfz4S will act as a negative regulator of Wnt/β-catenin signaling (Fig. 5).
Figure 5 Model for the modulation of Wnt/β-catenin pathway by extracellular domains of Frizzled receptors. We propose that a complex formed between Xfz4S and Wnt-3a/-8/-8b could be recognized by LRP and Wnt signaling could be activated in LRP dependent manner. Complexes between Xfz4S and Wnt-4/-5a/-11 would not be recognized by LRP and Xfz4S and Wnt/β-catenin signaling would not be activated by these Wnts.
Dual role of Frizzled ecto-domains: activation and repression
It has been shown that the secreted Frizzled related proteins (SFRPs) and Frizzled ecto-domains act by binding to Wnts and sequestering them in the extracellular space. Contrary to this view, we show that Xfz4S that resembles the ecto-domain of Frizzled receptor can act as a positive regulator of Wnt signaling with a specific group of Wnt ligands (Fig. 2B). We also show that the ecto-domain of Xfz8 can act synergistically with Wnt-5a in activating Wnt/β-catenin signaling in a LRP dependent manner (Fig. 3 and 4B). This seems to contradict a report in which expression of Xnr3 induced by Frizzled-8 and Wnt-5a was inhibited by ECD8 [20]. It is plausible however, that full length Frizzled-8 is more potent than ECD8 in activating Xnr3 in combination with Wnt-5a. This interpretation is supported by our finding that coinjection of ECD8 and Wnt-5a only induced partial secondary body axes, whereas coexpression of full length Frizzled-8 and Wnt-5a induced complete secondary axes in Xenopus embryos (Fig. 3C and data not shown). In the presence of Frizzled-8, ECD8 and Wnt-5a, Xnr-3 expression should be reduced compared to the combination Frizzled-8 and Wnt-5a.
Our results suggest that Frizzled ecto-domains may not exclusively act as inhibitors of Wnt signaling. Similar observation has been made in case of Drosophila Frizzled-2 [Dfz2; [22]]. A mutant Dfz2 lacking the carboxy-terminal cytoplasmic domain (Dfz2△C) can synergize with Wingless (Wg) in transmitting Wnt/β-catenin signaling. Although Dfz2△C retains the seven-transmembrane domains, which may play a role in this signaling, our results would suggest that a Dfz2 mutant containing only the ecto-domain may be sufficient to synergize with Wg in activating this pathway. It has also been reported that SFRP2 can antagonize SFRP1 function during metanephric kidney development. In this process SFRP1 inhibits Wnt-4 signaling whereas SFRP2 promotes it [23]. These observations suggest that SFRPs may activate or inhibit Wnt signaling in a context dependent manner. Such dual activities have also been described for proteins of the Dkk family. Dkk2 can activate Wnt/β-catenin signaling and it synergizes with Frizzled receptors as well as with LRP6 in activating this pathway; whereas Dkk1 is an inhibitor of Wnt signaling [24,25]. These data indicate that the activity of extracellular factors which modulate Wnt signaling activity is dependent on the type of Wnt ligand and the cellular context. The biological significance of such dual activity, however, is poorly understood and will be a priority for future work. Although it is assumed that the non-canonical Wnts such as Wnt-5a and Wnt-11 function in β-catenin independent manner, it is not clear, if these Wnts may have functions mediated by β-catenin in vivo. Overexpression of Wnt-5a has been shown to correlate with abnormal nuclear localization of β-catenin protein in phyllodes tumor and ectopic Wnt-11 can rescue axis structures in UV ventralized Xenopus embryos by activation of the Wnt/β-catenin pathway [26,27]. Maternal Wnt-11 has been shown to activate Wnt/β-catenin signaling required for axis specification in Xenopus whereas zygotic Wnt-11 regulates non-canonical Wnt signaling, which coordinates gastrulation movements later in development [28-30]. This indicates that the activities of Wnt ligands in activating the canonical or non-canonical Wnt signaling may be regulated by extracellular cofactors. Supporting this hypothesis, Exostosin, an enzyme necessary for heparan sulfate proteoglycans (HSPGs) biosynthesis and EGF-CFC protein FRL1 have been shown to modulate Wnt-11 activity [28]. We postulate that secreted Frizzled related proteins and Frizzled ecto-domains may regulate the activation of distinct downstream signaling pathways triggered by Wnts.
Conclusion
We conclude that the ecto-domains of Frizzled receptors may act both as positive and negative regulators of the Wnt/β-catenin signaling dependent on the Wnt ligand presented. Their activity may also depend on the cellular context. The dual activity of these secreted proteins adds a new level of regulation to Wnt signaling in the extracellular space.
Methods
Xenopus embryo manipulations
Xenopus eggs were obtained from females injected with 300 IU of human chorionic gonadotrophin (Sigma), and were fertilized in vitro. Eggs were dejellied with 2% cysteine hydrochloride pH 8 and embryos were microinjected in 1XMBS-H (88 mM NaCl, 1 mM KCl, 2.4 mM NaHCO3, 0.82 mM MgSO4, 0.41 mM CaCl2, 0.33 mM Ca(NO3)2, 10 mM HEPES pH 7.4, 10 μg/ml streptomycin sulfate and 10 μg/ml penicillin). The embryos were cultured in 0.1XMBS-H and staged according to Nieuwkoop and Faber (1967) [31].
Plasmid constructions and mRNA microinjections
Xfz4S cDNA was amplified from cDNA preparations of gastrula and neurula stages. The open reading frame of Xfz4S was amplified by PCR using following primers; 5'-ATGGGGGCAAGATCGCTGACCTTGTTGTAC-3' and 5'-CCTTGTGGTTTATAGGGAGAGGACACAGGC-3' and was cloned into pCS2+ plasmid. A part of the intron1 of Xfz4 (Xfz4-intronI, nt- 281-925 in Fig. 1A) used as an in situ probe to specifically detect the Xfz4S transcripts was amplified from the NF stage 19 cDNA preparation using the following primers: 5'-TTCACTCTACCAACGCGCAACTTACG-3' and 5'-GACACAGTCACTTTTTGTGGACGCTG-3' and was cloned into pCR-Blunt II-TOPO (Invitrogen). The ORF of Wnt-5a was amplified by PCR from pSP64T-Xwnt-5a [32] and was cloned into pCS2+ vector at EcoRI and XhoI sites. The extracellular amino terminus domain of human Frizzled 5 containing the first 233 amino acids was amplified by PCR using 5'-TTGCTGCTGCTCGGATCCGCCACCATGGCTC-3' and 5'-ATGGATCCCGTGCGCTCGTCGGCACTGAAG-3' primers and was cloned into pCS2+MT plasmid at BamHI site. Myc-tagged Fz4S and flag-tagged Wnt-5a were constructed by amplifying the respective ORFs by PCR and cloning them into pCS2+MT or pCS2+Flag plasmids (both gifts from Ralph Rupp). Both constructs contain the myc or flag tag at their C-terminus. All the constructs were verified by sequencing.
Capped mRNAs were synthesized from linearized plasmids using mMessage mMachine Kit (Ambion). Wnt-3a [33] (linearized with EcoRI, transcribed with SP6), Wnt-4 [34] and Xdsh-DIX [35] were linearized with SalI and transcribed with SP6. Wnt-5a, NXfz8 (ECD8) [36], Xdkk1 [37] and Xdsh-dd2 [38] and Xfz4S were linearized with NotI and transcribed with SP6. Wnt-8b [39], ΔC-LRP6 [9] and NXfz7 (ECD7) were linearized with Asp718 and transcribed with SP6. Synthetic mRNA from other constructs were prepared as follows: Wnt-8 (linearized with BamHI and transcribed with SP6) [40], Wnt-11 (linearized with EcoRI, transcribed with T7) [27], Hfz5 (linearized with HindIII, transcribed with SP6) [19] and NHfz5 (ECD5) was linearized with BstXI and transcribed with SP6.
RT-PCR
Total RNA was prepared from embryos or animal cap explants with Trizol® reagent (Invitrogen). First strand cDNA was synthesized with H minus M-MuLV reverse transcriptase (Fermentas) using random hexamers as primers. PCR was performed using standard conditions and the following sets of primers: Xfz4S-E1I1 (P1) '5-TTGTTGTACCTCCTGTGCTGCCTC-3' and '5-TGGTAGAGTGAAATGCGCAGCAGC-3' (271 bp, Tm 60°C and 29 cycles); Xfz4S-E2I2 (P2) '5-CATCAGGATCACCATGTGCCAG-3' and '5-GAAAGTAAACCCCCTGTGCTGAG-3' (277 bp, Tm 60°C, 29 cycles); Xnr-3 '5-TGAATCCACTTGTGCAGTTCC-3' and '5-GACAGTCTGTGTTACATGTCC-3' (233 bp, Tm 65°C, 29 cycles); ODC '5-GTCAATGATGGAGTGTATGGATC-3' and '5-TCCATTCCGCTCTCCTGAGCAC-3' (385 bp, Tm 65°C, 25 cycles).
In situ hybridization
Whole mount in situ hybridization and antisense probe preparation was carried out as described [41]. Digoxigenin labelled antisense RNA was synthesized from plasmid containing Xnr3 (linearized with EcoRI), pCR-Blunt II-TOPO – Xfz4-intronI and pCR-Blunt II-TOPO – Xfz4 (both linearized with BamHI) using T7 RNA polymerase. Digoxigenin labelled sense RNA for Xfz4-intronI was synthesizes by linearizing the plasmid with NotI and transcribing with SP6.
Co-immunoprecipitation
Xenopus embryos were injected with 500 pg myc-tagged Fz4S and 500 pg flag-tagged Wnt-5a mRNA at 2–4 cells stage. The embryos were grown until gastrula stage and protein was extracted in NP-40 lysis buffer (10 mM Tris-Hcl, pH 7.5, 100 mM NaCl, 2 mM EDTA, 1 mM EGTA, 0.5% NP-40, 5% glycerol with a cocktail of proteinase inhibitors). The embryo extract was incubated for 2 h either with 4 μg of anti-flag (M2, Sigma), 2 μg anti-myc (9E10, Calbiochem) or 2 μg of mouse IgG (Sigma) at 4°C with constant rotation. The samples were centrifuged and 30 μl of protein G beads (Pierce) was added to the supernatant. The beads were incubated with the protein extract for 2 h, centrifuged and washed four times with NP-40 lysis buffer. The immunoprecipitates were separated on 12% SDS-PAGE and were transferred to nitrocellulose membrane. For detection of immunoprecipitated proteins, the membranes were incubated with either anti-myc or anti-flag antibodies followed by incubation with peroxidase-conjugated secondary antibody. Bound secondary antibodies were visualized using SuperSignal west pico reagent (Pierce).
Competing interests
The author(s) declare that they have no competing interests.
Authors' contributions
RKS and HS designed the experiments. RKS performed most of the experiments and HS supervised the work. MK cloned the Xfz4 intron and established the genomic structure of Xfz4 gene. AM generated a new Xwnt-5a construct and performed in situ experiments. All authors contributed to writing and approved the final manuscript.
Acknowledgements
We thank X. He, J-C. Hsieh, P. S. Klein, M. Ku, R. T. Moon, C. Niehrs, U. Rothbächer, R. Rupp and S. Sokol for providing plasmids and reagents and U. Müller and K. Linsmeier for technical support. We thank Ana Cristina Silva for help with in situ, S. Cramton for critically reading the manuscript and P. Hausen for support. RKS is supported by a postdoctoral fellowship from the Faculty of Medicine, University of Heidelberg.
==== Refs
Wodarz A Nusse R Mechanisms of Wnt signaling in development Annu Rev Cell Dev Biol 1998 14 59 88 9891778 10.1146/annurev.cellbio.14.1.59
van de Wetering M de Lau W Clevers H WNT signaling and lymphocyte development Cell 2002 109 S13 S19 11983149 10.1016/S0092-8674(02)00709-2
Alonso L Fuchs E Stem cells in the skin: waste not, Wnt not Genes Dev 2003 17 1189 1200 12756224 10.1101/gad.1086903
Veeman MT Axelrod JD Moon RT A second canon. Functions and mechanisms of beta-catenin-independent Wnt signalling Dev Cell 2003 5 367 377 12967557 10.1016/S1534-5807(03)00266-1
Bhanot P Brink M Samos CH Hsieh JC Wang Y Macke JP Andrew D Nathans J Nusse R A new member of the frizzled family from Drosophila functions as a Wingless receptor Nature 1996 382 225 230 8717036 10.1038/382225a0
Yang-Snyder J Miller JR Brown JD Lai CJ Moon RT A frizzled homolog functions in a vertebrate Wnt signaling pathway Curr Biol 1996 6 1302 1306 8939578 10.1016/S0960-9822(02)70716-1
Hecht A Kemler R Curbing the nuclear activities of beta-catenin. Control over Wnt target gene expression EMBO Rep 2000 1 24 28 11256619 10.1093/embo-reports/kvd012
Wehrli M Dougan ST Caldwell K O'Keefe L Schwartz S Vaizel-Ohayon D Schejter E Tomlinson A DiNardo S arrow encodes an LDL-receptor-related protein essential for Wingless signalling Nature 2000 407 527 530 11029006 10.1038/35035110
Tamai K Semenov M Kato Y Spokony R Liu C Katsuyama Y Hess F Saint-Jeannet JP He X LDL-receptor-related proteins in Wnt signal transduction Nature 2000 407 530 535 11029007 10.1038/35035117
Pinson KI Brennan J Monkley S Avery BJ Skarnes WC An LDL-receptor-related protein mediates Wnt signalling in mice Nature 2000 407 535 538 11029008 10.1038/35035124
Mao J Wang J Liu B Pan W Farr GH 3rdFlynn C Yuan H Takada S Kimelman D Li L Wu D Low-density lipoprotein receptor-related protein-5 binds to Axin and regulates the canonical Wnt signaling pathway Mol Cell 2001 7 801 809 11336703 10.1016/S1097-2765(01)00224-6
Kikuchi A Regulation of beta-catenin signaling in the Wnt pathway Biochem Biophys Res Commun 2000 268 243 248 10679188 10.1006/bbrc.1999.1860
Aberle H Bauer A Stappert J Kispert A Kemler R beta-catenin is a target for the ubiquitin-proteasome pathway EMBO J 1997 16 3797 3804 9233789 10.1093/emboj/16.13.3797
Kawano Y Kypta R Secreted antagonists of the Wnt signalling pathway J Cell Sci 2003 116 2627 2634 12775774 10.1242/jcs.00623
Mao B Wu W Li Y Hoppe D Stannek P Glinka A Niehrs C LDL-receptor-related protein 6 is a receptor for Dickkopf proteins Nature 2001 411 321 325 11357136 10.1038/35077108
Semenov MV Tamai K Brott BK Kuhl M Sokol S He X Head inducer Dickkopf-1 is a ligand for Wnt coreceptor LRP6 Curr Biol 2001 11 951 961 11448771 10.1016/S0960-9822(01)00290-1
Sagara N Kirikoshi H Terasaki H Yasuhiko Y Toda G Shiokawa K Katoh M FZD4S, a splicing variant of frizzled-4, encodes a soluble-type positive regulator of the WNT signaling pathway Biochem Biophys Res Commun 2001 282 750 756 11401527 10.1006/bbrc.2001.4634
Shi DL Boucaut JC Xenopus frizzled 4 is a maternal mRNA and its zygotic expression is localized to the neuroectoderm and trunk lateral plate mesoderm Mech Dev 2000 94 243 245 10842080 10.1016/S0925-4773(00)00294-X
He X Saint-Jeannet JP Wang Y Nathans J Dawid I Varmus H A member of the Frizzled protein family mediating axis induction by Wnt-5A Science 1997 275 1652 1654 9054360 10.1126/science.275.5306.1652
Itoh K Sokol SY Axis determination by inhibition of Wnt signaling in Xenopus Genes Dev 1999 13 2328 2336 10485853 10.1101/gad.13.17.2328
Wharton KA Jr Runnin' with the Dvl: proteins that associate with Dsh/Dvl and their significance to Wnt signal transduction Dev Biol 2003 253 1 17 12490194 10.1006/dbio.2002.0869
Schweizer L Varmus H Wnt/Wingless signaling through beta-catenin requires the function of both LRP/Arrow and frizzled classes of receptors BMC Cell Biol 2003 4 4 12729465 10.1186/1471-2121-4-4
Yoshino K Rubin JS Higinbotham KG Uren A Anest V Plisov SY Perantoni AO Secreted Frizzled-related proteins can regulate metanephric development Mech Dev 2001 102 45 55 11287180 10.1016/S0925-4773(01)00282-9
Wu W Glinka A Delius H Niehrs C Mutual antagonism between dickkopf1 and dickkopf2 regulates Wnt/beta-catenin signalling Curr Biol 2000 10 1611 164 11137016 10.1016/S0960-9822(00)00868-X
Brott BK Sokol SY Regulation of Wnt/LRP signaling by distinct domains of Dickkopf proteins Mol Cell Biol 2002 22 6100 6110 12167704 10.1128/MCB.22.17.6100-6110.2002
Sawyer EJ Hanby AM Rowan AJ Gillett CE Thomas RE Poulsom R Lakhani SR Ellis IO Ellis P Tomlinson IP The Wnt pathway, epithelial-stromal interactions, and malignant progression in phyllodes tumours J Pathol 2002 196 437 444 11920740 10.1002/path.1067
Ku M Melton DA Xwnt-11: a maternally expressed Xenopus wnt gene Development 1993 119 1161 1173 8306880
Tao Q Yokota C Puck H Kofron M Birsoy B Yan D Asashima M Wylie CC Lin X Heasman J Maternal wnt11 activates the canonical wnt signaling pathway required for axis formation in Xenopus embryos Cell 2005 120 857 871 15797385 10.1016/j.cell.2005.01.013
Tada M Smith JC Xwnt11 is a target of Xenopus Brachyury: regulation of gastrulation movements via Dishevelled, but not through the canonical Wnt pathway Development 2000 127 2227 2238 10769246
Kuhl M Non-canonical Wnt signaling in Xenopus: regulation of axis formation and gastrulation Semin Cell Dev Biol 2002 13 243 249 12137733 10.1016/S1084-9521(02)00050-2
Nieuwkoop PD Faber J Normal Table of Xenopus laevis (Daudin) 1967 Amsterdam, North Holland Publishing Company
Moon RT Campbell RM Christian JL McGrew LL Shih J Fraser S Xwnt-5A: a maternal Wnt that affects morphogenetic movements after overexpression in embryos of Xenopus laevis Development 1993 119 97 111 8275867
Wolda SL Moody CJ Moon RT Overlapping expression of Xwnt-3A and Xwnt-1 in neural tissue of Xenopus laevis embryos Dev Biol 1993 155 46 57 8416844 10.1006/dbio.1993.1005
McGrew LL Otte AP Moon RT Analysis of Xwnt-4 in embryos of Xenopus laevis: a Wnt family member expressed in the brain and floor plate Development 1992 115 463 473 1425335
Rothbacher U Laurent MN Deardorff MA Klein PS Cho KW Fraser SE Dishevelled phosphorylation, subcellular localization and multimerization regulate its role in early embryogenesis EMBO J 2000 19 1010 1022 10698942 10.1093/emboj/19.5.1010
Deardorff MA Tan C Conrad LJ Klein PS Frizzled-8 is expressed in the Spemann organizer and plays a role in early morphogenesis Development 1998 125 2687 2700 9636083
Glinka A Wu W Delius H Monaghan AP Blumenstock C Niehrs C Dickkopf-1 is a member of a new family of secreted proteins and functions in head induction Nature 1998 391 357 362 9450748 10.1038/34848
Sokol SY Analysis of Dishevelled signalling pathways during Xenopus development Curr Biol 1996 6 1456 1467 8939601 10.1016/S0960-9822(96)00750-6
Cui Y Brown JD Moon RT Christian JL Xwnt-8b: a maternally expressed Xenopus Wnt gene with a potential role in establishing the dorsoventral axis Development 1995 121 2177 2186 7635061
Christian JL McMahon JA McMahon AP Moon RT Xwnt-8, a Xenopus Wnt-1/int-1-related gene responsive to mesoderm-inducing growth factors, may play a role in ventral mesodermal patterning during embryogenesis Development 1991 111 1045 1055 1879349
Epstein M Pillemer G Yelin R Yisraeli JK Fainsod A Patterning of the embryo along the anterior-posterior axis: the role of the caudal genes Development 1997 124 3805 3814 9367436
|
16236168
|
PMC1291380
|
CC BY
|
2021-01-04 16:39:15
|
no
|
Cell Commun Signal. 2005 Oct 19; 3:12
|
utf-8
|
Cell Commun Signal
| 2,005 |
10.1186/1478-811X-3-12
|
oa_comm
|
==== Front
Cost Eff Resour AllocCost effectiveness and resource allocation : C/E1478-7547BioMed Central London 1478-7547-3-101625962510.1186/1478-7547-3-10ResearchIs it worth offering a routine laparoscopic cholecystectomy in developing countries? A Thailand case study Teerawattananon Yot [email protected] Miranda [email protected] International health Policy Program, Bureau of Policy and Strategy, Ministry of Public Health, Nonthaburi, Thailand2 School of Medicine, Health Policy and Practice, University of East Anglia, Norwich, UK2005 31 10 2005 3 10 10 21 8 2005 31 10 2005 Copyright © 2005 Teerawattananon and Mugford; licensee BioMed Central Ltd.2005Teerawattananon and Mugford; licensee BioMed Central Ltd.This is an Open Access article distributed under the terms of the Creative Commons Attribution License (), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
Objective
The study aims to investigate whether laparoscopic cholecystectomy (LC) is a cost-effective strategy for managing gallbladder-stone disease compared to the conventional open cholecystectomy(OC) in a Thai setting.
Design and Setting
Using a societal perspective a cost-utility analysis was employed to measure programme cost and effectiveness of each management strategy. The costs borne by the hospital and patients were collected from Chiang Rai regional hospital while the clinical outcomes were summarised from a published systematic review of international and national literature. Incremental cost per Quality Adjusted Life Year (QALY) derived from a decision tree model.
Results
The results reveal that at base-case scenario the incremental cost per QALY of moving from OC to LC is 134,000 Baht under government perspective and 89,000 Baht under a societal perspective. However, the probabilities that LC outweighed OC are not greater than 95% until the ceiling ratio reaches 190,000 and 270,000 Baht per QALY using societal and government perspective respectively.
Conclusion
The economic evaluation results of management options for gallstone disease in Thailand differ from comparable previous studies conducted in developed countries which indicated that LC was a cost-saving strategy. Differences were due mainly to hospital costs of post operative inpatient care and value of lost working time. The LC option would be considered a cost-effective option for Thailand at a threshold of three times per capita gross domestic product recommended by the committee on the Millennium Development Goals.
==== Body
Introduction
It is widely accepted that laparoscopic cholecystectomy (LC) is the first-line treatment for uncomplicated gallstone disease in developed countries where up to 80% of all cholecystectomy is performed through laparoscopy [1-3].
In contrast to the conventional open cholecystectomy (OC), which is performed through an approximately 15-centimeter right sub-costal incision and commonly causes a serious degree of postoperative pain and longer hospital stay, LC is associated with a shorter hospitalization, more rapid return to work and better quality of life, at least in the short and intermediate term (3 years) after the operation[4].
The saving of hospital costs by LC is, however, not yet in agreement [5-8]] due to a higher operation cost which contributes around 60% of total hospital costs[9]. Furthermore, high complication rates were reported[10,11]. A meta-analysis also concludes that LC appeared to have a higher rate of common bile duct injuries[12].
In Thailand, since LC was first introduced in 1993, its adoption by healthcare providers has been relatively slow. By 2001, LC accounted for only 17% of the overall rate of cholecystectomy[13]. Some factors could explain the slow diffusion rate of the new procedure in Thailand. Firstly, the absence of financial incentives is believed to be a major cause. Under the capitation of the largest public insurance scheme, Universal Coverage Scheme (UC), OC has been reimbursed, but not LC. Thus, up to 10,000 Baht of co-payment by patients is needed if patients want laparoscopic surgery. Unsurprisingly, the proportion of UC patients undergoing LC, 13%, was the lowest across public health insurance schemes[13].
Patients under Civil Servant Medical Benefit Scheme (CSMBS) did not face the same financial disincentives. The CSMBS, covering government employees and their relatives, reimburses hospitals at a fixed rate for certain surgical procedures but, unfortunately, reimbursement rates set for OC and LC are alike, even though LC is associated with a shorter hospital stay. CSMBS reimburses the operation and hotel cost of hospital admission separately. A better socioeconomic status among CSMBS beneficiaries compared to UC may explain the greater LC uptake[14]. LC accounted for 28% of all cholecystectomy for these patients[13].
Secondly, the initial costs of setting up the equipment to perform LC and need for training for surgical team might have limited the ability of some hospitals to perform the procedure and these hospitals are likely to continue the status quo unless having strong support from Ministry of Public Health[13].
Given the lack of uptake and the greater potential cost-effectiveness, there is an urgent need to determine the value for money of LC compared to the traditional OC. Appropriate assessment of evidence may help to identify whether LC should be reimbursed under the Thai public insurance system. The aim of the present study is to investigate whether LC is a cost-effective strategy for managing gallbladder stone disease in the setting of Thailand. Although there are a number of economic evaluations on LC versus OC[6,7,15-18]], no single study was conducted in developing countries where cost burden may largely differ from those in western countries. The study is also conducted using societal perspective, which is recommended for public reimbursement[19], but rarely found in previous evaluations.
Materials and methods
Model
A decision tree validated by a group of experts in Thailand was used to model all of the clinically important outcomes of two different strategies for treating of gallbladder stones--pro LC versus pro OC policy (Figure 1). The start of the decision tree is the case of a patient who is eligible for surgery for gallbladder stones--cholelithiasis, but the management strategy of cholelithiasis will differ based on whether or not common bile duct (CBD) stones are presented. Surgeons do not know with certainty, which of their patients with cholelithiasis actually have CBD stones. This is accounted for by explicitly modelling the effects of each strategy for the two different situations--one where physician suspects the existing of CBD stones e.g. acute pancreatitis, duct dilatation of CBD, obstructive jaundice[2,20,21] and one where there are no suspected signs and symptoms of having CBD stones.
Figure 1 Decision tree illustrating the probable course of events for the management strategies being compared. A node in the represented by the box is the point of making decision between alternatives; nodes in the represented by the circle are points of events occurred.
The diagnosis of CBD stones or choledocholithiasis, could be done by ultrasonography, intravenous cholangiography, endoscopic retrograde cholangiography (ERCP) or Magnetic Resonance Image (MRI). Cholangiography and ERCP are most commonly used for definitive diagnosis in Thailand. Routine use of preoperative ERCP or intraoperative cholangiography (IOC) yields very little benefit for detection of CBD stones over and above that which is obtained with selective policies[16]. Hence, all patients without suspected CBD stones are directly assigned for LC or OC. Only those undergoing LC the result may be successful or unsuccessful (conversion from LC to OC). The final outcome is either cure or having bile-duct injuries.
For the patients with suspected CBD stones under pro LC policy, they will be assigned for ERCP or IOC depending on availability of the procedure. For those with positive IOC, open choledocholelithectomy is a choice of treatment. On the other hand, patients with positive ERCP undergo endoscopic sphincterotomy, and if it is unsuccessful, then open choledocholelithectomy is performed.
Under the pro OC policy the patients with suspected CBD stones receive IOC and if positive, open choledocholithectomy is the definitive treatment.
Estimation of parameter probabilities
All parameters for the model were obtained from a systematic review of national and international literature, searching of Thai- and English-language studies. We identified all English language articles published up to year 2004 from Medline and EMBASE using searching keyword "cholecystectomy". The Thai literature, including related Masters and Docteral's dissertations, were also identified using electronic and manual approach from database at Library of Mahidol University. We initially reviewed all identified abstracts. The full articles were considered if their objectives are about to quantify mortality, morbidity, compliance and treatment costs.
The mortality outcome of the laparoscopic and open strategies was considered to be equivalent[12]. A higher rate of common bile duct injuries, but better quality of life in short and intermediate terms were demonstrated for LC[4,16]. A cost-utility analysis in terms of Baht per Quality Adjusted Life Year (QALY) was, therefore, selected as the analytical approach.
Table 1 presents the probabilities for all clinical outcomes in the model and the utility values, as well as the range tested in sensitivity analysis. For example, the probability of finding patients with suspected signs and symptoms of having common bile duct stones was set at 0.3440 and its standard error (SE) was 0.0123[22,23]. With the overall incidence of CBD stones with among patients with cholelithiasis at 10%, the probability of having common bile-duct stones among suspected cases was set at 0.293 with SE of 0.045.
Table 1 Corresponding transitional probabilities of epidemiological variables and the utility values
Input Variables Point estimates (mean) Standard error for uncertainty analysis Data sources
Epidemiological variables
Probability of having suspected signs of CBD stones 0.3439 0.0123 22–24
Probability of having CBD stones among suspected cases 0.2929 0.0455 1
Proportion of ERCP available for patients who need it 0.5000 0.0498 Expert opinion
Probability of conversion from LC to OC 0.0550 0.0010 12, 24, 26
Probability of bile duct injury among patients undergoing LC 0.0050 0.0003 12
Probability of bile duct injury among patients conversed from LC to OC 0.0030 0.0017 Expert opinion
Probability of bile duct injury among patients undergoing OC 0.0024 0.0004 12
Probability of bile duct injury among patients undergoing open explored CBD 0.0010 0.0010 Expert opinion
Probability of retained CBD stones after undergoing ERCP 0.1279 0.0358 24–26
Utility variables
Utility of case with completed OC 0.80 0.02 16
Utility of case with completed LC 0.90 0.02 16
Utility of case with bile-duct injury in the first year 0.80 0.02 16
Utility of case with bile-duct injury in the subsequent twenty years 0.89 0.01 16
CBD = common bile duct
ERCP = endoscopic retrograde cholangiopancreatograpy
LC = laparoscopic cholecystectomy
OC = open cholecystectomy
Updating the meta-analysis conducted by Shea et al[12] the conversion rate from LC to OC was 0.055 (SE = 0.001) and the probability of bile duct injury among patients undergoing LC and OC were 0.0050 (SE = 0.0003) and 0.0024 (SE = 0.0024) respectively. A consensus from the Thai expert panel meeting to review the model indicated that a probability of bile duct injury among patients conversed from LC to OC should be something in between the probabilities of injury among patients undergoing OC and the probabilities of injury among patients undergoing LC. We assumed the probability of 0.0030 and SE of 0.0017. The panel also indicated that a probability of bile duct injury among patients undergoing open explored CBD should be the lowest. We set the rate and its standard error of 0.001. The meta-analysis of studies by Sangsubhan et al[24], Rhodes et al[25], and Konstadoulakis et al[26] found a probability of retained CBD stones after undergoing ERCP at 0.128 (SE = 0.0358).
Utility
Laparoscopic cholecystectomy was superior to open cholecystectomy in enhancing the quality of life for all eligible patients. The most comprehensive quality of life assessment study found from our review is of Cook et al[16]. They conducted a prospective assessment from 96 members of the general public in Melbourne using a standard questionnaire of which its questions included frequency and intensity of clinical factors e.g. pain, nausea, vomiting, identified from previous patient's interviews. We used their utility values and confidential intervals adjusted by 12-month values as our utility inputs (In fact, Topcu et al[4] indicated the difference in utility between LC and OC at 36 months. We took conservative estimation, assuming the difference of utility lasted by 12 months.)
Resource use parameters
In this study both direct and indirect costs borne by health care providers and households were collected in two ways: patient questionnaire interviews and data extraction from hospital case notes.
Chiang Rai regional hospital was purposively selected for costing study. This was the only public hospital that provided free of charge of LC and OC for UC beneficiaries, while other public hospitals offered only OC free. The prospective reviews of medical records of all patients undergoing open or laparoscopic cholecystectomy during September to November 2004 were performed. We also conducted a retrospective medical record review for rare conditions i.e. open choledocholithectomy, cases of conversion from LC to OC, treatments for CBD injuries, and endoscopic sphincterotomy, which had occurred in the past two years (October 2002-September 2004).
The case note review aimed to measure hospital resources consumed (e.g. clinician's time; type and number of investigations, drugs, and disposable equipments; length of hospital admission) in excess of usual preoperative, operative, and postoperative care. To ensure comparability of the groups, exclusion criteria were defined in such a way that an OC could be compared with a LC. Of a total of 80 medical records reviewed, 48 records (60%) of acute cholecystitis patients initially presented with peritonitis, sepsis, neoplasm, or co-existing conditions i.e. uncontrolled diabetes mellitus, uncontrolled hypertension were excluded from the analysis.
The valuation of the hospital costs, including capital and overhead costs, for each individual patient was determined from each department involved in the routine services for patients with gallstones disease.
We completed micro-cost analysis of 15 cases of OC, 17 cases of LC, 7 cases of open choledocholithectomy, 3 case of conversion from LC to OC, 3 case of treatment of CBD injuries, and 1 case of endoscopic sphincterotomy. The mean age was 60.8 years for patients underwent OC and 58.1 years for LC, there was no significant difference (using the student t-test, p > =0.05).
Thirty-two patients who had the operation in October and November 2004 were also contacted to determine indirect costs. Daily event and cost questionnaires were used to collect patient-specific information for estimation of other household expenses e.g. travel costs, food, accommodation and opportunity loss from providing informal care and visits of relatives and friends at hospital and home, patient's recovery time to full activity after surgery. We conducted face-to-face interviews with patients at the post-operation visit (2–4 weeks after operation). For the patients who did not get back to the hospital at all or having longer follow-up period than 4 weeks, we used telephone interview to collect that data.
Because of very low complication rates and no available information on indirect costs related to complications, we assumed in the model a similar indirect cost for patients with and without complications.
The costs were represented in the model in Thai Baht in 2004 (40 Baht = 1USD or 75 Baht = 1 GBP). All costs and outcomes occurred beyond one year were discounted using the same rate of 3.5%.
Uncertainty analysis
To determine if values within a plausible range for all input variables resulted in a different conclusion, we undertook probabilistic uncertainty analysis, assigning a beta distribution for all probability and utility parameters and gamma distribution for all cost parameters, and generating 1,000 rounds of simulations using Microsoft Excel® with macro function on all estimated quantities.
The cost-effectiveness acceptability curve based on the net benefit approach is also provided to present the relation between the values of the ceiling ratio (willingness to pay for a unit more of QALY) and probability of favouring each treatment strategy.
Results
Costs
Table 2 summarises all important cost parameters for use in the economic evaluation model. It is worth noting that the average hospital cost for OC in all 15 patients (9,355 Baht per OC case, SE = 717) was lower than the average cost for LC in all 17 cases (20,790 Baht per LC case, SE = 507). However, length of hospital stay and time to full recovery were markedly reduced in patients undergoing LC (mean 3.8 days, SE 0.4) compared to those having OC for LC (mean 6.1 days, SE 0.5) and, therefore, made a much lower indirect cost (mean 8,617 Baht for LC versus 14,484 Baht for OC).
Table 2 Costs (per patient) related to open and laparoscopic cholecystectomy, in 2004 Thai Baht, and used as inputs in the model
Open cholecystectomy Laparoscopic cholecystectomy
Variables Mean Standard error Mean Standard error
Direct costs
Cholecystectomy (pre-, intra-, and post-operation) 9,355 717 20,790 507
Conversion from Laparoscopic to open cholecystectomy 25,782 1,518
Intraoperative cholangiography (IOC) 1,502 154 1,502 154
Endoscopic retrograde cholangiopancreatography (ERCP) 2,011 120
Open choledocholithectomy 15,201 1,590 15,201 1,590
Endoscopic sphincterotomy 9,923 639
Treatment bile duct injury 12,068 2,926 12,068 2,926
Self-prescriptions and visiting private clinics (after discharge from hospital) 566 204 567 249
Indirect costs
Foods, accommodations, and lost working and spare times, on the part of relatives during admission 7,519 383 3,810 759
Lost working and spare times, on the part of relatives after discharging from hospital 2,945 1,409 2,693 650
Lost working and spare times, on the part of patients (on a whole course) 4,008 1,004 2,069 488
Transfer costs (i.e. sick compensations) 69 55 17 12
Cost-utility
Programme costs and outcome at base-case scenario for each treatment strategy are demonstrated in table 3. The programme costs are lower for pro OC policy in both using government's and societal viewpoints. However, the gap of programme costs (incremental cost) between the two strategies is 42% [(12,000-7,000)/12,000] smaller in the use of a societal viewpoint.
Table 3 Deterministic results from the model
Open cholecystectomy Laparoscopic cholecystectomy Incremental values
Programme cost using government perspective 11,000 23,000 12,000
Programme cost using societal perspective 26,000 33,000 7,000
Programme effectiveness (QALYs) 0.798 0.885 0.087
Cost per QALY using government perspective 134,000
Cost per QALY using societal perspective 89,000
Note: the costs and incremental values were given to nearest 1,000 Baht price level
Pro OC and LC strategy provide 0.798 and 0.885 QALYs, respectively. On the other hand, moving from a cheaper and lower effectiveness strategy, pro OC, to pro LC policy yields extra 0.087 QALYs.
When only direct costs were compared, an incremental cost per QALY of moving from OC to LC is 134,000 Baht. The incremental cost-effectiveness ratio decreases to 89,000 Baht per QALY when including indirect cost of a wider societal perspective. Moving from OC to LC would add a financial burden of 12,000 Baht per case to the government but offset the indirect costs of 5,000 Baht that is presently shouldered by the households.
Uncertainty analysis
The cost-effectiveness acceptability curves in figure 2 summarise the robustness of the model regarding uncertainty estimation of the programme cost and effect for each treatment strategy. At the zero ceiling ratio, indicating that no further resources would be allocated to healthcare, pro OC policy is a dominant strategy, particularly using a government perspective. When the ceiling ratios are greater than 90,000 and 140,000 using government and societal perspective respectively, pro LC policy becomes a preferable choice, offering a better chance of saving one QALY at given money. However, the probabilities that LC outweighed OC are not greater than 95% until the ceiling ratio reaches 190,000 and 270,000 Baht per QALY using societal and government perspective respectively.
Figure 2 Cost-effectiveness acceptability curves using net benefit approach.
Discussion
The study provides the same results using the two different viewpoints of the analysis; LC is a more expensive but likely to provide a better quality of life than OC. At base-case scenario the ratio of extra cost of LC to its extra utility gained varies between 89,000 and 134,000 Baht per QALY, depending on whether indirect costs are included.
To make the model as simple as possible and to focus only on a comparison between the use of OC and LC, the model has a limitation about retained CBD stones between the two strategies. Retained CBD stones could arise from a difference in the use of ERCP and IOC between the two approaches. However, estimating the sensitivity and specificity of IOC and ERCP is not straightforward since the two investigations and surgery are not performed simultaneously, so it is possible that the stones can migrate out of the CBD spontaneously in the interval between investigation and surgery, or that additional stones enter the CBD from the gallbladder. The review of literature by Urbach et al[27] indicates that both IOC and ERCP provided the same specificity in detecting CBD stones but ERPC had superior sensitivity to IOC (95% vs 89%). In the other words, IOC would provide additional 5 false negative cases compared to ERCP. The literature also reveals that not all retained stones would be a problem but only 15% of these would go on to cause clinical problems[21]. Measuring QALYs for the retained CBD stones is also problematic. We found no study assessing utility of retained CBD stones.
The results of the economic evaluation for management options for gallstone disease in Thailand are irrelevant to comparable previous studies conducted in developed countries of which Cook J et al[16] and Berggren et al[7] found that LC was a cost saving strategy in comparison to OC. The difference could be explained with two reasons. Firstly, a higher wage rate for both health professionals and patients in US and Australia caused a significant higher hospital admission costs and opportunity cost of taking sick-leave. Secondly, the indirect costs quantified by the two studies are an overestimation, especially when applied to the Thai context. Since, following guidance on estimating 'friction costs' of lost work[28], the studies estimated the costs by multiplying average employment costs with proportion of population in the work force, but our study found that only 50% and 53% of patients undergoing open and laparoscopic cholecystectomy were active workforce members. However, it may be important to add a comment that this approach does not value the time of people not in paid work, even though many are productive members of society, e.g. self employed farmers, or unpaid house works.
Similar to the previous study[9], an operation cost contributed to 75% of total hospital costs for LC compared to 20% for OC. More than two third (70%) of operation cost was from operative instruments, although the sample hospital was quite efficient in using these instruments since, where possible, reusable instruments were introduced. Thus, we believe there is a little room to make the operation cost smaller.
To allow for the extra quality of life gained from LC, however, additional funding would be required. The judgement about whether to advocate LC over OC in such situations would depend on what benefit could be obtained from the use of these extra resources elsewhere. However, a broad comparison across health care interventions is unlikely due to the fact that it would have had an enormous work of analysis of possible treatments.
Another approach is setting a ceiling value for health benefit that society is willing to pay. The committee for development of Millennium Development Goals recommends the use of three times of Gross Domestic Product (GDP) per capital as a threshold for the consideration in developing countries. This application would presently lead to a ceiling value in Thailand of 270,000 Baht. In this case, LC is a cost-effective intervention. If extra funds could not be obtained for the health system directly, resources would have been obtained from elsewhere e.g. a co-payment system.
Furthermore, economic evaluations are commonly criticised by decision makers for ignoring budget impacts, about which decision makers desperately concerned. Payers can get into financial difficulty if they adopt too many cost-effectiveness interventions[29] and affordability, which depends on the overall volume of patients, is therefore a prime concern. We projected the financial implication if pro LC policy is adopted in Thailand (table 4). Assuming that 80% of patients who need cholecystectomy eligible for LC, the government would require 96 million Baht for supporting pro LC policy. At the same time, households could save 40 million Baht from indirect medical care cost resulted in the net of 56 million Baht required by society as a whole.
Table 4 Estimated financial burden on government budget and off-set cost by households if pro LC policy was adopted
Type of Insurance Estimated cases in 2005 (a) Estimated cases in 2005 (b) Incremental financial burden of moving from OC to LC by government (c) Household's off-set cost from moving from OC to LC (d) Net financial burden to society (e)
UC 8,000 6,400 76,800,000 32,000,000 44,800,000
CSMBS 2,000 1,600 19,200,000 8,000,000 11,200,000
Total 10,000 8,000 96,000,000 40,000,000 56,000,000
UC = Universal Health Insurance Scheme
CSMBS = Civil Servant Medical Benefit Scheme
Calculation:
(b) = (a) × 0.8
(c) = (b) × 12,000
(d) = (b) × (12,000-7,000)
(e) = (c) - (d)
Acknowledgements
We appreciate grant support by National Health Security Office and World Health Organization country office, Thailand. The first author is currently supported by the World Health Organization under the Fellowship Program award to study at the University of East Anglia.
The authors would especially like to acknowledge and thank the review panel members namely Dr.Thavee Rattanachuake, Dr.Thaveesin Tanprayoon, Dr.Chaiwate Ratanaprisaan, and Dr.Thumrong Tragnawathakarn, whose expertise was invaluable throughout this study. We also thank to Miss Sanya Srirattana for her excellent fieldwork including interviews with study participants, and all the study participants for their time and contributions to this work.
Finally, the review comments on an earlier version of this paper by Dr.Chulaporn Limwattananon and an anonymous reviewer are gratefully acknowledged.
==== Refs
Legorreta AP Silber JH Costantino GN Kobylinski RW Zatz SL Increased cholecystectomy rate after the introduction of laparoscopic cholecystectomy.[see comment] JAMA 1993 270 1429 1432 8371441 10.1001/jama.270.12.1429
Beckingham IJ ABC of diseases of liver, pancreas, and biliary system. Gallstone disease BMJ 2001 322 91 94 11154626 10.1136/bmj.322.7278.91
Hobbs MS Mai Q Fletcher DR Ridout SC Knuiman MW Impact of laparoscopic cholecystectomy on hospital utilization ANZ Journal of Surgery 2004 74 222 228 15043732 10.1111/j.1445-2197.2004.02955.x
Topcu O Karakayali F Kuzu MA Ozdemir S Erverdi N Elhan A Aras N Comparison of long-term quality of life after laparoscopic and open cholecystectomy Surg Endosc 2002 17 291 295 12364987 10.1007/s00464-001-9231-2
Pietrafitta JJ Graner TJ Nervick JM Laparoscopic cholecystectomy: cost effectiveness and quality enhancement International Surgery 1994 79 195 201 7883491
McKellar DP Johnson RM Dutro JA Mellinger J Bernie WA Peoples JB Cost-effectiveness of laparoscopic cholecystectomy Surgical Endoscopy 1995 9 158 62; discussion 162-3 7597585 10.1007/BF00191958
Berggren U Zethraeus N Arvidsson D Haglund U Jonsson B A cost-minimization analysis of laparoscopic cholecystectomy versus open cholecystectomy American Journal of Surgery 1996 172 305 310 8873518 10.1016/S0002-9610(96)00197-3
Bosch F Wehrman U Saeger HD Kirch W Laparoscopic or open conventional cholecystectomy: clinical and economic considerations European Journal of Surgery 2002 168 270 277 12375608 10.1002/ejs.45
Traverso LW Hargrave K A prospective cost analysis of laparoscopic cholecystectomy American Journal of Surgery 1995 169 503 506 7747829 10.1016/S0002-9610(99)80205-0
Habu Y Matsui T Hayashi K Watanabe Y Kawai K A clinical decision analysis to assess therapeutic modalities for symptomatic gallstones with respect to patient's quality of life and cost-effectiveness Endoscopy 1992 24 709 716 1425464
Strasberg SM Clavien PA Overview of therapeutic modalities for the treatment of gallstone diseases American Journal of Surgery 1993 165 420 426 8480874
Shea JA Healey MJ Berlin JA Clarke JR Malet PF Staroscil RN Schwartz SJ Williams SV Mortality and complications associated with laparoscopic cholecystectomy: a meta-analysis Annuals of Surgery 1996 224 609 620 10.1097/00000658-199611000-00005
Teerawattananon Y Pannarunothai S Measuring the effects of reimbursement policy on utilization of high-cost health services in Thailand (draft) 2004 Norwich, University of East Anglia 15452013
Tangcharoesnathien V SSPS Thammathataree J Overview of health insurance systems Health Insurance Systems in Thailand 2001 Bangkok, Desire publisher 28 37
Bass EB Pitt HA Lillemoe KD Cost-effectiveness of laparoscopic cholecystectomy versus open cholecystectomy American Journal of Surgery 1993 165 466 471 8480883
Cook J Richardson J Street A A cost utility analysis of treatment options for gallstone disease: methodological issues and results Health Economics 1994 3 157 168 7921059
Srivastava A Srinivas G Misra MC Pandav CS Seenu V Goyal A Cost-effectiveness analysis of laparoscopic versus minilaparotomy cholecystectomy for gallstone disease. A randomized trial International Journal of Technology Assessment in Health Care 2001 17 497 502 11758294
Rosen MJ Malm JA Tarnoff M Zuccala K Ponsky JL Cost-effectiveness of ambulatory laparoscopic cholecystectomy Surgical Laparoscopy, Endoscopy & Percutaneous Techniques 2001 11 182 184 11444748 10.1097/00019509-200106000-00006
Drummond M Sculpher M Torrance G O'Brien B Stoddart G edition Medthod for the economic evaluation of health care programmes 2005 2 Oxford, Oxford University Press
Millat B Borie F Decker G Treatment of choledocholithiasis: therapeutic ERCP versus peroperative extraction during laparoscopic cholecystectomy Acta Gastroenterologica Belgica 2000 63 301 303
Metcalfe MS Ong T Bruening MH Iswariah H Wemyss-Holden SA Maddern GJ Is laparoscopic intraoperative cholangiogram a matter of routine? The American Journal of Surgery 2004 187 475 481 15041494 10.1016/j.amjsurg.2003.12.047
Moreaux J Traditional surgical Management of common bile duct stones: A prospective study during a 20-year experience Amarican journal of surgery 1995 169 220 226 10.1016/S0002-9610(99)80141-X
Duensing RA Williams RA Collins JC Wilson SE Common Bile Duct Stone Characteristics:Correlation With Treatment Choice During Laparoscopic Cholecystectomy Journal of Gastrointestinal Surgery 2000 4 6 12 10631357 10.1016/S1091-255X(00)80027-X
Sangsubhan C Tanphiphat C Tanprayoon T Chatamra K A prospective study of 500 consecutive laparoscopic cholecystectomies Chula Med Journal 1995 39 79 89
Rhodes M Sussman L Cohen L Lewis MP Randomised trial of laparoscopic exploration of common bile duct versus postoperative endoscopic retrograde cholangiopancreatography for common bile duct stones The Lancet 1998 251 159 161 10.1016/S0140-6736(97)09175-7
Konstadoulakis MM Antonakis PT Karatzikos G Alexakis N Leandros E Intraoperative findings and postoperative complications in laparoscopic cholecystectomy: the Greek experience with 5,539 patients in a single center Journal of Laparoendoscopic & Advanced Surgical Techniques 2004 14 31 36 15035842 10.1089/109264204322862333
Urbach DR Khajanchee YS Jobe BA Standage BA Hansen PD Swanstrom LL Cost-effective management of common bile duct stones: a decision analysis of the use of endoscopic retrograde cholangiopancreatography (ERCP), intraoperative cholangiography, and laparoscopic bile duct exploration Surgical Endoscopy 2001 15 4 13 11178753 10.1007/s004640000322
Koopmanschap MA van Ineveld BM Towards a new approach for estimating indirect costs of disease Social Science & Medicine 1992 34 1005 1010 1631600 10.1016/0277-9536(92)90131-9
Ubel PA Hirth RA Chernew ME Fendrick AM What is the price of life and why doesn't it increase at the rate of inflation? Arch Intern Med 2003 163 1637 1641 12885677 10.1001/archinte.163.14.1637
|
16259625
|
PMC1291381
|
CC BY
|
2021-01-04 16:39:14
|
no
|
Cost Eff Resour Alloc. 2005 Oct 31; 3:10
|
utf-8
|
Cost Eff Resour Alloc
| 2,005 |
10.1186/1478-7547-3-10
|
oa_comm
|
==== Front
Emerg Themes EpidemiolEmerging Themes in Epidemiology1742-7622BioMed Central London 1742-7622-2-111626908310.1186/1742-7622-2-11Analytic PerspectiveThe Bradford Hill considerations on causality: a counterfactual perspective Höfler Michael [email protected] Clinical Psychology and Epidemiology, Max Planck Institute of Psychiatry, Kraepelinstrasse 2-10, 80804 München, Germany2005 3 11 2005 2 11 11 10 6 2005 3 11 2005 Copyright © 2005 Höfler; licensee BioMed Central Ltd.2005Höfler; licensee BioMed Central Ltd.This is an Open Access article distributed under the terms of the Creative Commons Attribution License (), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
Bradford Hill's considerations published in 1965 had an enormous influence on attempts to separate causal from non-causal explanations of observed associations. These considerations were often applied as a checklist of criteria, although they were by no means intended to be used in this way by Hill himself. Hill, however, avoided defining explicitly what he meant by "causal effect".
This paper provides a fresh point of view on Hill's considerations from the perspective of counterfactual causality. I argue that counterfactual arguments strongly contribute to the question of when to apply the Hill considerations. Some of the considerations, however, involve many counterfactuals in a broader causal system, and their heuristic value decreases as the complexity of a system increases; the danger of misapplying them can be high. The impacts of these insights for study design and data analysis are discussed. The key analysis tool to assess the applicability of Hill's considerations is multiple bias modelling (Bayesian methods and Monte Carlo sensitivity analysis); these methods should be used much more frequently.
==== Body
Introduction
Sir Austin Bradford Hill (1897 – 1991) was an outstanding pioneer in medical statistics and epidemiology [1-4]. His summary of a lecture entitled "The environment and disease: Association or causation" [5] had an enormous impact on epidemiologists and medical researchers. Ironically, this paper became famous for something it was by no means intended to be [6,7]: a checklist for causal criteria (e.g. [8-10]).
Hill [5] provided nine considerations for assessing whether an observed association involved a causal component or not. These considerations were influenced by others before him [11,12]. He avoided defining explicitly what he meant by a causal effect, although seemingly he had the counterfactual conceptualisation mind. My core thesis in this paper is that counterfactual arguments contribute much to the question of when to apply Hill's considerations to a specific causal question. This is not to say that other conceptualisations of causality would not contribute to clarifying Hill's considerations, but the counterfactual model is the one that directly relates to many statistical methods [13,14], and it links the "metaphysical" side of causality to epidemiological practice. Moreover, I shall argue that some of Hill's considerations involve many counterfactuals in a broader causal system, and that the heuristic value of these considerations can be low.
Analysis
Counterfactual causality
Hill [5] avoided defining exactly what he meant with a causal effect:
"I have no wish, nor the skill to embark upon a philosophical discussion of the meaning of 'causation'."
However, it seems that he applied the counterfactual model because he then writes:
"... the decisive question is whether the frequency of the undesirable event B would be influenced by a change in the environmental feature A."
Counterfactual causality dates back at least to the 18th century Scottish philosopher David Hume [15] but only became standard in epidemiology from the 1980s. Being the inventor of randomised clinical trials [3,4], Hill was strongly influenced by the idea of randomised group assignment, which precludes confounding. The idea of randomisation, invented by R.A. Fisher's in the 1920s and 1930s was, in turn, stimulated by Hume [3]. As Fisher and Hill were friends for at least some years [3], it seems likely that Hill was strongly influenced by counterfactual thinking. Today, the counterfactual, or potential outcome, model of causality has become more or less standard in epidemiology, and it has been argued that counterfactual causality captures most aspects of causality in health sciences [13,14].
To define a counterfactual effect, imagine an individual i at a fixed time. Principally we assume that
(a) this individual could be assigned to both exposure levels we aim to compare (X = 0 and X = 1 respectively) and
(b) that the outcome Yi exists under both exposure levels (denoted by Yi0 and Yi1 respectively) [[14] and references therein].
The causal effect of X = 1 versus X = 0 within an individual i at the time of treatment or exposure assignment can be defined as [13-20]:
Yi1 - Yi0.
Note that the use of the difference measure is not exclusive – for strictly positive outcomes one can also use the ratio measure Yi1/Yi0. For a binary outcome, this definition means that the outcome event occurs under one exposure level but not under the other. Therefore, a causal effect of a binary event is a necessary condition without which the event would not have occurred; it is not necessarily a sufficient condition. Clearly, the outcome is not observable under at least one of the two exposure levels of interest. Thus, the outcome has to be estimated under the unobserved or counterfactual condition, known as the counterfactual or potential outcome.
According to Rothman [21], a comprehensive causal mechanism is defined as a set of factors that are jointly sufficient to induce a binary outcome event, and that are minimally sufficient; that is, under the omission of just one factor the outcome would change. Rothman [21] called this the sufficient -component cause model. A similar idea can be found in an earlier paper by Lewis [22]. Since several causal mechanisms are in line with the same specific counterfactual difference for a fixed individual at a fixed time, the sufficient-component cause model can be regarded as a finer version of the counterfactual model [[14], and references therein].
As there are often no objective criteria to determine individual counterfactual outcomes, the best option is usually to estimate population average effects. The population average effect is defined as the average of individual causal effects over all individuals in the target population on whom inference is to be made. The estimation of average causal effects in epidemiology is subject to various biases [23]. These biases are determined both by the study design and the mechanism that generates the data. In a randomised controlled trial (RCT), bias due to confounding cannot occur, but confounders might be distributed unequally across treatment levels by chance, especially in small samples. If compliance is perfect, there is no measurement error in the treatment. Other biases might still occur, however, such as bias due to measurement error in the outcome and selection bias (because individuals in the RCT might not represent all individuals in the target population). Observational studies are prone to all kinds of biases, and these depend on the causal mechanism underlying the data. For instance, bias due to confounding is determined by the factors that affect both exposure and outcome, and the distribution of these factors.
I shall demonstrate that most of Hill's considerations involve more than the X – Y association and biases in that association; their application depends on assumptions about a comprehensive causal system, of which the X – Y effect is just one component. I argue that the heuristic value of Hill's considerations converges to zero as the complexity of a causal system and the uncertainty about the true causal system increase.
The Bradford Hill considerations
The discussion of Hill's considerations is organised as follows: first, I use my own wording (in italics) to summarise the respective consideration. Hill's own argumentation is then briefly reviewed, followed by arguments of other authors (a subjective selection from the vast literature). I will then show which counterfactuals are involved in the application of a given consideration and what novel insights can be derived for the interpretation of study design and data analysis. To simplify the discussions, I will sometimes disregard random variation. Some of my arguments apply to several of Hill's considerations and I will occasionally not repeat them to avoid redundancy.
1. Strength of association
A strong association is more likely to have a causal component than is a modest association
Hill [5] illustrated this point with the high risk ratios for the association between exposure levels of smoking and incidence of lung cancer. However, he demonstrated with two counter-examples that the absence of a strong association does not rule out a causal effect. Hill acknowledged that the impression of strength of association depended on the index used for the magnitude of association [5].
Rothman and Greenland [[18], p.24] provided counter-examples for strong but non-causal relationships. Note that, unlike ratio measures, difference measures tend to be small unless there is nearly a one-to-one association between exposure and outcome [24]. The fundamental problem with the choice of an effect measure is that "neither relative risk nor any other measure of association is a biologically consistent feature of an association... it is a characteristic of a study population that depends on the relative prevalence of other causes" [[18], p.24]. Rothman and Poole [25] described how studies should be designed to detect weak effects. For Rothman and Greenland [[18], p. 25] the benefit of the consideration on strength was that strong associations could not be solely due to small biases, whether through modest confounding or other sources of bias.
The consideration on strength involves two main counterfactual questions about biases that have presumably produced the observed association (in terms of a pre-specified index): how strong would the association be expected to be as compared to the observed association if the data were free of bias? Would the interval estimate that properly accounts for not only random, but also systematic error (uncertainty about bias parameters such as misclassification probabilities) allow for the desired conclusion or not? (A desired conclusion might be the simple existence of a causal effect or a causal effect of at least a certain magnitude, for instance a two-fold increase in risk.)
Bias can be addressed with multiple bias modelling. Contemporary methods for multiple bias modelling include Bayesian methods and Monte Carlo sensitivity analysis (MCSA, which can be modified to be approximatively interpretable in a Bayesian manner under certain conditions [27]). These methods address the uncertainties about bias parameters by assigning prior distributions to them. Although Hill pointed out that non-random error was often under-estimated, these methods were hardly available in his time. With Bayesian and MCSA methods one can assess whether the observed magnitude of association is sufficiently high to allow for a certain conclusion. This requires a bias model including assumptions on which kinds of biases exist, how biases act together and which priors should be used for them. However, if one uses a bias model that addresses understood bias, the inference could still be distorted by misunderstood or unknown bias. Moreover, one can calculate which X – Y association and random error values would have allowed for the desired conclusion. One may also ask which priors on bias parameters would, if applied, have allowed for the desired conclusion ("reverse Bayesian analysis") and assess the probability that these priors are not counterfactual.
Clearly, high uncertainty about bias parameters requires larger associations than modest uncertainty does. In studies that control for several sources of bias, modest associations might still be indicative of a causal effect, whereas in more error-prone designs more bias and higher non-random error has to be taken into account. However, specifying a bias model can be a difficult task if the knowledge about biases is also limited. This uncertainty then carries over to the uncertainty on applying the consideration on strength. Despite this, there seems to be no alternative to multiple bias modelling when assessing which magnitude of association is necessary for the desired conclusion.
2. Consistency
A relationship is observed repeatedly
For Hill [5], the repeated observation of an association included "different persons, places, circumstances and time". The benefit of this rule was that consistently finding an association with different study designs (e.g. in both retrospective and prospective studies) reduced the probability that an association would be due to a "constant error or fallacy" in the same study design. On the other hand, he pointed out that shared flaws in different studies would tend to replicate the same wrong conclusion. Likewise, differing results in different investigations might indicate that some studies correctly showed a causal relationship, whereas others failed to identify it.
This point is explained by Rothman and Greenland [[18], p. 25]: causal agents might require that another condition was present; for instance, transfusion could lead to infection with the human immunodeficiency virus only if the virus was present. Now, according to the sufficient-component cause model [20,21], and as stated by Rothman and Greenland [[18], p. 25], whether and to what extent there is a causal effect on average depends on the prevalences of complementary causal factors.
Cox and Wermuth [[28], pp. 225] have added the consideration that an association that does not vary strongly across the values of intrinsic variables would be more likely to be causal. If an association were similar across individuals with different immutable properties, such as sex and birth date, the association would be more likely to have a stable substantive interpretation. Variables other than X and Y might change as a consequence of interventions among other factors in a comprehensive causal system. One should be careful when applying this guideline; effect heterogeneity depends on the choice of the effect measure. This choice should be based on a relevant substantive theory and on correspondence with the counterfactual and sufficient-component cause model (the latter two indicating that differences rather than ratios should be used); both may, however, contradict [29].
From the counterfactual perspective, the following questions arise when asking whether to apply the consideration on consistency:
a) If the causal effect was truly the same in all studies, would one expect to observe different associations in different studies (possibly involving different persons, places, circumstances and time)? To what degree would the associations be expected to differ?
b) If the causal effect varied across the studies, would one expect to observe equal or different associations? What magnitude of differences would one expect?
Note that in the presence of effect modifiers there exists no such thing as "the causal effect", the effect modifiers need to be fixed at suitable values. Also note that only a) or b) is actually counterfactual depending on whether the effect truly varies across the different studies or not. Answering these questions requires a comprehensive causal theory that indicates how different entities (individual factors, setting, time, etc.) act together in causing Y. Within such a causal system one can predict how the X – Y association should change if one used different persons, places, circumstances and times in different studies. As one can only observe associations this also involves bias, and bias might operate differently in different studies.
An observed pattern of association across the different studies that is in line with the expected pattern would provide evidence for an effect of X on Y if the underlying causal theory applies. Another pattern would indicate that there is either no effect of X on Y or that the supposed theory is false. In complex situations and bias-prone designs, the probability might be substantial that a causal theory does not include important features that change the expected X – Y association. Here, the uncertainty regarding whether or not to demand an association (or which magnitude of association) could be high, and so the consistency consideration might bring more harm than benefit.
3. Specificity
A factor influences specifically a particular outcome or population
For Hill [5], if one observed an association that was specific for an outcome or group of individuals, this was a strong argument for a causal effect. In the absence of specificity, Hill alludes to fallacies in applying this rule to conclude the absence of a causal effect: Diseases may have more than one cause (which Hill considered to be the predominant case). In turn, a factor might cause several diseases. According to Hill, the value of this rule lay in its combination with the strength of an association: For instance, among smokers, the risk of death from lung cancer should be elevated to a higher degree as compared to the risk of other causes of death. Hill's consideration on specificity for persons apparently contradicts his consideration on consistency, where repeatedly observing an association in different populations would increase the evidence for a causal effect.
Rothman and Greenland [[18], p. 25] have argued that when applying this rule one assumes that a cause had one single effect. This assumption is often meaningless; for instance, smoking has effects on several diseases. They considered the demanding of specificity "useless and misleading". Cox and Wermuth [[28], p. 226f.] pointed out that this rule applies to systems where quite specific processes act, rather than to systems where the variables involved represent aggregates of many characteristics. Weiss [30] has mentioned a situation in which it can be meaningful to require specificity with respect to an outcome: a theory could predict that an exposure affects a certain outcome, but does not affect other particular outcomes. He illustrated this with the example that wearing helmets should be protective specifically against head injury, not against injury of other parts of the body. If wearing helmets also protected against other injuries, so he argues, this could be indicative that the association is confounded by more careful riders tending to use helmets. He provided similar arguments for specificity of exposure and specificity with regard to individuals in whom a theory predicts an effect.
Counterfactual causality, and the logically equivalent causal graphs [19,30], generalise the argument of Weiss [30] and solve the problem of specificity with respect to other exposures and outcomes: outcomes other than the one under consideration (Y) must be related with the exposure (X) if they are either part of the causal chain between X and Y or a causal consequence of Y. Otherwise, they must not be associated with X. In the example above, other injuries are neither part of the causal chain between wearing helmets and head injury, nor the causal consequence of head injury; the association between wearing helmets and head injury might share the common cause of carefulness if wearing helmets were related to both head and other injury. Likewise, exposures other than X must be associated with Y if they belong to the causal chain between X and Y. Whether exposures that occur before X are associated with Y or not is not informative about causality between X and Y.
When applying this consideration, a bias model is required for each association in the entire causal system, involving the assessment of as many counterfactual differences as there are associations. A single wrong conclusion about the existence of a particular effect might still yield a graph that contradicts the theory and, thus, a wrong conclusion about the existence of the X – Y effect. RCTs only allow a small number of factors to be simultaneously randomised, and whether they are associated with one another is just a question of how the randomisation is done. The assessment of specificity with respect to other factors in RCTs is therefore limited. Cohort studies are more useful here, but confounding and measurement error in the exposure are of higher importance. To conclude, the consideration of specificity appears to be useful only when a causal system is simple and the knowledge about it is largely certain.
4. Temporality
The factor must precede the outcome it is assumed to affect
Hill [5] introduced this reflection with the proverb "Which is the cart and which is the horse?" For instance, he asked whether a particular diet triggered a certain disease or whether the disease led to subsequently altered dietary habits. According to Hill, temporal direction might be difficult to establish if a disease developed slowly and initial forms of disease were difficult to measure.
Considering individuals in whom X has occurred before Y, it is logically not possible that X would have changed if Y had changed, because X is fixed at the time when Y occurs (or not). Thus, Y cannot have caused X in these individuals. This is, indeed, the only sine qua non criterion for a counterfactual effect in a single individual [[7], p. 27, 11] – a point missed by Hill. Note that there is no logical link to individuals in whom Y has occurred before X. Among those, Y might or might not have caused X [[7], p. 25].
Even more confusion in applying this criterion arises when one aggregates information across several individuals. Some researchers believe that an association that is only observed in one direction is more likely to be causal than an association that is observed in both directions. If the presence of a certain disease (X) is associated with a higher subsequent incidence rate of another disease (Y), and if prior Y also predicts an elevated probability of subsequent onset of X, it is sometimes assumed that a shared vulnerability was the common cause of both associations. This possibility has, for instance, been discussed for the role of anxiety in the development of depression [31].
Applying this argument, however, requires a causal system that produces no Y – X association. This requires sufficient knowledge about shared risk factors of X and Y. To assess temporal order between X and Y, a longitudinal design is preferable to a design in which the temporal direction between X and Y has to be assessed retrospectively. In RCTs, temporal direction can be established without error.
5. Biological gradient
The outcome increases monotonically with increasing dose of exposure or according to a function predicted by a substantive theory
Hill [5] favoured linear relationships between exposure level and outcome, for instance, between the number of cigarettes smoked per day and the death rate from cancer. If the shape of the dose-response relationship were a more complex, especially a non-monotonic, function, this would require a more complex substantive explanation.
Others have been less demanding and more specific in their definition of a dose-response relation, requiring only a particular shape of relationship (not necessarily linear or monotonic), which is predicted from a substantive theory [[28], p. 225]. Rothman and Greenland [[18], p. 26] have argued that parts of J-shaped dose-response curves might be caused by the respective exposure levels while others might be due to confounding only. They also provided a counter-example for a non-causal dose-response association. To demand a dose-response relationship could be misleading if such an assumption contradicted substantive knowledge. No dose-response relationship in presumably causal effects has been found, for example, between the intake of inhaled corticosteroids and lung function among asthma patients [33] or in the pharmacotherapy of mental disorders [34]. Further examples and similar arguments as above had been provided previously by Lanes and Poole [35].
Counterfactual causality defines the difference between each pair of exposure levels as a distinct causal effect. The consideration on biological gradient is therefore again not a consideration on a specific causal difference but a consideration on a broader causal system involving several exposure levels. It requires a substantive theory that predicts how the outcome should change when the exposure varies over several levels. If there are k exposure levels, then this theory has to predict k-1 counterfactual differences. Some theories demand a gradient over the levels and others do not, while different theories might demand different gradients.
When applying this consideration, bias has to be properly corrected for each of the k-1 observed associations. If the observed sequence of associations over the exposure levels is in line with a theory and bias is properly addressed for each comparison, this provides evidence for the theory; otherwise, the theory, or at least one bias model, is false. Here, causal differences between specific exposure levels might in any case exist.
Several exposure levels are required to establish a dose-reponse relationship. On the other hand, the more exposure levels there are, the higher is the danger of mis-applying this consideration, because a single wrong conclusion (among k-1 possible wrong conclusions) about the existence of a specific causal difference might be sufficient for a wrong conclusion on the overall theory. RCTs are particularly useful for assessing dose-response relationships, because they avoid some biases that are sometimes difficult to correct for using other study designs.
6. Plausibility
The observed association can be plausibly explained by substantive matter (e.g. biological) explanations
For Hill [5], the presence of a biological explanation supported the drawing of a causal conclusion. On the other hand, in the absence of such a theory "the association we observe may be one new to science or medicine and we must not dismiss it too light-heartedly as just too odd" [5].
Cox and Wermuth [[28], p. 226] have added the extremely important point that a relationship that is predicted prospectively is much more convincing than one that is provided retrospectively; after observing an association, it is often easy to give a plausible explanation. According to Rothman and Greenland [[18], p. 26], the assessment of plausibility is subject to the prior beliefs of individual researchers. The weight of these prior beliefs could be balanced against the weight of the observed association in Bayesian inference. However, Bayesian analysis is not able to "transform plausibility into a causal criterion".
Based on the arguments of these authors there are two relevant counterfactual questions that researchers should ask themselves; they are related to plausibility, although they are not sufficient to clarify this consideration on their own:
a) If the observed association is in line with substantive knowledge, would you have assigned it a lower weight (relatively to the weight of substantive knowledge) if your observations had not been in line with substantive knowledge?
b) If the observed association is not in line with substantive knowledge, would you have assigned it a higher weight (relatively to the weight of substantive knowledge) if your observations had been in line with substantive knowledge?
Only if researchers are able to answer a) or b) (respectively) honestly with "yes" can one assume that the application of this consideration did not depend on the study results. Clearly, one cannot even be 100% confident that an answer is really honest for someone's own thoughts. The danger in applying this consideration is twofold: researchers might assign a higher weight to substantive knowledge if it agrees with their own prior opinion and might assign a lower weight otherwise. Substantive knowledge might be inconsistent and conflicting information could be weighted according to someone's assessment of the accuracy of the information. Likewise, scientists might choose the substantive knowledge they apply according to whether it fits to their prior opinions. The question remains how to weight prior opinions relatively to the observed results. This involves many design issues, such as the sample sizes in the present data and in other data, and the questions of how bias acted and was addressed in different studies.
7. Coherence
A causal conclusion should not fundamentally contradict present substantive knowledge
Hill [5] used the term "generally known facts" to indicate that the knowledge against which an association is evaluated has to be undisputable. Laboratory evidence that is in line with an association would underline a causal conclusion and help to identify the causal agent. Again, the absence of such knowledge would not be indicative of a non-causal explanation.
The difference in Hill's definitions of plausibility and coherence appears to be subtle [[7], p. 25]. Whereas plausibility is worded positively (an association that should be in line with substantive knowledge), coherence is verbalised negatively (an association that should not conflict with substantive knowledge). Rothman and Greenland [[7], p. 25] have drawn attention to the possibility that such conflicting knowledge might itself be wrong. Susser [11] has tried to retain this consideration by defining different subclasses of coherence depending on where knowledge comes from.
A subtle difference between coherence and plausibility is that plausibility asks: "Could you imagine a mechanism that, if it had truly operated (which could be counterfactual), would have produced results such as those observed in the data?" By contrast, coherence asks: "If you assume that the established theory is correct (i.e. not counterfactual), would the observed results fit into that theory?" Whereas the consideration of coherence would reject the observed result to be non-causal if it contradicted a predominant theory, plausibility leaves the researcher more room regarding which particular piece of substantive knowledge to evaluate the results against.
8. Experiment
Causation is more likely if evidence is based on randomised experiments
Hill [5] argued that a causal interpretation of an association from a non-experimental study was supported if a randomised prevention derived from the association confirmed the finding. For instance, after finding that certain events were related to the number of people smoking, one might forbid smoking to see whether the frequency of the events decrease consecutively.
To Rothman and Greenland [[7], p. 27], it has not been clear whether Hill meant evidence from animal or human experiments. Human experiments were hardly available in epidemiology, and results from animal experiments could not easily be applied to human beings. To Susser [11], Hill's examples suggested that he meant intervention and active change rather than research design. Both Susser [11] and Rothman and Greenland [[7], p. 27] stated that results from randomised experiments provided stronger evidence than results based on other study designs, but always had several possible explanations. Cox and Wermuth [[28], p. 225f.] relaxed that criterion by replacing the qualitative difference between experimental and non-experimental studies with the rather quantitative conception of "strength of intervention": an observed difference would be more likely to be causal if it followed a massive intervention. This is motivated by the possibility that a change following a modest intervention could result from the circumstances of a treatment rather than from the treatment itself. One might add that the Cox and Wermuth consideration requires a modest intervention to be precluded from having a strong influence – an assumption that is certainly context-dependent to a high degree.
In terms of counterfactual causality, the distinction between massive and modest interventions is irrelevant, because a causal effect is only defined for a fixed index and a fixed reference condition. Hence, if interpreted in terms of strength of intervention, this is again not a consideration on a specific causal difference, but rather a consideration on a comprehensive causal theory (as the one on biological gradient). Such a theory is required in order to decide what is a modest and what is a strong intervention.
If the consideration on experiment is interpreted in terms of avoiding some biases in estimating a specific causal effect by conducting an RCT, it should be generalised as follows: observed associations should equal the true counterfactual difference as closely as possible (despite random error). Bias is reduced either by using a study design that avoids major biases or by properly correcting for bias. Clearly, avoiding bias is preferable to correcting for it, but it is often impossible to avoid some biases. As already mentioned, in RCTs with perfect compliance, confounding cannot occur (although confounders might be distributed unequally by chance) and there is no measurement error in the exposure. However, bias due to measurement error could still occur in the outcome, and there may be bias due to selection, missing data, etc. [22]. Thus, Hill's original formulation [5] covered only one or two among a variety of possible biases.
Instead, two more general question arise: which study design is likely to validly identify a presumed causal effect? And, if the optimal study design is not possible, how can bias be accurately corrected for? As in the consideration on strength, this can be summarised by: which results would be expected to be observed if the data were free of any bias? A causal effect is more likely if, after bias adjustment, the interval estimate excludes the null value, and it is even more likely if the lower boundary is far from the null value. If adjustment is done properly, systematic error in the corrected interval estimate decreases if the knowledge about biases increases. As a consequence, one can hardly ever demonstrate a causal effect if biases are poorly understood – this is the case even in large samples, because the associated systematic error in the results would remain even while random error decreases.
9. Analogy
For analogous exposures and outcomes an effect has already been shown
Hill [5] wrote that it would be sometimes acceptable to "judge by analogy". He gives the following example:
"With the effects of thalamoide and rubella before us we would surely be ready to accept slighter but similar evidence with another drug or another viral disease in pregnancy."
Susser [11] interpreted Hill as someone claiming that "when one class of causal agents is known to have produced an effect, the standards for evidence that another agent of that class produces a similar effect can be reduced." Rothman and Greenland [[18], p. 27] retorted:
"Whatever insight might be derived from analogy is handicapped by the inventive imagination of scientists who can find analogies everywhere. At best, analogy provides a source of more elaborate hypothesis about the associations under study; absence of such analogies only reflects lack of imagination or lack of evidence."
When applying the consideration on analogy scientists should ask themselves: would you expect the same association if you used settings analogous to those in other studies (while taking biases into account, which may differ across the studies)? The term "analogous" suggests that the entities in external studies are only similar to those in the observed data (but not identical). This requires an additional modelling of the counterfactual effects of using analogous but not identical entities in different studies. This makes the application of the analogy consideration even more uncertain than the application of considerations on plausibility and coherence.
Conclusion
Hill himself used the terms "viewpoints" and "features to be considered" when evaluating an association. His aim was to unravel the question: "What aspects of this association should we especially consider before deciding that the most likely interpretation of it is causation?"[5] He expressed his ambivalence about the usefulness of his own considerations as follows:
"None of my nine viewpoints can bring indisputable evidence for or against the cause-and-effect hypothesis..."
Rothman and Greenland's conclusion [36] suggests that there are no causal criteria at all in epidemiology :
"Causal inference in epidemiology is better viewed as an exercise in measurement of an effect rather than as criterion-guided process for deciding whether an effect is present or not."
The usefulness of the counterfactual approximation of Hill's considerations is that their heuristic value can be assessed by answering counterfactual questions. I have argued that the application of seven of the nine considerations (consistency, specificity, temporality, biological gradient, plausibility, coherence and analogy) involves comprehensive causal theories. Complex causal systems comprise many counterfactuals and assumptions about biases. If complexity becomes very large, the uncertainty regarding whether or not to apply a given consideration can be expected to approach a decision made by coin toss. Thus, with increasing complexity, the heuristic value of Hill's considerations diminishes.
Here, an original argument of Hill [5] becomes of particularly important: the required amount of evidence for a causal effect should depend on the possible consequences of interventions derived from causal conclusions. If a causal conclusion needed an action that brought about more harm if wrongly taken than benefit if rightly taken, a correspondingly high amount of evidence would be required. If the relationship between benefit and harm were converse, less evidence would be necessary. The major tool to assess the applicability of these considerations is multiple bias modelling. Multiple bias models should be much more frequently used. Moreover, the decision as to whether or not to apply one of these considerations is always implicitly based on one or several multiple bias models. For instance, demanding an association of at least a certain magnitude is logically equivalent to the "true bias model" being part of the set of multiple bias models in which priors on bias parameters would require at least this magnitude of association to be observed.
One may ask the counterfactual question of how epidemiology and medical research would have developped if Hill had been more explicit in recommending when to apply each of his considerations. I am far from claiming to be able to answer this question, but I consider my speculation being worth mentioning. In their paper entitled "The missed lessons of Sir Austin Bradford Hill" [6] Phillips and Goodman reviewed malpractices denounced by Hill that were still being made later in practice: over-emphasis of statistical tests, systematic error being under-estimated and cost/benefit trade-offs being disregarded in intervention decisions. Hill's considerations were misused as "causal criteria", and they were taught more often than more sound causal conceptions [6]. There is no reason to believe that more explicit recommendations on when to apply his considerations would have been better heeded; the cautionary notes that Hill actually made were largely ignored.
My own experience is that scientific recommendations are widely followed if they provide easy guidance; recommendations that call for complex action are frequently ignored. My guess is that this is due to many researchers' desire for simple and globally applicable answers. This desire leads to misinterpretation of scientific texts and to taking individual statements out of their context. More pessimistically, the question of which guidance is followed depends on which guidelines are in line with the desired answer. Therefore, it seems likely that, even if Hill's paper had not been published, scientists' desire for simple answers would have caused another paper to be written or to be misinterpreted in the same way as happened with Hill's [5] article.
List of abbreviations
MCSA: Monte Carlo sensitivity analyses
RCT: randomised and controlled trial
Competing interests
The author(s) declare that they have no competing interests.
Acknowledgements
I wish to thank Evelyn Alvarenga for language editing.
==== Refs
Armitage P Before and after Bradford Hill: Some trends in medical statistics J Roy Stat Soc A 1995 158 143 153
Doll R Sir Austin Bradford Hill: A personal view of his contribution to epidemiology J Roy Stat Soc A 1995 158 155 163
Armitage P Fisher, Bradford Hill, and randomisation Int J Epidemiol 2003 32 925 928 14681247 10.1093/ije/dyg286
Chalmers I Fisher and Bradford Hill: theory and pragmatism? Int J Epidemiol 2003 32 922 924 14681246 10.1093/ije/dyg295
Hill AB The environment and disease: Association or causation? Proceed Roy Soc Medicine – London 1965 58 295 300
Phillips CV Goodman KJ The missed lessons of Sir Austin Bradford Hill Epidemiol Perspect & Innov 1965 1 3 10.1186/1742-5573-1-3
Goodman KJ Phillips CV The Hill criteria of causation Encyclopedia of Statistics in Behavioral Sciences 2005 London: Wiley
Hussain SP Harris CC Molecular epidemiology of human cancer Toxicol Lett 1998 103 219 225 10022257 10.1016/S0378-4274(98)00310-5
Spitzer WO Bias versus causality: Interpreting recent evidence of oral contraceptive studies Am J Obstet Gynecol 1998 179 S43 S50 9753310 10.1053/ob.1998.v179.a93059
Naschitz JE Kovaleva J Shaviv N Rennert G Yeshurun D Vascular disorders preceding diagnosis of cancer: distingushing the causal relationship based on the Bradford-Hill guidelines Angiology 2003 54 11 17 12593491
Susser M What is a cause and how do we know one? A grammar for pragmatic epidemiology Amer J Epidemiol 1991 7 635 648 2018019
Morabia A On the origin of Hill's causal criteria Epidemiol 1991 5 367 369
Little RJA Rubin DB Causal effects in clinical and epidemiological studies via potential outcomes: Concepts and analytical approaches Annu Rev Public Health 2000 21 121 145 10884949 10.1146/annurev.publhealth.21.1.121
Höfler M Causal inference based on counterfactuals BMC Med Res Methodol 2005 5 18 15904499 10.1186/1471-2288-5-28
Holland PW Statistics and causal inference J Amer Stat Ass 1986 81 945 962
Rubin DB Estimating causal effects of treatments in randomised and nonrandomised studies J Educ Psychol 1974 66 688 701
Rosenbaum PR Rubin DB The central role of the propensity score in observational studies for causal inference Biometrika 1983 70 41 55
Rothman KJ Greenland S eds Modern Epidemiology 1998 2 Philadelphia: Lippincott Williams & Wilkins
Pearl J Causality – Models, reasoning and inference 2002 Cambridge: Cambridge University Press
Maldonado G Greenland S Estimating causal effects Int J Epidemiol 2002 31 422 429 11980807 10.1093/ije/31.2.422
Rothman KJ Causes Amer J Epidemiol 1976 104 587 592 998606
Lewis D Causation. Journal of Philosophy 1973 70 556 567 4684392
Maclure M Schneeweiβ S Causation of bias: the episcope Epidemiology 2001 12 114 122 11138805 10.1097/00001648-200101000-00019
Kraemer HC Kazdin AE Offord DE Kessler RC Jensen PS Kupfer DJ Measuring the potency of risk factors for clinical or policy significance Psychol Meth 1999 4 257 271 10.1037/1082-989X.4.3.257
Rothman KJ Poole C A strengthening programme for weak associations Int J Epidemiol 1988 17 955 959 3225112
Rosenbaum PR Observational Studies 2002 2 New York: Springer
Greenland S Multiple-bias modelling for analysis of observational data J Roy Stat Soc 2005 168 267 291 10.1111/j.1467-985X.2004.00349.x
Cox DR Wermuth N Multivariate Dependencies Models, Analyses and Interpretation 1996 London: Chapman and Hall
Greenland S Basic problems in interaction assessment Envir Health Perspect 1993 101 59 66
Weiss N Can the "specificity" of an association be rehabilitated as a basis for supporting a cuasal hypothesis? Epidemiology 2002 13 6 8 11805580 10.1097/00001648-200201000-00003
Greenland S Pearl J Robins JM Causal diagrams for epidemiological research Epidemiology 1999 10 37 48 9888278 10.1097/00001648-199901000-00005
Mineka S Watson D Clark LA Comorbodity of anxiety and unipolar mood disorders Annu Rev Psychol 1998 49 377 412 9496627 10.1146/annurev.psych.49.1.377
O'Sullivan S Cormican L Murphy M Poulter LW Conor MB Effects of varying doses of flutocasone propionate on the physiology and bronchial wall immunopathology in mild-to-moderate asthma Chest 2002 122 1966 1972 12475834 10.1378/chest.122.6.1966
Melmon KL Morelli HF Hoffman BB Niederenberg DW Eds Clinical pharmacology: Basic principles in therapeutics 1992 New York: McGraw-Hill
Lanes SF Poole C Truth in packaging' The unwrapping of epidemiological research J Occupat Medic 1984 8 571 574
Rothman KJ Greenland S Causation and causal inference in epidemiology Am J Publ Health 2005 95 S144 150 10.2105/AJPH.2004.059204
|
16269083
|
PMC1291382
|
CC BY
|
2021-01-04 16:38:27
|
no
|
Emerg Themes Epidemiol. 2005 Nov 3; 2:11
|
utf-8
|
Emerg Themes Epidemiol
| 2,005 |
10.1186/1742-7622-2-11
|
oa_comm
|
==== Front
Filaria JFilaria Journal1475-2883BioMed Central London 1475-2883-4-111627447410.1186/1475-2883-4-11ResearchA flow cytometry based method for studying embryogenesis and immune reactivity to embryogenic stages in filarial parasites Sahu Bikash Ranjan [email protected] Alok Das [email protected] Arindam [email protected] Pradip K [email protected] Balachandran [email protected] Division of Immunology, Regional Medical Research Centre, Indian Council of Medical Research, Chandrasekarpur, Bhubaneswar, 751023, India2 Division of Microbiology, National Institute of Cholera and Enteric Diseases, Kolkata, 70000, India2005 7 11 2005 4 11 11 18 4 2005 7 11 2005 Copyright © 2005 Sahu et al; licensee BioMed Central Ltd.2005Sahu et al; licensee BioMed Central Ltd.This is an Open Access article distributed under the terms of the Creative Commons Attribution License (), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
Background
In the absence of intermediate animal hosts, the process of embryogenesis leading to fecundity of adult female filarial worms is very critical for persistence of these obligate parasites in human communities. Embryogenesis in adult female filarial parasites involves fertilization of eggs or oocytes by sperms and their subsequent development into motile microfilariae inside the uterine cavity of worms. Development of assays for monitoring embryogenesis in adult female worms is a critical requirement in filariasis research – filarial worms are known to harbour endosymbionts such as Wolbachia which play a significant role in fecundity. Tetracycline or doxycycline treatment of the infected hosts effectively eliminates the endosymbionts resulting in inhibition of embryogenesis in female worms. Currently, inhibition of embryogenesis in adult filarial worms can be monitored only by microscopic examination of in vitro harvested intrauterine stages.
Methods
Adult female filarial worms of bovine filarial parasites, Setaria digitata were collected from the peritoneum of infected animals and intrauterine stages were harvested in culture medium and were analyzed for forward and side scatter by flowcytometry using a BD FACS Calibur. Different populations were gated, sorted and identified by phase microscopy. Binding of biotinylated lectins to intra uterine stages was monitored using FITC labeled Avidin and monitored by flow cytometry of gated populations. Similarly, binding of antibodies in human filarial sera to intrauterine stages was monitored using FITC labeled anti-human immunoglobulins.
Results
The forward and side scatter for intrauterine stages delineated 3 distinct populations labeled as R1, R2 and R3. The three populations were sorted and identified to be a) fully stretched microfilariae, b) early and c) late developmental stages of eggs respectively. Lectins such as Wheat Germ agglutinin or Concanavalin-A were found to bind strongly to egg stages and less prominently to intra-uterine microfilariae. Similarly the binding of antibodies in filarial sera to the three intra-uterine stages could also be precisely quantified.
Conclusion
The manuscript reports a novel flow cytometry based method to monitor progression of embryogenesis in adult filarial worms. Apart from relative quantification of different intra uterine developmental stages, the assay allows quantitative binding of lectins and antibodies to each of the intrauterine stages. It may now be possible to quantify levels of antibodies in infected and immune hosts to monitor anti-fecundity immunity in filariasis – the assay can thus be used as a powerful tool for drug development and in immunological studies in human and experimental filariasis.
==== Body
Background
Lymphatic filariasis causes debilitating chronic hydrocele and/or lymphoedema in about 40 million people world wide – nearly 120 million people are found infected with the nematodes, about 90% with W.bancrofti and the rest with B.malayi, mostly in tropical countries. Infective larvae (L3) from mosquitoes enter the mammalian host and develop into male and female adult stage parasites in the lymphatics. After mating the adult female worms release thousands of microfilariae (Mf) that enter the blood circulation for further development in mosquitoes. In the absence of intermediate animal hosts, the process of embryogenesis leading to fecundity of adult female worms is very critical for persistence of these obligate parasites in human communities.
Morphologically, different intrauterine developmental stages are discernable in the uterine cavity of adult female worms. Eggs or oocytes after fertilization with sperms transform into motile microfilariae and are released by the adult female worms [1]. Currently, tools are not available to quantify the different developmental stages of embryogenesis other than approximate scoring by microscopy [2,3]. Development of precise assays for monitoring embryogenesis in adult female worms have the potential to address crucial issues in filariasis research – filarial worms are known to harbour endosymbionts such as Wolbachia, which play a significant role in fecundity of adult filarial worms [3,4]. Tetracycline or doxycycline treatment of the infected hosts effectively eliminates the endosymbionts resulting in inhibition of embryogenesis in female worms [5]. Inhibition of embryogenesis in infected human hosts can be scored only by monitoring decrease/loss of peripheral microfilaraemia-lymphatic dwelling adult stage parasites are not accessible for study. However in experimental animal models the adult female worms can be harvested and dissected in vitro and the intrauterine stages can be approximately scored by microscopy [2,3]. In this communication we describe a flow cytometry based method for studying embryogenesis in adult female filarial worms. The utility of this method for quantifying binding of lectins and antibodies to different intra uterine stages of filarial parasites has also been evaluated.
Methods
Preparation of intra-uterine stages for flow cytometry
Adult female filarial worms, Setaria digitata were collected from the peritoneum of cattle at a nearby abattoir in sterile alpha – MEM containing 1% glucose, Penicillin-100 units/ml, Streptomycin-100 μg/ml Gentamycin-50 μg/ml and Amphotericin-B 2.5 μg/ml and transported to the laboratory and used on the same day. Individual worms taken in a petridish were washed three times (x3) in sterile medium and dissected into small pieces in about 5 ml of medium and incubated at 37°C for 30 mins. The large pieces were removed and the medium containing intra-uterine stages were harvested, washed in MEM and the final pellet of cells suspended in 1 ml of sheath fluid and analyzed using a flowcytometer (FACSCalibur, Becton Dickinson, USA) using Cell Quest Pro software. The data (5000 events) were acquired for forward and side scatter using the following settings: FSC, voltage E00 and SSC, voltage 340. The three populations were gated and sorted using a FACS sorter under moderate fidelity settings. The sorted suspensions were centrifuged on to microscopic slides using a cytospin and observed under a phase contrast microscope for identification of organisms in the gated populations.
Collection of blood for sera from human filariasis cases
Blood for sera were collected from patients with chronic filariasis (elephantiasis/hydrocoele) and microfilariae carriers from filariasis endemic areas near Bhubaneswar as described by us earlier [6]. Sera stored at -20°C were diluted in PBS with 1% BSA and used for binding to intra-uterine stages as described above.
Preparation of immune sera against intrauterine stages
Five Mastomys coucha were immunized with three doses (15 days apart) of intra-uterine stages (50,000 cells per dose) in complete Freund's adjuvant and blood for sera was collected between days 40–45 and tested for antibody reactivity to intrauterine stages as described above.
Purification of circulating microfilariae from infected blood
Microfilariae (Mf) from human blood were purified using 5 μM pore polycarbonate membranes as described by us earlier [6]. Nocturnal blood samples collected from W.bancrofti infected Mf carriers or day time blood samples from S.digitata infected cattle were used for purification – Mf washed in Tris-EDTA buffer were taken in sheath fluid and analyzed by flow cytometry for forward and side scatter as described above.
Lectin binding assay
A suspension (500 μl) containing about 10,000 intra uterine stages in PBS 7.2 with 1% BSA were mixed with 500 μl of 4 μg/ml or 1 μg/ml of biotinylated WGA (L-5142, Sigma Chemical Co, USA). Simlarly two dilutions, 500 or 2000 fold diluted biotinylated Con-A (BA-16, Bangalore Genie, India) were used and the suspensions were incubated at RT for 45 min. The cells were then washed x3 in PBS and taken in 0.5 ml of PBS with 1% BSA to which 0.5 ml of 250 fold diluted Avidin – FITC (S-3762, Sigma Chemical Co, USA) was added and incubated for 30 min at RT. The suspension was washed again thrice in PBS and samples taken in 1 ml of sheath fluid and 5000 events were acquired. The three populations were gated and fluorescence intensity was read using a 488 nm laser. The percentage reactivity of lectins in comparison to Avidin-FITC controls was calculated using CellQuest Pro software.
Antibody binding assay
A suspension (500 μl) containing about 10,000 IU stages in PBS-BSA were mixed with 500 μl of 50 fold diluted human or Mastomys sera (normal as well as immunized animals) and incubated at RT for 45 min. The cells were then washed x3 and taken in 0.5 ml of PBS-BSA to which 0.5 ml of 250 fold diluted FITC labeled anti-human Immunoglobulin (F-6506, Sigma Chemical Co.USA) or 50 fold diluted FITC labeled anti-mouse IgG was added and incubated for 30 min at RT. The suspension was washed x3 in PBS and organisms were taken in 1 ml of sheath fluid for acquiring 5000 events in FACS Calibur. The three populations were gated and fluorescence intensity was read using a 488 nm laser. The percentage reactivity as well as mean fluorescence intensity were calculated for antibody binding and compared with FITC labeled second antibody conjugate controls using CellQuest Pro software.
Results
Forward and Side scatter and identity of stages
Figs 1a &1c show the dot plots of intrauterine stages in forward and side scatter for two representative adult worms; respective contour maps are shown in Fig 1b &1d. Three distinct populations of organisms, R1, R2, and R3 could be identified. The three populations were consistently found in several worms although the relative numbers in each scatter varied between worms as shown in Fig 1. Purified mf of S.digitata and W.bancrofti (from blood) by using nucleopore membrane filtration scattered exclusively in the R1 region indicating that organisms in this region are microfilariae (Fig 2a and 2b respectively). The organisms in the three scatter groups were purified by sorting them in a FACS Calibur sorter and examined under a phase contrast microscope: Figs 10 &11 show organisms in the unsorted population which contain a mixture of early and late developmental stages of eggs as well as fully stretched Mf, while Fig 12 and 13 reveal pure Mf of S.digitata in the sorted R1 population confirming the data presented in fig 2a and 2b above. Organisms in R2 gate were found to be early developmental stages of eggs (Fig 8) while late developmental stages were found in the R3 population (Fig 9).
Figure 1 Three populations R1, R2 and R3 in forward (FSC, voltage E00) and side scatter (SSC, voltage 340) of intra-uterine stages; two representative worms by Flow cytometry are shown: 1a and 1c – Dot plots; 1b and 1d – respective contour maps; the number of events in individual gates for IU stages are indicated below respective figures.
Figure 2 Dot plots for purified Mf of S.digitata (2a) and W.bancrofti (2b) in forward and side scatter.
Figure 8 Sorted R2 population showing early developmental stages of eggs.
Figure 9 Sorted R3 cells of late developmental stages of eggs.
Figure 10 Phase contrast microscopic image of unsorted population of intrauterine stages.
Figure 11 Phase contrast microscopic image of unsorted population of intrauterine stages.
Figure 12 Sorted R1 cells showing pure Mf.
Figure 13 Sorted R1 cells showing pure Mf.
Lectin binding to intra-uterine stages
Two lectins, Wheat Germ Agglutinin (WGA) and Conconavalin-A (Con-A) that specifically react primarily with N-Acetyl-D-Glucosamine and D-mannose residues respectively (which are present ubiquitously in several parasites including filarial parasites), were chosen as markers in this study to evaluate the flow cytometry based assay procedure. The intrauterine stages incubated with biotinylated WGA or Con-A were analysed for single colour fluorescence using 488 nm laser in the three gated population. The background reactivity and mean fluorescence reactivity of avidin-FITC (controls) for the three gated populations was minimal. The specific binding of WGA to the three populations is shown in Fig 7a–f. There was a dose dependant binding of WGA to intra-uterine stages – lower concentrations Fig 7d,e &7f) bound proportionately less than higher concentrations as shown in histograms in Fig (7a,b &7c). Similarly a dose dependent binding of Con-A to intrauterine stages could also be demonstrated (Fig 3a–f). Both the lectins bound significantly (> 95%) to early and late developmental stages of eggs (R2 and R3 respectively) – their reactivity to intra-uterine Mf (R1) was, however, not high.
Figure 3 Con-A binding to intra-uterine stages: single colour analysis using 488 nm laser; 3 a,b and c: Con-A (500 fold diluted) reactivity to R1,R2 and R3 gated populations. 3 d, e and f: Con-A (2000 fold diluted) reactivity to R1, R2 and R3 gated populations. Green line: Avidin-FITC control; coloured shaded areas: specific reactivity of Con-A. Numbers shown on top left and top right on histograms represent geometric mean intensity of fluorescence for control and Con-A respectively.
Figure 7 Binding of WGA to intra-uterine stages: single parameter analysis using 488 nm laser; 7 a,b and c: WGA (2 μg/ml) reactivity to R1,R2 and R3 gated populations. 7 d,e and f: WGA (0.5 μg/ml) reactivity to R1,R2 and R3 gated populations. Green line: Avidin-FITC control; coloured shaded areas show specific reactivity of WGA. Numbers shown on top left and top right on histograms represent geometric mean intensity of fluorescence for control and WGA respectively
Antibody binding to Intra-uterine stages
The binding of antibodies in human Bancroftian filariasis and in S.digitata immune Mastomys sera to intrauterine developmental stages could also be studied. The results of single colour fluorescence using 488 nm laser in the three-gated populations in two Mastomys sera are shown in Fig 4. The background binding of anti-mouse IgG-FITC conjugate to the three gated populations R1, R2, and R3 was very minimal. The binding profile of pre-immune sera were similar to conjugate controls while significant binding of antibodies to intrauterine eggs (R2 and R3 populations) could be demonstrated in both the immune sera (Fig 4a–f), in addition, similar reactivity was shown in three other immune Mastomys sera also (data not shown). Binding of antibodies in human filariasis sera to intrauterine stages could also be shown by the assay. The assay for two sera of elephantiasis cases is shown in Fig 5a–f. Significant binding of antibodies to intrauterine stages, particularly to egg stages could be demonstrated.
Figure 4 Antibodies binding in sera of M.coucha to three gated populations (R1,R2 and R3) of intra-uterine stages. Fig 4 a,b and c for one animal and Fig 4 d,e and f for a second animal are shown in the histogram. Blue lines: Anti-mouse IgG-FITC conjugate control; red lines: Pre-immune sera; Coloured shaded areas: antibody reactivity in respectiveimmunized sera. Numbers shown on top left and top right on histograms represent geometric mean intensity of fluorescence for pre and post-immunized sera respectively.
Figure 5 Antibody binding in two patients with chronic filariasis: Histograms showing reactivity to the three gated populations (R1,R2 and R3) of intra-uterine stages. Fig 5 a,b and c for one human serum and Fig 5 d,e and f for another patient. Blue line: Anti-human IgG-FITC control; Colour shaded areas: Antibody reactivity in test sera; Numbers shown on top left and top right on histograms represent geometric mean intensity of fluorescence for control and test sample respectively.
Although antibodies in human sera did not bind well to intra-uterine Mf, very significant binding of antibodies to a small select population of Mf was demonstrable by the assay (Fig 5a &5d). The mean antibody levels in nine patients with chronic filariasis were quantified (Fig 6). The levels were expressed as percentage reactivity and mean geometric fluorescence intensity as shown in Fig 6a &6b respectively. The antibody and lectin binding reported in this study are primarily restricted to reactivity of these molecules to the surface of intra-uterine stages since fresh live stages collected from adult female worms were used for the assays. Fixation with formaldehyde and permeabilising the cells resulted in antibody binding to intracellular components also (data not shown).
Figure 6 Antibody binding in chronic filariasis cases; Fig 6 (a): Mean percentage reactivity of antibodies in nine subjects to different intra-uterine stages. Fig 6 (b): Mean Geometric fluorescent intensity for the same sera depicted in 6a. Striped bars: Anti-human IgG-FITC conjugate control; Open bars: test samples.
Discussion
Embryogenesis is central to the biology of helminths and could be an effective target for devising intervention strategies for blocking transmission of parasitic helminths from mammalian hosts. Immunological studies in human as well as experimental filariasis have been largely directed towards host responses against infective larvae, microfilariae and adult stage parasites [7,8]. However, very few studies have addressed antigens expressed in/on intrauterine developing stage [9]. We consider the non-availability of sensitive assays to undertake studies on antibody responses to intrauterine stages as one of the reasons for such a limited number of studies. The large size of intrauterine stages (about 50 μM for egg stages and about 160 μM for Mf) do not limit the applicability of this novel assay since no special modification in the flowcytometer is required, the dimensions of standard flowcell provided by the manufacturers in FACSCalibur model was sufficient to perform the assays. We envisage a wide application for this new flowcytometry-based method of monitoring embryogenesis as well as antibody binding to intra-uterine stages in filariasis research: a) removal of endosymbionts in human hosts by doxycycline requires administration of the drug over a period of several weeks (5) and there is a need for developing newer anti-rickettsial drugs for blocking embryogenesis/fecundity in adult filarial worms; the new method described in this communication can be effectively used for screening a large number of potential compounds/drugs; b) although doxycycline and tetracycline have been demonstrated to eliminate endosymbionts in adult filarial worms, the precise intrauterine stage at which embryogenesis is blocked is not known and can now be studied with this new assay; c) cytokines such as IL-4 and molecules of innate immunity such as TLR-4 have been demonstrated to play a role in regulating production of microfilaraemia in experimental hosts [4,10]. Adult filarial worms can now be implanted in mice made deficient for specific cytokine gene expression or transgenic for such host molecules to monitor their role in embryogenesis; d) several filarial antigens have been cloned, sequenced and expressed in recent years and some have been used as putative vaccine candidates in experimental models [11]; e) high reactivity of antibodies in human filariasis sera to a select population of Mf (Fig 5a &5b) could be due to polymorphic antigens expressed on Mf sheath described by us several years ago (12). The novel procedure described here could assist in selectively sorting such reactive population of Mf for genetic analysis of polymorphic filarial antigens. The effect of induction of immune response to such recombinant molecules on embryogenesis can be studied using the assay described in this communication; e) filariasis in human communities presents with a variety of clinical and parasitological features. Infection is characterized by presence of circulating Mf and/or antigenemia and diseased subjects display acute and/or chronic clinical manifestations. Studies on immune responses of these patients to different developmental stages viz., infective larvae, adult stage parasite and Mf have indicated significant differences between the groups [summarized in [7]]. The new assay reported in this communication can be expected to now allow quantification of antibody responses to different intra-uterine stages of filarial parasites in human filariasis. The principles of flow cytometry have been recently used for separation and monitoring death/survival of another nematode C.elegans [13]. It may now be possible to study programmed cell death of developing intra-uterine stages in filarial parasites.
A bovine filarial parasite S.digitata has been used for establishment of the new assay system in this study. The assay could also be used for microfilariae of W.bancrofti (purified from infected endemic subjects). It is however, essential to evaluate the utility of this flowcytometry based method for other more commonly used filarial parasites, viz., Brugia malayi, Brugia pahangi, Dirofilaria immitis, Litomosoides sigmodontis etc.
Conclusion
The manuscript reports a novel flow cytometry based method to monitor progression of embryogenesis in adult filarial worms. Apart from relative quantification of different intra uterine developmental stages, the assay allows quantitative binding of lectins and antibodies to each of the intrauterine stages. It may now be possible to monitor levels of antibodies in infected as well as immune hosts to intra-uterine developmental stages that could become a parameter for monitoring anti-fecundity immunity in filariasis, the assay can therefore be used as a powerful tool for drug development and in immunological studies in human and experimental filariasis.
Competing interests
The author(s) declare that they have no competing interests.
Authors' contributions
BRS: Conducted significant part of the experimental work including the assays using flowcytometer, compiled the data and wrote the draft manuscript.
ADM: Conducted part of the experimental work including the assays using flowcytometer, performed the immunoassays.
AM: Purified the stages using the sorter of flowcytometer, performed cytospin and captured phase contrast images for the sorted population.
PKD: Participated in experiments involving the use the flowcytometer sorter and use of confocal/ phase contrast microscopy
BR: Conceived the idea, designed experiments, interpreted the data and completed the manuscript
Acknowledgements
The Regional Medical Research Centre, Bhubaneswar is supported by grants from the Indian Council of Medical Research, New Delhi. The authors thank Prof. Rick Maizels, University of Edinburgh, UK for critically reading the manuscript. BRS is supported by an Indo-German grant of ICMR and ADM is supported with a Junior Research Fellowship from University Grants Commission, New Delhi. The authors thank the Director of the center, Dr.S.K.Kar for sustained support.
==== Refs
Scott AL Nutman TB Lymphatic – dwelling filariae Lymphatic filariasis 2000 Imperial College Press, London 5 39
Pfarr KM Fischer K Hoerauf A Involvement of Toll like receptor 4 in the embryogenesis of the rodent filaria Litomosoides sigmodontis Med Microbiol Immunol 2003 192 53 56 12592564
Chirgwin SR Coleman SU Porthouse KH Nowling JM Punkosdy GA Klei TR Removal of Wolbachia from Brugia pahangi is closely linked to worm death and fecundity but does not result in altered lymphatic lesion formation in Mongolian Gerbils(Meriones unguiculatus) Infection and Immunity 2003 71 6986 6994 14638788 10.1128/IAI.71.12.6986-6994.2003
Taylor MJ Hoerauf A Wolbachia bacteria of filarial nematodes Parasitol Today 1999 15 437 442 10511685 10.1016/S0169-4758(99)01533-1
Hoerauf A Mand S Adjei O Fleischer B Buttner DW Depletion of Wolbachia endobacteria in Onchocerca volvulus by doxycycline and microfilaridermia after ivermectin treatment The Lancet 2001 357 1415 1416 11356444 10.1016/S0140-6736(00)04581-5
Sahoo PK Babu Geddam JJ Satapathy AK Mohanty MC Ravindran B Bancroftian filariasis: prevalence of antigenaemia and endemic normals in Orissa, India Transactions of the Royal Society of Tropical Medicine and Hygiene 2000 94 515 517 11132379 10.1016/S0035-9203(00)90070-1
Ravindran B Satapathy AK Sahoo PK Mohanty MC Protective immunity in human lymphatic filariasis: problems and prospects Medical Microbiology and Immunology 2003 192 41 468 12592562
Lawrence RA Immunity to filarial nematodes Vet Parasitol 2001 12 33 44 11522404 10.1016/S0304-4017(01)00481-2
Triteeraprapab S Richie TL Tuan RS Shepley KJ Dinman JD Neubert TA Scott AL Molecular cloning of a gene expressed during early embryonic development in Onchocerca volvulus Mol Biochem Parasitol 1995 69 161 71 7770081 10.1016/0166-6851(94)00187-R
Devaney E Gillan V Wheatley I Jenson J O'Conner R Balmer P Interleukin-4 influences the production of microfilariae in a mouse model of Brugia infection Parasite Immunol 2002 24 29 37 11856444 10.1046/j.0141-9838.2001.00433.x
Maizels RM Blaxter M Scott AL Immunological genomics of Brugia malayi: filarial genes implicated in immune evasion and protective immunity Parasite Immunology 2001 23 327 344 11472553 10.1046/j.1365-3024.2001.00397.x
Ravindran B Satapathy AK Sahoo PK Bancroftian filariasis-Differential reactivity of anti-sheath antibodies in microfilariae carriers Parasite Immunology 1994 16 321 323 7970869
Gill MS Olsen A Sampayo JN Lithgow GJ An automated high-thoroughput assay for survival of the nematode C.elegans Free Radical Biology and Medicine 2003 35 558 565 12957648 10.1016/S0891-5849(03)00328-9
|
16274474
|
PMC1291383
|
CC BY
|
2021-01-04 16:38:24
|
no
|
Filaria J. 2005 Nov 7; 4:11
|
utf-8
|
Filaria J
| 2,005 |
10.1186/1475-2883-4-11
|
oa_comm
|
==== Front
Harm Reduct JHarm Reduction Journal1477-7517BioMed Central London 1477-7517-2-221624889310.1186/1477-7517-2-22ReviewEffectiveness of harm reduction programmes for injecting drug users in Dhaka city Azim Tasnim [email protected] Najmul [email protected] Robert [email protected] ICDDR, B: Centre for Health and Population Research, Bangladesh, Mohakhali, Dhaka 1212, Bangladesh2 CARE, Bangladesh, Dhaka Field Office, 49/1, Babar Rd, Mohammedpur, Dhaka 1207, Bangladesh3 Family Health International, Bangladesh Country Office, House #5, Rd. #5, Gulshan-2, Dhaka 1212, Bangladesh2005 25 10 2005 2 22 22 27 11 2004 25 10 2005 Copyright © 2005 Azim et al; licensee BioMed Central Ltd.2005Azim et al; licensee BioMed Central Ltd.This is an Open Access article distributed under the terms of the Creative Commons Attribution License (), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
This paper provides a brief overview of the harm reduction programme for injecting drug users (IDU) of CARE, Bangladesh in Dhaka city and uses data from surveillance and a focussed research study on a cohort of IDU, to evaluate the programme. The harm reduction programme in Dhaka is run by CARE, Bangladesh and includes needle/syringe exchange, awareness raising on HIV/STI, abscess management, condom distribution and advocacy with different groups of people. The needle/syringe exchange programme (NEP) has been in place since 1998, the 2nd Generation Surveillance in Bangladesh is being conducted since 1998, and an in-depth cohort study, started in 2002, is being conducted in two areas of Dhaka city with approximately 500 IDU under CARE's NEP who are being followed bi-annually to assess risk behaviour, incidence of HIV, hepatitis C and syphilis. As the surveillance and the cohort study are both closely associated with the NEP of CARE, Bangladesh, these data can be used to monitor the NEP.
==== Body
Review
Harm Reduction Programme for IDU in Dhaka city
The number of injecting drug users (IDU) in Bangladesh is not known. In a Rapid Situation Assessment conducted in 2002 [1] IDU were found in 20/24 sites studied and although this assessment was not an exercise in enumerating the number of drug users in the country, it was apparent that the maximum number of IDU were in the capital city, Dhaka. CARE, Bangladesh started a needle/syringe exchange programme (NEP) in Dhaka in 1998 covering approximately 880 IDU. This NEP has expanded over the years and in 2004, IDU in 19 districts of Bangladesh were reached by CARE, Bangladesh and at the same time in Dhaka the number of IDU that were being reached had increased to approximately 4400, with 3900 being regular attendees. The NEP services in Dhaka are are coordinated through a central field office and provided through 21 Drop-In Centres (DIC) which are located within or close to communities of IDU and through out reach workers in the field. The services provided include needle/syringe exchange in the field, DIC based clinical services for the management of abscesses and sexually transmitted infections (STIs), drug detoxification camps, condom distribution, education and awareness on the harmful effects of drugs, safe injections, HIV/AIDS, STIs, other blood borne infections. Presently, the average number of needles/syringes given to each IDU in Dhaka is five per week and condoms are being received by 15% of IDU. Of the registered IDU, 55% are married however, there is no specific programme to reach the sex partners of IDU, rather IDU are encouraged to bring their sex partners to the DIC for STI management. In the last year 87 sex partners received STI treatment from the DIC.
In addition to these services, advocacy constitutes a major part of CARE, Bangladesh's intervention programme for IDU. Advocacy is done with members of the community, law enforcing agencies, policy makers from the government, general community leaders, NGOs, and Development Partners as and when required. Special local level advocacy meetings are held at the community level when problems are faced in the field.
In Dhaka, harm reduction services to IDU are provided almost exclusively by CARE, Bangladesh. Other organisations are providing detoxification and some rehabilitation services, and some also provide clinical management for STIs but their clientele are primarily non-injecting drug users.
Data for evaluating the NEP
National Surveillance
Second Generation Surveillance for HIV is being conducted in Bangladesh since 1998 and IDU are one of the population groups sampled in surveillance and they have been accessed from several cities [2-7]. The surveillance system in Bangladesh has the serological and Behavioural Surveillance System (BSS) components, which are conducted in tandem on similar groups, but not the same individuals and the methodologies for sampling for the two systems are different. The serological system sampled IDU exclusively through the NEP of CARE, Bangladesh and most of the IDU sampled in BSS were members of the NEP. Given that the surveillance is so closely linked to the intervention programmes, data from the surveillance system can be used to monitor intervention programmes for IDU, i.e. the NEP of CARE, Bangladesh, as that has the only ongoing harm reduction programme with IDU in Dhaka. Five rounds of surveillance have been completed and as IDU have been sampled from the NEP since the 2nd round (1999–2000), data from the 2nd to the 5th (2003–2004) rounds have been used in this evaluation.
Serological Surveillance
The serological surveillance in Bangladesh accesses IDU through intervention programmes [6] and the reason for this is that blood samples cannot be collected in the field setting and results of syphilis tests and treatment for syphilis can only be provided to the IDU in a clinical setting. For serological surveillance, IDU have been accessed annually through harm reduction NGOs with extensive outreach into the community (i.e. CARE, Bangladesh) since the 2nd round (1999–2000) and since the 3rd round (2000–2001) this has been exclusive source of IDU as previous attempts at accessing IDU through detoxification clinics yielded very few IDU.
Serological surveillance has recorded a rise in HIV prevalence over the rounds in IDU from the NEP in Central Bangladesh with prevalence rising from at 1.4% in 1999–2000 (2nd round) to 4.0% in 2002 (4th round) and remaining at 4% during the 5th round (2003–2004) (p < 0.001) [3,8]. Although in the 5th round the HIV prevalence has remained at 4%, in one particular neighbourhood in the same city, the prevalence was recorded at 8.9% (95% CI: 5.4–14.4%) [7]. These findings show that HIV is gradually rising within the IDU in this city and at present most of the cases are localised to one neighbourhood, which may be the epicentre of the epidemic. High rates of hepatitis C have been recorded in the IDU from the very beginning of surveillance and these rates have remained high throughout (66.5% in 2nd round, 62.3% in the 4th round and 59.2% in the 5th round) [7,8]. High hepatitis C rates suggest risky injection and drug sharing behaviours [9]. In contrast to HIV and hepatitis C, active syphilis rates declined significantly over the rounds (9.3% in the 2nd round to1.2% in the 5th round, p < 0.001) [7] suggesting that interventions for sexual risk are possibly being addressed successfully.
Behavioural Surveillance System
A two-stage cluster sampling was used in BSS [6]. In the first stage a time location systematic random sampling method was employed to select primary sampling units (PSUs) and in the second stage the numbers of respondents to be interviewed from each PSU was calculated. The PSU for IDU were spots where at least three IDU were found in a specific time frame. Through this method of sampling IDU selected included both those who were in and out of intervention programmes although in the 5th round most IDU sampled were members of the NEP of CARE, Bangladesh. The data used for evaluation are from the 4th and 5th rounds of BSS.
During the 4th round of BSS conducted in 2002, 44.5% (95% CI: 34.6–54.8%) of the IDU sampled from Central city A were under the NEP of CARE, Bangladesh and this increased to 88.3% (95% CI: 84.4–91.4%) during the 5th round in 2003–2004 (p < 0.001). Analyses of data from the 4th round of BSS showed that IDU in interventions practice safer behaviours than those out of intervention [10] and fewer IDU within interventions shared needles/syringes and reported STIs than IDU outside of interventions [11] whereas more IDU within interventions who reported STIs in the last year sought treatment [11]. These findings suggested that the NEP might be having some preventive effect. However, although, coverage increased between the 4th and 5th rounds of surveillance, borrowing of used needles/syringes by IDU during the last week also increased from 65.8% (95% CI: 60.1–71.1%) in 2002 to 86% (95% CI: 82.4–89.0%) in 2004 (p < 0.001). Unfortunately BSS cannot provide reasons for changes in behaviours seen but field observations at the time of surveillance suggested that considerable disruption occurred in the field due to "cleaning" drives by the community and law enforcement agencies which drove the IDU underground from their regular shooting spots. Nine such observations were recorded during mapping and interviewing for BSS over a three month period.
IDU are not only at risk of acquiring HIV through injection sharing but also through unprotected sex [12] which can also allow transmission to other non-IDU partners in their sexual network [13-15]. In Bangladesh, BSS data show that IDU in Central city A are sexually active and in the 5th round of BSS, in the last year 34.5% (95% CI: 29.4–40.0%) bought sex from female sex workers, 2.3% (95% CI: 1.2–4.6%) from male sex workers and 35.7% (95% CI: 29.6–42.3%) had regular non-commercial female sex partners. Although the proportion of IDU who bought sex declined between the 4th and 5th rounds of surveillance (p < 0.001) fewer IDU reported using condoms in the last sex act with female sex workers in the 5th round 15.7% (95% CI: 10.3–23.3%) compared to the 4th round (29.3%, 95% CI: 23.5–35.9%) of surveillance (p = 0.005). These findings indicate that sexual behaviour in those IDU involved in commercial sex has become riskier.
As the BSS sampled IDU from the same areas of the city as the serological surveillance, secondary analyses of the 5th round of BSS data was done to assess whether there were behavioural factors associated with the localised epidemic of HIV recorded by serological surveillance in IDU in a neighbourhood of Central city A (unpublished observations). BSS data from IDU were compared between the neighbourhood with the concentrated HIV epidemic (referred to as area 1) and the rest of the city. In brief, such comparisons revealed that IDU from area 1 were from a lower socio-economic status being significantly less educated, having a lower monthly average income and more commonly living on the street (p ≤ 0.001 for all) than IDU from the rest of the city. In area 1, IDU took more injections on average in the last day and in the last week (p < 0.001 for both) compared to the IDU in the rest of the city although no differences were observed in injection sharing behaviour between the two areas. However, among those IDU who shared in the last week and said they had shared with different partners, the mean number of persons that shared injections were higher in area 1 than in the rest of the city (p = 0.026). Moreover, IDU in area 1 had a lower risk perception than those in the rest of the city. Therefore, there appear to be neighbourhood factors that enhance vulnerability of the IDU to HIV. From these data it is not possible to gage what these factors may be but in other countries factors such as neighbourhood socio-economic status have been shown to play a role in fuelling the HIV epidemic [16,17].
Research study on a cohort of IDU
A research study on a cohort of IDU in two areas of Dhaka city within the NEP of CARE, Bangladesh is ongoing. This study has been focussing on the incidence of HIV, hepatitis C and syphilis and risk behaviours of IDU within the NEP. Preliminary findings obtained at the baseline of the cohort study in 2002 showed that there is a localised epidemic (5.9%; 95% CI: 4.2–8.1%) within the cohort of 561 IDU and that in one of the two areas under study the prevalence is 8.0% (95% CI: 5.7–11.1% [18], the same neighbourhood that has recorded an epidemic during the 5th round of surveillance.
Baseline data from the IDU within the cohort showed that 27.7% (95% CI: 24.1–31.5%) and 34.5% (95% CI: 30.7–38.6%) IDU borrowed and lent used needles/syringes, respectively, in the last week despite being in the NEP [18]. Almost half the IDU (48.3%; 95% CI: 44.4–52.6%) obtained their needles/syringes from the NEP as well as drug stores. The reasons for obtaining needle/syringes outside the NEP and for sharing needles/syringes included not having access to outreach workers while injecting (13.2%; 95% CI: 10.0–17.4%), not having needles/syringes with themselves at the time of injection (16%; 95% CI: 12.4–20.4%), not having adequate knowledge about HIV/AIDS (14.2%; 95% CI: 10.8–18.4%), inadequate supply of needle/syringe by the NEP (3.1%; 95% CI: 1.7–5.6%) and a combination of these (33.2%; 95% CI: 28.3–38.5%). In addition, the IDU complained of harassment by law enforcers and the community, poor abscess management, inadequate access to medical treatment and detoxification services. Similar to the BSS data, a significant proportion of IDU in the cohort also reported unsafe sexual behaviour (data not shown).
Response of CARE, Bangladesh to the data
In response to data from surveillance and the cohort study, CARE, Bangladesh increased its coverage of IDU in Dhaka city and in the neighbourhood where the epidemic was recorded and took some specific actions to improve their programme [19]. To increase the likelihood of having access to outreach workers at the time of injection, the number of outreach workers in the field for distributing needles/syringes were increased from 12 in a month in January 2003 to 100 in December 2003; the number of shifts during which needles/syringes are distributed were increased from one to two daily; and needles/syringes exchange in the field was supervised more closely. To reduce the harassment that IDU face from law enforcing agencies, CARE, Bangladesh increased its advocacy efforts with law enforcers, community and policy makers. Since the study started, monthly visits of IDU in the two study DICs of the IDU increased between January and October 2003 from 337 to 1118 in area A and 548 to 1217 in area B. Similarly, the numbers of IDU whose abscesses were managed increased during the same time at those DICs (from 83 to 278 in area A and 127 to 219 in area B). A hospital willing to provide medical services to IDU such as management of severe abscesses, orthopaedic services, surgery, etc was identified to which IDU are now referred.
Through the cohort study 40 HIV positive IDU have been identified and these HIV positive IDU have raised new demands mainly related to their clinical management and maintenance of confidentiality. Counsellors and a physician from ICDDR,B were appointed especially for their needs and CARE, Bangladesh formed a support group of HIV positive IDU and their partners. CARE, Bangladesh also opened a new DIC for the HIV positive IDU, which is also open for any IDU wishing to attend so as not to discriminate between the HIV positive and negative IDU. Through this DIC medical care and counselling has become much easier and more regular. So far there has been no breach in confidentiality.
Conclusion
At present the HIV prevalence is still low in IDU but and although in some aspects sex practices may have become safer, injection sharing behaviour has become riskier. Given the high levels of continued risk behaviour and hepatitis C prevalence, the frequent disturbances in the field interfering with the programme activities and that one area of Dhaka city is already experiencing a localised epidemic of HIV, if all out efforts from all quarters are not made, the epidemic is likely to spread in the near future. Available data suggest that the NEP in Dhaka is responsive to the data generated but the responses have been insufficient in changing their injection sharing practices. The harm reduction programme has to develop a more comprehensive response and as there are neighbourhood differences the programme also needs to address a broader environment, which may be making the IDU more vulnerable in that neighbourhood. Consideration of structural interventions may be necessary if an HIV epidemic is to be averted.
Authors' contributions
TA was responsible for the surveillance and cohort study on IDU and drafted the manuscript.
NH was responsible for managing the NEP of CARE, Bangladesh and was the counterpart for the cohort study on IDU from CARE, Bangladesh.
RK provided technical input to the Behavioural Surveillance and helped draft the manuscript.
All authors read and approved the final manuscript.
Acknowledgements
Surveillance was conducted by ICDDR,B, on behalf of the Govt. of Bangladesh, with support of grants from UNAIDS, from the Department for International Development (DfID), UK grant number BGH 9800 552/582/002, Govt. of Bangladesh/ DfID/IDA, credit number 3441 BD and USAID/FHI IMPACT #HRN-A-00-97-00017-00. The cohort study was conducted by ICDDR,B in close collaboration with CARE, Bangladesh and was funded by AusAID. The ICDDR,B acknowledges with gratitude the commitment of the different development partners to the Centre's research efforts. DfID funded CARE, Bangladesh for the intervention programme with IDU in Dhaka.
We acknowledge with gratitude all investigators of surveillance and the cohort study, individuals who are part of CARE, Bangladesh's IDU intervention programm e and all IDU who were part of these studies and programme.
==== Refs
Panda S Mallick PS Karim MA Sharifuzzaman M Ahmed AHT Baatsen P National assessment of situation and responses to opiod/opiate use in Bangladesh (NASROB) 2002 Dhaka, FHI/CARE/HASAB
GoB HIV in Bangladesh: where is it going? 2001 Dhaka, National AIDS/STD Program, Directorate General of Health Services, Ministry of Health and Family Welfare, Govt. of Bangladesh
GoB HIV in Bangladesh: is time running out? 2003 Dhaka, National AIDS/STD Program, Directorate General of Health Services, Ministry of Health and Family Welfare, Govt. of Bangladesh
GoB Report on the Second National Expanded HIV Surveillance, 1999-2000, Bangladesh 2000 , National AIDS/STD Program, Directorate General of Health Services, Ministry of Health and Family Welfare, Govt. of Bangladesh
GoB Report on the sero-surveillance and behavioural surveillance on STD and AIDS in Bangladesh, 1998-1999. 2000 , National AIDS/STD Program, Directorate General of Health Services, Ministry of Health and Family Welfare, Govt. of Bangladesh
GoB National HIV serological and behaviural surveillance, 2002, Bangladesh: fourth round technical report 2004 Dhaka, National AIDS/STD Programme, Directorate General of Health Services, Ministry of Health and Family Welfare, Govt. of the Peoples Republic of Bangladesh 15504471
GoB HIV in Bangladesh: the present scenario 2004 Dhaka, National AIDS/STD Programme, Directorate General of Health Services, Ministry of Health and Family Welfare, Govt. of Bangladesh 15504471
Azim T Alam MS Rahman M Sarker MS Ahmed G Khan MR Rahman S Rahman ASMM Sack DA Impending concentrated HIV epidemic among injecting drug users in Central Bangladesh International Journal of STD & AIDS 2004 15 280 282 15075028 10.1258/095646204773557875
Garten JG Lai S Zhang J Liu W Chen J Vlahov D Yu X Rapid transmission of hepatitis C virus among young injecting heroin users in Southern China Int J Epidemiol 2004 33 182 188 15075167 10.1093/ije/dyh019
Jenkins C Rahman H Saidel T Jana S Hussain AM Measuring the impact of needle exchange programs among injecting drug users through the National Behavioural Surveillance in Bangladesh AIDS Education and Prevention 2001 13 452 461 11718444 10.1521/aeap.13.5.452.24141
MAP R AIDS in Asia: face the facts. A comprehensive analysis of the AIDS epidemic in Asia 2004 Bangkok, Mapping the AIDS Pandemic Network
Strathdee SA Sherman SG The role of sexual transmission of HIV infection among injection and non-injection drug users J Urban Health 2003 80 iii7 iii14 14713667
Pisani E Dadun Suchaya PK Kamil O Jazan S Sexual behaviour among injection drug users in 3 Indonesian cities carries a high potential for HIV spread to noninjectors J Acquir Immune Defic Syndr 2003 34 403 406 14615658
Saidel TJ Des Jarlais DC Peerapatanapokin W Dorabjee J Singh S Brown T Potential impact of HIV among IDUs on heterosexual transmission in Asian settings: scenarios from the Asian epidemic model Int J Drug Policy 2003 14 63 74 10.1016/S0955-3959(02)00209-8
Panda S Chatterjee A Bhattacharya SK Manna B Singh PN Sarkar S Naik TN Chakrabarti S Detels R Transmission of HIV from injecting drug users to their wives in India International Journal of STD & AIDS 2000 11 468 473 10919490 10.1258/0956462001916137
Galea S Ahern J Vlahov D Contextual determinants of drug use risk behavior: a theoretical framework J Urban Health 2003 80, suppl. 3 iii50 iii58 14713671
Rhodes T Singer M Bourgois P Friedman SR Strathdee SA The social structural production of HIV risk among injecting drug users Soc Sci Med 2005 61 1026 1044 15955404 10.1016/j.socscimed.2004.12.024
Azim T Chowdhury E Hossain N Rahman M Khan R Ahmed G Sarker MS Faruque O Jana S Baseline characteristics of a cohort of injecting drug users in an intervention programme in Bangladesh: 20-24 April 2004; Melbourne, Australia. 2004
Faruque M Azim T Hussein N Chowdhury E Bosu A Enhancing efforts at needle/syringe exchange: the response of an intervention program to ongoing research: 20-24 April 2004; Melbourne. 2004
|
16248893
|
PMC1291384
|
CC BY
|
2021-01-04 16:36:50
|
no
|
Harm Reduct J. 2005 Oct 25; 2:22
|
utf-8
|
Harm Reduct J
| 2,005 |
10.1186/1477-7517-2-22
|
oa_comm
|
==== Front
Head Face MedHead & Face Medicine1746-160XBioMed Central London 1746-160X-1-111627448010.1186/1746-160X-1-11ResearchEffects of co-administered dexamethasone and diclofenac potassium on pain, swelling and trismus following third molar surgery Bamgbose Babatunde Olamide [email protected] Jelili Adisa [email protected] Wasiu Lanre [email protected] Akinola Ladipo [email protected] Godwin Toyin [email protected] Mobolanle Olugbemiga [email protected] Department of Oral and Maxillofacial Surgery, Lagos University Teaching Hospital, P.M.B 12003, Lagos, Nigeria2 Department of Oral and Maxillofacial Surgery, College of Medicine, University of Lagos, P.M.B 12003, Lagos, Nigeria2005 7 11 2005 1 11 11 17 6 2005 7 11 2005 Copyright © 2005 Bamgbose et al; licensee BioMed Central Ltd.2005Bamgbose et al; licensee BioMed Central Ltd.This is an Open Access article distributed under the terms of the Creative Commons Attribution License (), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
Background
The apparent interactions between the mechanisms of action of non-steroidal anti-inflammatory drugs (NSAIDS) and steroids suggest that co-therapy may provide beneficial inflammatory and pain relief in the absence of side effects. The aim of the study was to compare the effect of co-administered dexamethasone and diclofenac potassium (diclofenac K) with diclofenac K alone on the postoperative pain, swelling and trismus after surgical removal of third molars.
Patients and Methods
A prospective randomized double-blind study was conducted at the Department of Oral and Maxillofacial Surgery, Lagos University Teaching Hospital, Nigeria. A total of 100 patients were randomly allocated to two treatment groups of dexamethasone (prophylactic 8 mg and postoperative 4 mg IV) and diclofenac K (50 mg Oral before and after surgery), and diclofenac K alone (as with first group). The overall analgesic efficacy of the drug combinations was assessed postoperatively by determination of pain intensity using a category rating scale. Facial swelling was measured using a tape measure placed from tragus to gonion to tragus, while interincisal mouth-opening of patients was measured using a vernier calibrated caliper pre-operatively and post-operatively.
Results
Co-administration of dexamethasone and diclofenac K was significantly superior to diclofenac alone for the relief of pain (P < 0.05), and facial swelling up to post-operative 48 hour (P < 0.05). However, there was no significant difference for trismus relief between the two medication protocols (P > 0.05).
Conclusion
This study illustrates enhanced effects of co-administered dexamethasone and diclofenac K on short-term post-operative pain and swelling, compared to diclofenac potassium alone in third molar surgery.
==== Body
Introduction
Surgical removal of wisdom teeth under local anaesthesia is widely carried out in general dental practice and in many institutional surgery clinics, occupying an appreciable amount of clinical time [1,2]. This procedure is usually associated with postoperative pain, swelling, and trismus [1-4] as direct and immediate consequences of the surgical procedure [5,6]. The adverse effects of the wisdom tooth surgery on the quality of life has been reported to show a three-fold increase in patients who experience pain, swelling and trismus alone or in combinations; compared to those who were asymptomatic [5-7]. Many clinicians have, thus, emphasized the necessity for better pain, swelling and trismus control in patients who undergo third molar surgery [8,9].
The introduction of non-steroidal anti-inflammatory drugs (NSAIDs, e.g. diclofenac potassium and ibuprofen) has significantly altered the management of postoperative pain in dentistry and medicine. There are 2 possible mechanisms for the efficacy of NSAIDs when administered prior to surgical trauma. The first may simply be a pharmacokinetic advantage. By administering the NSAIDs prior to pain onset, drug absorption would have begun and therapeutic blood level will be present at the time of pain onset. Second, the presence of a cyclooxygenase inhibitor at the surgical site may limit the production of prostaglandins and prostacyclins associated with hyperalgesia and edema [10,11]. The use of corticosteroids (e.g. dexamethasone, betamethasone) is another preventive strategy for limiting postoperative edema and trismus following third molar extractions. Postoperative swelling and edema may be due in part to the conversion of phospholipids to arachidonic acid by phospholipase A2, and the resultant production of leukotrienes, prostacyclins, prostaglandins and thromboxane A2, acting as mediators of the inflammatory response. The use of steroids may inhibit the initial step in this process [12]. Clinical trials in oral surgery have also supported the hypothesis that preemptive NSAIDs and corticosteroids are effective in delaying and preventing many postoperative sequelae [10]. The apparent interactions between the mechanisms of action of non-steroidal anti-inflammatory drugs (NSAIDS) and steroids suggests that co-therapy may provide beneficial inflammatory and pain relief in the absence of side effects.
The aim of the study was, thus, to compare the effect of co-administered dexamethasone-diclofenac potassium with diclofenac potassium alone, on the postoperative management of pain, swelling and trismus following removal of impacted lower third molars.
Patients and methods
Patients who attended the Oral and Maxillofacial surgery clinic of the Lagos University Teaching Hospital, requiring surgical removal of unilateral or bilateral (at least 15 days between the two surgical procedures), impacted mandibular third molar teeth under local anaesthesia were included. The study protocol and the informed consent forms were approved by the research and ethics committee of the hospital. The study protocol was explained to the patients in detail after which consent was obtained. Patients were randomly allocated into two groups in a double-blind fashion by using prepared randomizations in sealed envelopes. Criteria for exclusion of patients included: renal or hepatic disease, blood dyscrasia, previous or present gastric ulcers, heart disease, known hypersensitivities, allergies, or idiosyncratic reactions to any study medications, pregnancy and lactation. In addition, patients who had taken analgesics or anti-inflammatory drugs within 24 hours before surgery were excluded from the study. All selected candidates were free of pain and other inflammatory symptoms that included swelling, hyperemia and decreased mouth opening at the time of surgery.
In Group I, patients were given a combination dexamethasone and diclofenac potassium (diclofenac K). Group II comprised of patients who were given diclofenac K alone.
The degree of surgical difficulty was assessed using Winter's and Terence Ward lines and Pell-Gregory criteria [13]. Oral perioperative antibiotics (500 mg ampicillin-cloxacillin, SmithKline Beecham, England and 400 mg metronidazole, Aventis Pharm. Int., Switzerland) were administered to all patients 30 minutes before surgery.
Operative procedure
Surgical extraction of the third molars was carried with buccal guttering technique after adequate elevation and reflection of buccal mucoperiosteal flap under local anaesthesia (2% lignocaine hydrochloride with 1:100,000 adrenaline). Tooth delivery was followed by meticulous irrigation of the surgical site with physiologic saline (0.9%). The three-sided mucoperiosteal flap was repositioned and sutured. A single operator performed all surgical procedures.
Pain measurement
Preoperative pain was assessed using a four-point Category Rating Scale [14,15] Accordingly, pain was recorded as: "0-no pain" (patient experiences no discomfort), "1-mild pain" (almost unnoticeable pain), "2-moderate pain" (noticeable pain, but patient can still engage in routine daily activities), and "3-severe pain" (very noticeable pain which disturbs the patient's daily routine). For each patient, the appropriate score was recorded in the questionnaire by one operator at 48 h and by the patient on a daily basis for 5 days. Before leaving the clinic, the operator ensured that all patients were thoroughly instructed how to complete the pain self-assessment diary and when to take medications.
Measurement of facial width
As no published method satisfies all criteria for assessing facial swelling, we decided to use a measuring tape to measure facial width and swelling in one-dimension only. Facial width (swelling) was measured using a measuring tape. The reference points used were the tip of tragus of left and right ears, with the gonium in between. A single operator, repeating the procedure three times on each patient, made the measurements. The average of measurements was then taken (in cm) and recorded. The measurements were carried out just before the surgery and at post-operative days 1, 2, and 7. Postoperative swelling was expressed as a percentage increase in facial width.
Measurement of mouth-opening ability
A vernier-calibrated sliding caliper was used to measure maximum interincisal mouth-opening ability of the patient at the commencement of the procedure. The reference point used was incisal edge of the maxillary central incisor and incisal edge of mandibular central incisor at maximum opening available.
The measurements were made in triplicate and the average was recorded in millimetres (mm). The measurement was carried out just before the surgery and at post-operative days 1, 2, and 7. Postoperative trismus was measured as a percentage decrease in mouth opening.
Medication
The operator supplied the medications to ensure compliance. Dexamethasone (Epil Pharmaceuticals, China) was given parenterally 8 mg 30 minutes preoperatively, and then 4 mg 6 hours postoperatively in two doses. Diclofenac K (Novartis, Switzerland) was given 50 mg, 30 minutes preoperatively; and thereafter 50 mg, 2 times daily for five days. All patients were placed on a five-day antibiotic regimen (500 mg Ampicillin-cloxacillin, SmithKline Beecham, England; 4 times daily and 400 mg Metronidazole, Aventis Switzerland, 3 times daily). All the medications were administered orally, except dexamethasone, which was administered parenterally.
Statistical Analysis
Data was analyzed using SPSS for windows (v11.5, SPSS Inc, Chicago, IL) statistical software package. One-way analysis of variance, student's t-test and χ2 were used for repeated measures for category rating scale, interincisal opening and facial swelling. The level of significance was set at P < 0.05.
Results
A total of 100 patients (equally distributed into groups I and II) who completed the study were included in the analysis. The mean age of the participants was 27.9 ± 5.2 years (range, 19–45 years; group I: 29.8 ± 5.3 years and group II: 26.1 ± 4.5 years). The male-to-female ratio was 1:1.1. The radiographic analysis of the type of impactions showed that mesio-angular impaction constituted 51.0% of cases, followed by disto-angular impaction (21.0%, Table 1).
Table 1 Types of impactions
Types Frequency (%)
Mesioangular 51 (51.0)
Distoangular 21 (21.0)
Horizontal 16(16.0)
Vertical 12(12.0)
Total 100 (100)
Table 2 presents postoperative pain intensity, facial swelling and maximal mouth opening on post-operative days 1 and 2. For group I, the mean pain score on days 1 and 2 was significantly lower than that for group II (p < .05, Table 2). Co-administration of Dexamethasone-diclofenac K led to a significant reduction in both postoperative pain and swelling on Days 1 and 2 when compared with diclofenac K alone (P < 0.05).
Table 2 Pain intensity, facial swelling and mouth opening at Days 1 and 2 (D1, D2) in group I (Dexamethasone-diclofecac K) and group II (Diclofenac only). Values are expressed as mean ± SD
Group N Pain intensity Facial Swelling Mouth Opening
D1 D2 D1 D2 D1 D2
Group I 50 0.62 ± 0.6* 0.5 ± 0.51* 30.9 ± 1.6* 31.0 ± 1.58* 36.0 ± 11.2 38.1 ± 10.05
Group II 50 1.64 ± 0.9* 1.3 ± 0.62* 31.7 ± 1.6* 32.04 ± 1.5* 39.2 ± 11.3 36.0 ± 10.02
*p < 0.05
Although there was no significant difference between the treatment groups with regard to reduction in mouth opening, the "trismus" scores of group I (0.31 mm) were lower than those of group II (3.19 mm) between 24 h and 48 h.
By the post-operative 7th day, all symptoms had restored to the preoperative level in both groups. Neither groups demonstrated any adverse reaction, side effect or other complications (e.g., tendency for bleeding) during the follow-up period.
Discussion
By pharmacologically controlling the extent of the inflammatory process, the intensity or severity of postoperative sequelae such as pain, swelling and trismus, may be reduced [16,17]. One technique that has been proposed for reduction of postoperative inflammation is the administration of corticosteroids [16]. Cortisol and the synthetic analogue of cortisol have the capacity to interfere with the physiologic processes of inflammation and, thus, suppress the development of local fever, redness, swelling and tenderness by which inflammation is recognized [16]. Another technique is to control the synthesis of prostaglandins. Prostaglandins play a major role in the induction of pain, inflammation, and fever [3,11]. The reduction of biosynthesis of prostaglandins by inhibition of the cyclo-oxygenase enzyme system is considered an important mechanism of action of NSAIDs. When administered preoperatively, NSAIDs have been shown to be particularly effective in combating postoperative pain [3,11].
Preventive strategies for postoperative management of pain and inflammation are based on the known ability of NSAIDs to block the arachidonic acid cascade. When NSAIDs are administered preoperatively, absorption and distribution of the medication may occur before the initiation of tissue trauma, the ensuing synthesis of prostaglandins and the subsequent inflammatory response. Prevention of the inflammatory response may decrease the sequelae of tissue trauma; especially the accompanying pain [11]. Diclofenac K has been shown to be useful in controlling postoperative pain after removal of third molars [18]. Diclofenac K is known to possess both analgesic and anti-inflammatory effect. Due to its anti-inflammatory effects [18], the administration of dexamethasone may synergize the anti-inflammatory effect of cataflam and contribute to the reduction of inflammatory exudates as well as edema and pain. Therefore the co-administration of diclofenac K and dexamethasone may be expected to reduce post-operative pain more than that achieved with diclofenac K alone [18].
The present study assessed the clinical effect dexamethasone-diclofenac K combination and diclofenac K alone on pain, facial swelling and trismus. Regardless of the drug combination used, the pattern of postoperative pain has been reported to increase between the post-operative days 1 and 3, after which the symptoms subside gradually within one week [19-22]. Our results confirm this observation. The comparison of pain intensity between the dexamethasone-diclofenac K group and diclofenac K group showed significant difference between the two groups (P < 0.05), indicating an enhanced analgesic effect of diclofenac K when administered in combination with dexamethasone. This finding corroborates with those of previous reports [3,23-26]. Schultze-Mosgau et al [25] investigated the combined use of ibuprofen and methylprednisolone for pain relief, concluding that this combination has good analgesic and anti-inflammatory action. It has also been reported that a single dose administration of a glucocorticoid reduces tissue levels of bradykinin and suppresses circulating levels of cortisol and beta-endorphin [25]. As known, bradykinin and kallidin are the two kinins that act independently as well as synergistically with products of the arachidonic acid cascade to produce both hyperalgesia as well as increased vascular permeability [26].
Post-surgical facial edema is difficult to quantify accurately, since it requires a three-dimensional measurement with an irregular, convex surface and can manifest itself internally as well as externally. Over the years, numerous researchers have tried various techniques in an effort to objectively measure edema [23,26], most of which are indirect assessments of the altered contours of skin surface. Measurement tools mentioned in the literature have included visual analog scales, trismus recordings, standardized stereo-radiographic or photographic measurements, computerized tomography, modified face bow devices, ultrasonography, facial plethysmographs, or various other means of taking direct facial measurements [23,26]. In the present study, a single measurement from the tip of tragus to gonion to the tip of contralateral tragus was taken. The recordings were made in triplicate and the average was recorded. It is noteworthy to mention herein that the cheek swelling following third molar surgery is diffuse in different planes and is very difficult to measure accurately. The co-administration of dexamethasone diclofenac K preoperatively and postoperatively, produced a clear reduction in postoperative pain and cheek swelling. The mean increase in facial swelling in days 1 and 2 in Group I (dexamethasone- diclofenac K combination) was significantly less than that of Group II (diclofenac K only). This result shows that co-administration of dexamethasone diclofenac K also enhances the control of postoperative facial swelling [24,27-29].
Independent T-test did not show any significant difference in reduction of mouth opening (trismus) between the study groups (P > 0.05). While this observation does not corroborate with those of previous reports [21,22,25], the enhanced effect of steroids on mouth-opening may be observed clinically. In the present study, the mean reduction in mouth-opening between the post-operative days 1 and 2 were 0.31 mm and 3.19 mm, for groups I (dexamethasone and diclofenac K) and II (diclofenac K only), respectively. This shows a "clinically significant" difference in the interincisal distance. These results indicate a positive clinical association between the adjunct use of dexamethasone and postoperative recovery of trismus in third molar surgery.
The time course for pain and facial swelling findings described in the present study are in agreement with those of a recent multicenter trial indicating similar symptoms that reached a maximum at Days 1 or 2 postoperatively and generally resolved at Day 7 [30,31].
The potency and dosage of dexamethasone within the first 24 h (total of 16 mg, including pre-operative dose) was adequate to enhance the efficacy of diclofenac K. It appears that steroids are preferably administered pre-operatively, extending the coverage up to 24–48 hours after surgery. Intravenous administration of dexamethasone, as done in the present study, enhances earlier bioavailability in comparison to oral administration [9]. Such treatment with high dosage does not impair adrenal function. Additionally, intravenous administration of dexamethasone prior to third molar surgery bears no detrimental impact on wound healing, even in patients predicted to be at high risk for delayed clinical recovery [9].
Conclusion
This study illustrates enhanced effects of co-administered dexamethasone and diclofenac K on short-term post-operative pain and swelling, compared to diclofenac potassium alone in third molar surgery.
Competing interests
The author(s) declare that they have no competing interests.
Authors' contributions
BBO and JAA conceived the study. BBO and WLA coordinated the write-up, did literature search and submission of the article. ALL, GTA, and MOO participated in the writing of the manuscript. All the authors read and approved the final manuscript.
Acknowledgements
We are grateful to Dr. Zafer C. Cehreli (Department of Paediatric Dentistry, Faculty of Dentistry, Hacettepe University, Ankara, Turkey) for his efforts and support in the final preparation of this manuscript for Head & Face Medicine.
==== Refs
Thomas D Walker R Smith A Shepherd J The provision of oral surgery services in England and Wales 1984–1991 Br Dent J 1994 176 215 219 8167064 10.1038/sj.bdj.4808417
Antila H Lehtinen R Heinaro A Lansineva A Salonen M Successful Pain Management by Finnish Oral Surgeons Oral Surg Oral Med Oral Pathol 1992 74 19 23 1508502
van der Westhuijzen AJ Roelofse JA Grotepass FW Becker PJ Randomized double-blind comparison of tiaprofenic acid and diclophenac sodium after third molar surgery Oral Surg Oral Med Oral Pathol 1994 78 557 566 7838460
Seymour RA Kelly PJ Hawkesford JE The efficacy of ketoprofen and paracetamol (acetaminophen) in post-operative pain after third molar surgery Br J Clin Pharmacol 1996 41 581 585 8799525 10.1046/j.1365-2125.1996.34015.x
Ruta DA Bissias E Ogston S Ogden GR Assessing health outcomes after extraction of third molars: postopeartive symptom severity (PoSSe) scale Br J Oral Maxillofac Surg 2000 38 480 487 11010778 10.1054/bjom.2000.0339
McGrath C Comfort MB Lo ECM Luo Y Changes in life quality following third molar surgery- the immediate postoperative period Br Dent J 2003 194 265 268 12658303 10.1038/sj.bdj.4809930
Slade GD Foy SP Shugars DA Phillips C White RP Jr The impact of third molar symptoms, pain and swelling on oral health-related quality of life J Oral Maxillofac Surg 2004 62 1118 1124 15346364 10.1016/j.joms.2003.11.014
Odgen GR Third molar surgery and postoperative pain relief Br Dent J 2003 194 261 (comment) 10.1038/sj.bdj.4809928
Tiwana PS Foy SP Shugars DA Marciani RD Conrad SM Phillips C White RP The impact of intravenous corticosteroid with third molar surgery in patients at high risk for delayed health-related quality of life and clinical recovery J Oral Maxillofac Surg 2005 63 55 62 15635558 10.1016/j.joms.2004.01.029
Moore PA Brar P Smiga ER Costello BJ Preemptive rofecoxib and dexamethasone for prevention of pain and trismus following third molar surgery Oral Surg Oral Med Oral Pathol Radiol Endod 2005 99 E1 7 10.1016/j.tripleo.2004.08.028
Jackson DL Moore PA Hargreaves KM Preoperative nonsteroidal anti-inflammatory medication for the prevention of postoperative dental pain JADA 1989 119 641 647 2691542
Hirschman JV Some principles of systemic glucocorticoid therapy Clin Exp Dermatol 1986 11 27 46 3708890
Akinwande JA Mandibular Third Molar Impaction- A comparison of two methods for predicting surgical difficulty Niger Dent J 1991 10 3 7
Rodrigo MR Rosenquist JB Cheung LK Paracetamol and difflunisal for pain relief following third molar surgery in Hong Kong Chinese Int J Oral Maxillofac Surg 1987 16 566 571 3116111 10.1016/S0901-5027(87)80107-8
Ong KS Seymour RA Pain Measurement in humans Surg J R Coll Surg Edinb Irel 2004 2 15 27
Ito U Reulen HJ Tomita H Ikeda J Saito J Maechara T Formation and propagation of brain edema fluid around human brain metastases Acta Neurochirur (Wien) 1998 90 35 41 10.1007/BF01541264
Mense S Sensitization of group IV muscle receptor to bradykinin by 5-hydroxytryptamin and prostaglandin E2 Brain Res 1981 225 95 105 6271342 10.1016/0006-8993(81)90320-6
Matthews RW Sully CM Levers BGH The efficacy of diclofenac sodium with and without paracetamol in the control of postsurgical dental pain Br Dent J 1984 157 357 359 6394041 10.1038/sj.bdj.4805486
Sisk AL Bonnington GJ Ga A Evaluation of methylprednisolone and flurbiprofen for inhibition of the postoperative inflammatory response Oral Surg Oral Med Oral Pathol 1985 60 137 145 3862020
Seymour RA The use of pain scales in assessing the efficacy of analgesics in post-operative dental pain Eur J Clin Pharmacol 1982 23 441 444 7151849 10.1007/BF00605995
Hyrkas T Ylipaavalniemi P Oikarinen VJ Paakkari I Pre-operative intravenous diclofenac for post-operative pain prevention in outpatients Br J Oral and Maxillofac Surg 1993 31 351 354 8286287 10.1016/0266-4356(93)90189-4
Neupert EA Lee JW Philput CB Gordon JR The evaluation of dexamethasone for reduction of postsurgical sequelae of third molar removal J Oral Maxillofac Surg 1992 50 1177 1182 1403273
Roger EA Roger RT A review of perioperative corticosteroid use in dentoalveolar surgery Oral Surg Oral Med Oral Pathol 2000 90 406 415
Ross R White CP Evaluation of hydrocortisone in prevention of postoperative complications after oral surgery: a preliminary report Journal of Oral Surgery 1958 16 220 226 13526058
Schultze-Mosgau S Schmelzeisen R Frolich JC Schmele H Use of ibuprofen and methylprednisolone for the prevention of pain and swelling after removal of impacted third molars J Oral Maxillofac Surg 1995 53 2 7 7799116 10.1016/0278-2391(95)90486-7
Troullos ES Hargreaves KM Buttler DP Dionne RA Comparison of nonsteroidal anti-inflammatory drugs, ibuprofen and flurbiprofen with methylprednisolone and placebo for acute pain, swelling and trismus J Oral Maxillofac Surg 1990 48 945 952 2395047
Beirne OR Hollander B The effect of methylprednisolone on pain, trismus, and swelling after removal of third molars Oral Surg Oral Med Oral Pathol 1986 61 134 138 3457335 10.1016/0030-4220(86)90173-8
Beirne OR Evaluation of dexamethasone for reduction of postsurgical sequelae of third molar removal (Discussion) J Oral Maxillofac Surg 1992 50 1182 1183
Huffman G Use of methylprednisolone succinate to reduce postoperative edema after removal of impacted third molar J Oral Surg 1977 35 198 202 264521
White RP JrShugars DA Shafer DM Laskin DM Buckley MJ Philips C Recovery after third molar surgery: clinical and health-related quality of life outcomes J Oral Maxillofac Surg 2003 61 535 544 12730831 10.1053/joms.2003.50106
Conrad SM Blakey GH Shugars DA Marciani RD Philips C White RP Jr Patients' perception of recovery after third molar surgery J Oral Maxillofac Surg 1999 57 1288 1294 10555792 10.1016/S0278-2391(99)90861-3
|
16274480
|
PMC1291385
|
CC BY
|
2021-01-04 16:37:14
|
no
|
Head Face Med. 2005 Nov 7; 1:11
|
utf-8
|
Head Face Med
| 2,005 |
10.1186/1746-160X-1-11
|
oa_comm
|
==== Front
Health Qual Life OutcomesHealth and Quality of Life Outcomes1477-7525BioMed Central London 1477-7525-3-671625963010.1186/1477-7525-3-67ResearchThe reliability, validity, and preliminary responsiveness of the Eye Allergy Patient Impact Questionnaire (EAPIQ) Alexander Michael [email protected] William [email protected] Patricia [email protected] John [email protected] Caroline [email protected] Jeff [email protected] Rob [email protected] Linda [email protected] Niagara Clinical Research, 5673 North Street, Niagara Falls, Ont L2G1J4, Canada2 Southern California Research, 27800 Medical Center Road, Suite 240, Mission Viejo, CA 92691, USA3 Allergan, Inc., Ettlingen GmbH, Pforzheimer Str. 160, Ettlingen 76275, Germany4 Allergan, Inc., 2525 Dupont Drive, Irvine, CA 92612, USA5 CT Burk, Inc., 1337 Cerritos Drive, Laguna Beach, CA 92651, USA6 Mapi Values Ltd, Adelphi Mill, Grimshaw Lane, Bollington, Macclesfield, Cheshire SK10 5JB, UK2005 31 10 2005 3 67 67 18 8 2005 31 10 2005 Copyright © 2005 Alexander et al; licensee BioMed Central Ltd.2005Alexander et al; licensee BioMed Central Ltd.This is an Open Access article distributed under the terms of the Creative Commons Attribution License (), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
Background
The Eye Allergy Patient Impact Questionnaire (EAPIQ) was developed based on a pilot study conducted in the US and focus groups with eye allergy sufferers in Europe. The purpose of this study was to present the results of the psychometric validation of the EAPIQ.
Methods
One hundred forty six patients from two allergy clinics completed the EAPIQ twice over a two-week period during the fall and winter allergy seasons, along with concurrent measures of health status, work productivity, and utility. Construct validity, reliability (internal consistency and test-retest), concurrent, known-group, and clinical validities, and responsiveness of the EAPIQ were assessed. Known-group validity was assessed by comparing EAPIQ scale scores between patients grouped according to their self-rating of ocular allergy severity (no symptoms, very mild, mild, moderate, severe, very severe). Clinical validity was assessed by assessing differences in EAPIQ scores between groups of patients rated by their clinician as non-symptomatic, mild, moderate, and severe.
Results and Discussion
Results from the validation study suggested the deletion of 14 of 43 items (including embedded questions) that required patients to complete the percentage of time they were troubled by something (daily activity limitations/emotional troubles). These items yielded a significant amount of missing or inconsistent data (50%). The resulting factor analysis suggested four domains: symptoms, daily life impact, psychosocial impact, and treatment satisfaction. When included as separate scales, the symptom-bother and symptom-frequency scales were highly correlated (> 0.9). As a consequence, and due to superior discriminative validity, the symptom bother and frequency items were summed. All items met the tests for item convergent validity (item-scale correlation = 0.4). The success rate for item discriminant validity testing was 97% (item-scale correlation greater with own scale than with any other). The criterion for internal consistency reliability (alpha coefficient ≥ 0.70) was met for all EAPIQ scales (range 0.89–0.93), as was the criterion for test-retest reliability (intraclass correlation [ICC] ≥ 0.70). Largely moderate correlations between the scales of the EAPIQ and the mini Rhinoconjunctivitis Quality of Life Questionnaire (miniRQLQ) and low correlations with the Health Utilities Index 2/3 (HUI2/3) were indicative of satisfactory concurrent validity. The EAPIQ symptoms, Daily Life Impact, and Psychosocial Impact scales were able to distinguish between patients differing in eye allergy symptom severity, as rated by patients and clinicians, providing evidence of satisfactory known-group and clinical validities, respectively. Preliminary analyses indicated the EAPIQ Symptoms, Daily Life Impact, and Psychosocial Impact scales to be responsive to changes in eye allergies.
Conclusion
Following item reduction, construct validity, reliability, concurrent validity, known-group validity, and preliminary responsiveness were satisfactory for the EAPIQ in this population of ocular allergy patients.
Patient functioningocular allergypsychometric validationEAPIQpatient reported outcomes
==== Body
Background
The term ocular allergy is used to describe a number of distinct disease entities, ranging from allergic conjunctivitis, a relatively mild condition, to keratoconjunctivitis, a sight-threatening condition affecting the cornea [1]. All could be described as atopic conditions affecting the conjunctiva and the surrounding structures of the eye, including the eyelids. Underlying immune mechanisms responsible have not been clarified, but it is believed that IgE related mast cell and eosinophil mediated inflammation leads to the release of mast cell mediators and toxic eosinophil granule proteins and enzymes. Ocular allergy affects approximately 15% of the world population, and its incidence is increasing in industrialised countries [2]. Elsewhere it has been reported that approximately 20% of the population in temperate climates suffer from allergic rhinoconjunctivitis [3]. Patients suffering from ocular allergy might experience such symptoms as red, itchy, burning, swollen or dry eyes in differing degrees of severity and duration. Some patients might only be affected for a few weeks, while others may experience symptoms continuously throughout the year. Thus, ocular allergy potentially affects patients in their daily life activities, thereby impacting their health-related quality of life (HRQoL). In particular, people suffering from ocular allergy may be limited in performing daily activities such as reading, computer work, and going outside.
In order to treat patients effectively it is necessary to know which treatments work best and which treatments patients prefer to use. Patient reported outcomes instruments can be used to determine which drugs have the greatest effect on patient reported HRQoL, treatment satisfaction, and work productivity. Given the plethora of drugs on the market, patient reported outcomes data can provide patients and clinicians guidance as to which treatments are most beneficial for ocular allergy patients.
The EAPIQ (Appendix [see Additional file 1]), an ocular allergy-specific questionnaire, was recently developed to evaluate the impact of eye allergies on patient functioning and daily activities, and to assess patient satisfaction with treatment, for use in clinical trials. In addition, three questionnaires measuring the HRQoL of patients with ocular allergy have been identified in the literature. They are as follows: the Rhinoconjunctivitis Quality of Life Questionnaire (RQLQ) with standardised activities, the miniRQLQ and the Allergic Conjunctivitis Quality of Life Questionnaire. Previous versions of the EAPIQ have been used in studies in Europe and US, and results have been presented as posters [4,5]. The objective of the present study was to further validate the questionnaire by investigating the psychometric properties of the EAPIQ in a sample of patients with ocular allergies, in US and Canada.
Methods
Subjects and study design
This was an observational validation study involving patients with ocular allergy (data collected between October 2002 and March 2003). There were 146 ocular allergy patients in two allergy clinics in US and Canada. All 146 patients were administered the EAPIQ, and the two concurrent measures at baseline seven to ten days later, 79 of these patients were administered the EAPIQ a second time in addition to the Health Change questionnaire (for the assessment of test retest reliability).
The patients were stratified by the clinicians into four groups based on the severity of their symptoms based on their clinic experience: 'no current symptoms' (n = 34), 'mild symptoms' (n = 40), 'moderate symptoms' (n = 42), and 'severe symptoms' (n = 30).
Measures
The following measures were administered during the study:
EAPIQ (Appendix [see Additional file 1])
A patient perspective questionnaire consisting of 49 items developed to measure ocular allergy symptoms and their impact on HRQoL, work productivity and treatment satisfaction. The EAPIQ was developed at Allergan from ocular allergy related questions derived from the mini Rhinoconjunctivitis QOL Questionnaire (mini RQLQ, Juniper et al.2000). Its structure, format, and layout was patterned after questions from the Ocular Surface Disease Index (OSDI, Walt et al.). The EAPIQ was presented to four patient focus groups (n = 10 in each group) in 2001 in UK, France, Italy and Sweden where language specific questionnaires (controlled by forward-backward translations) were generated for non- English groups. Patients were asked to comment on layout, structure, and clarity of questions. Based on these focus groups, the EAPIQ was restructured and questions were rephrased to be more patient friendly and concise. Further validation of the EAPIQ was conducted using the revised questionnaire at two allergy clinics in the US and Canada (146 patients).
Of the 49 original items in the questionnaire, the 43 items assessing symptoms (1–12), the impact of symptoms on HRQoL (items 18–31) and treatment satisfaction (items 32–34) were included in the item reduction and psychometric validation analyses. Six items assessing healthcare resource utilisation (item 13), work status (items 14 and 15), work productivity (items 16 and 17), and activity limitations (item 35) were not assessed in the analyses described because they require categorical and non-Likert type responses. Scores for the EAPIQ scales are transformed to give a minimum score of 0 and a maximum score of 100. Higher scores indicate a greater impact on health due to worse symptoms.
Mini Rhinoconjunctivitis Quality of Life Questionnaire (miniRQLQ)
A 14-item self-administered questionnaire developed by Elizabeth Juniper (MCSP, MSc) to measure the problems that adults with rhinoconjunctivitis experience in their day-to-day lives [6]. The miniRQLQ has five domains: activity limitations, practical problems, nose symptoms, eye symptoms and non-nose/eye symptoms.
Health Utilities Index (HUI2/3)
A health status and preference-based health-related quality of life measure suitable for use in clinical and population studies [7]. This 17 item self-administered questionnaire consists of seven attributes: sensation (vision, hearing, speech), mobility, emotion, cognition, self-care, pain, and fertility. The fertility dimension was excluded.
Health Status Change Questionnaire-Short Form
Administered at follow up, this questionnaire use six questions to assess the extent of any health change in the patient 7–10 days after the baseline assessment. Responses were used to categorise patients' health as 'better', 'stable', or 'worse' in order to assess the responsiveness of the EAPIQ.
Analyses
Exploratory Factor Analysis (principal components analysis) with the number of factors left free was performed to categorise each item to its respective domains. The methodology used thereafter utilised the information gained from the factor analysis. The number of factors selected was determined by the number of factors that provided more than a 0.5 step in eigen value, ± 2 factors. Consideration was also made of the number of factors with eigen values > 1.0. Items were considered for deletion if they loaded on two or more factors, or had a correlation of less than 0.40 with their own factor, or had a high (> 0.70) floor or ceiling effect (based on item-descriptive statistics). However, if items were found to have substantial face or content validity they may still be retained, regardless of the factor analysis results.
The EAPIQ was then assessed for the following psychometric properties: item convergent validity [8,9], item discriminant validity [10], internal consistency reliability [11,12], test retest reliability [13], floor and ceiling effects, scale-scale correlations, concurrent validity [14], known-group validity, clinical validity and responsiveness [14], all defined in Table 1.
Table 1 Purpose of psychometric tests
Property Purpose
Item convergent validity To assess an item's correlation with its own hypothesized sub-scale score (satisfied if correlation achieved is ≥ 0.40)
Item discriminant validity To assess whether an item considered in isolation has a higher correlation with its hypothesized scale than with other scales in the questionnaire
Internal consistency reliability To evaluate the extent to which individual items of the instrument are consistent to one another and reflect an underlying scheme or construct (satisfied if Cronbach's alpha coefficient = 0.70 is achieved)
Test-retest reliability Assesses the extent to which the measure yields the same results in repeated applications in an unchanged population. The intra-class correlation coefficient (ICC) was used as a measurement of test-retest reliability, and was assessed in patients who reported their health status to be stable between baseline (week 0) and study end (7 to 10 days later) (satisfied if an ICC coefficient = 0.70 is achieved)
Floor and ceiling effects Refer to a high percentage of patients scoring the lowest score possible and a high percentage of patients achieving the highest score possible, respectively. High baseline floor or ceiling effects are indicative of a scale that is limited in its responsiveness to clinical change. Minimal floor and ceiling effects are therefore recommended. For the EAPIQ scales a percentage of 20% at floor or at ceiling was considered a significant effect
Scale-scale correlations To determine whether the concepts measured in the individual scales (domains) of the EAPIQ were distinct and that none of the domains were redundant
Concurrent validity Concurrent validity was supported if the EAPIQ sub-scales were substantially correlated (≥ 0.40), with miniRQLQ sub-scales measuring similar concepts. Conversely, sub-scales measuring unrelated concepts should be poorly correlated. As a generic measure of health status the HUI2/3 was expected to be less strongly correlated with the EAPIQ scales
Known-group validity Differences in EAPIQ scores were expected among groups of patients known to differ in their patient-evaluated health status
Clinical validity Clinical validity assesses the ability of scores to discriminate among groups of patients defined according to clinical severity. Patients who have a good clinical status at baseline should score well in the questionnaire, and patients who have a poor clinical status at baseline should score poorly
Responsiveness Responsiveness refers to the ability of a measure to reflect underlying change. Preliminary responsiveness of the EAPIQ was assessed by comparing EAPIQ scores in those patients who report a change in their health status over the two-week period. Patients who were assessed at baseline and two weeks later were stratified by their report of worsening, no change and improvement in their 'overall health', 'all allergies', and 'eye allergy' symptoms, over the 7 to 10 days
Known-group validity was assessed by comparing EAPIQ scale scores between patients grouped according to their self-rating of ocular allergy severity. These patient-rated severity subgroups were compared using analysis of variance (ANOVA) on baseline data. It was hypothesised that patients in more severe groups would have worse (higher) EAPIQ scores, with the exception of the Treatment Satisfaction Scale.
Clinical validity was assessed by comparing EAPIQ scores according to the clinician report of ocular allergy severity. Severity was assessed using a single item measure that asks clinicians to rate the patient's ocular allergy as non-symptomatic, mild, moderate, or severe. Scores for non-symptomatic, mild, moderate and severe groups were expected to differ significantly from one another when compared using the Analysis of Variance (ANOVA) test.
In a preliminary analysis of responsiveness, changes in EAPIQ scores between baseline (week 0) and follow up (7–10 days after baseline) were compared among groups of patients who rated themselves as 'better', 'stable' or 'worsened' in terms of 'eye allergies', 'all allergies', and 'overall health' (as assessed using the Health Change Questionnaire). As a disease specific measure of allergy symptoms and wellbeing, EAPIQ scores were expected to be most sensitive to changes in 'eye allergies', and least sensitive to changes in 'overall health'.
Changes in EAPIQ scores were defined as small, moderate, or large using effect sizes (ES), as defined by Kazis [15]. Kazis proposed that an effect size between 0.20 and 0.49 are considered small, 0.50 to 0.79 are moderate, and 0.80 or above are large. It was hypothesised that those participants who reported an improved or worsened health status over the two weeks would show corresponding changes in EAPIQ scores, and those who reported an unchanged health status would have no significant change in their EAPIQ scores.
Statistical Analysis Software (SAS Institute Inc., Cary, NC) was used in the factor analysis assessment, clinical and known-group validity, and responsiveness over time. Multi-trait Analysis Program-Revised (MAP-R) [16] software was used for the assessment of psychometrics (internal consistency reliability, item convergent/divergent validity, floor/ceiling effects, scale/scale correlations). A significance level of 0.05 was used for all tests unless otherwise stated.
Results
One hundred and forty six patients were recruited. Demographic and clinical characteristics for the patient population at baseline are presented in Table 2.
Table 2 Demographic and clinical characteristics
Characteristic n (%) or mean
Gender n (%)
Male 34 (23.29)
Female 99 (67.81)
Missing data 13 (8.90)
Age
Mean 41.4
Standard deviation 13.3
Range 18.0–76.0
Missing data 1
Ethnicity n (%)
Caucasian 114 (81.43)
African-American 2 (1.43)
Hispanic/Spanish American 10 (7.14)
Asian/Oriental/Pacific is. 6 (4.29)
Other 8 (5.71)
Missing data 6 (4.11)
Highest level of education n (%)
High school or less 5 (3.62)
High school diploma 22 (15.94)
Some college 0 (0.00)
College degree 32 (23.19)
Graduate/postgraduate 47 (34.06)
Other 32 (23.19)
Missing Data 8 (5.48)
Current work status n (%)
Working (FT/PT) 102 (71.33)
Retired – ill health 4 (2.80)
Retired – age 6 (4.20)
Never in paid employment 2 (1.40)
Unemployed/searching 12 (8.39)
Other 17 (11.89)
Missing Data 3 (2.05)
Domestic situation n (%)
Living alone 15 (14.56)
Living with husband/partner 51 (49.51)
Living with children 9 (8.74)
Living with family/friends 25 (24.27)
Other 3 (2.91)
Missing Data 43 (29.45)
Patient perceived severity of ocular allergy n (%)
I don't have eye allergy symptoms 26 (17.81)
Very mild 10 (6.85)
Mild 34 (23.29)
Moderate 42 (28.77)
Severe 26 (17.82)
Very Severe 7 (4.79)
Missing data 1 (0.68)
Currently taking dry eye medication n (%)
Yes 93 (63.40)
No 52 (35.62)
Missing data 1 (0.68)
Construct validity
Fourteen items in the EAPIQ asked for the percentage of time the patients were troubled while carrying out daily activities. Responses for these items were often missing or were inconsistent with responses for corresponding 'level of bother' items. Consequently, the 14 frequency of bother items (the second part of questions 18 to 31) were deleted. Principal Components Analysis (PCA) was then conducted on the remaining items using Varimax, Promax, and Oblimin rotation methods. Items 11 and 12 ('Please rate to what extent you usually suffer from eye allergy symptoms in relation to OVERALL allergy symptoms' and 'How many days in the past week have you been free from allergy symptoms', respectively) were deleted because they did not load on any of the factors.
In addition, items 23 'Trouble with putting on/wearing make-up' and 31 'Troubled by feeling uncomfortable in business settings' were excluded from further analyses because of the high frequency of missing data for these items. The high frequency of missing data for these items is likely due to a large number of patients (for example, men) who do not wear makeup or who do not work. As these two items provided important information about patients for whom there is relevance, the items have been retained as single items instead of being part of any scale scores.
The relative merits of assessing symptom-bother in a scale separate from symptom-frequency were assessed. Each of the symptom-bother items was highly correlated to its corresponding symptom-frequency item (range: r = 0.85–0.90), suggesting redundancy. Furthermore, when MAP-R analysis was performed with the symptom-frequency and symptom-bother items as two separate scales, the two scales correlated very highly (r = 0.90) with each other, again suggesting redundancy. Known-group validity testing suggested the superior discriminative ability of the symptom-frequency scale (F = 44.63 vs. 39.63). However, when symptom-bother and -frequency items were summed, discriminative validity was superior for the summed measure (F = 45.29). As a result, symptom-bother and symptom-frequency items were summed in the scoring, reducing 10 items to five in the final factor analysis and psychometric analyses.
To summarise, 16 items were dropped from the questionnaire (items 12, 13, and the second part of questions 18–31), two items were excluded from further analyses but retained as single item measures (items 23 and 31), and five symptom-frequency items (items 1–5)_were combined with five symptom-bother items (items 1–6) in the scoring. Thus 20 items were included in the final factor analysis. The final factor analysis resulted in four domains being established: Daily Life Impact (eight items), Psychosocial Impact (four items), Symptoms (five items) and Treatment Satisfaction (three items). Standardised regression coefficients are presented in Table 3. There were five items that loaded on more than one factor. These items were assigned to scales based on a qualitative assessment of their content (face validity).
Table 3 Final rotated factor pattern, Oblimin rotation method (Standardized Regression Coefficients).
Factor 1: Daily Life Impact Factor 2: Psychosocial Impact Factor 3: Symptoms Factor 4: Treatment Satisfaction
22 Troubled with concentrating on daily tasks 0.89986 -0.12534 0.01571 0.13486
26 Troubled by feeling irritable 0.82799 0.27741 -0.19979 -0.15545
25 Troubled by feeling frustrated/angry 0.62794 0.45009 -0.08244 0.04161
24 Troubled by feeling tired/fatigued 0.61900 0.01189 0.22953 0.01391
21 Troubled with sleeping 0.58228 0.19613 0.06456 -0.05398
20 Troubled with going outdoors 0.55971 0.17176 0.19318 0.08847
18 Troubled with reading 0.53845 -0.07626 0.47037 0.09245
19 Troubled with driving 0.50325 -0.03648 0.45243 0.05192
29 Troubled by feeling less attractive -0.16437 0.82591 0.26851 0.06665
30 Troubled by feeling uncomfortable in social settings 0.15673 0.80984 -0.00459 -0.12471
28 Troubled by feeling helpless 0.19658 0.70135 -0.08085 0.12504
27 Troubled by feeling embarrassed 0.00565 0.69426 0.22569 0.12197
3 Red eyes -0.09104 0.14292 0.80942 0.08124
2 Water eyes 0.10791 0.27184 0.56104 -0.13643
1 Swollen / puffy 0.01604 0.36873 0.54369 -0.10554
5 Dry eyes 0.49205 -0.24760 0.51892 -0.12212
4 Itchy / burning eyes 0.46646 0.10921 0.45380 0.00162
32 Satisfaction with eye drops 0.01473 -0.00548 -0.02048 0.93682
34 Satisfaction with comfort of eye drops 0.00802 0.01457 -0.02372 0.93502
33 Satisfaction with how quickly eye drops improved 0.00569 0.04313 -0.02938 0.90542
Results of tests of item convergent validity, item discriminant validity, reliability, and floor and ceiling effects are presented in Table 4. All items met the criterion for item convergent validity (item-scale correlations of ≥ 0.40), and 90.7% of item-scale correlations (adjusted for overlap) were higher with the item's own scale than with any other EAPIQ scale (criterion for item discriminant validity). Only three items (items 5 'Dry eyes', 24 'Troubled by feeling tired/fatigued', and 26 'Troubled by feeling irritable') correlated slightly higher with a scale other than their own, as compared to the correlation with their own scale.
Table 4 Results of tests of item convergent validity, item discriminant validity, reliability, and floor and ceiling effects for the EAPIQ (total sample)
Item-level Scale-level Reliability Scale-level
EAPIQ scale No. of Items Convergent validitya Discriminant validityb Internal consistency Test-retest Floor effects Ceiling effects
Range of correlations Success rate (%) Success rate (%) Cronbach's alpha ICC % %
Symptomsc 5 0.53–0.77 100 90 0.84 0.75 11.3 0.7
Daily Life Impactc 6 0.57–0.78 100 83.3 0.88 0.84 15.5 0.0
Psychosocial Impactc 6 0.58–0.75 100 91.7 0.88 0.85 28.2 0.0
Treatment Satisfactiond 3 0.84–0.86 100 100 0.93 0.72 1.7 0.0
a Percentage of item-scale correlations ≥ 0.40.
b Percentage of item-scale correlations (adjusted for overlap) higher with the item's own scale than with any other EAPIQ scale
c Sample size of 142 patients who completed more than half of the items in the Daily Life Impact, Psychosocial Impact and Symptoms scales.
d Sample size of 119 patients who completed more than half of the items in the Daily Life Impact, Psychosocial Impact and Symptoms and Satisfaction scales.
All scales demonstrated excellent internal consistency reliability, with alpha coefficients ranging from 0.84 to 0.93. In addition, all scales surpassed the 0.70 criterion for test-retest reliability [ICC coefficients ranged from 0.72 to 0.85]. These results demonstrate satisfactory reliability for the EAPIQ multi-item scales.
There were no significant ceiling effects (percentage scoring at ceiling ranged from 0% to 0.7%) for any of the EAPIQ scales when assessed for the total cross sectional sample. When floor effects were assessed in the total cross sectional sample there were significant floor effects (> 20%) for the Psychosocial Impact scale only (28.2% scoring at floor). Patients with 'no eye allergy symptoms', are expected to score at floor. When these patients were excluded, there were no significant floor effects (2.7% of scoring at floor for the Symptoms scale and 17.3% of scoring at floor for the Psychosocial Impact scale).
Concurrent validity
EAPIQ Symptoms, Daily Life Impact, and Psychosocial Impact scores all correlated significantly with the miniRQLQ scores (P < 0.0001 for all). The correlations were moderate, ranging from r = 0.34 to r = 0.85. There was a low, statistically significant correlation between EAPIQ Treatment Satisfaction scores and miniRQLQ Eye Symptoms scores (r = 0.24, P = 0.0090). The EAPIQ Treatment Satisfaction Scale did not correlate significantly with any of the other miniRQLQ scales.
Correlations between the EAPIQ scales and the items of the HUI2/3 were low (0.20<r<0.45) or negligible and not statistically significant. These lower correlations were expected because the HUI2/3 is a generic health status measure, whereas the EAPIQ and the miniRQLQ are measures specific to ocular allergies.
Comparison of EAPIQ scores according to patient demographics
Scores from female subjects were significantly higher than those from male subjects for the EAPIQ Symptoms (F = 9.58, P = 0.0024), Daily Life Impact (F = 10.02, P = 0.0019), and Psychosocial Impact (F = 14.66, P = 0.0002) scales (Figure 1). Treatment Satisfaction scores did not differ by gender (F = 1.11, P = 0.2940).
Figure 1 EAPIQ scale scores at baseline by gender. *Mean EAPIQ scores with 95% Confidence Interval (n) **Overall ANOVA results found statistically significant differences between groups (P < 0.01)
None of the EAPIQ scale scores correlated significantly with age, or with years of suffering from eye allergy symptoms. Patients taking medication for their eye allergy symptoms had significantly higher scores than those not taking medication for the EAPIQ scales of Symptoms (F = 9.10, P = 0.0030), Daily Life Impact (F = 8.31, P = 0.0046), and Psychosocial Impact (F = 6.92, P = 0.0095) (Figure 2). Treatment Satisfaction scores were not compared between these two groups, as individuals not on treatment did not complete the questions corresponding to the Treatment Satisfaction scores.
Figure 2 Comparison of EAPIQ scores at baseline between patients taking medication and those not taking medication for their eye allegy symptoms. *Mean EAPIQ scores with 95% Confidence Interval **Overall ANOVA results found statistically significant differences between groups (P < 0.01) ***Except for the Symptoms scale, for which n = 92
Known-group validity
Known-group validity estimates how well the questionnaire discriminates between groups. The results from the ANOVA test showed that the EAPIQ Symptoms (F = 27.96, P < 0.0001), Daily Life Impact (F = 16.88, P < 0.0001), and Psychosocial Impact (F = 14.97, P < 0.0001) scales distinguish between patients who rated themselves as having no allergy symptoms versus different grades of symptom severity (very mild, mild, moderate, severe, and very severe). (Figure 3).
Figure 3 Known groups validity: EAPIQ scale scores at baseline by patient rating of ocular allergy severity. *Mean EAPIQ scores with 95% Confidence Interval (n) **Overall ANOVA results found statistically significant differences between groups (P < 0.0001)
EAPIQ Treatment Satisfaction scores did not differ significantly between groups of varying patient-rated severity. The result was expected since Treatment Satisfaction is not expected to change with severity.
Clinical validity
The patients' clinicians were asked to rate each patient as having either no eye allergy symptoms, mild eye allergy symptoms, moderate eye allergy symptoms or severe eye allergy symptoms. EAPIQ scale scores were compared among these four groups. Results from the ANOVA test showed that the EAPIQ scales of Symptoms (F = 46.95, P < 0.0001), Daily Life Impact (F = 34.55, P < 0.0001), and Psychosocial Impact (F = 24.83, P < 0.0001) distinguished with statistical significance between the patients in the no allergy symptoms category and the different severity groups as rated by clinicians (Figure 4). As expected, Treatment Satisfaction scores did not differ significantly between clinician-rated severity groups.
Figure 4 Clinical validity: EAPIQ scale scores at baseline by clinician rating of ocular allergy severity. *Mean EAPIQ scores with 95% Confidence Interval (n) **Overall ANOVA results found statistically significant differences between groups (P < 0.0001)
Responsiveness
For comparisons among the 'better', 'stable' and 'worsened' groups according to all 3 health change items, small sample sizes (N<20) in the 'better' and 'worsened' groups warrant cautious interpretation.
The EAPIQ is responsive to changes in eye allergies. For all EAPIQ scales, scores worsened (ES range: 0.26 to 0.50) for patients who reported a deterioration of their eye allergy, improved (ES range: -0.10 to -0.56) for the group with 'better' eye allergies, and showed negligible or small change (ES range: -0.05 to 0.20) in patients who reported 'stable' eye allergies (Figure 5). However, sample sizes were not large enough to make statistical comparisons among the groups.
Figure 5 Responsiveness: change over time in EAPIQ scales by change in eye allergies. *Paired t-tests found statistically significant change over time within groups (P < 0.05) **Mean change in EAPIQ scores with 95% Confidence Interval (n)
Findings were similar when comparing changes in EAPIQ scale scores according to change in 'all allergies', and 'overall health'.
Discussion
Based on the results of this psychometric evaluation, the EAPIQ was found to be reliable, valid, and responsive. Following item reduction, scaling assumptions were met satisfactorily for the 4 multi-item scales and most items. The results provide evidence of the psychometric integrity of the EAPIQ within the studied eye allergy population and support its use in patients with eye allergies.
Findings suggest that asking patients to write in their responses can lead to inconsistent responses or missing data. Having items which required patients to rate the frequency with which they were bothered by their eye allergies while carrying out activities, in addition to their level of bother, proved confusing for some subjects. Furthermore the frequency of bother items which required patients to write in their responses also had high levels of missing data. Consequently, frequency items were deleted from the questionnaire and excluded from the remainder of the analyses.
High correlations between the symptoms 'intensity of bother' items (items 6–10) and their corresponding 'frequency' items (items 1–5) suggest redundancy. Clinicians view frequency as the more pertinent index of severity, whereas patients consider intensity of bother to be of greater salience. Consequently, instead of deleting either the 'intensity of bother' or 'frequency of bother' items, they were combined in the scoring. The 'intensity of bother' and 'frequency of bother' items were summed, not multiplied because known-group validity testing indicated superior discriminative ability for combining the items by summation rather than multiplication.
Items 23 and 31, pertaining to wearing makeup and working in business settings, respectively, were not included in the psychometric validation analyses due to high levels of missing data. However, these items provide important information for whom these items are relevant. Therefore, these items are scored as single items (not forming part of any scale score) rather than being excluded from the questionnaire entirely.
The final factor analysis suggested four domains: Daily Life Impact (eight items), Psychosocial Impact (four items), Symptoms (five items) and Treatment Satisfaction (three items), and two single items (wearing makeup and working in business settings). There were five items that loaded on two factors and these double loadings were logical in terms of content validity. For example, 'troubled with reading' item loaded on both the Daily Life Impact' factor (regression coefficient of 0.54), and the Symptoms factor (regression coefficient of 0.47). In terms of face validity, 'trouble with reading' is expected to be part of the Daily Life Impact factor. Since reading may be affected by the severity of eye allergy symptoms, it is logical that it also loads on the Symptoms factor.
The four multi-item scales of the EAPIQ were psychometrically robust; in that, all scales demonstrated excellent item convergent validity, excellent internal consistency reliability, and satisfactory test-retest reliability. All but three items satisfied the requirements for item discriminant validity. Floor and ceiling effects were satisfactory for the EAPIQ scales when patients with 'no current symptoms' who would be expected to score at floor were excluded from analysis.
Moderate correlations between the Symptoms, Daily Life Impact and Psychosocial Impact scales indicate that these three scales are measuring concepts that are related but distinguishable and not redundant. The low correlations between the Treatment Satisfaction scale and the remaining EAPIQ scales are also in line with expectations, as satisfaction as a concept is not expected to be strongly associated with symptom severity or impact on either the daily life or psychosocial factors.
When correlations were examined between the EAPIQ scales and the concurrent measures, correlations between similar scales were moderate or high, confirming the concurrent validity of the EAPIQ. Correlations between the concurrent measures and the treatment satisfaction scale were low, as expected since treatment satisfaction is not generally related to symptom severity or disease impact.
The EAPIQ scales correlated higher with the scales of the miniRQLQ than with the items of the HUI2/3. This finding was expected since the HUI2/3 is a generic health status measure, whereas the EAPIQ and the miniRQLQ are specific to ocular allergies. Thus, the EAPIQ demonstrated satisfactory concurrent validity.
The EAPIQ Symptoms, Daily Life Impact and Psychosocial Impact scales were able to distinguish between varying levels of patient-reported symptom severity. Higher scores (indicating worse health) were observed for more severe symptom severity, confirming the known-group validity of the EAPIQ. In line with expectations, Treatment Satisfaction scores did not change with varying degrees of patient-perceived symptom severity.
The EAPIQ Symptoms, Daily Life Impact, and Psychosocial Impact were also able to discriminate between varying levels of clinician-rated symptom severity, as evident by higher scale scores (indicating worse health) for patients with more severe symptom severity. Treatment Satisfaction scores did not change with varying degrees of clinician-reported symptom severity. These findings demonstrate the clinical validity of the EAPIQ.
Analyses of responsiveness were exploratory and should be interpreted with caution due to small sample sizes in the 'better' and 'worse' groups. For patients who reported their eye allergies as better, worsened, or stable between baseline and follow up, there were corresponding changes in scores for the scales of Symptoms, Daily Life Impact, Psychosocial Impact and Treatment Satisfaction. However, changes in EAPIQ scores in 'worsened' and 'better' groups were not consistently statistically significant.
Similar patterns were found for the change in EAPIQ scores according to 'change in overall heath' and 'change in eye allergies' even though the EAPIQ change scores for patients who rated their status as 'stable' were larger. These results suggest that the EAPIQ is responsive to change over time, but further study in a larger population is required to verify this assertion.
Minimal important differences or minimal important change over time were not examined for the EAPIQ scales. Knowledge of minimal important differences are important for interpreting the meaning of HRQoL results, thus, in future studies some attempt should be made to define minimal important differences for the EAPIQ.
Conclusion
Results suggest that the EAPIQ is a reliable and valid measure and is appropriate for use in studies of ocular allergy. The EAPIQ is appropriate for assessing ocular allergy symptoms and their impact on patients' daily lives and psychosocial functioning, in addition to satisfaction with treatment. Further research investigating responsiveness to change over time and minimal important difference for the EAPIQ is warranted. Use of the EAPIQ in future research may contribute to a greater understanding of the impact of ocular allergies on patients' lives and ultimately lead to the use of treatments that improve functioning in the areas that are of greatest importance to the patients themselves.
Authors' contributions
PB, JW CB and LA conceived the study and participated in the design of the study. MA, WB, PB and CB coordinated the data collection. RA and LA oversaw the statistical analysis and drafted the manuscript. All authors were involved in item reduction decisions. All authors read and approved the final manuscript.
Supplementary Material
Additional File 1
Appendix: Eye Allergy Patient Impact Questionnaire
Click here for file
Acknowledgements
This research was supported by Allergan Pharmaceuticals.
==== Refs
Bhargava A Jackson WB El-Defrawy SR Ocular allergy disease Drugs Today 1998 34 957 971 14743264
Wood B New treatments to relieve ocular allergies Rev Optom 1999 136 124 135
Knight A The role of levocabastine in the treatment of allergic rhinoconjunctivitis Br J Clin Pract 1994 48 139 143 7913336
Walt J Wojcik A Buchholz P Initial Development and Validation of the Eye Allergy Patient Impact Questionnaire (EAPIQ) Poster presented at the Annual Meeting of the International Society for Quality of Life Research (ISOQOL) October 30 – November 2 2002 Orlando Florida, USA
Buchholz P Walt J Lorenz DG Burk C Lee J Patient Impact of Allergic Conjunctivitis as measured by the Eye Allergy Patient Impact Questionnaire (EAPIQ) Poster presented at the Annual Meeting of the Association for Research in Vision and Opthalmology, (ARVO); May 4–9, 2003 Ft. Lauderdale, FL
Juniper EF Thompson AK Ferrie PJ Roberts JN Development and validation of the mini Rhinoconjunctivitis Quality of Life Questionnaire Clin Exp Allergy 2000 30 132 140 10606940 10.1046/j.1365-2222.2000.00668.x
Feeny DH Torrance GW Furlong WJ Spilker B Health Utilities Index Quality of Life and Pharmacoeconomics in Clinical Trials 1996 Second Philadelphia: Lippincott-Raven Press 239 252
Ware JEJ Snow KK Kosinski M Gandek B SF-36 Health Survey Manual and Interpretation Guide 1993 Boston; USA: New England Medical Center
Hays RD Hayashi T Beyond internal consistency reliability: Rationale and user's guide for Multitait Analysis Program on the microcomputer Behav Res Methods 1990 22 167 175
Campbell DT Fiske JL Convergent and discriminant validation by the Multitrait multimethod matrix Psychol Bull 1959 56 81 105 13634291
Cronbach LJ Coefficient alpha and the internal structure of tests Psychometrika 1951 16 297 334
Nunnaly JC Bernstein IR Psychometric Theory 1994 3 New York: McGraw-Hill. Chapter 7 The Assessment of Reliability 248 292
Chassany O Sagnier P Marquis P Fulleton S Aaronson N Patient-Reported outcomes: the example of health related quality of life – a European guidance document for the improved integration of health related quality of life assessment in the drug regulatory process Drug Inf J 2002 36 209 238
Hays R Anderson R Reviki DA Staquet MJ, Hays RD, Fayers PM Assessing reliability and validity of measurement in clinical trials Quality of Life Assessment in Clinical Trials 1998 New York: Oxford University Press
Kazis LE Anderson JJ Meenan RF Effect sizes for interpreting changes in health status Med Care 1989 27 S178 S189 2646488
Ware JE Hams WJ Gandek B Rogers B Reese PR MAP-R for Windows: multi-trait/multi-item analysis porgram-revised Users guide 1997 Boston: health Assessment Lab
|
16259630
|
PMC1291386
|
CC BY
|
2021-01-04 16:38:13
|
no
|
Health Qual Life Outcomes. 2005 Oct 31; 3:67
|
utf-8
|
Health Qual Life Outcomes
| 2,005 |
10.1186/1477-7525-3-67
|
oa_comm
|
==== Front
Int J Health GeogrInternational Journal of Health Geographics1476-072XBioMed Central London 1476-072X-4-281627766110.1186/1476-072X-4-28MethodologyA national, geographic database of CDC-funded HIV prevention services: development challenges and potential applications Hanchette Carol L [email protected] Deborah A [email protected] Aisha [email protected] Kieran J [email protected] Mark [email protected] Department of Geography and Geosciences, University of Louisville, Louisville KY, USA2 RTI International, Research Triangle Park, NC, USA3 Centers for Disease Control and Prevention, Atlanta, GA, USA4 Department of Interdisciplinary Health Studies, Western Michigan University, Kalamazoo, MI, USA2005 8 11 2005 4 28 28 7 9 2005 8 11 2005 Copyright © 2005 Hanchette et al; licensee BioMed Central Ltd.2005Hanchette et al; licensee BioMed Central Ltd.This is an Open Access article distributed under the terms of the Creative Commons Attribution License (), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
Background
From 2000–2002, the Centers for Disease Control and Prevention (CDC) funded a study that was designed to improve the information available to program planners about the geographic distribution of CDC-funded HIV prevention services provided by community-based organizations (CBOs). Program managers at CDC recognized the potential of a geographic information system (GIS) to organize and analyze information about HIV prevention services and they made GIS a critical component of the study design. The primary objective of this study was to construct a national, geographically-referenced database of HIV prevention services provided by CDC-funded CBOs. We designed a survey instrument to collect information about the geographic service areas where CBOs provided HIV prevention services, then collected data from CBOs that received CDC funding for these services during fiscal year 2000. We developed a GIS database to link questionnaire responses with GIS map layers in a manner that would incorporate overlapping geographies, risk populations and prevention services. We collected geographic service area data in two formats: 1) geopolitical boundaries and 2) geographic distance.
Results
The survey response rate was 70.3%, i.e. 1,020 of 1,450 community-based organizations responded. The number of HIV prevention programs administered by each CBO ranged from 1 to 23. The survey provided information about 3,028 prevention programs, including descriptions of intervention types, risk populations, race and ethnicity, CBO location and geographic service area. We incorporated this information into a large GIS database, the HIV Prevention Services Database. The use of geopolitical boundaries provided more accurate results than geographic distance. The use of a reference map with the questionnaire improved completeness, accuracy and precision of service area data.
Conclusion
The survey instrument design and database development procedures that we used for this study successfully met our objective. The development of the HIV Prevention Services Database for CDC is an important step toward the implementation of a spatial decision support system. Due to the costs involved in a nationwide survey such as this, we recommend that future data collection efforts use Web-based survey methodologies that incorporate interactive maps.
==== Body
Background
The Centers for Disease Control and Prevention (CDC) is among the nation's leading sources of funding for programs to prevent the spread of human immunodeficiency virus (HIV) and acquired immune deficiency syndrome (AIDS). Yet, until recently, CDC had limited information about the geographic distribution of these programs and the extent to which funded services are accessible to the populations at greatest risk of contracting HIV/AIDS.
From 2000–2002, CDC funded a study that was designed to improve the information available to program planners about the geographic distribution of CDC-funded HIV prevention services provided by community-based organizations (CBOs). Researchers and program managers at CDC recognized the potential of a geographic information system (GIS) to organize and analyze information about HIV prevention services. They made GIS a critical component of the study design.
The primary objective of this study was to construct a national, geographically-referenced database of HIV prevention services provided by CDC-funded community-based organizations. To achieve this objective, we were faced with three challenges:
1. to obtain information about geographic service areas across the entire United States and its territories;
2. to design a survey instrument in a manner that encouraged service providers to give accurate information about service area geography; and
3. to design and maintain a database that could handle overlapping geographies, risk populations and prevention services and still be user-friendly for program managers at CDC.
This paper describes how we addressed each of these challenges to achieve our major objective.
Geographic analyses and health services research
Geography has been a critical component of health care analysis for many decades. In their classic text on health services research, Joseph and Phillips provided a discussion of health care delivery systems around the world, clearly defined the meanings of concepts such as "access" and "utilization," and presented many spatial methods that have been used to analyze them [1]. The text was written when GIS was in its infancy (although developments in computer cartography had been proceeding for several years), but the authors demonstrated how mapping could be used to gain an understanding of the spatial organization of health care providers, such as physicians, and the spatial hierarchy of hospital facilities. They reported on a wide range of methods for measuring accessibility to health services, many of them borrowed from economic geography. These include the location quotient and coefficient of localization, which provide general measures of regional distribution and inequity. Many of the methods used to assess regional accessibility fall into the "distance decay" category, i.e. they involve measurements of cumulative distances between health services and neighbourhoods or other population units and operate under the assumption that distance is a barrier or deterrent to health-seeking behaviour. Other topics covered were health planning and strategies for locating new hospitals and health services.
Many of these methods, such as the computation of physician-patient ratios and location quotients, were demonstrated in Rickets et al. [2]. Rural-urban classification systems were discussed and health care shortage areas mapped. Most of these methods can be applied with geographic information systems.
A geographic information system is an information management system that contains spatially referenced data. Clarke has referred to GIS as 1) a toolbox, 2) an information system, and 3) an approach to science. As a toolbox, a GIS is a software package that contains a variety of tools and functions for processing, mapping and analyzing spatial data. As an information system, it contains a series of databases with observations about features and other entities with known locations. As an "approach to science" it involves the study of the scientific disciplines, such as geography and cartography that have contributed to the development of GIS technology [3].
As an information system, geography thus is the common denominator for disparate data types. Mapping the location of prevention services in relation to HIV incidence, for example, could graphically demonstrate possible gaps in service availability and suggest priorities for locating new service sites. Because maps make complex data more accessible to both experts and non-experts, they can facilitate discussion about issues of access and service needs.
While GIS data are often viewed as maps, the spatial relationships among objects and features in a GIS make it much more powerful than a mere mapping tool. The main reason for this is that spatial data in a GIS are structured in a manner that maintains topological relationships among features such as points, lines and areas. These relationships include adjacency, containment, and connectivity and they allow GIS users to perform analyses among features in a single map layer or multiple map layers [4]. For example, distance could be computed easily between hospitals in one map layer and census blocks in another map layer, to measure accessibility of population to hospitals using a distance decay function. GIS technology can also support spatial query, analysis and modeling functions. Techniques include buffer zone analysis that can estimate the number of persons living within a specified distance of a resource (such as a test site), or the distribution of HIV prevention services in relation to the number of persons living with AIDS.
GIS technology has lead to the enhancement of existing techniques and development of new methods for analyzing health services. GIS can be used to create health service regions based on spatial data using such methods as Thiessen polygons and flow mapping [2]. Spatial interaction models, mathematical programming and GIS network analyses use street data and distance measurements to model flows among patients and health services and are used to allocate patients to services, route emergency vehicles, or strategically locate new facilities [5]. Many of these processes are iterative and can be run many times to examine a host of different scenarios. Of course, GIS analyses would be rather limited without the widespread availability of digital spatial data and health data with geographic identifiers [6].
GIS is increasingly used in public health and health services research. Since 1994, the National Center for Health Statistics has published its bimonthly electronic report, Public Health GIS News and Information. In 1999, the Journal for Public Health Management and Practice devoted two issues entirely to GIS applications. In May, 2003, the Annual Review of Public Health published five articles that, together, constituted a "minisymposium" on the use of GIS in public health. Two issues of the Journal of Medical Systems were devoted to GIS in 2004. GIS has been recognized as an emerging technology in the field of public health informatics [7]. Additionally, new journals, such as the International Journal of Health Geographics have emerged to meet the demand for research in medical informatics and geographic analyses of health issues.
The CDC has made a commitment to utilizing new technologies to improve health information [8]. The potential of GIS has also been recognized by the Department of Health and Human Services (DHHS). The Healthy People 2010 Objective 23-3 is to "increase the proportion of all major national, state, and local health data systems that use geocoding to promote nationwide use of geographic information systems (GIS) at all levels" [9].
Several recent studies have involved the use of GIS in health services research [10-12]. Of renown also are the Dartmouth atlases of health care, which use data from health care claims databases to map and analyze geographic aspects of health care in the U.S. [13,14]. Most geographic studies are more localized, however, such as an analysis of the accessibility of HIV services in Toronto neighbourhoods or GIS-based assessment of physician shortages took place in a 9-county area of Illinois [15,10].
One of the most important potential uses of GIS technology in evaluation and planning of health care services is as a spatial decision support system (SDSS). SDSS provide information to planners and program managers that allow them to make important decisions about resource allocation. They require the development of a spatially-enabled database, a database management system, and a set of analytical tools, such as those found with most GIS software, for solving problems [16]. The first step in the development of a spatial decision-support system is to develop a spatially-enabled database.
The result of our work has been the development of such a database – a dynamic, national spatial database of locations and corresponding geographic service areas of CDC-funded CBOs that provide HIV prevention services. We termed this database the HIV Prevention Services Database. This database is maintained in a GIS and has enormous potential to provide information to program managers for decision making.
Methods
We collected data via a questionnaire that was mailed to all HIV prevention service providers funded by CDC during fiscal year 2000. Service providers included those funded directly by CDC and those funded indirectly through cooperative agreements with state or local health departments. While most HIV prevention service providers were CBOs, in some cases, state and local health departments were respondents, describing services that they provided themselves rather than through contracts with CBOs.
The questionnaire prompted respondents to provide information about the following: 1) descriptions of prevention interventions; 2) descriptions of persons served by the intervention; and 3) the location of service delivery and the geographic area in which persons served live. Particularly problematic was the issue of how to ask questions about geographic service areas. While locations of CBOs themselves generally have mappable addresses, how does one ask questions that can provide accurate information about the delineation of geographic service areas? Should respondents draw service areas on a map (to be digitized later) or should they indicate which standard geographic units (e.g. county, ZIP code) best describe their service area? We discuss these issues later in this section.
We pretested a draft questionnaire with HIV prevention providers in Raleigh and Durham, North Carolina. Following revisions suggested by the pretest, we conducted a pilot test in San Diego with HIV prevention program managers in six CBOs. After completing the questionnaire, the program managers participated in debriefing interviews in which they described how they interpreted questions and chose responses and discussed any difficulties they encountered with the instrument. Other revisions were suggested during an expert panel meeting convened at CDC to discuss the survey instrument, database design issues, and analysis.
To maximize compatibility of survey data with other current and planned CDC data collection efforts, response categories for intervention type and persons served were consistent with those of CDC's Evaluation Guidance [17]. Using response options shown in Table 1, the following types of data were collected for each prevention program: 1) intervention type, 2) risk population, 3) race and ethnicity and 4) funding source. Multiple responses were allowed for intervention type, risk population, race and ethnicity of the individuals served. For funding source data, we asked respondents whether prevention programs were funded directly by CDC, indirectly through a state or local health department, or both. Although this information was available from CDC and health department data at the CBO level, it was included on the survey instrument to see whether funding sources for specific prevention programs could be identified when respondents received funding from multiple sources.
Table 1 Response categories for interventions, risk populations and race/ethnicity of persons served. Data for the following response categories were collected by the survey. Multiple responses were allowed for all categories
Intervention Type Risk Populations Race and Ethnicity
• Individual-level interventions
• Group-level interventions
• Street and community outreach
• Prevention case management
• Community-level interventions
• Health communications/public information
• Counseling, testing, referral, and partner notification • Men who have sex with men (MSM)
• MSM/intravenous drug users (IDU) (and other drug users)
• IDU
• Heterosexual
• Mother with/at risk for HIV
• General public • African American
• American Indian or Alaska Native
• Asian
• Native Hawaiian or Other Pacific Islander
• Hispanic or Latino
• White
• More than one race*
• Race unknown
Service area definitions
Data describing intervention types and persons served, combined with the address of responding CBOs, would by itself yield valuable information about the locations of services being provided with CDC funds for specific populations. However, the intent of this study was to describe service area as well as service location. Geographic service areas can be defined in several ways, each of which has ramifications in terms of data collection issues and analyses:
1. Patient origin. The service area is defined by compiling actual addresses for persons served. Although this approach provides very precise data, it also involves concerns about respondent burden, confidentiality, and data quality. Many HIV prevention programs do not collect address information; consequently, this approach was not feasible.
2. Geographic distance. The service area is defined by the maximum distance from which persons served come to the service. Distance measures are relatively simple in terms of data collection and management. However, because service areas rarely correspond to circular areas described by distance measures, the resulting data can be of relatively poor quality. In some cases, distance measures are converted to administrative units that fall within the specified distance (e.g. all counties that are entirely or partially within a 50-mile radius).
3. Geopolitical boundaries. The service area is defined by naming the states, counties, cities, ZIP codes or other administrative units in which services are provided. These units are familiar to most persons and may already be used by respondents in planning and describing their activities. However, geopolitical units may not correspond to service areas that are defined in terms of neighborhoods, and they are sometimes imprecise, such as when a city boundary spans county lines [18].
Based on discussions among the project team and findings from the pilot test, the study team decided to collect service area data in the form of both geographic distance measures and geopolitical units. We did this because there was no clear precedent as to which method would provide the most useful information, and this would provide us with an opportunity to test both. The questionnaire provided respondents with a cascading set of geopolitical unit responses, from which they could list multiple responses at one or more levels of specificity, i.e., multiple counties or a county with additional cities. This list was, for the most part, geographically hierarchical. If, for example, the respondent checked the box for "entire state" and listed "Missouri," no other geographic area response for Missouri was necessary. Response options for distance included six choices ranging from less than 5 miles to more than 25 miles. Response options for service areas are shown in Table 2.
Table 2 Service area response options. Service area data were collected in the form of geopolitical units and distance measures. Geopolitical unit responses were listed in hierarchical order. A customized map accompanied each survey to increase accuracy and completeness of responses.
Geopolitical Description
• An entire state or territory, or multiple states or territories (Please list the states served:)
• An entire county or island, or multiple counties or islands, but an area smaller than an entire state or territory (Please list the counties served:)
• An entire city/town or multiple cities/towns, but an area smaller than an entire county (Please list the cities and town served:)
• A ZIP code or multiple ZIP codes, i.e. an area smaller than an entire city/town (Please list the ZIP codes served:)
• Tribal lands (Please list the tribal lands served:)
Distance Specification
• < 5 miles (specify)
• 5-10 miles
• 10-15 miles
• 15-20 miles
• 20-25 miles
• > 25 miles (specify)
We defined "service area" as the location of persons actually served. In some of the health services literature, this is referred to as "market area" [19]. This may differ from the target area, for which services were planned. For CDC-funded services, the concept of "service area" provided more useful information. The question was phrased in terms of where persons served live, although we did not ask respondents to consult actual address records when choosing their response. For street and community outreach activities, we instructed respondents to describe the area in which the intervention took place because these activities may be directed at transient populations or persons who congregate in a specific area without necessarily living there.
Our instructions to respondents about service area were to specify it as "the area where the majority (roughly 80%) of people receiving this prevention program live," or, for street and community outreach, "where the majority of activities took place." This wording was intended to avoid responses that were skewed toward large service areas by a small number of service users or activities outside the usual service area. In the pilot test, the study team found that this wording elicited responses that more closely represented actual activities.
Reference map
Each survey package included a custom-made one-page color reference map created for that CBO and generated by an automated GIS routine. The map showed two views of the area surrounding the CBO's location: one identifying cities, counties, and major roads within a 30-mile radius; the other showing a more detailed view of ZIP codes and towns within a 5-mile radius. In both views, we plotted concentric circles at set distances to provide a spatial frame of reference. We based our decision to include these maps on the San Diego pilot test, in which respondents completed service area questions twice: first without a reference map, and then with it. Using a reference map improved data quality in several ways:
1. Completeness. Respondents named more cities served when looking at a map that included names of all cities in the county.
2. Accuracy. Estimates of distance from the CBO location were more accurate when respondents consulted a map showing distance in 5-mile increments.
3. Precision. Respondents described service areas in terms of specific ZIP codes within the city rather than the entire city when using a map showing ZIP code boundaries.
Results
Surveys were mailed in July 2000. The initial universe was 1,562 CBOs. A number of CBO records in the database were later identified as duplicates or ineligibles (e.g., a CBO that did not provide HIV prevention services in fiscal year 2000), with a resulting survey population of 1,450 CBOs. With follow-up measures such as postcard reminders and callbacks, the survey had an overall response rate of 70.3 percent. In other words, 1,020 of 1,450 CBOs responded to the survey. The number of HIV prevention programs administered by each of these CBOs ranged from 1 to 23. Of the 1,020 CBOs, 432 reported having only one CDC-funded prevention program, but the majority of responding CBOs had more than one. In all, information about 3,028 prevention programs was provided by the survey. We maintained all survey data, actions, and responses in a Microsoft Access control system that was designed specifically for this project.
Figure 1 shows the location of each of the 1450 CBOs in the survey population and their response status. Triangles represent CBOs that did not respond. Particularly notable are the number of non-responses in Illinois and Montana. Montana CBOs were identified late in the data collection process and may not have had enough time to return surveys before that phase of the project ended. In Illinois, the State Health Department acted as an intermediary for the survey and the lack of direct contact for follow-up is likely to have reduced response rates. In this map, non-responses are drawn over responses, which accounts for the pattern present in many of the northeastern cities.
Figure 1 CBO response to HIV prevention service area survey. This map shows the location of all CBOs that received CDC funding for HIV prevention services in 2000. CBOs that responded are shown in red; green triangles indicate a non-response.
Response rates varied substantially among states, as shown in Figure 2. In the majority of states, 60 to 80 percent of CBOs responded. Higher response rates occurred in some of the Plains states, Utah, the Ohio Valley region, and pockets of the southeastern and northeastern U.S. Eight states/territories had response rates less than or equal to 50%. Response rates are particularly unstable for areas with few CBOs, where responses from just one or two CBOs dramatically influenced the response rate.
Figure 2 Response rates by state. This map shows the CBO response rate, by state and/or territory. Darker shades indicate higher response rates. White indicates no response.
We assigned geographic Federal Information Processing Standard (FIPS) codes to service area responses. FIPS codes were developed by federal government agencies to standardize coding for states, counties and other legal and statistical geographic entities. We used the FIPS codes to link the survey data to GIS maps.
Coded surveys were processed by professional data entry staff. Data entry staff wrote a data entry program specifically for this project that included verification, cleaning, and other quality control measures. All data were double-entered and verified. The results of the data entry process were two large text files, one that contained more general CBO information and one that contained all of the HIV prevention program survey responses.
Database design
We converted the text files from the data entry process to a series of 10 Microsoft Access 2000 tables. These 10 tables were developed to normalize the data, (i.e., group them into tables in a formalized procedure to eliminate duplication of information and provide flexibility in table structure for future additions or changes) and to allow linkage to GIS map files via GIS software. Full details of the database design are described in a separate report to CDC and are beyond the scope of this paper [20]. We discuss three important tables in the database, however. These are shown in the Figure 3 schematic, which uses a fictitious CBO. Field names have been changed for readability.
Figure 3 Database tables and their linkages. This diagram shows how the three main tables in the HIV Prevention Services Database were linked, or related. CBOs are linked to the GEOGAREA and PROGRAM tables by CBO-ID. The GEOGAREA table contains one record per geographic unit per CBO. This table contains the FIPS codes necessary for linkage to a GIS (map) database.
The first table, CBO, contains a master list of CBOs and includes the following information: CBO identifier; name, address and contact person; and information about survey responses. This information was used for survey administration and geocoding.
The FIPS codes for all geographic service area entities (i.e. state, county, city/town, ZIP code and/or Indian Reservation) were stored in a second table, GEOGAREA. This table contained four fields: 1) the CBO identifier, 2) the program identifier (many CBOs had multiple programs), 3) the FIPS type (e.g. state, county, city, ZIP, tribal lands) and 4) the actual FIPS codes. Each geographic unit that represented a portion or all of a service area for a particular program was stored as a single record. In the Figure 3 example, the Fictitious CBO was assigned an identifier of 15015. The service area of this CBO's Program #1 covered three zip codes. Data in the FIPS_TYPE field indicate which GIS base map (i.e., state, county, city, ZIP code, or reservation) to link to. For example, a FIPS_TYPE of 4 indicates that the linkage is to the national ZIP code map layer. The values in the FIPS_CODE field are actual ZIP codes, which can be queried and displayed with the GIS.
A third important table, PROGRAM contains all of the non-geographic information for each program, i.e. intervention type, risk populations, race/ethnicity and funding source. This table is linked, via a combination CBO/program identifier, to the geographic tables. This table also stored values provided for the last survey question, in which respondents were asked to indicate the distance within which the majority of people served lived.
Geospatial data development
We used a suite of Environmental Systems Research Institute, Inc. (ESRI, Redlands, CA) GIS software products for all spatial data processing and analyses, including ArcGIS 8.12, ArcMap and ArcCatalog. However, the final product of this research – a dynamic spatially-enabled database to be used by CDC program managers – was set up for use in ArcView 8.
We integrated the survey data in the Access database with a series of GIS map layers for subsequent mapping and analysis. These included U.S. states and territories, counties, cities and towns, American Indian reservations, and ZIP code area boundaries. These GIS map layers were derived from two sources: 1) generalized U.S. Census Topologically Integrated Geographic Encoding and Referencing (TIGER) 2000 Arc/Info export files, obtained from the U.S. Census web site; and 2) the ESRI Data and Map series, Version 8.1, which came bundled with ESRI software.
We made some enhancements to the original map layers to incorporate all survey responses. The ZIP code area boundary layer includes some small buffer polygons of ZIP code points that were added for this project. A number of the cities and towns that were identified by the survey participants did not exist in the city/town map layer, so we augmented the places map layer with places found in the online U.S. Geological Survey (USGS) Geographic Names Information System(GNIS). Lastly, a special areas layer was created manually from other background data sets for a few areas specified by survey participants that did not match any of the other background layers.
All responses about geographic services areas were matched to one or more of the geographic boundary files described above. A different procedure was used to develop map layers of CBO and program locations. The CBO and PROGRAM tables in the Access database contain addresses for CBOs and their programs. These addresses were used to derive the CBO and program point locations in geographic coordinates (i.e., latitude and longitude), such as those displayed in Figure 1. CBO and program addresses were address matched by a vendor. Response codes were linked to the address matched CBO data, so response status (i.e., whether the CBO responded to survey or not) of each CBO could be queried and mapped.
The GEOGAREA table contains information about all geographic entities that were indicated, by respondents, to be part of a geographic service area. Responses about geographic distance were stored in the PROGRAM table and linked to the map layer of program locations. The ArcView Buffer Wizard was used to buffer each program point by the corresponding distance estimate to create a new map layer showing service areas based on distance.
The primary result of this project is the HIV Prevention Services Database, a dynamic, spatially-enabled database that provides CDC with a wealth of information about HIV prevention services that it funds, and a large potential for geographic modeling, analyses, and mapping. This database handles overlapping geographies, risk populations and prevention services. In order to make it user-friendly for CDC program managers, we provided CDC with an ArcView (.mxd) application that automatically loads all of the spatial data (i.e. shapefiles) and tables needed for analysis and mapping. Relationships among tables (i.e. "joins") needed for query and analysis are also maintained in this ArcView application. Due to the wide range of potential database queries, we developed a Visual Basic for Applications (VBA) query tool that makes it easy for users to structure a query based on intervention type, race/ethnicity and risk population. With this tool, the user has the option of mapping CBO and program locations, plus their geographic service areas. The query tool interface is shown in Figure 4.
Figure 4 Query tool interface. This tool allows users to structure a query based on intervention type, race/ethnicity and risk population, then map corresponding CBO and program locations and/or geographic service areas.
The HIV Prevention Services Database presents a wide range of query and display capabilities, based on responses to the survey. For example, Figure 5 shows the geographic service areas of all programs that provide interventions to Hispanic/Latino populations. While the map is national in scale, the zoom and query functions in a GIS allow users to examine geographic areas at any scale. This query was based on program responses to questions about intervention type and populations served.
Figure 5 HIV prevention services to Hispanics or Latinos. This map is the result of a query to the HIV Prevention Services Database. It shows all areas where HIV prevention services are provided to Hispanics/Latinos. Service areas are drawn in pink. Program locations are represented by red dots. Green triangles represent the locations of CBOs that did not respond to the survey.
Queries can be based on geography as well. The question, "Which HIV prevention services are being provided by CDC-funded CBOs in the state of Rhode Island" would produce a map of Rhode Island CBO locations and their service areas and a wide range of information about types of service and risk populations in connected database tables.
The HIV Prevention Services Database is being used by CDC researchers. One analysis has focused on the geographic distribution of services at the national level and another on services to specific populations, such as African Americans.
Discussion
We successfully developed a national geographic database of CDC-funded HIV prevention services; however, we did encounter several challenges during the data development phase of the project. Some of these were related to data quality and integrity issues. We describe some of these challenges because they are likely to be encountered in other efforts to develop spatially-enabled data.
Validity of statewide service areas
Many CBOs indicated that they provided prevention services to an entire state. In many cases, this did not seem feasible, and concerns were raised about the integrity of these responses. We developed a set of procedures for confirming the validity of state responses that included 1) using the distance values in the last survey question for validation; 2) a consideration of the size of the state (statewide coverage in Rhode Island is more feasible than that of Texas, for example); 3) examining the type of intervention (e.g. prevention case management vs. health communications;) and 4) telephone callbacks to CBO program administrators by CDC staff.
Nonexistent geographic entities
In some cases, geographic entities provided by survey respondents simply could not be located in a geospatial database or even an atlas or gazetteer. The most common of these were the ZIP codes. Some CBOs provided ZIP codes that did not exist in the United States Post Office database. Thus, for some CBOs, service area data are missing or incomplete.
Polygon data not available for some zip codes
Some of the ZIP codes identified by survey respondents did not exist in the ZIP code polygon (area) GIS map layer, but did exist in another GIS layer of points only (i.e., represented by a singlelatitude/longitude coordinate). We made the assumption that these "point only" ZIP codes represented very small ZIP code areas. These ZIP codes were given "area" coverage through the creation of 0.1-mile buffers around their representative points.
Miscoding of geographic entities by survey data processors
We used a series of GIS queries and logical consistency checks to identify data anomalies. Each time an inconsistency was noted, we examined the original surveys. In a handful of cases, the coders had misinterpreted the respondent's handwriting, and we made corrections.
In spite of some of the challenges we encountered, we successfully developed methods to obtain primary information about CDC-funded HIV prevention services in the U.S. and its territories and were able to develop a dynamic GIS database of CBO locations, service areas and prevention services that is being used by CDC staff to perform analysis and make program decisions. This database, the HIV Prevention Services Database, was delivered to CDC in May 2002.
We need to caution, however, that the response rate to the survey was 70%. While this is a high response rate, we realize that program and service area data are missing for 430 CBOs. Any comprehensive analysis of service provision must take into account the locations of these non-responding CBOs. Additionally, CBO-provided services funded by CDC by no means make up the total of HIV prevention services in the U.S.
The CDC HIV Prevention Services Database was developed to enable CDC researchers to plan and evaluate CBO-provided HIV prevention services. While CDC is using this database primarily to identify gaps and overlaps in service, its potential is broad and includes the following applications:
• geographic analyses of HIV-prevention services to racial and ethnic minorities,
• an examination of HIV prevention services in the context of health disparities as a follow-up to work by Krieger et al. [21],
• use by local health agencies to determine how to integrate CDC-funded services with other community prevention services,
• regional studies of HIV prevention services for areas such as Appalachian Regional Commission counties, and
• analysis of CDC-funding levels vs. assessment of need, based on HIV/AIDS rates and risk populations.
Conclusion
The development of the HIV Prevention Services Database is a step in the right direction in terms of meeting Healthy People 2010 Objective 23-3 and, ultimately, in the development of a spatial decision support system. We have demonstrated the feasibility of constructing such a valuable, spatially-enabled database nationwide and have successfully transferred the data and technology to CDC for internal use. In terms of defining geographic service areas, we feel that information about geopolitical/administrative units was more useful than distance information although the distance information provided us with a means of conducting logical consistency checks on political/administrative unit responses.
The program data collected for this project were for prevention services provided during fiscal year 2000. We strongly recommend that the HIV Prevention Services Database be updated and maintained on a regular basis. Because of the expense of conducting such a large mail survey, we recommend that future data collection efforts use Web-based survey methodologies that incorporate interactive maps for the delineation of survey areas. These methodologies are being used increasingly in health, social sciences, and educational research [22].
List of abbreviations
AIDS: acquired immunodeficiency syndrome
CBO: community-based organization
CDC: Centers for Disease Control and Prevention
DHHS: Department of Health and Human Service
FIPS: federal information processing standard
GIS: geographic information system
HIV: human immunodeficiency virus
SDSS: spatial decision support system
VBA: Visual Basic for Applications
Authors' contributions
CLH and DAG are responsible for the study design. They managed the GIS and survey components of the study, respectively, executed the project, and wrote the final manuscript. This project was the "brain child" of AG. AG and KJF served as CDC Technical Monitors for the duration of the project and directed the study design and implementation. MB developed the GIS database and contributed to the more technical aspects of the final manuscript.
Notes
1. The HIV Prevention Services Database is in the public domain and is available at no cost for instruction or research purposes.
2. At the time of this writing, ArcGIS 9 was in use.
Acknowledgements
This project was made possible by funding from the Centers for Disease Control and Prevention, Office of Program Planning and Evaluation. DHHS Contract No. 282-98-0022, Task 10. Justine Allpress, of RTI International, produced Figures 1, 2, and 5.
==== Refs
Joseph AE Phillips DR Accessibility and Utilization: Geographical Perspectives on Health Care 1984 New York: Harper & Row
Rickets TC Savitz LA Gesler WM Osborne DM Eds Geographic Methods for Health Services Research: A Focus on the Rural – Urban Continuum 1994 Lanham, Maryland: University Press of America
Clarke KC Getting Started with Geographic Information Systems 1999 Upper Saddle River, NJ: Prentice-Hall
Vine MF Degnan D Hanchette C Geographic information systems: their use in environmental epidemiologic research Environ Health Perspect 1997 105 598 605 9288494
Cromley EK McLafferty SL GIS and Public Health 2002 New York: Guilford Press
Lee CV Irving JL Sources of spatial data for community health planning J Public Health Manag Pract 1999 4 7 22 10538418
Hanchette CL O'Carroll PW, Yasnoff WA, Ward ME, Ripp LH, Martin EL Geographic information systems Public Health Informatics and Information Systems 2002 New York: Springer-Verlag Publishers 431 466
Centers for Disease Control and Prevention (CDC) Utilizing new technologies to provide credible health information
Department of Health and Human Services (DHHS) Healthy People 2010: Understanding and Improving Health 2000 2 Washington, DC: U.S. Government Printing Office
Luo W Using a GIS-based floating catchment method to assess areas with shortage of physicians Health Place 2004 10 1 11 14637284 10.1016/S1353-8292(02)00067-9
McLafferty S Grady S Prenatal care need and access: a GIS analysis J Med Syst 2004 28 321 333 15446621 10.1023/B:JOMS.0000032848.76032.28
Phillips RL Kinman EL Schnitzer PG Lindbloom EJ Ewigman B Using geographic information systems to understand health care access Arch Fam Med 2000 9 971 978 11115195 10.1001/archfami.9.10.971
Wennberg JE Cooper MM (Eds.) The Dartmouth Atlas of Health Care 1998 1998 Chicago: American Hospital Publishing, Inc
Dartmouth Medical School Center for the Evaluative Clinical Sciences The Dartmouth Atlas Project
Fulcher C Kaukinen C Mapping and visualizing the location of HIV service providers: an exploratory spatial analysis of Toronto neighborhoods AIDS Care 2005 17 386 396 15832887 10.1080/09540120512331314312
McLafferty SL GIS and health care Annu Rev Public Health 2003 24 25 42 12668754 10.1146/annurev.publhealth.24.012902.141012
Centers for Disease Control and Prevention (CDC) Evaluating CDC-funded Health Department HIV Prevention Programs Guidance Atlanta 2001 1
Simpson K DesHarnais S Jacobs A Menapace A Rickets TC, Savitz LA, Gesler WM, Osborne DN Methods for defining medical service areas Geographic Methods for Health Services Research: A Focus on the Rural – Urban Continuum 1994 Lanham, MD: University Press of America 207 232
Rushton G Methods to evaluate geographic access to health services J Public Health Manag Pract 1999 5 93 100 10537836
Centers for Disease Control and Prevention (CDC) Geoanalysis of HIV Prevention Services Provided by CDC-Funded Community-Based Organizations (CBOs): Final Report DHHS Contract Number 282-98-0022 Atlanta 2002
Krieger N Waterman PD Chen JT Soobader MJ Subramanian SV Monitoring socioeconomic inequalities in sexually transmitted infections, tuberculosis, and violence: geocoding and choice of area-based socioeconomic measures – the public health disparities geocoding project (US) Public Health Rep 2003 118 240 60 12766219
McLafferty SL Clifford NJ, Valentine G Conducting questionnaire surveys Key Methods in Geography 2003 London: Sage Publications 87 100
|
16277661
|
PMC1291387
|
CC BY
|
2021-01-04 16:39:06
|
no
|
Int J Health Geogr. 2005 Nov 8; 4:28
|
utf-8
|
Int J Health Geogr
| 2,005 |
10.1186/1476-072X-4-28
|
oa_comm
|
==== Front
Immun AgeingImmunity & ageing : I & A1742-4933BioMed Central London 1742-4933-2-131626908010.1186/1742-4933-2-13ResearchHuman leukocyte antigen class I, class II, and tumor necrosis factor-alpha polymorphisms in a healthy elder Mexican Mestizo population Soto-Vega Elena [email protected] Yvonne [email protected] Luis [email protected] Department of Immunology and Rheumatology, Instituto Nacional de Ciencias Médicas y Nutrición Salvador Zubirán. Mexico City, Mexico2005 3 11 2005 2 13 13 16 8 2005 3 11 2005 Copyright © 2005 Soto-Vega et al; licensee BioMed Central Ltd.2005Soto-Vega et al; licensee BioMed Central Ltd.This is an Open Access article distributed under the terms of the Creative Commons Attribution License (), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
Background
There is strong evidence that an individual's genetic background is an important predisposing factor to longevity. In the present study we analysed the frequency of HLA class I, class II, as well as the TNF-α -308 polymorphism that may be related to an increased life span in Mexican Mestizo healthy elders.
Results
HLA typing was performed by polymerase chain reaction sequence specific oligonucleotide (PCR SSO) reverse dot blot. The TNF-α -308 polymorphism was assessed by PCR restriction fragment length polymorphism. A significant increased frequency of HLA-DRB1*11 was found in elderly women whereas this allele was not present in elderly males. The TNF2 allele was also increased in the elder group when compared to young controls. The frequencies of the remaining alleles tested were not statistically different among groups.
Conclusion
These data suggest an ethnicity independent tendency of HLA-DRB1*11 in elder females to increase life span and a possible role of the TNF2 allele with the successful remodelling of senescent immune system.
==== Body
Background
Ageing is associated with a complex remodelling of the immune system often in the direction of apparently decreased immune competence. However, it has been proposed that longevity is related to optimal function of the immune system because some genetic determinants for successful ageing might reside in those polymorphisms of genes that regulate immune responses e.g. major histocompatibility complex (MHC) [1,2]. The studies of life span in humans have resulted in different, and even controversial, associations of HLA class I, class II and class III genes with old age. Thus, longevity has been shown to be related with the selection of HLA alleles or haplotypes being the more frequently associated the A1 B8 Cw7 DR3 in Caucasians [3,4], the increased or decreased frequency of HLA-B40 [5-7], the augmented frequency of HLA-DR11 in elder women in Caucasian as well as in Japanese populations [8,9] among other alleles. Immunogenetic studies in aged, healthy Mexican Mestizo population have not been yet performed.
Mexican Mestizo individuals have a proportion of 56% Native American Indian genes, 40% Caucasian genes, and 4% African genes [10]. Mestizo population represents a complex mixture of European and American native inhabitants, and constitute the core of the Mexican population. This complex genetic process began in the 16th century and has been expanded during the course of time and continues to be a dynamic process, e.g., African population inhabited the Americas almost 200 years later. On the other hand, aged people in the Mexican population have increased from a life expectancy of 63 years in the 1970's up to 70 years nowadays.
Healthy elderly people show a pro-inflammatory phenotype with increased levels of cytokines such as TNF-α [11]. The TNF-α gene is located within the HLA class III region. This cytokine is involved in tissue remodelling, epithelial cell barrier permeability, macrophage activation, recruitment of inflammatory infiltrate, and up-regulation of adhesion molecules, among other functions. TNF-α also determines the strength, effectiveness and duration of local and systemic inflammatory responses [12].
TNF-α promoter contains numerous polymorphic sites, which are possible targets for transcription factors. The polymorphism at position -308 is defined by de substitution of a G by an A, where the presence of G defines the common allele TNF1 and A defines the TNF2 variant, which is less frequent [13]. Some studies have shown that individuals bearing the TNF2 polymorphism are higher TNF-α producers than those who bear the TNF1 variant [14].
The aim of this study was to evaluate the class I, class II HLA genotypes as well as the TNF-α -308 polymorphism with successful ageing in Mexican Mestizo.
Results and Discussion
Of the 71 elders, 38 (53.5%) were females and 33 (46.5%) males. Results of the observed and expected antigen and genotype frequencies for HLA-B, -DRB1 and -DQB1 loci were consistent with those predicted by the Hardy-Weinberg equilibrium. It is worth mentioning that the frequencies identified in our population are similar from those reported in previous studies performed in Mexican Mestizo populations [10].
Gene frequencies at the HLA-DRB1 locus were not statistically significant different among groups. The most frequent allele in both, young subjects and elders, was the HLA-DRB1*04. The high resolution typing revealed that the more frequent alleles in the elderly group were HLA-DRB1*0802, DRB1*0701, DRB1*0404 and DRB1*0407, whereas HLA-DRB1*0802, DRB1*0701, DRB1*1101 and DRB1*0404 were the more frequent in the young group, being these alleles not statistically significant.
When the differences between aged males and females were analysed, a decreased frequency of HLA-DRB1*11 in males was found when compared to females, for no one in the former group was HLA-DRB1*11 (p = 0.002; OR = 0.18; CI 95% = 0.06–0.53; p = 0.024 by Bonferroni's correction). On the other hand, when elderly males were compared to young males for HLA-DRB1*11, a diminished frequency was found although not statistically significant when Bonferroni's correction was applied. The high resolution typing showed a decreased frequency of the HLA-DRB1*1101 allele in elderly females when compared to the other HLA-DRB1*11 alleles from young women and, again, this difference was not significant by Bonferroni's correction.
Our data must be balanced against the small number of subjects studied. However, we found that, interestingly, no old man was HLA-DRB1*11, which may suggest that survival being gender-dependent [15,16]. This finding for Mexican Mestizo women is in agreement with previous reports in different populations, which might indicate that the contribution of this allele to an increase in the life span is independent of ethnicity [8].
Concerning the HLA-B and HLA-DQB1 alleles, no differences were observed among elders and young controls nor when compared by gender. The high resolution typing of HLA-DQB1 showed that in the elderly group the most frequent subtypes were HLA-DQB1*0302, DQB1*0402 and DQB1*0501 whereas in the young group were the HLA-DQB1*0302, DQB1*0301, DQB1*0402 and DQB1*0201. Finally, the generic typing of the HLA-B locus showed that the more frequent alleles in both groups were HLA-B*35 and HLA -B*15. A decreased tendency in the HLA-B*14 frequency in the elder group was also observed when compared to that of the control group, although this was not statistically significant.
The analysis of the TNF-α -308 promoter polymorphism showed a diminished frequency of the TNF1 allele accompanied by an increased one in the TNF2 allele in elders. The heterozygote genotype TNF1/2 was more common in the elderly group (9.02%) than in the control group (1.5%) (Table 1). Serum levels of TNF-α showed a slight increase in the elderly group although this difference was not statistically significant (data not shown). Interestingly, it was found that 58% of the TNF2 individuals were HLA-DR3 and HLA-DR4. The -308A TNF-α polymorphism has been shown to vary between individuals. These variants have been associated with certain HLA-DR alleles, e.g., HLA-DR2 subjects are considered low TNF- producers while those bearing HLA-DR3 and HLA-DR4 alleles produce high levels of this cytokine [13,14,17-19]
Table 1 TNF-α -308 polymorphism in elderly and young adults.
Healthy elders (n = 72) Young controls (n = 198) Statistics
gf (n = 144) gf (n = 110) P OR CI 95%
TNF1 0.909 0.985 NS - -
TNF2 0.090 0.015 0.001 0.33 0.14–0.73
gf = gene frequency; OR = Odds ratio; CI = Confidence interval; NS: non significant.
The TNF2 genotype has been related to an increased cytokine transcription rate [20]. Furthermore, it has been proposed that individuals who are heterozygous for this polymorphism possess an optimal inflammatory response that protects them against age-related neurodegeneration [21]. Although the inflammatory process has been related to chronic illnesses, incapacity, and death, it is interesting that in healthy old people this pro-inflammatory phenotype may be involved in the remodeling of the cytokine network which contributes to the successful ageing process [22]. Notwithstanding, lack of such association has also been reported [23,24].
Conclusion
Longevity studies must ideally include healthy centenarians, as they represent the extreme of human old age. Most studies carried out on centenarians come from either Europe or Asiatic countries such as Japan. It is precisely in these regions that the largest numbers of centenarians can be found. This is closely related on the one hand, with the social, cultural, and economic conditions of developed countries and, on the other, with biological factors including the long history of their ethnic groups, which has permitted a natural selection of genes favouring longevity [25]. By contrast, the ethnic group known as Mexican Mestizo is, in evolutionary terms, a recent one barely 500 years old. This lapse is certainly not enough to establish an allele selection or, even, modify completely the structure of previous haplotypes by the genetic admixture process that could account to change certain HLA clusters of Mexican natives. On the other hand, infectious diseases imported by conquerors indeed, have influenced the selection of HLA repertoire in our region. The cultural and socio-economic conditions of our geographical zone (Mexico and Central America) have contributed to a marked increase in life expectancy, particularly over the last 50 years, even though this is still far removed from that of developed countries. It is worth mentioning that the mean age of our study group is at least 10 years greater than the current life expectancy for Mexico and, hence, they represent the oldest population that have enjoyed successful ageing within our country.
Certainly, a great diversity of genes and factors influence successful ageing. Single nucleotide polymorphisms located in cytokine gene promoters have been demonstrated to affect the binding of transcription factors and, hence, its gene expression. Genetic variants that determine an increased production of anti-inflammatory cytokines or a decreased one of pro-inflammatory cytokines have been associated with successful ageing, suggesting a role in the control of the inflammatory state in the attainment of increased life span [12,26,27].
Moreover, the study of polymorphic genes on the X chromosome do not have to be procrastinated for there are some – albeit relatively old – evidence that they are critical in the genetic regulation of the immune response and thus could be of utmost importance for conditioning the life span expectancy [28-30].
Methods
Subjects
A total of 71 healthy elders were studied, age ranged from 80 to 96 years (mean 86.2 years). The control samples were obtained from 99 young (from 21 – 54 years; mean 35.2 years) healthy individuals unrelated to elders. All subjects were Mexican Mestizo which is defined as an individual who was born in Mexico and is descendent from mixed racial ancestry of the native Americans of such region with individuals from Europe (mainly Spain) or Africa, all of them living in Mexico City. A complete social and medical history and physical examination was performed. Laboratory tests included complete blood cell count, erythrocyte sedimentation rate (Westergreen), immunoglobulin levels (IgG, IgA, IgM), serum electrolytes, glucose, creatinine, urea, alkaline phosphatase, aspartate transaminase, total bilirubin, and total cholesterol. All subjects were informed about the objectives and methods of the study. Work has been done according to current law.
DNA isolation
Genomic DNA was extracted from 5 mL of peripheral whole blood employing the Wizard genomic DNA purification kit (Promega, Madison, WI) according to manufacturer's instructions. Isolated DNA was quantified by spectrophotometry and adjusted to a concentration of 100 ng/μL, and stored at -70°C until use.
HLA Typing
Generic HLA-DRB1, DQB1, and HLA-B typing was performed by polymerase chain reaction sequence specific oligonucleotide (PCR-SSO) reverse dot blot using the Dynal RELI SSO system (Hoffman-La Roche Ltd. and Roche Molecular Systems, Inc., Alameda, CA) as described previously [31]. High resolution typing of HLA-DRB1 and DQB1 loci was done by sequence based typing method (SBT). The primers' sequences and PCR conditions for amplification of polymorphic exons were obtained from protocols of the 13th International Histocompatibility Workshop (Seattle, 2002).
TNF-α-308 polymorphism
Genotyping for the TNF-α -308 polymorphism was performed using a PCR fragment amplified using the forward primer 5'-AGG CAA TAG GTT TTG AGG GCC AT-3' and the reverse primer 5'-TCC TCC CTG CTC CGA TTC CG-3' to create a restriction site for the NcoI endonuclease (Fermentas, Hanover, MD) according to previous reports [32]. Digestion products were analysed by photo typing in 2% agarose gels and stained with ethidium bromide. Only 55 healthy young subjects were analysed for this polymorphism.
TNF-α serum levels
Serum from each subject was obtained from 5 mL whole blood and stored at -70°C until tested. An ELISA kit was used for the measurement of TNF-α levels according to manufacturer's instructions (R&D Systems, Minneapolis, MN).
Statistics
Allele frequencies were evaluated by gene count and 2 × 2 contingency tables. Statistical differences of allele frequencies between elders and controls were done employing chi square test and Yate's correction. Obtained P values were subjected to Bonferroni's correction. Odds ratio was calculated for healthy elders' carriers of specific alleles. Data were tested for the goodness of fit between the observed and expected genotype values and their fit to Hardy-Weinberg equilibrium.
List of Abbreviations
HLA Human Leukocyte Antigen
TNF Tumor Necrosis Factor
PCR Polymerase Chain Reaction
PCR SSO Polymerase Chain Reaction Sequence Specific Oligonucleotide
Acknowledgements
E. Soto-Vega was supported by a scholarship from the Consejo Nacional de Ciencia y Tecnología (CONACYT), Mexico. This work was part of the PhD degree thesis of E. Soto-Vega, Universidad Nacional Autónoma de México (UNAM).
Authors thank Dr. G. Vargas-Alarcón for kindly providing the genotypic data from healthy young controls and the Department of Geriatrics from our Institute for the elders' clinical evaluation.
==== Refs
Caruso C Candore G Colonna Romano G Lio D Bonafè M Valensin S Franceschi C Immunogenetics of longevity complex polymorphism relevant to the control of human longevity? A review of literature data Mech Ageing and Dev 2001 122 445 462 10.1016/S0047-6374(00)00255-4
Candore G Lio D Colonna Romano G Caruso C Pathogenesis of autoimmune diseases associated with 8.1 ancestral haplotype: effect of multiple gene interactions Autoimmunity Rev 2002 1 29 35 12849055 10.1016/S1568-9972(01)00004-0
Proust J Moulias R Fumeron F Bekkhoucha F Busson M Schimd M Hors J HLA and longevity Tissue Antigens 1982 19 168 173 7089955
Rea I Middleton D Is the phenotypic combination A1B8Cw7DR3 a marker for male longevity? J Am Geriatr Soc 1994 42 978 923 8064108
Bender BS Nagel JE Adler WH Andres R Absolute peripheral blood lymphocyte count and subsequent mortality of elderly men. The Baltimore longitudinal study of aging J Am Geriatr Soc 1986 34 649 654 3734313
Hansen HE Sparck JV Larsen SO An examination of HLA frequencies in three age groups Tissue Antigens 1977 10 49 55 70855
Lagaay AM D'Amaro J Ligthart GJ Schreuder GM van Rood JJ Hijmans W Longevity and heredity in humans. Association with human leukocyte antigen phenotype Ann NY Acad Sci 1991 621 78 89 1859103
Ivanova R Hénon N Lepage V Charron D Vicaut E Schächter F HLA-DR alleles display sex-dependent effects on survival and discriminate between individual and familial longevity Hum Mol Genetics 1998 2 187 194 10.1093/hmg/7.2.187
Takata H Suzuki M Ishii T Sekiguchi S Iri H Influence of major histocompatibility complex region genes on human longevity among Okinawan-Japanese centenarians and nonagenarians Lancet 1987 2 824 826 2889033 10.1016/S0140-6736(87)91015-4
Lisker R Pérez-Briceño R Granados J Babinsky V de Rubens J Armendares S Buentello L Gene frequencies and admixture estimates in Mexico City population Am J Phys Anthropol 1986 71 203 207 3099584
Franceschi C Monti D Sansoni P Cossarizza A The immunology of exceptional individuals: the lesson of centenarians Immunol Today 1995 16 12 16 7880382 10.1016/0167-5699(95)80064-6
Bidwell J Keen L Gallagher G Kimberly R Huizinga T McDermott MF Oksenberg J McNicholl J Pociot F Hardt C D'Alfonso Cytokine gene polymorphism in human disease: on-line databases Genes Immunity 1999 1 3 19 11197303 10.1038/sj.gene.6363645
Wilson AG Symons JA McDowell TL McDevitt HO Duff GW Effects of a polymorphism in the human tumor necrosis factor-α promoter on transcriptional activation Proc Natl Acad Sci USA 1997 94 3195 3199 9096369 10.1073/pnas.94.7.3195
Hajeer AH Hutchinson IV Influence of TNF alpha gene polymorphisms on TNF alpha production and disease Hum Immunol 2001 62 1191 1199 11704281 10.1016/S0198-8859(01)00322-6
Franceschi C Motta L Valensin S Rapisarda R Franzone A Berardelli M Motta M Monti D Bonafé M Ferruci L Deiana L Pes GM Carru C Desole MS Barbi C Sartoni G Gemelli C lescai F Oliviere F Marchegiani F Cardelli M Cavallone L Gueresi P Cossarizza A Trojano L Pini G Sansoni P Passeri G Lisa R Spazzafumo L Amadio L Iunta S Stecconi R Morresi R Viticchi C Mattace R De Benedictis G Baggio G Do men and women follow different trajectories to reach extreme longevity? Italian Multicenter Study on Centenarians (IMUSCE) Aging (Milano) 2000 12 77 84 10902049
Caruso C Candore G Colona Romano G Lio D Bonafé M Valensi S Franceschi C HLA, aging, and longevity: A critical reappraisal Hum Immunol 2000 61 942 949 11053639 10.1016/S0198-8859(00)00168-3
Jacob CO Fronek Z Lewis GD Koo M Hansen JA Mc Devitt HO Heritable major histocompatibility complex class II associated differences in production of tumor necrosis factor-α: relevance to genetic predisposition to systemic lupus erythematosus Proc Natl Acad Sci USA 1990 87 1233 1237 2105500
Bouma G Crusius JBA Poo MO Kolkman JJ von Bloomberg BME Kostense PJ Giphart MJ Schreuder GMT Meuwissen SGM Peña AS Secretion of tumor necrosis factor α and lymphotoxin α in the relation to polimorphisms in the TNF genes and HLA-DR alleles. Relevance for inflammatory bowel disease Scand J Immunol 1996 43 456 463 8668926 10.1046/j.1365-3083.1996.d01-65.x
Abraham LJ French NAH Dawkins RL Polymorphic MHC ancestral haplotypes affect the activity of tumor necrosis factor-alpha Clin Exp Immunol 1993 92 14 18 8096802
Fargion S Valenti L Dongiovanni P Fracanzani AL TNF alpha promoter polymorphisms Methods Mol Med 2004 98 47 58 15064432
McCusker SM Curran MD Dynan KB McCullagh CD Urquhart DD Middleton D Patterson CC McIlroy SP Passmore AP Association between polymorphism in regulatory region of gene encoding tumor necrosis factor alpha and risk of Alzheimer's disease and vascular dementia: a case-control study Lancet 2001 357 436 439 11273064 10.1016/S0140-6736(00)04008-3
Franceschi C Ottaviani E Stress, inflammation and natural immunity in the aging process: a new theory Aging (Milano) 1997 9 30 31 9358873
Ross OA Curran MD Rea IM Hyland P Duggan O Barnett CR Annett K Patterson C Barnett YA Middleton D HLA haplotypes and TNF polymorphism do not associated with longevity in the Irish Mech Ageing Dev 2003 124 563 567 12714268 10.1016/S0047-6374(03)00037-X
Wang XY Hurme M Jylha M Hervonen A Lack of association between human longevity and polymorphisms of IL-1 cluster, IL-6, IL-10 and TNF-alpha genes in Finnish nonagenarians Mech Ageing Dev 2001 123 29 38 11640949 10.1016/S0047-6374(01)00338-4
Price P Witt C Allcock R Sayer D Garlepp M Kok CC French M Mallal S Christiansen F The genetic basis for the association of the 8.1 ancestral haplotype (A1, B8, DR3) with multiple immunopathological diseases Immunol Rev 1999 167 257 274 10319267
Lio D Scola L Crivello A Colonna-Romano G Candore G Bonafe M Cavallone L Marchegiceni F Olivieri F Franceschi C Caruso C Inflammation, genetics, and longevity: further studies on the protective effects in men of IL-10 -1082 promoter SNP and its interaction with TNF-α -308 promoter SNP J Med Genet 2003 40 296 299 12676903 10.1136/jmg.40.4.296
Kroeger KM Carville KS Abraham LJ The -308 tumor necrosis factor-a promoter polymorphism effects transcription Mol Immunol 1997 34 391 399 9293772 10.1016/S0161-5890(97)00052-7
Zeicher M Mozes E Lonai P Lymphocytes alloantigens associated with X-chromosome linked immune response genes Proc Natl Acad Sci USA 1977 74 721 724 265535
Litwin SD Does the X chromosome have a special role in immune biology? Med Hypotheses 1982 9 313 323 6815439 10.1016/0306-9877(82)90161-X
Jansson L Holmdahl R Genes on the X chromosome affect development of collagen-induced arthritis in mice Clin Exp Immunol 1993 94 459 465 8252807
Bignon JD Fernandez-Viña MA Charon D Protocols of the 12th International histocompatibility Workshop for typing of HLA class II alleles by DNA amplification by polymerase chain reaction (PCR and hybridization with sequence specific oligonucleotide probes (SSOP) Genetic Diversity of HLA, functional and Medical implications 1997 I Paris: EDK
Wilson AG di Giovine FS Blakemore AI Duff GW Single base polymorphism in the human Tumor Necrosis Factor (TNF) alpha gene detectable by NcoI restriction of PCR product Hum Mol Gen 1992 1 353 358 1363876
|
16269080
|
PMC1291388
|
CC BY
|
2021-01-04 16:36:31
|
no
|
Immun Ageing. 2005 Nov 3; 2:13
|
utf-8
|
Immun Ageing
| 2,005 |
10.1186/1742-4933-2-13
|
oa_comm
|
==== Front
J Ethnobiol EthnomedJournal of Ethnobiology and Ethnomedicine1746-4269BioMed Central London 1746-4269-1-91627091110.1186/1746-4269-1-9ResearchKnowledge and use of medicinal plants by local specialists in an region of Atlantic Forest in the state of Pernambuco (Northeastern Brazil) Gazzaneo Luiz Rodrigo Saldanha [email protected] Lucena Reinaldo Farias Paiva [email protected] Albuquerque Ulysses Paulino [email protected] Departamento de Biologia, Área de Botânica, Laboratório de Etnobotânica Aplicada, Universidade Federal Rural de Pernambuco, Rua Dom Manoel de Medeiros s/n, Dois Irmãos, Recife, Pernambuco, 52171-030, Brazil2005 1 11 2005 1 9 9 1 9 2005 1 11 2005 Copyright © 2005 Gazzaneo et al; licensee BioMed Central Ltd.2005Gazzaneo et al; licensee BioMed Central Ltd.This is an Open Access article distributed under the terms of the Creative Commons Attribution License (), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
The study of local knowledge about natural resources is becoming increasingly important in defining strategies and actions for conservation or recuperation of residual forests. This study therefore sought to: collect information from local populations concerning the use of Atlantic Forest medicinal plants; verify the sources of medicinal plants used; determine the relative importance of the species surveyed, and; calculate the informant consensus factor in relation to medicinal plant use. Data was obtained using semi-structured forms to record the interviewee's personal information and topics related to the medicinal use of specific plants. The material collected represent 125 plants, distributed among 61 botanical families, with little participation of native plants. This study demonstrated that local people tend to agree with each other in terms of the plants used to treat blood-related problems, but cite a much more diverse group of plants to treat problems related to the respiratory and digestive systems – two important categories in studies undertaken in different parts of the world. The local medicinal flora is largely based on plants that are either cultivated or obtained from anthropogenic zones, possibly due to the use and access restrictions of the legally protected neighboring forest. Despite these restrictions, the species with the highest use-value by this community was Pithecellobium cochliocarpum (Gomez) Macb., a native plant of the Atlantic Forest.
ethnobotanyAtlantic Forestmedicinal plantstraditional knowledge
==== Body
Introduction
Fifteen percent of Brazil was once covered by Atlantic Forest, and although less than 5% of the original forest remains today [1], it is still one of the highest biodiversity areas on the planet [2]. This forest also demonstrates extremely high rates of endemism: up to 74.4% for bromeliads, 55% for trees, and 64% for palms [1,3], and we can infer that there is yet much to be studied and discovered within this ecosystem, especially in terms of its useful resources.
The Atlantic Forest ecosystem has been the victim of successive economic cycles since colonial times (brazilwood, sugarcane, mining, coffee, and cattle), especially in areas very near the coast. Today, large cities (50% of the Brazilian population) are located in areas once covered by these forests, and 80% of the GNP is produced there [1]. For these reasons, innumerous species have disappeared since Brazil's discovery, and continue to disappear without having been recorded; and with them, important information related to ecology, pharmacology, botany, and other fields, may vanish completely.
Rural communities are considered to be the most neglected area in terms of ethnobotanical studies [4]. Despite the availability of modern medicines, most agro-cultural communities (either by choice or for lack of economic resources) still use and detain an extensive pharmacopoeia of native plants [4].
The study of local knowledge about natural resources is becoming increasingly important in defining strategies and actions for conservation or recuperation of residual forests. This study therefore sought to: collect information from local populations concerning the use of Atlantic Forest medicinal plants; verify the sources of medicinal plants used; determine the relative importance of the species surveyed, and; calculate the informant consensus factor in relation to medicinal plant use.
Materials and methods
The study area and its inhabitants
This field study was undertaken with the aid of the "Três Ladeiras" community in the municipality of Igarassu, Pernambuco State, in northeastern Brazil. Igarassu is located in the mesoregion of the Recife metropolitan area (the state capital) and in the Itamaracá microregion (Figure 1). The municipal seat is located 26 km from the state capital [5] at 20 m above sea level (7° 50' 20"S and 35° 00' 10"W). The municipality has a total area of 304.2 km2, a population of 72,990 (219.9 inhabitants/km2), and it is 74.9% urbanized. The regional climate is humid, with a mean annual temperature of 27°C. Average annual rainfall is 2,000 mm, with a moderate water deficit in the summer [6].
Figure 1 Place of ethnobotanical data collection of the medicinal plants cited by the population in the municipality of Igarassu (Northeast Brazil).
The Rural Inhabitants
The community studied is part of the "Usina São José" sugarcane refinery, which employs most of the male population. The community is surrounded by Atlantic Forest fragments that belong to the refinery's ecologic reserve [7] (Fig. 1). The fragments occupy a total area of 210 hectares, with altitudes varying from 50 to 140 m. The number of people employed by the refinery oscillates during the year, increasing and decreasing according to the seasons and the periods of land preparation, planting, and harvesting. It is common to find families cultivating small plantations that serve as a nutritional and/or economic reserve between periods of employment activity. There is a 323.3 ha forest southeast of the refinery, which represents 0.80% of the total area of the municipality. The forest is part of the Botafogo River Basin source protection area, according to law N° 9860, dated August 12, 1986. The purpose of this reserve is to protect the landscape, soil, and the water basin [8].
The community views itself as a coherent group, and is formed basically by active or retired workers of the local sugar cane refineries. The women of the community carry out almost all of the activities related to caring for the family and their property. Their are two public schools and a small Peruvian cherry ("acerola") plantation that produces fruit pulp and provides job opportunities for youths during the harvest season. There is no basic sanitation, and a single health care center treats only simple health problems; patients requiring more intensive care are removed to hospitals in Igarassu.
Data Collection
Information on the use of medicinal plants was collected for one year (2003) with the help of rural dwellers of the "Três Ladeiras" community. During this period, door-to-door visits were made in order to attempt to identify local people with a specialized knowledge of medicinal plant use [9]. As such, the sampling was intentionally non-random [10], under the assumption that local specialists would provide more specific and higher quality information concerning medicinal plants. Data was obtained using semi-structured forms to record the interviewee's personal information and topics related to the medicinal use of specific plants. During the first contacts with the local population, a specialist was identified by the inhabitants themselves. A specialist is defined as "a person recognized by the community as having deep knowledge about the uses of native and/or introduced plants in manufacturing remedies and in promoting cures". Using the snowball method [11], names of other specialists were then obtained. Six local specialists were eventually interviewed (three men and three women), aged 51–102. All the people identified as specialists had been living in the community for at least 30 years and were frequently sough out by other community members for advice on the use of plants. The men had all been woodsmen or hunters, although these activities are currently prohibited. The women were housewives who had acquired a broad knowledge of medicinal plants by way of experimentation and trading information with relatives or people from the surrounding communities
Data Analysis
The first step employed in the data analysis calculating the informant consensus factor (ICF) [12]. ICF values will be low (near 0) if plants are chosen randomly, or if informants do not exchange information about their use. Values will be high (near 1) if there is a well-defined selection criterion in the community and/or if information is exchanged between informants.
The ICF is calculated as follows: number of use citations in each category (nur) minus the number of species used (nt), divided by the number of use citations in each category minus one:
All citations were placed into one of 14 categories: undefined pains or illnesses; skin and subcutaneous tissues; diseases of the endocrine glands, metabolism, and nutrition; blood and hematopoietic organs; skeletal, muscle, and connective tissues; infectious and parasite-related diseases; neoplasies; problems of the circulatory system; problems of the digestive system; problems of the genitourinary system; problems of the nervous system; problems of the respiratory system; problems of the sensorial system – ear; and problems of the sensorial system – eye.
The use value (adapted from the proposal of Phillips et al. [13]), a quantitative method that demonstrates the relative importance of species known locally, was also calculated:
UV = ΣU/n
where: UV = use value of a species; U = number of citations per species; n = number of informants
All of the material collected was processed, identified with the aid of specialists, and subsequently deposited in the PEUFR herbarium of the Biology Department of the Federal Rural University of Pernambuco. All material was collected with the help of local informants.
Results
Local specialists and medicinal plants
The plants collected represent 125 plants, distributed among 61 botanical families (e.g. see additional file 1). Lamiaceae (12 spp.) was best represented in terms of the number of species, followed by Asteraceae (6 spp.), Myrtaceae (5 spp.), Arecaceae (5 spp.), and Poaceae (5 spp.). When analyzing the number of citations for the plant parts used to prepare local remedies, a preference for the use of leaves becomes noticeable. The use of teas had the highest relative value (42%), followed by the use of syrups (20%), and in natura (16%).
Plants cited are listed by collection locality (anthropogenic zones or forest) and degree of management (cultivated or non-cultivated). It was observed that 54.5% of the species used are cultivated, while native and/or weed plants composed the non-cultivated category. In terms of where the plants were collected, 82.7% were from areas backyards and small farms, while only 17.3% were gathered from inside the forest. Most of the non-cultivated plants were weed, with little participation of native plants.
The species with the highest use-value was Pithecellobium cochliocarpum (Gomez) Macbr, popularly known as "barbatenon", with an UV of 1.8. Its main importance resides in its attributed healing effects on wounds. It is used externally, in the form of an alcoholic bark extract. Alpinia zerumbet (Pers.) Burt. ex R. M. Smith ("colônia") had the second highest UV (1.6). This species is frequently used as an ornamental plant in home gardens. Its main medicinal uses are to treat coughs (flower, leaf, and/or a root syrup), and headaches (a lightly heated leaf is applied to the forehead).
Two species shared the third highest use value (1.3): Hymenaea martiana Hayne ("jatobá") and Aeolanthus suaveolens Mart. ex Spreng. ("macassá"). H. martiana was cited for the treatment of blows to the body, inflammations, and rheumatism. It is principally prepared as an alcoholic extract of the fruit, using the local distilled drink ("cachaça") or wine. A. suaveolens is well known among the interviewees as being capable of treating general pains, especially earaches (for which it is applied in natura, as an earplug).
Cited for the treatment of inflammations, Schinus terebinthifolius Raddi ("aroeira"), Boerhavia diffusa L ("pega-pinto"), and Borreria verticillata G. F. W. Meyer ("vassoura de botão") all had a use value of 1.1. Protium heptaphyllum (Aubl.) March. ("amescla") is used to treat tooth and headaches, and Solanum panicultum L. ("jurubeba branca") to treat several ailments (such as indigestion, anemia, hepatitis, kidney, and spleen problems); both had use-values of 1.
Consensus factor among specialists
The highest ICF values were linked to problems related to the blood and to hematopoietic organs (1.0), and for problems of the sensorial system – ear (0.60). The use category with the lowest ICF value was "diseases of the endocrine glands, metabolism, and nutrition" (0.14). The ICF for infectious and parasite-related diseases was strikingly low (0.27) – the result of the use of a number of species that was almost as varied as the number of citations. A more detailed description of each category follows (see Table 1).
Table 1 Informant Consensus Factor by corporal systems categories or diseases.
Category Species (%) All Species Use citations (%) All use citations ICF
Undefined pains or illnesses 42 32.3 68 19.88 0.38
Skin and subcutaneous tissues 16 12.3 34 9.9 0.54
Diseases of the endocrine glands, metabolism and nutrition 7 5.3 8 2.3 0.14
Blood and hematopoietic organs 1 0.7 5 1.4 1.00
Skeletal, muscle and connective tissues 13 10.0 23 6.7 0.45
Infectious and parasite-related diseases 14 10.7 19 5.5 0.27
Neoplasies 1 0.7 1 0.2 0.00
Problems of the circulatory system 7 5.3 10 2.9 0.33
Problems of the digestive system 36 27.6 60 17.5 0.40
Problems of the genitourinary system 24 18.4 34 9.9 0.30
Problems of the nervous system 3 2.3 4 1.16 0.33
Problems of the respiratory system 38 29.2 65 19.0 0.42
Problems of the sensorial system (ear) 3 2.3 6 1.7 0.60
Problems of the sensorial system (eye) 3 2.3 5 1.4 0.50
Blood and hematopoietic organs
This category had the highest ICF value (1.0). There were five citations for a single species, demonstrating total agreement on selection criteria among the informants. The main problem treated within this category was hemorrhaging, often caused by accidents involving the perforating or cutting instruments used in sugarcane culture and in subsistence farming. Blood loss is diminished by applying a cloth soaked in a tincture made from the bark of Pithecellobium cochliocarpum (Gomez) Macbr ("barbatenon").
Problems of the sensorial system – ear
This group obtained the second highest ICF value (0.60). Of all of the plants cited by informants (127), 2.3% (three species) are used to treat medical problems within this group. Only one of the interviewees did not cite an ear-related disease. Earaches were cited by all of the specialists, and there was general agreement in the citations of Aeolanthus suaveolens Mart. ex Spreng. as the main remedy. There was also agreement on the use of Plectranthus sp. ("hortelã grande") and Ruta graveolens L ("arruda"). In all cases, the manner of usage was an earplug made from the leaves (in natura).
Skin and subcutaneous tissues
With a total of 34 citations (10% of the total), this category had an ICF of 0.54. Its main representative – as in the first category cited – was Pithecellobium cochliocarpum (Gomez) Macbr ("barbatenon"), a plant of great importance to the local population.
Problems of the sensorial system – eye
This is a heterogeneous group, with an ICF of 0.50. Of the five use-citations, three species are recommended for "tired eyesight" and conjunctivitis. The most cited phytotherapeutic resource was Ocimum basilicum L. ("manjericão"), with three citations. Recipes mention the use of an aqueous extract of this plant (obtained as a tea) as an eyewash.
Skeletal, muscle, and connective tissues
With an ICF of 0.45, the 23 citations of this group included 10% (13 species) of the total number of plants cited in this study. The most important representatives of this group were Hymenaea martiana Hayne ("jatobá"), with four citations, and Caesalpinia ferrea Mart. ex Tul. ("jucá"), with three citations. It is interesting to note that in both species the fruit is used, and it is prepared in a very similar fashion – wine is employed to obtain an alcoholic solution for internal use.
Problems of the respiratory system
This category had the second highest number of citations (65), only slightly less than the group of undefined pains and illnesses (68). The ICF of this category was 0.42. It is important to note that approximately 32% (42) of the total number of species cited were included within this group, demonstrating a great diversity in the knowledge of medicinal plants for the treatment of respiratory problems. The remedies used are mainly administered in the form of syrups to treat coughing, breathlessness, asthma, bronchitis, and the common cold – maladies that especially affect children, either due to the fragility of their immunological system, or to problems related to their immature respiratory system.
According to the data collected, sugar is often added to the syrup preparation – for in addition to its preservative properties (it increases the osmolality of the solution), its sweet taste masks the bitter and unpalatable taste of some of the herbs.
The documented use of a wide variety of species is in great part due to the practice of using innumerous herbs (sometimes over 15) in recipes for a single syrup, and by the belief that the greater the number of herbs (always in odd numbers) the better the syrup. Another reason for the large number of citations is the frequency of cases of breathlessness and asthmatic problems when the sugarcane fields surrounding the community are burned during the pre-harvest preparations.
Problems of the digestive system
This category includes all problems related to organs directly or indirectly linked to digestion, including the teeth and gums. The ICF value for this group was 0.40, considered low. However, if we consider only the citations for toothaches, the ICF value would be nearly 1.0, since there was great agreement on the use of Aeolanthus suaveolens Mart. ex Spreng. ("macassá") to treat this complaint.
Undefined pains or illnesses
This category includes all citations for undefined pains, and for diseases with unspecific symptoms. The ICF value was 0.38. This group had the highest number of use-citations (68 of a total of 342) as well as species used (42). This result is compatible with the category's diversity of problems.
Once again, Aeolanthus suaveolens Mart. ex Spreng. ("macassá") stands out, followed by "verga morta" (an undetermined species) and Protium heptaphillum (Aubl.) March. ("amescla"). All of these plants have analgesic uses. Pfaffia glomerata (Spreng.) Pederson ("acônito") had the highest number of citations among plants used to treat fevers. Ruta graveolens L. ("arruda") stood out for its importance as an anti-inflammatory agent.
Problems of the circulatory system
Heart diseases and blood pressure are examples of problems within this category. A total of seven species were cited, with 10 uses. The ICF for this category was low (0.33). Strokes received great attention from the interviewees, totaling 60% of the citations. Argemone mexicana L. ("cardo santo") was the species most cited.
Problems of the nervous system
An ICF of 0.33 was observed for this category. Fainting and insomnia were the most frequent citations. Half of the citations referred to the calming effects attributed to a tea made from orange tree flowers or leaves (Citrus sinensis (L.) Burmann), which is also frequently used to treat insomnia.
Problems of the genitourinary system
With 34 citations, this category comprises 24 plant species, with an ICF of 0.30. Women in the community commonly use sitting baths made with tea from the bark of Boerhavia diffusa L ("pega-pinto") or Pithecellobium cochliocarpum (Gomez) Macbr ("barbatenon").
Infectious and parasite-related diseases
The ICF for this category was 0.27. The common cold was frequently cited, followed by worms and hepatitis. A total of 19 disease citations were recorded, and 14 plants species were indicated as cures for these diseases, or as treatments for their symptoms.
Diseases of the endocrine glands, metabolism, and nutrition
The most cited diseases were diabetes (metabolism) and anemia (nutrition). Seven plants were cited for eight diseases, and the ICF was correspondingly very low (0.14).
Neoplasies
As only a single citation was obtained (for cancer) and therefore this category's ICF appears as 0, as this index requires citations from at least two informants.
Discussion
There seems to be a tendency for a few plant families to stand out in any pharmacopoeia. In a study by Almeida & Albuquerque [14], the family Lamiaceae was classified as the richest in species citations. The families Lamiaceae and Asteraceae stood out in the studies of Bennett & Prance [15], for together they represented 21% of the 216 species surveyed. Hanazaki et al. [2] likewise found that these two families had the greatest number of species cited by fishing communities and in other coastal areas of Atlantic Forest in São Paulo State [16]. These plants are normally herbaceous species that can either be cultivated or occur as weeds. Preference for their use may be related as much to their ready availability, for they are common in disturbed areas [17], as to factors related to their biological activity. Based on evidence and availability theory, Stepp & Moerman [18] suggest that these plants concentrate very active biological compounds as a function of their habit or of their life strategies. Much evidence has accumulated indicating that chemical and ecological factors orient the selection and use of medicinal plants in local communities in all parts of the world [cf. [19,20]].
Another interesting factor is that the cited species generally are indicated to the same illnesses in other localities [2,3,14,17,23,25,28,29]. Moreover, observing the lists of species of other studies, it can be evidenced that the most important species vary of region for region [14,20,23,27], although to verify a set of species, in special the cultivated ones, common in several works, as: Plectranthus spp., Mentha spp. and Ocimum spp.
Another relevant observation relates to the tendency of communities living near humid forests to use plant leaves. In a study undertaken in Barra do Piraí (state of Rio de Janeiro) with merchants and traditional medicine users, Parente & Rosa [21] observed that leaves were the most used category of plant parts. In this study, leaves were used in 42% of the time – over twice the number of citations for the second most used part (the root, at 18%). Similar observations had already been recorded for other communities near forested areas, where vegetation is always green and leaves are abundant [16,22]. On the other hand, communities in dry regions tend to focus their attention on plant parts that are continuously available, such as the bark [23,14]. As the plants in these localities are regularly exposed to long periods of drought and lose their leaves, bark and roots are more often used. This observed difference in usage of plant parts in different areas should be more closely investigated. Zschocke et al. [24] suggested, for example, that the use of bark is much more common in tropical regions than in temperate zones. However, as noted above, even in tropical regions the use of bark seems to be most strongly associated with arid and semi-arid zones, reflecting the resources most readily available to these communities. Additionally, the preference towards leaves places more emphasis on the preparation of medicinal teas. Parente & Rosa [21] also found high percentages for teas (51%), baths (39%), and other forms (liquid extracts, infusions, and in natura – 10%) due to the great preference for leaves in preparing remedies.
The emphasis on cultivated plants observed in this study was not unexpected. However, it might also have been expected that a larger proportion of the plants cited would be native species in view of the fact that the informants are all specialists having an excellent knowledge of the native resources. Nonetheless, the proportion of non-cultivated to cultivated species used varies among different communities that have access to the resources in the Atlantic Forest [2,25,16]. Amorozo [22] observed a similar situation in the community of Barcarena (state of Pará), where 57.2% of the plants used were cultivated, while only 42.8% were not. Yet, in the same study, Amorozo [22] demonstrated that in the Santo Antonio do Leverger community (state of Mato Grosso) the percentages invert themselves, with non-cultivated plants totaling 56.2% and cultivated plants 43.8%. This result may be due to the fact that Santo Antonio de Leverger is undergoing a process of modernization, resulting in additional anthropogenic alterations in the natural areas where many medicinal plants grow, and causing the devaluation of local culture as well as the loss of traditional practices such as plant cultivation.
In regards to the population studied here, their preference for using plants from anthropogenic zones may be explained by the establishment of the "Usina São José" forest reserve, which does not permit plant collection. This would force the local specialists to rely more on weed and cultivated plants. Local specialists informed us that, in function of the prohibition, they prevent to collect plants in the fragment forest. In this sense, they "opted" to using non-native plants. The ample use of plants derived from disturbed areas is quite common in many parts of the world [26], including areas of Atlantic Forest [17]. This may be explained by easy access to these species, especially when there is little cultural resistance to using such plants. Albuquerque et al. [27] counters this idea when he cites the preference of a rural northeastern Brazilian community for native species even when they have easy and rapid access to substitutes in disturbed areas.
In general, the results presented here give rise to two questions. First, what are the implications of the local use of a large number of non-native species? A mixture of native and introduced species in the pharmacopoeias of tropical zone communities is quite common [15], and in view of the complexity of the local knowledge, cannot be simply explained. On one hand, Begossi et al. [25] view the question from a conservationist's point of view, as the use of cultivated species reduces pressure on native forest products. On the other hand, in the case of the present study, the use of exotic and cultivated plants is a result of the need for alternatives to native (prohibited) resources in spite of their great importance to this community. This suggests that a contextual analysis of this phenomenon must take into account not only the proportion of species used (native as well as introduced), but also local preferences.
The second question relates to the transmission of local knowledge. The enforced restriction of the access to native plants may result, in the long run, to the abandonment of their use, and thus a loss of local knowledge. This is particularly serious in the present study, as the most knowledgeable people are very old and the young have not shown much interest in this accumulated knowledge. This will surely affect the resilience of the local knowledge [25]. Additionally, the opportunities for these specialists to invent and experiment with new native resources will be lost, as well as the possibility for scientists to investigate new bio-active products. In agreement with observations in other Atlantic Forest communities [28-30], we consider that "the process of household Forest uses, of economic alternatives, and of environmental education may be closely tied and may serve as a vehicle for conservation, and key individuals from each community should be included in management programs" [25].
Conclusion
This study demonstrated that local specialists in the Atlantic Forest community studied tend to agree with each other in terms of the plants used to treat blood-related problems, but cite a much more diverse group of plants to treat problems related to the respiratory and digestive systems – two important categories in studies undertaken in different parts of the world. In the case of the community examined here, the extensive repertory of plants employed to treat respiratory problems seems to be a response to problems related to the extensive use of fire in local agricultural practices. The local medicinal flora is largely based on plants that are either cultivated or obtained from anthropogenic zones, possibly due to the use and access restrictions of the legally protected neighboring forest. Despite these restrictions, the species with the highest use-value by this community was Pithecellobium cochliocarpum (Gomez) Macb., a native plant of the Atlantic Forest.
Supplementary Material
Additional File 1
List of Plants used in the community of "Três Ladeiras" in the municipality of Igarassu (Pernambuco, Northeast Brazil): indications and Use Value. C = cultivated. NC = non-cultivated.
Click here for file
==== Refs
Consórcio Mata Atlântica, Universidade Estadual de Campinas Reserva da Biosfera da Mata Atlântica 1992 Plano de Ação. Referências Básicas
Hanazaki N Tamashiro JY Leitão-Filho HF Begossi A Diversity of plant uses in two Caiçaras communities from the Atlantic Forest coast, Brazil Biodiversity and Conservation 2000 9 597 615 10.1023/A:1008920301824
Rossato SC Leitão-Filho H Begossi A Ethnobotany of caiçaras of the Atlantic Forest Coast (Brasil) Economic Botany 1999 53 387 395
Prance GT What is ethnobotany today? Journal of Ethnopharmacology 1991 32 209 216 1881159 10.1016/0378-8741(91)90120-3
FIAM – Fundação de Desenvolvimento Municipal do Interior de Pernambuco Perfil Municipal de Igarassu, Pernambuco 1997
FIDEM Diagnóstico das Reservas Ecológicas: Região Metropolitana do Grande Recife Recife 2001
SECTAM Diagnósticos das Reservas Ecológicas da Região Metropolitana do Recife Recife 2001
FIDEM Monitoramento das Reservas Ecológicas da Região Metropolitana do Grande Recife Recife 1993
Davis A Wagner JR Who Knows? On the Importance of Identifying "Experts" When Researching Local Ecological Knowledge Human Ecology 2003 31 463 489 10.1023/A:1025075923297
Albuquerque UP Lucena RFP Albuquerque UP, Lucena RFP Seleção e escolha dos informantes Métodos e técnicas na pesquisa etnobotânica 2004 Recife: Editora NUPEEA 19 35
Bailey K Methods of social research 1994 4 New York: The Free Press
Heinrich M Ankli A Frei B Weimann C Sticher O Medicinal plants in Mexico: Healers'consensus and cultural importance Social Science and Medicine 1998 47 1863 1875
Phillips O Gentry AH Reynel C Wilki P Gávez-Durand CB Quantitative ethnobotany and Amazonian conservation Conservation Biology 1994 8 225 248 10.1046/j.1523-1739.1994.08010225.x
Almeida CFCBR Albuquerque UP Uso e conservação de plantas e animais medicinais no Estado de Pernambuco (Nordeste do Brasil): um estudo de caso Interciencia 2002 27 276 285
Bennett BC Prance GT Introduced plants in the indigenous pharmacopoeia of Northern South America Economic Botany 2000 54 90 102
Di Stasi LC Oliveira GP Carvalhaes MA Queiroz-Junior M Tien OS Kakinami SH Reis MS Medicinal plants popularly used in the Brazilian Tropical Atlantic Forest Fitoterapia 2002 73 69 91 11864767 10.1016/S0367-326X(01)00362-8
Voeks RA Tropical Forest healers and habitat preference Economic Botany 1996 50 382 400
Stepp JR Moerman DE The importance of weeds in ethnopharmacology Journal of Ethnopharmacology 2001 75 19 23 11282438 10.1016/S0378-8741(00)00385-8
Stepp JR The role of weeds as sources of pharmaceuticals Journal of Ethnopharmacology 2004 92 163 166 15137997 10.1016/j.jep.2004.03.002
Almeida CFCBR Silva TCL Amorim ELC Albuquerque UP Life strategy and chemical composition as predictors of the selection of medicinal plants from the caatinga (Northeast Brazil) Journal of Arid Environments 2005 62 127 142 10.1016/j.jaridenv.2004.09.020
Parente CET Rosa MMT Plantas comercializadas como medicinais no Município de Barra do Piraí, RJ Rodriguésia 2001 52 47 59
Amorozo MCM Uso e diversidade de plantas medicinais em Santo Antônio de Leverger, MT, Brasil Acta Botânica Brasílica 2002 16 189 203
Albuquerque UP Andrade LHC Uso de recursos vegetais da caatinga: o caso do agreste do Estado de Pernambuco (Nordeste do Brasil) Interciencia 2002 27 336 345
Zschocke S Rabe T Staden J Plant part substitution – a way to conserve endangered medicinal plants? Journal of Ethnopharmacology 2000 71 281 292.18 10904175 10.1016/S0378-8741(00)00186-0
Begossi A Hanazaki N Tamashiro JY Medicinal Plants in the Atlantic Forest (Brasil): Knowledge, Use and Conservation Human Ecology 2002 30 281 299 10.1023/A:1016564217719
Anklin A Sticher O Heinrich M Medical ethnobotany of the Yucatec Maya: Healter's consensus as a quantitative criterion Economic Botany 1999 53 144 160
Albuquerque UP Andrade LHC Silva ACO Use of plant resources in a seasonal dry forest (northeastern Brazil) Acta Botânica Brasilic 2005 19 27 38
Figueiredo GM Leitão-Filho HF Begossi A Ethnobotany of Atlantic Forest coastal communities: diversity of plant uses in Gamboa (Itacuruçá Island, Brazil) Human Ecology 1993 21 419 430 10.1007/BF00891142
Figueiredo GM Leitão-Filho HF Begossi A Ethnobotany of Atlantic Forest Coastal Communities: II. Diversity of plant uses at Sepetiba Bay (SE Brasil) Human Ecology 1997 25 353 360 10.1023/A:1021934408466
Hanazaki N Leitão-Filho HF Begossi A Uso de recursos na Mata Atlântica: o caso da Ponta do Almada (Ubatuda, Brasil) Interciencia 1996 21 268 276
|
16270911
|
PMC1291389
|
CC BY
|
2021-01-04 16:40:12
|
no
|
J Ethnobiol Ethnomed. 2005 Nov 1; 1:9
|
utf-8
|
J Ethnobiol Ethnomed
| 2,005 |
10.1186/1746-4269-1-9
|
oa_comm
|
==== Front
J Inflamm (Lond)Journal of Inflammation (London, England)1476-9255BioMed Central London 1476-9255-2-131625963210.1186/1476-9255-2-13ResearchReversal of experimental colitis disease activity in mice following administration of an adenoviral IL-10 vector Sasaki Makoto [email protected] J Michael [email protected] Merilyn H [email protected] Paul [email protected] Yuping [email protected] Tomoaki [email protected] Takashi [email protected] J Steven [email protected] Department of Molecular and Cellular Physiology, LSU Health Sciences Center, Shreveport, LA, 71130-3932, USA2 Department of Cell Biology and Anatomy, LSU Health Sciences Center, Shreveport, LA, 71130-3932, USA3 Department of Gastroenterology, LSU Health Sciences Center, Shreveport, LA, 71130-39322, USA4 Department of Obstetrics and Gynecology, LSU Health Sciences Center, Shreveport, LA, 71130-39322, USA5 Department of Internal Medicine and Bioregulation, Nagoya City University Graduate School of Medical Sciences, Nagoya, Japan2005 31 10 2005 2 13 13 13 6 2005 31 10 2005 Copyright © 2005 Sasaki et al; licensee BioMed Central Ltd.2005Sasaki et al; licensee BioMed Central Ltd.This is an Open Access article distributed under the terms of the Creative Commons Attribution License (), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
Genetic deficiency in the expression of interleukin-10 (IL-10) is associated with the onset and progression of experimental inflammatory bowel disease (IBD). The clinical significance of IL-10 expression is supported by studies showing that immune-augmentation of IL-10 prevents inflammation and mucosal damage in animal models of colitis and in human colitis. Interleukin-10 (IL-10), an endogenous anti-inflammatory and immunomodulating cytokine, has been shown to prevent some inflammation and injury in animal and clinical studies, but the efficacy of IL-10 treatment remains unsatisfactory. We found that intra-peritoneal administration of adenoviral IL-10 to mice significantly reversed colitis induced by administration of 3% DSS (dextran sulfate), a common model of colitis. Adenoviral IL-10 (Ad-IL10) transfected mice developed high levels of IL-10 (394 +/- 136 pg/ml) within the peritoneal cavity where the adenovirus was expressed. Importantly, when given on day 4 (after the induction of colitis w/DSS), Ad-IL10 significantly reduced disease activity and weight loss and completely prevented histopathologic injury to the colon at day 10. Mechanistically, compared to Ad-null and DSS treated mice, Ad-IL10 and DSS-treated mice were able to suppress the expression of MAdCAM-1, an endothelial adhesion molecule associated with IBD. Our results suggest that Ad-IL10 (adenoviral IL-10) gene therapy of the intestine or peritoneum may be useful in the clinical treatment of IBD, since we demonstrated that this vector can reverse the course of an existing gut inflammation and markers of inflammation.
==== Body
I. Introduction
Endothelial cell adhesion molecules ('ECAMs') play essential roles in the development of chronic inflammation by recruiting leukocytes, especially lymphocytes, to tissues. ECAMs support several forms of leukocyte adhesion including rolling, firm adhesion and extravasation [1]. Infiltration of tissues by leukocytes is a common hallmark of many chronic inflammatory states that include the inflammatory bowel diseases (IBD), ulcerative colitis (UC), and Crohn's disease (CD). In the setting of IBD, the expression of ECAMs like ICAM-1, VCAM-1, and MAdCAM-1 is observed in experimental models of colitis, and also within the inflamed human colon in Crohn's disease and ulcerative colitis [2-6].
Among the adhesion molecules up-regulated in IBD, MAdCAM-1, the mucosal cell adhesion molecule, is thought to be preeminent in the development of chronic gut inflammation. MAdCAM-1 is normally expressed in the gut, and its expression is dramatically amplified during inflammation [2,3]. The functional significance of increased appearance of MAdCAM-1 in IBD is supported by several reports which show that immunoneutralization of either MAdCAM-1 or its ligand, the α4β7 integrin, attenuate inflammation and mucosal damage in animal models of colitis [7-9]. However, since monoclonal antibodies directed against other ECAMs, particularly VCAM-1, can as well reduce disease activity in colitis models, the literature suggests that MAdCAM-1 is probably necessary, but insufficient for the maximal penetrance of experimental and probably also clinical IBD [10-13].
Based on these findings, it is apparent that a better understanding of the mechanisms regulating ECAM expression, especially that of MAdCAM-1, might help to devise improved therapies for colitis.
It has been suggested that pathologic activation of the mucosal immune system in response to antigens is a key factor in the pathogenesis of IBD. Furthermore, changes in leukocyte migration and cytokine production appear to contribute to the perpetuation of IBD [14]. Based on modern advances, recombinant anti-inflammatory cytokines (i.e. IL-10) treatment is now being developed for experimental colitis and human IBD. IL-10 produced by macrophages and monocytes appears to limit chronic inflammation [15-17], through several mechanisms including inhibition of the release of several inflammatory factors (IL-1, IL-6, IL-12, TNF-α, GM-CSF, GCSF), suppression of cell adhesive determinants (MHC class II molecule, β7), and by blocking ICAM-1 induction [18-24]. Conversely, IL-10 gene-knockout mice develop a chronic colitis that is extremely similar to IBD [25]. IL-10 treatment can reduce inflammation in several models of colitis and human IBD [26,18-34]. However, the clinical efficacy of systemically administered IL-10 for patients with mild to moderately active Crohn's disease has not been as effective as hoped [31-34]. Furthermore the efficacy of IL-10 administration in mouse colitis models is variable and model-specific [35].
We have previously described that exogenous IL-10 in vitro can block the expression of MAdCAM-1 in response to TNF-α, and attenuates lymphocyte adhesion to lymphatic node derived endothelium under cytokine stimulating conditions via NF-kB inhibition [36]. The purpose of the current study was to show that induction of endothelial expression of IL-10 through an IL-10 expression vector attenuates MAdCAM-1 expression in response to TNF-α and optimistically suggests the possibility of targeted Th2-cytokine gene therapy in IBD.
II. Results
A. Measurement of human IL-10 concentration in lavage fluids from the transfected peritoneum
To screen for the efficacy of adenovirus mediated production of IL-10 in transfected mice, we measured the IL-10 concentration in the lavaged peritoneum in untreated, in adeno-'null' treated mice and in adeno-IL-10 transfected mice. There was no detectable human IL-10 signal in the non-transfected lavage fluid (control), nor was any mouse IL-10 detected (data not shown). However, the lavage fluid from the adenoviral IL-10 transfected mice showed a large and signficant increase in the IL-10 concentration (395 ± 136 pg/ml at 48 h after IL-10 gene transfection (Figure 1). Importantly, IL-10 was not detected in serum samples from these mice.
Figure 1 IL-10 concentration in lavage fluids from the transfected peritoneum. ELISA measurement of IL-10 in peritoneal lavage fluids from control shows a very high level of expression of IL-10 at approximately 400 pg/ml. No IL-10 was detected in lavage fluids of control or adeno-null mice (n = 5).
B. Reduced disease activity in adeno IL-10 gene transfected mice
A combinatorial index of disease, or disease acticvity index (DAI), defined as stool blood, stool form, and weight loss [37] was used to analyze the therapeutic benefit of adenoviral gene delivery. We found that compared to adeno-null or untreated mice, adenoviral IL-10 gene transfection after induction of clinical disease reversed the course of the disease induced by DSS (Figure 2).
Figure 2 Disease activity in mice with experimental colitis given adenoviral IL-10 gene. Disease activity in mice in which experimental colitis was induced by feeding 3% DSS was significantly attenuated at days 7–10 when adenoviral administration of IL-10 was given on day 6. Disease activity in control mice continued at the same level as the adeno-null mice on DSS. Disease activity was slightly higher in adeno-null mice which was significant at day 8, suggesting that adenoviral infection produces some inflammation. This is important to note since Ad-IL-10 still promotes protection despite the tendency towards higher inflammation.
C. Body weight change in adeno IL-10 gene transfected mice during colitis
The anti-inflammatory effect of adenoviral IL-10 gene transfer to mice was analyzed in experimental colitis induced by feeding of oral 3% dextran sulfate (DSS, MW~40–50 kD) over the course of 10 days, and body weight recorded daily. Feeding behaviour was not altered (measured by the weight of consumed food pellets, data not shown). Body weight change in response to DSS was significantly different from adeno-null mice at days 8, 9 and 10 but not different from adeno IL-10 treated mice (Figure 3) consistent with a rescue from progressive weight seen in untreated mice.
Figure 3 Body weight of mice in DSS colitis. Adeno-IL-10 mice did not lose any body weight over the course of DSS colitis, but adeno-null mice lost significantly more weight than adeno-IL-10 transfected mice (n = 5).
D. Colon shortening in DSS colitis and adenoviral IL-10
Animals fed DSS exhibited significant colon shortening compared to controls, an effect which was eliminated by adenoviral IL-10 gene transfer (Figure 4).
Figure 4 Adeno-IL-10 blocks colon shortening induced by DSS colitis. Adenoviral IL-10 adminstration significantly reduced the colon shortening produced by 3% DSS colitis (n = 5).
E. Adenoviral IL-10 significantly lowers histopathology score in DSS colitis
Perhaps the most remarkable finding in this study was the histopathologic effect of adeno-IL-10 on gut histopathology. Animals which had received adenoviral IL-10 vector showed virtually no evidence of any inflammation (Figure 5c), although adeno-null animals showed significant injury in response to DSS (Figure 5b) compared to controls (Figure 5a). Interestingly, the level of inflammation measured histopathologically in adenoviral IL-10 treated mice given DSS was actually lower than that measured for controls and may suggest that within the normal gut, there is a persistent, low basal level of inflammation which is normal, but that this mild inflammation can be suppressed by additional supplementation with Th2 cytokines e.g. IL-10 (Figure 6).
Figure 5 Colon histology for adenoviral transfected mice given DSS colitis. Figure 5A shows control colons with normal histopathology, 5B shows extensive regional leukocytic infiltration of the colon; see inset. This leukocyte infiltration is completely absent in adenoviral IL-10 treated mice which show normal or even improved morphology.
Figure 6 Analysis of histopathology in adenoviral transfected DSS colitis model. Compared to control mice, adeno-null treated mice exhibited significantly worse histopathology; whereas adeno-IL-10 treated mice had completely normal histology.
F. Immunohistochemistry for MAdCAM-1
Staining of colon sections for the presence of MAdCAM-1 showed occasional staining in control treated sections. In the null adenovirus treated mice receiving DSS, colon sections showed a strong and obvious increase in MAdCAM-1 positive staining (indicated by white arrows in Figure 7b) over controls (Figure 7a), which is not observed in adeno-IL-10/DSS treated samples (Figure 7c). Image analysis revealed a large and significant increase in vessel staining from 40.33 +/- 2.79 (n = 38) in controls to 399 +/- 58.5 (n = 49); this was significantly reduced by treatment with adeno-IL-10 (79.4 +/- 22.8, n = 12) (p < 0.05, Dunnetts test).
Figure 7 Adenoviral IL-10 reduces MAdCAM-1 expression in experimental colitis. Figure 7A (control) shows only sporadic and weak staining for MAdCAM-1. Figure 7B shows intense MAdCAM-1 staining in adeno-null + DSS-treated colon sections, unlike Adeno-IL10 + DSS-treated sections (Figure 7C) which lack strong MAdCAM staining, and much more closely resemble controls.
III. Discussion
Experimental colitis produced by DSS is thought to share many important characteristics with forms of human inflammatory bowel disease. We have previously shown that a pre-emptive induction of interleukin-10 (using a plasmid based expression vector) within endothelial cells will significantly attenuate the expression of MAdCAM-1, an important adhesive determinant which contributes to the development of human IBD, in response to TNF-a [38]. These effects may be due to enhanced endothelial barrier function [39], or to effects on adhesion molecules e.g. MAdCAM-1 and other endothelial cell adhesion molecules [4]. This is further supported by in vivo studies where animals genetically deficient in IL-10 develop spontaneous colitis with many of the characteristics of human IBD and clinical studies where IL-10 has shown some benefit in the treatment of human IBD [40].
Although many experimental therapies have been shown to be effective at preventing the induction of experimental colitis, it has of course proven more difficult to reduce an existing level of inflammatory bowel disease, since the disease process may be highly complex and difficult to control by altering a single mediator. However, models which can demonstrate effective attenuation of existing disease may provide the most relevant and important models of how human disease can be treated [41].
We showed that an adenoviral IL-10 expression vector is capable of producing very high levels of IL-10 within the peritoneal compartment, the bulk of which appears to remain confined to the peritoneal cavity, since IL-10 is not detected in plasma or serum samples following adenoviral transfection. Expression of MAdCAM-1 has also been reported in the brain, and in the heart; based on these findings, it has now been suggested that MAdCAM-1 might play roles in chronic inflammation of these organs as well [42,43].
In normal biology and especially during active inflammatory bowel disease, MAdCAM-1 may be essential to the lymphocyte homing to mucosa associated lymphoid tissue (MALT) [5,44]. Since MAdCAM-1 is normally expressed mainly within the gut microvasculature, and is dramatically amplified during IBD, it has been suggested that increased MAdCAM-1 expression contributes to the etiology of IBD through its ability to direct homing of lymphocytes to the gut. This notion is well supported by several reports that show that antibodies directed against either MAdCAM-1, or its lymphocyte ligand, the α4β7 integrin, will significantly attenuate several indices of gut damage in experimental models of colitis [8,46]. Furthermore, clinical studies conducted by Feagan et. al (2005) indicate that a humanized antibody against α4β7, an important MAdCAM-1 ligand administered to patients with active ulcerative colitis, effectively reduced the severity of the disease in comparison to those patients who received the placebo [47].
Several studies have indicated that T helper (Th1) immune responses have important roles in the development of IBD [48-50]. Moreover, dysregulation of cytokine networks is involved in Th1-dominant immune responses in IBD. Among the Th1 cytokines, TNF-α is thought to be perhaps the most important cytokine responsible for driving the onset and evolution of IBD. Because of this prime role of TNF-α in IBD, anti-TNF-α antibody therapy has been very successfully used in IBD to reduce both colonic injury and expression of ECAMs in IBD [51].
IL-10, a cytokine produced by activated macrophages and Th2-type T cells, has crucial inhibitory effects on the Th-1 type immune response, as well as on the antigen-presenting function of monocytes and macrophages [15,16]. IL-10 appears to play an important role in preventing the onset of IBD, since animals deficient in IL-10 develop colitis spontaneously, and low levels of IL-10 are positively correlated with recurrences of Crohn's disease [25,52]. However, unlike TNF-α based therapies, administration of recombinant IL-10 (rIL-10) shows poor efficacy. This may reflect the fact that TNF-α therapies for IBD are aimed at efficiently clearing TNF-α, while IL-10 therapies must increase IL-10 and recombinant IL-10 is likely too rapidly cleared from the circulation after in vivo administration to produce a uniform protection [53]. On the other hand, IL-10 gene transfer technology has been used with some success in models of colitis, however its efficacy is variable. One reason for this variability may be that the final serum IL-10 concentration of gene-transfected mice is below the threshold level needed for gut protection [53,54]. Therefore targeting of the IL-10 gene to the inflamed colon or its compartment should ideally exploit tissue (i.e. gut) specific promoters to control selective organ gene transfer technology, endothelial specific promoters and also organ specific intra-arterial injection of vector to activate some genes in specific locations [55].
Administration of IL-10 in vitro prevents TNF-α stimulated expression of MAdCAM-1 and also blocks lymphocyte adhesion on endothelial cells to the same level as dexamethasone treatment, currently a key therapy in IBD [36]. While it has been previously shown that delivery of IL-10 to the endothelium in vitro is protective against TNF-α [36], in vivo administration of IL-10 may be less effective [33]. Therefore methods like endothelial gene transfection in vivo may effectively maintain adequate IL-10 concentrations at the endothelial surface to finally achieve protection not obtained with intravenous IL-10 administration.
The most important index of efficacy for gene mediated recombinant IL-10 delivery in IBD is the effective inhibition of the lymphocyte-endothelium interaction mediated by MAdCAM-1. In this experiment, IL-10 induction in the endothelium efficiently blocked TNF-α induced MAdCAM-1 expression and α4β7-dependent lymphocyte adhesion on SVEC endothelial cells. Although we have not used tissue specific promoters, their use might permit even greater organ selective transgene delivery.
Our findings suggest that lymphatic or gut endothelial transfection with Th2 cytokines like IL-10 may be an effective method to reduce important symptoms associated with IBD.
IV. Experimental procedures
A. Adenoviral IL-10 gene transfer
Adenoviruses
The AdvIL-10 construct was a generous gift from Thomas Ritter, Institute of Medical Immunology, Charite-Campus Mitte, Humboldt University, Berlin, Germany. The control Ad-null construct, consisting of an E1a deleted Ad with no CMV promoter and no transgene cassette, was provided by Canji, Inc. (ZZNB; San Diego, CA). High titer adenoviral stocks were propagated in 293 cells and purified by cesium chloride gradient centrifugation. Banded virus was removed, desalted by dialysis in storage buffer (1 M sucrose, 5 mM alpha-cyclodextrin (Sigma) in PBS), and stored in small aliquots at -80°C. Repeated freeze/thaw cycles of the Ad stocks were avoided. Viral stocks and infected cells were handled only in a Class II laminar flow hood and maintained in a CO2 incubator designated for that purpose. The concentration of total viral particle numbers (PN) was determined by measuring the absorbance of the stocks at 260 nm. Infectious PNs were determined by measuring the concentration of viral hexon protein-positive 293 cells after a 48-h infection period. Multiplicity of infection (m.o.i.) was determined using an Adeno-X Rapid Titer Kit (Clonetics).
B. Evaluation of Clinical Colitis
The mice were C57B6 mice, males which were obtained at 6–8 weeks of age, and used at 8–10 weeks of age, with an average weight of 23 g at the beginning of the experiments. Mice were fed either water or 3% DSS as previously described, [56]. In all animals, weight, stool blood, presence of gross blood and stool consistency were determined daily as previously described [37]. Disease activity index (DAI) was determined by combining scores of a) weight loss b) stool consistency and c) bleeding (divided by 3). Each score was determined as follows, change in weight (0:<1%, 1: 1–5%, 2: 5–10%, 4:>15%), stool blood (0: negative, 2: positive) or gross bleeding (4), and stool consistency (0: normal, 2: loose stools, 4: diarrhea) as previously described [57]. Bodyweight loss was calculated as the percent difference between the original bodyweight and the actual bodyweight on any particular day. Typically in DSS colitis animals will lose 10–15% body weight over the course of 10 days. The appearance of diarrhea is defined as mucus/fecal material adherent to anal fur. The presence or absence of diarrhea was scored as either 1 or 0, respectively, and the cumulative score for diarrhea was calculated by adding the score for each day and dividing by the number of days of exposure. Rectal bleeding was defined as diarrhea containing visible blood/mucus or gross rectal bleeding and scored as described for diarrhea. Occult blood was detected using the 'Coloscreen' (Helena Laboratories, Beaumont, TX). At the end of these studies mice were anesthetized with high dose ketamine/xylazine (200 ul/animal) with carbon dioxide asphyxia prior to collection of tissues.
C. MadCAM-1 Immunohistochemistry
3 mm sections of tissue were frozen in OCT embedding compound and 15 um frozen sections collected onto 1% gelatin coated slides. Sections were incubated in 1:100 diluted primary anti-mouse MAdCAM-1 antibody in 0.1% milk powder in PBS for 12 h, washed 3× in this buffer, incubated in 1:1000 goat anti-rat Cy3 labeled antibody for 1 h, washed 3× and then mounted in Vectashield (Vectorlabs, Burlingame, CA). Images were analyzed for vessel staining (area) using the Image-J software package (NIH, Bethesda, MD, ).
E. Morphological analysis
Mice were killed on day 10 of the experiment, organs were removed and fixed in 3.7% phosphate buffered formaldehyde. Sections of the distal colons were cut into 1 cm pieces and then embedded in epon/aryldite (Ted Pella). General histological assessment and scoring was carried out on sections stained using haematoxylin and eosin.
F. Histological scoring
Histological scoring was performed on operator blinded sections using the standardized histological point system described by Cooper et al., which is used routinely for histological scoring of IBD severity [57]. A score of 0 reflects normal epithelium, without blunting, normal crypt appearance, low monocyte infiltration, and low or absent neutrophil infiltration. Three serial sections of five to six different sites of the colon (accounting for up to 18 sections per mouse) were examined at 200 × magnification; the most affected part was scored, ulceration being considered the worst lesion. A score of 1 indicates loss of single epithelial cells, mild blunting of the epithelium, single inflammatory cell infiltration of crypts, slight monocyte and neutrophil infiltrate; a score of 2 signifies loss of multiple epithelial cells (in patches), obvious flattening of the epithelia, cryptitis, and a moderate increase in monocytes and neutrophils; a score of 3 indicates frank epithelial ulceration with crypt abscesses and a marked increase in monocyte/neutrophils.
G. Statistical analysis
All values are expressed as mean ± SD. Data were analyzed using multiple comparisons. Probability (P) values of <0.05 were considered significant.
Competing interests
The author(s) declare that they have no competing interests.
Authors' contributions
Author 1 (MS) carried out the animal studies, Author 2 (JMM) prepared the adenovirus used in these studies. Author 3 (MHJ) and 6 (TA) participated in visual sample processing and analysis. Authors 4 (PJ) and 5 (YW) helped conceive and design animal studies. Author 7 (TJ), 5 (YW) and 8 (JSA) conceived and designed the study.
Acknowledgements
The authors would like to acknowledge Ms. Nicole Walker for her outstanding editorial assistance with this manuscript.
==== Refs
Salmi M Jalkanen S Endothelial ligands and homing of mucosal leukocytes in extraintestinal manifestations of IBD Inflamm Bowel Dis 1998 4 149 56 9589300
Connor EM Eppihimer MJ Morise Z Granger DN Grisham MB Expression of mucosal addressin cell adhesion molecule-1 (MAdCAM-1) in acute and chronic inflammation J Leukoc Biol 1999 65 349 355 10080539
Shigematsu T Specian RD Wolf RE Grisham MB Granger DN MAdCAM mediates lymphocyte-endothelial cell adhesion in a murine model of chronic colitis Am J Physiol Gastrointest Liver Physiol 2001 281 G1309 G1315 11668040
Oshima T Jordan P Grisham MB Alexander JS Jennings M Sasaki M TNF-alpha induced endothelial MAdCAM-1 expression is regulated by exogenous, not endogenous nitric oxide BMC Gastroenterol 2001 1 5 11481030 10.1186/1471-230X-1-5
Briskin M Winsor-Hines D Shyjan A Cochran N Bloom S Wilson J Human mucosal addressin cell adhesion molecule-1 is preferentially expressed in intestinal tract and associated lymphoid tissue Am J Pathol 1997 151 97 110 9212736
Souza HS Elia CC Spencer J MacDonald TT Expression of lymphocyte-endothelial receptor-ligand pairs, alpha4beta7/MAdCAM-1 and OX40/OX40 ligand in the colon and jejunum of patients with inflammatory bowel disease Gut 1999 45 856 863 10562584
Fong S Jones S Renz ME Chiu HH Ryan AM Presta LG Mucosal addressin cell adhesion molecule-1 (MAdCAM-1). Its binding motif for alpha 4 beta 7 and role in experimental colitis Immunol Res 1997 16 299 311 9379078
Kato S Hokari R Matsuzaki K Iwai A Kawaguchi A Nagao S Amelioration of murine experimental colitis by inhibition of mucosal addressin cell adhesion molecule-1 J Pharmacol Exp Ther 2000 295 183 189 10991977
Vainer B Nielsen OH The influence of adhesion molecules in inflammatory bowel diseases Ugeskr Laeger 1997 159 3767 3771 9214052
Bendjelloul F Maly P Mandys V Jirkovska M Prokesova L Tuckova L Intercellular adhesion molecule-1 (ICAM-1) deficiency protects mice against severe forms of experimentally induced colitis Clin Exp Immunol 2000 119 57 63 10606964 10.1046/j.1365-2249.2000.01090.x
Hamamoto N Maemura K Hirata I Murano M Sasaki S Katsu K Inhibition of dextran sulphate sodium (DSS)-induced colitis in mice by intracolonically administered antibodies against adhesion molecules (endothelial leucocyte adhesion molecule-1 (ELAM-1) or intercellular adhesion molecule-1 (ICAM-1)) Clin Exp Immunol 1999 117 462 468 10469048 10.1046/j.1365-2249.1999.00985.x
Sans M Panes J Ardite E Elizalde JI Arce Y Elena M VCAM-1 and ICAM-1 mediate leukocyte-endothelial cell adhesion in rat experimental colitis Gastroenterology 1999 116 874 883 10092309
Taniguchi T Tsukada H Nakamura H Kodama M Fukuda K Saito T Effects of the anti-ICAM-1 monoclonal antibody on dextran sodium sulphate-induced colitis in rats J Gastroenterol Hepatol 1998 13 945 949 9794195
Neurath MF Schurmann G Immunopathogenesis of inflammatory bowel diseases Chirurg 2000 71 30 40 10662999
Fiorentino DF Bond MW Mosmann TR Two types of mouse T helper cell. IV. Th2 clones secrete a factor that inhibits cytokine production by Th1 clones J Exp Med 1989 170 2081 2095 2531194 10.1084/jem.170.6.2081
Fiorentino DF Zlotnik A Mosmann TR Howard M O'Garra A IL-10 inhibits cytokine production by activated macrophages J Immunol 1991 147 3815 3822 1940369
Del Prete G De Carli M Almerigogna F Giudizi MG Biagiotti R Romagnani S Human IL-10 is produced by both type 1 helper (Th1) and type 2 helper (Th2) T cell clones and inhibits their antigen-specific proliferation and cytokine production J Immunol 1993 150 353 360 8419468
Schreiber S Heinig T Thiele HG Raedler A Immunoregulatory role of interleukin 10 in patients with inflammatory bowel disease Gastroenterology 1995 108 1434 1444 7729636 10.1016/0016-5085(95)90692-4
Bogdan C Nathan C Modulation of macrophage function by transforming growth factor beta, interleukin-4, and interleukin-10 Ann N Y Acad Sci 1993 685 713 739 8363277
Bogdan C Vodovotz Y Nathan C Macrophage deactivation by interleukin 10 J Exp Med 1991 174 1549 1555 1744584 10.1084/jem.174.6.1549
de Waal MR Abrams J Bennett B Figdor CG de Vries JE Interleukin 10(IL-10) inhibits cytokine synthesis by human monocytes:an autoregulatory role of IL-10 produced by monocytes J Exp Med 1991 174 1209 1220 1940799 10.1084/jem.174.5.1209
Wang P Wu P Siegel MI Egan RW Billah MM IL-10 inhibits transcription of cytokine genes in human peripheral blood mononuclear cells J Immunol 1994 153 811 816 8021515
Wang P Wu P Anthes JC Siegel MI Egan RW Billah MM Interleukin-10 inhibits interleukin-8 production in human neutrophils Blood 1994 83 2678 2683 8167346
Willems F Marchant A Delville JP Gerard C Delvaux A Velu T Interleukin-10 inhibits B7 and intercellular adhesion molecule-1 expression on human monocytes Eur J Immunol 1994 24 1007 1009 7512027
Kuhn R Lohler J Rennick D Rajewsky K Muller W Interleukin-10-deficient mice develop chronic enterocolitis Cell 1993 75 263 274 8402911 10.1016/0092-8674(93)80068-P
Tomoyose M Mitsuyama K Ishida H Toyonaga A Tanikawa K Role of interleukin-10 in a murine model of dextran sulfate sodium-induced colitis Scand J Gastroenterol 1998 33 435 440 9605267 10.1080/00365529850171080
Powrie F Leach MW Mauze S Menon S Caddle LB Coffman RL Inhibition of Th1 responses prevents inflammatory bowel disease in SCID mice reconstituted with CD45RBhi CD4+ T cells Immunity 1994 1 553 562 7600284 10.1016/1074-7613(94)90045-0
Powrie F Leach MW Mauze S Caddle LB Coffman RL Phenotypically distinct subsets of CD4+ T cells induce or protect from chronic intestinal inflammation in C. B-17 SCID mice Int Immunol 1993 5 1461 1471 7903159
Herfarth HH Mohanty SP Rath HC Tonkonogy S Sartor RB Interleukin 10 suppresses experimental chronic, granulomatous inflammation induced by bacterial cell wall polymers Gut 1996 39 836 845 9038666
Herfarth HH Bocker U Janardhanam R Sartor RB Subtherapeutic corticosteroids potentiate the ability of interleukin 10 to prevent chronic inflammation in rats Gastroenterology 1998 115 856 865 9753488 10.1016/S0016-5085(98)70257-4
van Deventer SJ Elson CO Fedorak RN Multiple doses of intravenous interleukin 10 in steroid-refractory Crohn's disease. Crohn's Disease Study Group Gastroenterology 1997 113 383 389 9247454 10.1053/gast.1997.v113.pm9247454
Fedorak RN Gangl A Elson CO Rutgeerts P Schreiber S Wild G Recombinant human interleukin 10 in the treatment of patients with mild to moderately active Crohn's disease. The Interleukin 10 Inflammatory Bowel Disease Cooperative Study Group Gastroenterology 2000 119 1473 1482 11113068 10.1053/gast.2000.20229
Schreiber S Fedorak RN Nielsen OH Wild G Williams CN Nikolaus S Safety and efficacy of recombinant human interleukin 10 in chronic active Crohn's disease. Crohn's Disease IL-10 Cooperative Study Group Gastroenterology 2000 119 1461 1472 11113067 10.1053/gast.2000.20196
Sands BE Therapy of inflammatory bowel disease Gastroenterology 2000 118 S68 S82 10868899
Ribbons KA Thompson JH Liu X Pennline K Clark DA Miller MJ Anti-inflammatory properties of interleukin-10 administration in hapten-induced colitis Eur J Pharmacol 1997 323 245 254 9128846 10.1016/S0014-2999(97)00017-4
Oshima T Pavlick K Grisham MB Jordan P Manas K Joh T Itoh M Alexander JS Glucocorticoids and IL-10, but not 6-MP, 5-ASA or sulfasalazine block endothelial expression of MAdCAM-1: implications for inflammatory bowel disease therapy Aliment Pharmacol Ther 2001 15 1211 8 11472325 10.1046/j.1365-2036.2001.01048.x
Dieleman LA Pena AS Meuwissen SG van Rees EP Role of animal models for the pathogenesis and treatment of inflammatory bowel disease Scand J Gastroenterol Suppl 1997 223 99 104 9200314
Sasaki M Jordan P Houghton J Meng X Itoh M Joh T Alexander JS Transfection of IL-10 expression vectors into endothelial cultures attenuates alpha4beta7-dependent lymphocyte adhesion mediated by MAdCAM-1 BMC Gastroenterol 2003 20 3 12625840 10.1186/1471-230X-3-3
Oshima T Laroux FS Coe LL Morise Z Kawachi S Bauer P Grisham MB Specian RD Carter P Jennings S Granger DN Joh T Alexander JS Interferon-gamma and interleukin-10 reciprocally regulate endothelial junction integrity and barrier function Microvasc Res 2001 61 130 43 11162203 10.1006/mvre.2000.2288
Braat H Peppelenbosch MP Hommes DW Interleukin-10-based therapy for inflammatory bowel disease Expert Opin Biol Ther 2003 3 725 31 12880373 10.1517/14712598.3.5.725
Li MC He SH IL-10 and its related cytokines for treatment of inflammatory bowel disease World J Gastroenterol 2004 10 620 5 14991925
Schaible UE Vestweber D Butcher EG Stehle T Simon MM Expression of endothelial cell adhesion molecules in joints and heart during Borrelia burgdorferi infection of mice Cell Adhes Commun 1994 2 465 479 7538017
Kanwar JR Kanwar RK Wang D Krissansen GW Prevention of a chronic progressive form of experimental autoimmune encephalomyelitis by an antibody against mucosal addressin cell adhesion molecule-1, given early in the course of disease progression Immunol Cell Biol 2000 78 641 645 11114975 10.1046/j.1440-1711.2000.00947.x
Rott LS Briskin MJ Butcher EC Expression of alpha4beta7 and E-selectin ligand by circulating memory B cells: implications for targeted trafficking to mucosal and systemic sites J Leukoc Biol 2000 68 807 14 11129647
Komoto S Miura S Koseki S Goto M Hachimura S Fujimori H Hokari R Hara Y Ogino T Watanabe C Nagata H Kaminogawa S Hibi T Ishii H Effect of specific antigen stimulation on intraepithelial lymphocyte migration to small intestinal mucosa Clin Exp Immunol 2005 140 249 57 15807848 10.1111/j.1365-2249.2005.02761.x
Picarella D Hurlbut P Rottman J Shi X Butcher E Ringler DJ Monoclonal antibodies specific for beta 7 integrin and mucosal addressin cell adhesion molecule-1 (MAdCAM-1) reduce inflammation in the colon of scid mice reconstituted with CD45RBhigh CD4+ T cells J Immunol 1997 158 2099 106 9036954
Feagan BG Greenberg GR Wild G Fedorak RN Pare P McDonald JW Dube R Cohen A Steinhart AH Landau S Aguzzi RA Fox IH Vandervoort MK Treatment of ulcerative colitis with a humanized antibody to the alpha4beta7 integrin N Engl J Med 2005 352 2499 507 15958805 10.1056/NEJMoa042982
Fuss IJ Marth T Neurath MF Pearlstein GR Jain A Strober W Anti-interleukin 12 treatment regulates apoptosis of Th1 T cells in experimental colitis in mice Gastroenterology 1999 117 1078 1088 10535870 10.1016/S0016-5085(99)70392-6
Sartor RB Pathogenetic and clinical relevance of cytokines in inflammatory bowel disease Immunol Res 1991 10 465 471 1955773
Simpson SJ Hollander GA Mizoguchi E Allen D Bhan AK Wang B Expression of pro-inflammatory cytokines by TCR alpha beta+ and TCR gamma delta+ T cells in an experimental model of colitis Eur J Immunol 1997 27 17 25 9021993
Baert FJ D'Haens GR Peeters M Hiele MI Schaible TF Shealy D Tumor necrosis factor alpha antibody (infliximab) therapy profoundly down-regulates the inflammation in Crohn's ileocolitis Gastroenterology 1999 116 22 28 9869598
Meresse B Rutgeerts P Malchow H Dubucquoi S Dessaint JP Cohard M Low ileal interleukin 10 concentrations are predictive of endoscopic recurrence in patients with Crohn's disease Gut 2002 50 25 28 11772962 10.1136/gut.50.1.25
Chernoff AE Granowitz EV Shapiro L Vannier E Lonnemann G Angel JB A randomized, controlled trial of IL-10 in humans. Inhibition of inflammatory cytokine production and immune responses J Immunol 1995 154 5492 5499 7730651
Barbara G Xing Z Hogaboam CM Gauldie J Collins SM Interleukin 10 gene transfer prevents experimental colitis in rats Gut 2000 46 344 349 10673295 10.1136/gut.46.3.344
Sferra TJ McNeely D Johnson PR Gene transfer to the intestinal tract: a new approach using selective injection of the superior mesenteric artery Hum Gene Ther 1997 8 681 687 9113508
Sasaki M Bharwani S Jordan P Elrod JW Grisham MB Jackson TH Lefer DJ Alexander JS Increased disease activity in eNOS-deficient mice in experimental colitis Free Radic Biol Med 2003 15 1679 87 14680690 10.1016/j.freeradbiomed.2003.09.016
Cooper HS Murthy SN Shah RS Sedergran DJ Clinicopathologic study of dextran sulfate sodium experimental murine colitis Lab Invest 1993 69 238 49 8350599
|
16259632
|
PMC1291390
|
CC BY
|
2021-01-04 16:36:23
|
no
|
J Inflamm (Lond). 2005 Oct 31; 2:13
|
utf-8
|
J Inflamm (Lond)
| 2,005 |
10.1186/1476-9255-2-13
|
oa_comm
|
==== Front
Lipids Health DisLipids in Health and Disease1476-511XBioMed Central London 1476-511X-4-261625577210.1186/1476-511X-4-26ResearchTreatment of dyslipidemia with lovastatin and ezetimibe in an adolescent with cholesterol ester storage disease Tadiboyina Venu T [email protected] Dora M [email protected] Brooke A [email protected] Jian [email protected] Robert A [email protected] Department of Medicine, Schulich School of Medicine and Dentistry, University of Western Ontario, London, ON, N6A 5C1, Canada 2 Vascular Biology Group and Blackburn Cardiovascular Genetics, Laboratory, Robarts Research Institute, London, ON, N6A 5K8, Canada 2005 28 10 2005 4 26 26 8 10 2005 28 10 2005 Copyright © 2005 Tadiboyina et al; licensee BioMed Central Ltd.2005Tadiboyina et al; licensee BioMed Central Ltd.This is an Open Access article distributed under the terms of the Creative Commons Attribution License (), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
Background
Cholesterol ester storage disease (CESD) is an autosomal recessive illness that results from mutations in the LIPA gene encoding lysosomal acid lipase. CESD patients present in childhood with hepatomegaly and dyslipidemia characterized by elevated total and low-density lipoprotein cholesterol (LDL-C), with elevated triglycerides and depressed high-density lipoprotein cholesterol (HDL-C). Usual treatment includes a low fat diet and a statin drug.
Results
In an 18-year old with CESD, we documented compound heterozygosity for two LIPA mutations: a novel frameshift nonsense mutation and a deletion of exon 8. The patient had been treated with escalating doses of lovastatin for ~80 months, with ~15% decline in mean LDL-C. The addition of ezetimibe 10 mg to lovastatin 40 mg resulted in an additional ~16% decline in mean LDL-C.
Conclusion
These preliminary anecdotal findings in a CESD patient with novel LIPA mutations support the longer term safety of statins in an adolescent patient and provide new data about the potential efficacy and tolerability of ezetimibe in this patient group.
==== Body
Background
Cholesteryl ester storage disease (CESD; MIM 278000) is an autosomal recessive disorder caused by a deficiency of lysosomal acid lipase (LAL; acid cholesteryl hydrolase; EC 3.1.1.13). LAL is responsible for the intralysosomal hydrolysis of cholesteryl esters (CE) and triglycerides (TG) [1]. Patients with CESD present in childhood with hepatomegaly, hypercholesterolemia and hypertriglyceridemia; most are diagnosed by age 20 [1]. Reduced LAL activity is detectable in peripheral blood leukocytes, cultured skin fibroblasts and liver homogenates [1]. More recently mutational screening of the human LAL gene (LIPA) [2-4] has been used for diagnosis. Wolman disease (WD; MIM 278000) also results from mutations in LIPA. WD is characterized by early death (usually before age 6 months) and widespread intracellular storage of both CE and TG, mainly in liver, adrenal glands and intestine [1]. In vitro catalytic activity was decreased ~200-fold in WD fibroblasts, but only ~50-fold in CESD fibroblasts [5], showing correlation with the differences in phenotypic severity [6,7].
Defective LAL activity results in decreased free intracellular cholesterol [6,7]. Because intracellular free cholesterol normally inhibits 3-hydroxy-3-methylglutaryl-CoA (HMG-CoA) reductase, cholesterol biosynthesis is increased in CESD patients. Thus, pharmacological inhibition of HMG-CoA reductase with statins would seem to be a reasonable approach to restrain the increased cholesterol biosynthesis in CESD. While the plasma lipoprotein response to statins among CESD patients has been variable [8-14], these drugs are considered to be the anti-dyslipidemia agents of choice in CESD.
Ezetimibe is a novel type of lipid-lowering medication that prevents the absorption of cholesterol and plant sterols at the small intestinal brush border by interfering with the activity of the NPC1L1 receptor [15-18]. Ezetimibe has been used in adult hypercholesterolemic patients either as monotherapy [19] or in combination with statins [20-25]. The rates of myopathy and serum transaminase elevations in ezetimibe-treated patients appeared to be comparable to those in placebo-treated patients [19-25]. We now present our experience with the combination of lovastatin and ezetimibe treatment in an 18 year old male with CESD.
Results
Patient history
A three-year old boy presented to his paediatrician for assessment of a pruritic abdominal rash. His birth and infancy had been unremarkable, with normal growth and development. There was no consanguinity; both parents and two older sisters were all healthy. At age 3, hepatosplenomegaly was noted on abdominal examination and was confirmed by ultrasound. No diagnosis was made and he was monitored periodically. At age 8, he was admitted to the hospital with gastroenteritis. Light microscopy of a liver biopsy showed increased intracytoplasmic glycogen and small lipid droplets in hepatocytes. Electron microscopy showed membrane-bound lipid droplets with small electron dense granules. A working diagnosis of glycogen storage disease type III (DeBrancher disease) was made, but skin fibroblast Debrancher activity was normal.
At age 10, hepatomegaly persisted and a second liver biopsy was taken. Light microscopy showed altered lobular architecture of the hepatic parenchyma with distended hepatocytes containing cytoplasmic granules and vacuoles with mild periportal fibrosis. Fibroblast acid lipase activity was found to be 7% of normal, confirming the diagnosis of CESD. Plasma concentrations of total cholesterol (TC), triglycerides (TG), low-density lipoprotein cholesterol (LDL-C) were each above the 95th percentile for age and sex at 7.49, 3.23 and 5.59 mmol/L, respectively, while plasma high-density lipoprotein cholesterol (HDL-C) was below the 5th percentile at 0.46 mmol/L; he had combined hyperlipidemia (hypercholesterolemia, hypertriglyceridemia, hypoalphalipoproteinemia and hyperbetalipoproteinemia). After 12 months, a low fat diet was started (Figure 1).
Figure 1 Plasma lipoprotein responses to treatment. The graph shows plasma lipoproteins measured serially in the proband over ~96 months. Abbreviations: TC, total cholesterol; LDL-C, low-density lipoprotein cholesterol; TG, triglycerides; HDL-C, high-density lipoprotein cholesterol; Diet, fat restricted to 30% of total calories, L10, L20 and L40 for lovastatin 10, 20 and 40 mg daily, respectively; E, ezetimibe 10 mg daily.
After 6 months of diet alone, lovastatin 10 mg daily was added. Because of rising plasma concentrations of TC and LDL-C, the dose of lovastatin was increased to 20 mg after ~22 months and increased again to 40 mg after a further ~8 months. Ezetimibe 10 mg per day was added after a further ~40 months and the combination of lovastatin 40 mg and ezetimibe 10 mg daily has continued for 12 months. Serum asparagine transaminase (AST) and creatine kinase (CK) were measured concurrently with the lipoproteins.
The proband's lipoprotein profile since age 10 is summarized in Figure 1. Mean ± standard deviation (SD) lipoprotein concentrations on each phase of treatment were determined from a minimum of three values. From baseline concentrations, the diet was associated with a 5.3% decrease in TC (6.67 ± 0.79 to 6.32 ± 0.26 mmol/L), a 27.3% decrease in TG (2.75 ± 0.48 to 1.99 ± 0.50 mmol/L), a 2.2% increase in LDL-C (4.63 ± 0.77 to 4.73 ± 0.53 mmol/L), a 7.8% increase in HDL-C (0.64 ± 0.12 to 0.69 ± 0.04 mmol/L), and a decrease in TC:HDL-C ratio of 11.5% (10.4 ± 0.3 to 9.2 ± 0.5). When pooled data over ~70 months from all 19 determinations on lovastatin were compared with diet alone, statin treatment was associated with a further 13.5% decrease in TC (to 5.47 ± 0.42 mmol/L), a 12.6% increase in TG (to 2.24 ± 0.36 mmol/L), a 15.9% decrease in LDL-C (to 3.98 ± 0.41 mmol/L), an 8.7% decrease in HDL-C (to 0.63 ± 0.08 mmol/L), and a 5.5% decrease in TC:HDL-C ratio (to 8.7 ± 0.2). Finally, when pooled data over 12 months from all four determinations on ezetimibe plus lovastatin were compared with lovastatin monotherapy, the drug combination was associated with a further 9.4% decrease in TC (to 4.96 ± 0.60 mmol/L), a 3.1% decrease in TG (to 2.17 ± 0.20 mmol/L), an 15.8% decrease in LDL-C (to 3.35 ± 0.46 mmol/L), no change in HDL-C (0.63 ± 0.08 mmol/L), and a 9.1% decrease in TC:HDL-C ratio (to 7.9 ± 0.3). Unpaired t-tests showed that the TC and LDL-C concentrations were significantly different for the period with lovastatin monotherapy compared to the period with combination therapy (P < 0.05). Also, there were no deviations of plasma CK and AST above the upper limit of normal for any treatment period. Finally, liver and spleen size evaluated clinically were reduced compared to baseline over the treatment period with statin and then later statin plus ezetimibe; specifically, while the liver edge was palpable 5 cm below the right costal margin before drug treatment it was not palpable at the most recent clinical assessment.
Molecular genetic studies
Genomic DNA sequencing of the LIPA gene revealed that the proband had two mutations (Figure 2). The first was a T insertion in exon 6 at codon 178 that shifted the reading frame (Figure 2A) and caused a premature termination at codon 190 (FS A178-X190). The second was G>A change at the last nucleotide of exon 8 (Q277), which resulted in a silent mutation at the amino acid level (Figure 2B). The patient was heterozygous for both mutations. In order to determine the chromosomal phase of the two LIPA mutations, sequencing of exon 6 and exon 8 from the proband's mother's genomic DNA revealed that she was a simple heterozygote for the frameshift mutation, confirming the that two mutations in the proband were on different chromosomes. Reverse transcriptase PCR amplification of LIPA from the proband, followed by sequence analysis of the partial cDNA spanning exon 5 through exon 10 revealed an abnormal sequence in which the entire exon 8 had been deleted (Figure 2C).
Figure 2 Nucleotide sequence analysis of LIPA. The nucleotide sequences, codon numbers, single letter amino acid codes for the deduced protein sequence are shown in each panel. Panel A shows normal LIPA genomic sequence above and the sequence from the proband's genomic DNA below. The inserted nucleotide is indicated by the arrow and the shifted reading frame is suggested by the presence of two peaks at each position following the insertion. Panel B shows normal LIPA genomic sequence above and the sequence from the proband's genomic DNA below, with abnormal sequence italicized. The single base nonsense mutation is indicated by the arrow. Panel C shows normal LIPA cDNA sequence from a single copy cloned source derived from a normal individual, spanning part of exon 7, all of exon 8 and then part of exon 9. The lower part of the panel shows the cDNA sequence for one of the proband's alleles, in which exon 8 has been deleted in frame. This confirms that the mutation at the intron-exon boundary of exon 8 affected RNA splicing.
Discussion
The untreated lipoprotein profile of our CESD patient revealed not only a combined hypercholesterolemia and hypertriglyceridemia, but also a severe hypoalphalipoproteinemia, indicating that mutations in LIPA are a rare genetic cause of complex dyslipidemia. Of course, this is in the context of numerous systemic abnormalities, specifically hepatomegaly. We also showed improvement of the plasma lipoprotein profile with low fat diet, with further improvement with statin monotherapy and even further improvement with the addition of ezetimibe in combination with the statin.
Lovastatin has been shown to be safe and effective in treating hypercholesterolemia over the long term in adults [26,27]. The ~80 month treatment period for our proband was among the longest time spans for any of our young patients with respect to duration of statin therapy. Over this period, the patient's hepatomegaly improved clinically and the AST and CK have remained stable. The addition of ezetimibe was associated with further improvement of plasma lipoproteins, and was also well tolerated in combination with statin treatment.
Statins have previously been successfully used in adolescents with CESD with some variability in reported efficacy [8-14]. This may be explained by genetic heterogeneity in response to lovastatin or by underlying differences in the factors responsible for the hyperlipidemia [3] Statins block the conversion of 3-hydroxymethylglutaryl-coenzyme A (HMGCoA) to mevalonate, a rate limiting step in cholesterol biosynthesis. This results in an increase in the number and activity of LDL receptors on the hepatocyte membrane, and the rate of LDL catabolism increases. In patients with CESD, the increased activity of the LDL receptors theoretically could lead to increased accumulation of cholesteryl esters in the liver [4]. However, Ginsberg et al. [2] showed no change in hepatic cholesteryl ester accumulation after 8 months of lovastatin 40 mg daily in a 9 year old girl with CESD. Furthermore, our patient had reduced hepatomegaly clinically, suggesting that cholesteryl ester accumulation in the liver was unlikely.
Our findings also suggest that ezetimibe may be a useful treatment in patients with CESD. Ezetimibe interferes with the normal function of the NPC1L1 gene product, which regulates sterol absorption in the small intestine [15-17]. This is thought to result in depletion of hepatic cholesterol and upregulation of hepatic LDL receptors. The mean plasma LDL-C reduction seen with ezetimibe is ~20%, and this has been remarkably consistent across patient subgroups defined by age, gender, ethnic background and concomitant use of other lipid regulating agents, such as statin drugs [20-25]. Inter-individual genetic variation may also play a role in the response to ezetimibe; for instance, a subset of individuals with a particular NPC1L1 haplotype appears to have a larger plasma LDL-C response [28]. Combination therapy for hypercholesterolemia may allow more patients to achieve target plasma TC and LDL-C goals. We observed that coadministration of ezetimibe with lovastatin resulted in reduction in plasma LDL-C concentration of 16% compared to lovastatin alone. Ezetimibe appeared to be well tolerated by our patient and there were no reported adverse effects. Ezetimibe added to lovastatin did not result in an increase in muscle enzymes. In fact, serum CK (mean ± SD) actually decreased from 185 ± 156 with statin monotherapy to 120 ± 68 during combination treatment with the statin and ezetimibe. AST levels never exceeded the upper limit of normal at any time. The complementary mechanism of action of ezetimibe and statins may offer a new treatment alternative for dyslipidemia management in CESD patients.
Methods
Genomic DNA analysis
Genomic DNA was isolated from whole blood obtained from patients (Puregene, GentraSystem, Minneapolis, MN). The entire coding region and intron-exon boundaries of LAL gene were amplified using custom primer pairs shown in table 1. PCR amplifications were carried out in a 50 μl mixture containing 32 pmole of each primer, 0.2 mM of each dATP, dCTP, dGTP and dTTP, 1.5 mM MgCl2, 50 mM KCl, 20 mM Tris-HCl (pH 8.4), 2.5 units of Taq platinum DNA polymerase (Invitrogen, Mississauga, ON). 30 cycles were performed consisting of denaturation at 94°C, annealing at 56°C and extension at 72°C for 30 s each, followed by extension for 10 min at 72°C. PCR products were purified from agarose (QIAQuick Gel Extraction Kit, Qiagen, Mississauga, ON). Direct DNA sequence results were analyzed with an ABI automated sequencer 3730 and were read with ABI Sequence Navigator software (both from PE Biosystems, Mississauga, ON).
Table 1 Primers used to amplify coding regions of LIPA
Exon Primer sequences
1 Forward: 5' AGC GCT AAA CAG CTT GCT AG
Reverse: 5' CTT GCT GAA GGC ACC AGC
2 Forward: 5' GGC TGG AGT CAT TTG TTT CA
Reverse: 5' AGA ATC ACT TGA GCC CCT GA
3 Forward: 5' GCC TGG AGA ACA TAG TTT ATC TGC
Reverse: 5' TTA GAT GAC TCT TGT CCT TAC TTC
4 Forward: 5' ATG TGA GTA CAT CAC TAT GTC
Reverse: 5' CTC ATA CAA CTT CAG AGT TAC
5 Forward: 5' TTC CCA GCT GTG TTT AGT TTG TG
Reverse: 5' GAC TAA ATG TTA CCA ACA TTC C
6 Forward: 5' GTG TTA GGG CAC ACG GAA GT
Reverse: 5' GTG TGC AGG AAA CGA CAG G
7 Forward: 5' GCA TCC TGA TTT GAT GTC CA
Reverse: 5' CAT AAG AAG GTG ACC ACA GTC AG
8 Forward: 5' TGG CTC TAG TTT TTA GTG CTT TGA
Reverse: 5' GGA CTC TGG GGA AGA AAA CC
9 Forward: 5' TTC TGT GTC AGG TGG TAG CTG
Reverse: 5' TGG ACT GAT GGA AAA CAA ACA
10 Forward: 5' CTC CAC AGC TAG TGG CGA TT
Reverse: 5' CAC ACA ATT CTT TGG GCC TAT
Reverse transcriptase polymerase chain reaction and cDNA sequence analysis
Total RNA was isolated from the proband using PAXgene Blood RNA Tube and Blood RNA Kit (Qiagen, Mississauga, ON). 2.5 ml blood was used for RNA isolation according to manufacturer's instruction. 100 ng of total RNA was used for first strand cDNA synthesis (SuperScript First Strand Synthesis System, Invitrogen, Mississauga, ON). 2 μl of first strand reaction was used for amplify partial cDNA sequence of the LIPA gene spanning exon 5 to 10 with primer pair 5' AAT ATG ACC TAC CAG CTT CCA, 3' GTA AGC AAA CAC ATT TTC ACA. PCR products were gel purified, sequenced and read as described above.
Authors' contributions
VT: data analysis, manuscript preparation and approval
DML: data analysis, manuscript preparation, manuscript approval
BAM: data collection, patient interaction, manuscript approval
JW: sequencing, data analysis, editing, manuscript approval
RAH: experimental design, manuscript preparation and approval
Acknowledgements
Supported by the Jacob J. Wolfe Distinguished Medical Research Chair, the Edith Schulich Vinet Canada Research Chair (Tier I) in Human Genetics, a Career Investigator award from the Heart and Stroke Foundation of Ontario, and operating grants from the Canadian Institutes for Health Research, the Heart and Stroke Foundation of Ontario, the Ontario Research and Development Challenge Fund (Project #0507) and by Genome Canada. The first author was supported by a grant from the Summer Research Training Program (SRTP) from the Schulich School of Medicine and Dentistry, University of Western Ontario.
==== Refs
Schmitz G Assmann G Scriver CR, Sly WS, Valle D Acid lipase deficiency The Metabolic Basis of Inherited Disease 1989 New York: McGraw Hill 1623 1644
Anderson RA Sando GN Cloning and expression of cDNA encoding human lysosomal acid lipase/cholesteryl ester hydrolase: similarities to gastric and lingual lipases J Biol Chem 1991 266 22479 22484 1718995
Koch G Lalley PA McAvoy M Shows TB Assignment of LIPA, associated with human acid lipase deficiency, to human chromosome 10 and comparative assignment to mouse chromosome 19 Somatic Cell Genet 1981 7 345 358 7292252 10.1007/BF01538859
Anderson RA Rao N Byrum RS Rothschild CB Bowden DW Hayworth R Pettenati M In situ localization of the genetic locus encoding the lysosomal acid lipase/cholesteryl esterase (LIPA) deficient in Wolman disease to chromosome 10q23.2-q23.3 Genomics 1993 15 245 247 8432549 10.1006/geno.1993.1052
Burton BK Reed SP Acid lipase cross-reacting material in Wolman disease and cholesterol ester storage disease Am J Hum Genet 1981 33 203 208 6782865
Aslanidis C Ries S Fehringer P Buchler C Klima H Schmitz G Genetic and biochemical evidence that CESD and Wolman disease are distinguished by residual lysosomal acid lipase activity Genomics 1996 33 85 93 8617513 10.1006/geno.1996.0162
Ries S Aslanidis C Fehringer P Carel JC Gendrel D Schmitz G A new mutation in the gene for lysosomal acid lipase leads to Wolman disease in an African kindred J Lipid Res 1996 37 1761 1765 8864960
Ginsberg HN Le NA Short MP Ramakrishnan R Desnick RJ Suppression of apolipoprotein B production during treatment of cholesteryl ester storage disease with lovastatin: implications for regulation of apolipoprotein B synthesis J Clin Invest 1987 80 1692 1697 3680522
Tarantino MD McNamara DJ Granstrom P Ellefson RD Unger EC Udall JN Jr Lovastatin therapy for cholesterol ester storage disease in two sisters J Pediatr 1991 118 131 135 1986080
Glueck CJ Lichtenstein P Tracy T Speirs J Safety and efficacy of treatment of pediatric cholesteryl ester storage disease with lovastatin Pediatr Res 1992 32 559 565 1480458
Levy R Ostlund RE JrSchonfeld G Wong P Semenkovich CF Cholesteryl ester storage disease: complex molecular effects of chronic lovastatin therapy J Lipid Res 1992 33 1005 1015 1431581
Leone L Ippoliti PF Antonicelli R Use of simvastatin plus cholestyramine in the treatment of lysosomal acid lipase deficiency J Pediatr 1991 119 1008 1009 1801793
Rassoul F Richter V Lohse P Naumann A Purschwitz K Keller E Long-term administration of the HMG-CoA reductase inhibitor lovastatin in two patients with cholesteryl ester storage disease Int J Clin Pharmacol Ther 2001 39 199 204 11380065
Di Bisceglie AM Ishak KG Rabin L Hoeg JM Cholesteryl ester storage disease: hepatopathology and effects of therapy with lovastatin Hepatology 1990 11 764 772 2347551
Altmann SW Davis HR JrZhu LJ Yao X Hoos LM Tetzloff G Iyer SP Maguire M Golovko A Zeng M Niemann-Pick C1 Like 1 protein is critical for intestinal cholesterol absorption Science 2004 303 1201 1204 14976318 10.1126/science.1093131
Smart EJ De Rose RA Farber SA Annexin 2-caveolin 1 complex is a target of ezetimibe and regulates intestinal cholesterol transport Proc Natl Acad Sci USA 2004 101 3450 3455 14985510 10.1073/pnas.0400441101
Garcia-Calvo M Lisnock J Bull HG Hawes BE Burnett DA Braun MP Crona JH Davis HR JrDean DC Detmers PA Graziano MP Hughes M Macintyre DE Ogawa A O'neill KA Iyer SP Shevell DE Smith MM Tang YS Makarewicz AM Ujjainwalla F Altmann SW Chapman KT Thornberry NA The target of ezetimibe is Niemann-Pick C1-Like 1 (NPC1L1) Proc Natl Acad Sci USA 2005 102 8132 8137 15928087 10.1073/pnas.0500269102
Kosoglou T Statkevich P Johnson-Levonas AO Paolini JF Bergman AJ Alton KB Ezetimibe: a review of its metabolism, pharmacokinetics and drug interactions Clin Pharmacokinet 2005 44 467 494 15871634
Knopp RHGH Truitt T Lipka LJ Lebeaut AP Suresh R Veltri EP Ezetimibe reduces low-density lipoprotein cholesterol: results of a phase III, randomized, double blind, placebo-controlled trial Atherosclerosis 2001 2 38
Davidson MH McGarry T Bettis R Melani L Lipka LJ LeBeaut AP Suresh R Sun S Veltri EP Ezetimibe coadministered with simvastatin in patients with primary hypercholesterolemia J Am Coll Cardiol 2002 40 2125 2134 12505224 10.1016/S0735-1097(02)02610-4
Ballantyne CM Houri J Notarbartolo A Melani L Lipka LJ Suresh R Sun S LeBeaut AP Sager PT Veltri EP Effect of ezetimibe coadministered with atorvastatin in 628 patients with primary hypercholesterolemia: a prospective, randomized, double-blind trial Circulation 2003 107 2409 2415 12719279 10.1161/01.CIR.0000068312.21969.C8
Kerzner B Corbelli J Sharp S Lipka LJ Melani L LeBeaut A Suresh R Mukhopadhyay P Veltri EP Efficacy and safety of ezetimibe coadministered with lovastatin in primary hypercholesterolemia Am J Cardiol 2003 91 418 424 12586255 10.1016/S0002-9149(02)03236-8
Dujovne CA Ettinger MP McNeer JF Lipka LJ LeBeaut AP Suresh R Yang B Veltri EP Efficacy and safety of a potent new selective cholesterol absorption inhibitor, ezetimibe, in patients with primary hypercholesterolemia Am J Cardiol 2002 90 1092 1097 12423709 10.1016/S0002-9149(02)02798-4
Knopp RH Gitter H Truitt T Bays H Manion CV Lipka LJ LeBeaut AP Suresh R Yang B Veltri EP Effects of ezetimibe, a new cholesterol absorption inhibitor, on plasma lipids in patients with primary hypercholesterolemia Eur Heart J 2003 24 729 741 12713767 10.1016/S0195-668X(02)00807-2
Melani L Mills R Hassman D Lipetz R Lipka L LeBeaut A Suresh R Mukhopadhyay P Veltri E Efficacy and safety of ezetimibe coadministered with pravastatin in patients with primary hypercholesterolemia: a prospective, randomized, double-blind trial Eur Heart J 2003 24 717 728 12713766 10.1016/S0195-668X(02)00803-5
Bradford RH Shear CL Chremos AN Franklin FA Nash DT Hurley DP Dujovne CA Pool JL Schnaper H Hesney M Expanded clinical evaluation of lovastatin (EXCEL) study results: III. Efficacy in modifying lipoproteins and implications for managing patients with moderate hypercholesterolemia Am J Med 1991 91 18S 24S 1867232 10.1016/0002-9343(91)90052-Y
Mantell G Lipid lowering drugs in atherosclerosis: the HMG-CoA reductase inhibitors Clin Exp Hypertens 1989 11 927 941
Hegele RA Guy J Ban MR Wang J NPC1L1 haplotype is associated with inter-individual variation in plasma low-density lipoprotein response to ezetimibe Lipids Health Dis 2005 4 16 16098225 10.1186/1476-511X-4-16
|
16255772
|
PMC1291391
|
CC BY
|
2021-01-04 16:39:19
|
no
|
Lipids Health Dis. 2005 Oct 28; 4:26
|
utf-8
|
Lipids Health Dis
| 2,005 |
10.1186/1476-511X-4-26
|
oa_comm
|
==== Front
Lipids Health DisLipids in Health and Disease1476-511XBioMed Central London 1476-511X-4-271626291110.1186/1476-511X-4-27ResearchAssociation between lifestyle factors and plasma adiponectin levels in Japanese men Tsukinoki Rumi [email protected] Kanehisa [email protected] Kunio [email protected] Department of Social and Environmental Medicine, Osaka University School of Medicine, 2-2 Yamada-oka, Suita, Osaka 565-0871, JAPAN2005 2 11 2005 4 27 27 28 9 2005 2 11 2005 Copyright © 2005 Tsukinoki et al; licensee BioMed Central Ltd.2005Tsukinoki et al; licensee BioMed Central Ltd.This is an Open Access article distributed under the terms of the Creative Commons Attribution License (), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
Background
Adiponectin is an adipocyte-specific protein that plays a role in obesity, insulin resistant, lipid metabolism, and anti-inflammation. Hypoadiponectinemia may be associated with a higher risk for type 2 diabetes and cardiovascular disease. Some studies suggest that adiponectin levels are modulated by lifestyle factors, but little is known about the associations between lifestyle factors and plasma adiponectin levels in Japanese people. We therefore investigated the associations between lifestyle factors and plasma adiponectin levels in general Japanese men.
Methods
The subjects were 202 Japanese male workers who participated in an annual health check. They provided details about anthropometrical data, blood collection, their use of prescribed medication, and the clinical history of their families. They also completed a self-administered questionnaire about their lifestyles.
Results
Subjects with plasma adiponectin levels below 4.0 μg/ml had significantly lower levels of HDL cholesterol and higher levels of BMI, SBP, DBP, total cholesterol, FBG, and platelets than did subjects with higher adiponectin levels. In multiple logistic regression after multiple adjustment, a plasma adiponectin level below 4.0 μg/ml was significantly associated with smoking (odds ratio [OR] = 2.08, 95% confidence interval [CI] = 1.01–4.30), a daily diet rich in deep-yellow vegetables (OR = 0.25, 95% CI= 0.07–0.91), frequent eating out (OR = 2.45, 95% CI = 1.19–5.08), and physical exercise two or more times a week (OR = 0.21, 95% CI = 0.06–0.74).
Conclusion
Our findings show that adiponectin levels in general Japanese men are independently related to smoking, dietary factors, and physical exercise. We think that lifestyle habits might independently modulate adiponectin levels and that adiponectin might be the useful biomarker helping people to avoid developing type 2 diabetes and cardiovascular disease by modifying their lifestyles.
adiponectinsmokingdietary factorphysical exercisegeneral Japanese men
==== Body
Background
Adiponectin, an adipocyte-specific protein and one of the adipocytekines, is a 244-amino acid peptide with a structure highly homologous to complement factor C1q, collagen VIII, and collagen X [1,2]. Identified in the human adipose tissue cDNA library, it is encoded by adipose most abundant gene transcript 1 (apM1) [1] and is found in high concentrations in the peripheral circulation [2]. Adiponectin expression reflects peroxisome proliferators-activated receptor γ (PPAR-γ) [3,4], and is associated with the expression of tumor necrosis factor-α (TNF-α) [5]. Adiponectin expression is reduced in obesity individuals [2], and it is associated with lipid metabolism [6-8]. It modulates insulin action and resistance [9,10], and low adiponectin levels predict the development of type 2 diabetes [7,11-13]. And adiponectin play role in anti-inflammatory factor, and it is also related to the development of atherosclosis, hypertension, and coronary heart disease [14-17], and some reports show that adiponectin levels are associated with the inflammatory factors C-reactive protein (CRP), TNF-α, interleukin-6 (IL-6) and fibrinogen [18-20]. Male Japanese patients with hypoadiponectinemia (<4.0 μg/ml) show a significant 2-fold increase in the prevalence of coronary artery disease (CAD), independent of well-known CAD risk factors [14], and adiponectin levels below 4 μg/ml are closely associated with the clinical phenotype of the metabolic syndrome in Japanese men [21].
Although cross-sectional studies and studies in weigh-loss programs suggest that adiponectin levels are modulated by lifestyle factors such as nutritional variables, moderate alcohol intake, and smoking [22-27], little is known about the associations between lifestyle factors and plasma adiponectin levels in Japanese [27]. Iwashima et al. have shown in a study of 98 healthy Japanese men and 233 Japanese men with hypertension, diabetes, and hyperlipidemia that adiponectin levels are associated with habitual smoking [27]. Therefore, we investigated cross-sectionally the associations between lifestyle factors and plasma adiponectin levels in general Japanese men.
Methods
Subjects
The subjects were 202 Japanese male workers at a metal plant who were participating in an annual health check for employees in October 2003 (Figure 1). Anthropometric dates and blood samples were collected from each participant by trained medical staff. All subjects completed a questionnaire that asked for the worker's medical history and family clinical history, and 195 of them (96.5%) also completed a self-administered detailed questionnaire about lifestyle habits, mental stress, occupational status, etc. Subjects who received medication for diabetes, hypertension, or cancer were excluded from the study. This study was approved by the Ethics Committee at Osaka University School of Medicine, and written informed consent was obtained from all subjects.
Figure 1 A flow-chart of the sampling procedure for 202 Japanese male factory workers.
Assessment of anthropometrical data
The height (without shoes) of each subject was measured in centimeters, and weight (without shoes and in light clothing) was measured in kilograms (TANITA). Body mass index (BMI) was calculated as weight in kilograms over height in meters squared.
Blood pressure was measured, by trained nurses using a digital blood-pressure monitor (Inteli Sense, HEM-907, OMRON) on the right arm, twice with the subject in the sitting position and with at least 5 min rest between the two measurements. The values used in this study were the average of the two measurements.
Evaluation of lifestyle factors and medical history
Lifestyle habits were assessed by using a self-administered questionnaire that asked about physical activity, habitual dietary intake, alcohol drinking habits, and smoking. The physical activity of the subjects was assessed by asking them about physical exercise, hours of walking on weekdays, sleeping hours, and hours of TV-watching, work, and physical activity in leisure time. Habitual dietary intake was assessed by asking about the usual average consumption of 15 foods and about 5 behaviors: the frequency of eating out, the meal for weigh loss, whether they ate within the 2 hours before bedtime, whether they eat too much, whether they eat too fast. The frequency of food consumption was queried using four categories: everyday, often, sometimes, and never. Alcohol drinking habits were assessed by ask about drinking frequency and alcohol consumption per occasion. Subjects were asked about smoking status and were classified into three categories: current smoker, ex-smoker, and nonsmoker. Current smokers and ex-smokers were asked about the number of cigarettes smoked each day and about how many years they had been smoking.
Public health nurses questioned all 202 of the subjects about their medical history and their family clinical history.
Measurement of Biochemical Variables
All blood data except leptin and adiponectin levels were measured in one laboratory (Shionogi Institute for Medical Science, Japan). Within two hours after blood samples for adiponectin and leptin were obtained, they were centrifuged at 3000 rpm for 25 min at -4°C before being stored at -80°C until the levels of the two adipocytekines were assayed. The laboratory-measured values of the serum lipids came from the Cholesterol Reference Method Laboratory Network, and the standardization values came from the Center for Disease Control and the Prevention/Cholesterol Reference Method Laboratory Network [28].
Plasma adiponectin concentration was determined in duplicate with an ELIZA assay (Otsuka Assay Institute, Japan) [2].
Statistical Analysis
Plasma adiponectin concentrations were classified into two categories: <4.0 μg/ml (hypoadiponectinemia) and ≥4.0 μg/ml [12,19]. The statistics listed in Table 1 that described adiponectin levels, clinical characteristics, and lifestyle habits are means ± the standard deviation (SD). The chi-square test was used to compare dichotomous variables, and t testing was used to compare means between the two groups classified according to adiponectin level. Associations between adiponectin levels and clinical factors in these two groups were examined by using an age-adjusted partial correlation coefficient.
Table 1 Clinical Characteristics of 202 Japanese men by adiponectin levels
Adiponectin Levels
All < 4.0 μg/mL ≥ 4.0 μg/mL P*
N 202 77 125
Adiponectin (μg/mL)
Mean 4.9 ± 2.2 2.9 ± 0.8 6.1 ± 1.9 .000
Median (Min, Max) 4.5 (0.58, 15.30) 3.0 (0.58, 3.96) 5.7 (4.02, 15.30) .000
Age (years) 42.0 ± 10.3 42.3 ± 9.1 41.7 ± 11.0 .670
BMI (kg/m2) 23.6 ± 2.8 24.3 ± 2.9 23.2 ± 2.6 .005
SBP (mmHg) 126.0 ± 14.2 128.8 ± 16.2 124.2 ± 12.5 .035
DBP (mmHg) 75.7 ± 12.0 78.9 ± 13.4 73.7 ± 10.5 .004
Total chol. (mg/dL) 203.6 ± 36.0 212.1 ± 37.1 198.4 ± 34.5 .008
HDL chol. (mg/dL) 59.9 ± 14.2 55.3 ± 11.9 63.8 ± 14.7 .000
LDL chol. (mg/dL) 134.0 ± 34.2 139.5 ± 33.3 129.9 ± 34.5 .053
FBG (mg/dL) † 89.7 ± 13.1 92.5 ± 16.9 88.0 ± 9.6 .000
Platelet (× 104 cells/μL) 23.6 ± 5.0 25.0 ± 5.5 22.6 ± 4.5 .002
Data are shown in mean ± SD.
* t-test
†log transformed
BMI, body mass index; SBP, systolic blood pressure; DBP, diastolic blood pressure; HDL, high-density lipoprotein; LDL, low-density lipoprotein; FBG, Fasting plasma glucose
Multiple logistic regression analysis models were used to evaluate the relations between hypoadiponectinemia and lifestyle habits. The dependent variable was the presence or absence of hypoadiponectinemia (<4.0 μg/ml), and the independent variables were four lifestyles habits: eating many deep-yellow vegetables, frequency of eating out, physical exercise, and smoking. In our model we adjusted for age, BMI, total cholesterol level, high-density lipoprotein (HDL) cholesterol level, hypertension, hyperglycemia, and family clinical history. Age was classified into four categories: 20–29 years, 30–39 years, 40–49 years, and 50–59 years. The subjects were divided into three categories according to BMI (≤20 kg/m2, 20.1–24.9 kg/m2, ≥25 kg/m2) and total cholesterol level (<160 mg/dl, 160–219.9 mg/dl, ≥220 mg/dl) and into two categories according to the presence or absence of hypertension (systolic blood pressure ≥130 mmHg and/or diastolic blood pressure ≥85 mmHg), low HDL cholesterol levels (<40 mg/dl), and hyperglycemia (a fasting blood glucose level ≥110 mg/dl). The total numbers of family members with a clinical history of diabetes mellitus, hypertension, stroke, heart disease, gout, or cancer (yes = 1, no = 0) were divided into four categories: 0, 1, 2, and 3+. All p values presented are two-tailed, and p <0.05 was considered statistically significant. Statistical analyses were performed using the statistical software SPSS version 11.5 (Texas Instruments, Chicago, IL) [29]
Results
Statistics describing the clinical characteristics and adiponectin levels of the hypoadiponectinemic and normoadiponectinemic groups are listed in Table 1. The hypoadiponectinemia group had significantly higher levels of BMI, systolic blood pressure (SBP), diastolic blood pressure (DBP), total cholesterol, fasting blood glucose (FBG) and platelets and had significantly lower levels of HDL cholesterol. The number of family members with a clinical history was also greater in the hypoadiponectinemia group (1.25 ± 1.39 points /6 points) than in the normoadiponectinemic group (1.0 ± 1.03 points/6 points) (p < 0.18, t-test).
The partial correlation of adiponectin levels with selected anthropometric and biochemical factors are shown in Table 2. After adjustment for age, adiponectin levels were negatively correlated with BMI, SBP, DBP, total cholesterol, FBG, and platelet level and were positively correlated with HDL cholesterol in all participants. After adjustment for age and BMI, adiponectin levels were significantly correlated with lipid and platelet levels and trended to associate with DBP and FBG.
Table 2 Partial correlation of plasma adiponectin with anthropometric and biochemical factors (n = 202)
r(Age-adjusted) p r(Age,BMI-adjusted) p
BMI -0.29 **
SBP -0.16 * -0.05 0.46
DBP -0.22 ** -0.11 0.11
Total chol. -0.27 ** -0.20 *
HDL chol. 0.34 ** 0.30 **
FBG† -0.19 * -0.13 *
Platelet -0.18 * -0.17 0.07
*p < 0.05, **p < 0.01
†:log transformed
Relations between lifestyle habits and plasma adiponectin levels can be seen in Table 3. Fewer than one fifth of the subjects who filled out the lifestyle questionnaire exercised at least twice a week, but the number of them that were hypoadiponectinemic was significantly smaller than the number that were normoadiponectinemic (2.1% versus 15.4% of the subjects who filled out the lifestyle questionnaire). The hypoadiponectinemic current smokers were 18.5% of the subjects who filled out the lifestyle questionnaire and the normoadiponectemic ex-smokers and nonsmokers were 39.0% of the subjects who filled out the lifestyle questionnaire. Eating many deep-yellow vegetables (p < 0.075) and the frequency of eating out (p < 0.053) trended to be associated with adiponectin levels but alcohol drinking habits did not. The results listed in Table 3 thus indicate that smoking, eating many deep-yellow vegetables, eating out frequently, and getting physical exercise are hypoadiponectinemia-related lifestyle factors.
Table 3 Characteristics of Lifestyle habits by adiponectin levels (n = 195)
Adiponectin
<4.0 μg/mL ≥4.0 μg/mL P*
N (%) N (%) N (%)
Physical exercise
≥ Twice /week 34 (17.4) 4 (2.1) 30 (15.4) .000
Hours of walking in weekdays
≥30 hr/day 103 (53.1) 35 (18.0) 68 (35.1) .373
Smoking
Nonsmoker, Ex-smoker 113 (57.9) 37 (19.0) 76 (39.0) .134
Current smoker 82 (42.1) 36 (18.5) 46 (23.6)
Alcohol drinking
None, Seldom 67 (34.4) 23 (11.8) 44 (22.6) .537
Many deep-yellow vegetables
Everyday 25 (12.8) 5 (2.6) 20 (10.3) .075
Many snack
Everyday 113 (57.9) 38 (19.5) 75 (38.5) .231
A lot of meat
Everyday, often 40 (20.5) 13 (6.7) 27(13.8) .583
Many deep-fry
Everyday, often 49 (25.1) 15 (7.7) 34 (17.4) .307
Many salty foods
Everyday, often 27 (13.8) 9 (4.6) 18 (9.2) .675
Frequency of eating out
< once a day 109 (55.9) 34 (17.4) 75 (38.5) .053
Sleeping hours
7 to 8 hr/day 47 (24.1) 14 (7.2) 33 (16.9) .231
* χ2 – test
We next used multiple logistic regression analysis to examine the association between hypoadiponectinemia and these four hypoadiponectinemia -related lifestyle factors (Table 4). After adjustment for multiple variables (age, BMI, hypertension, total cholesterol, HDL cholesterol, hyperglycemia, platelet level, and the number of family members with a clinical history), the risk of hypoadiponectinemia was significantly decreased by frequent physical exercise (= twice a week; Odds ratio [OR] = 0.21, 95% confidence interval [CI]= 0.06–0.74) and the frequent eating of many deep-yellow vegetables (everyday; OR = 0.25, 95%CI = 0.07–0.91) and was significantly increased by smoking (current smoker; OR = 2.08, 95% CI = 1.01–4.30) and by eating out frequently (=once a day; OR = 2.45, 95% CI = 1.19–5.08).
Table 4 Results of multiple logistic regression analysis of the association between hypoadiponectinemia and various lifestyle factors (n = 195). Dependent variable: hypoadiponectinemia (adiponectin <4.0 μg/ml = 1, ≥4.0 μg/ml = 0)
Variables OR 95% CI p
Physical exercise 0.21 (0.06 – 0.74) 0.015
(twice a week or more often = 1; once a week, a few times a month, never = 0)
Smoking 2.08 (1.01 – 4.30) 0.047
(smoker = 1; nonsmoker, ex-smoker = 0)
Many deep-yellow vegetables 0.25 (0.07 – 0.91) 0.035
(everyday = 1; often, sometimes, never = 0)
Frequency of eating out 2.45 (1.19 – 5.08) 0.016
(3 times a day, once a day = 1; a few days a week, a few days a month, never = 0)
OR: odds ratio, 95% CI: 95% confidence interval
Adjusted for age, BMI, hypertension, total cholesterol, HDL cholesterol, hyperglycemia, platelet count, and the number of family history (diabetes mellitus, gout, stroke, hypertension, heart disease, or cancer).
Discussion
We showed that that, in general Japanese men, eating out once a day or more and smoking are independently associated with a higher risk of hypoadiponectinemia, whereas getting physical exercise at least twice a week and eating many deep-yellow vegetables daily are significantly associated with higher adiponectin concentrations. We suggest that physical activity, dietary factors, and smoking are independently related to plasma adiponectin levels in general Japanese men. This is consistent with the results of a few earlier studies indicating that lifestyle factors may modulate adiponectin levels in the general male population.
We found that smoking is independently associated with hypoadiponectinemia in general Japanese men. Other investigators have reported that smoking is associated with adiponectin levels in healthy Japanese men as well as in Japanese men with hypertension, diabetes, and hyperlipidemia and that adiponectin levels are significantly lower in smokers after multiple adjustment. Furthermore, oxidative stress and nicotine reduce the expression and secretion of adiponectin in cultured mouse 3T3-L1 adipocytes [27]. Many studies also suggest that smoking induces inflammatory factors (TNF-α, CRP, IL-6, fibrinogen, etc.) that are risk factors for atherosclerosis and cardiovascular diseases [30,31]. Adiponectin accumulates in the subendothelial space of injured vascular walls and inhibits the transformation of macrophages to foam cells [15,32]. Our findings and those of previous studies suggest that adiponectin levels are a useful biomarker for evaluating the effects of smoking on the risk of atherosclerosis and cardiovascular diseases in the general Japanese population.
Our results also showed that daily diets rich in deep-yellow vegetables are associated with a significantly lower risk of hypoadiponectinemia and that eating out once or more a day is associated with a significantly higher risk of hypoadiponectinemia. Some previous reports have suggested that dietary factors are related to adiponectin levels in human beings. They showed that a Mediterranean-style diet and a high-fiber diet with a low caloric and low glycemic load were associated with higher adiponectin levels. Esposito et al. reported that a weight-loss program that included exercise increased the plasma adiponectin levels of obese women in a randomized trial [22]. They also showed that lower adiponectin levels were associated with diets including whole-grain products, legumes, fruits, vegetables, fish, and olive oil. Pischon et al. found in a cross-sectional study of 532 men without a history of cardiovascular disease that a diet with a high glycemic load was significantly associated with lower adiponectin levels that and carbohydrate intake tended to be associated with lower adiponectin levels [25]. Qi et al. showed in a cross-sectional analysis of 780 diabetic patients that diets low in glycemic load and high in fiber might increase plasma adiponectin concentrations [24]. A higher frequency of eating out has also been found to be associated with adverse nutritional consequences related to increased obesity [33]. Dietary factors are closely related to obesity and the development of type 2 diabetes and cardiovascular disease. We suggest that dietary factors independently modulate adiponectin levels in general Japanese men and that improving dietary factors can increase adiponectin levels and thereby reduce the risk of developing type 2 diabetes and cardiovascular disease.
We observed an independent association of more frequent physical exercise with higher adiponectin levels in general Japanese men. Adiponectin stimulates glucose utilization and fatty-acid oxidation by activating AMP kinase in muscle and enhances insulin action [34,35]. Exercise also improves glucose utilization and fatty-acid oxidation by activating AMP kinase in muscle [36]. Previous studies have evaluated small samples of the participants in exercise programs or weight-loss programs for the short term, and this has led to variability in results showing a relation between adiponectin and exercise [37-41]. Yokoyama et al. reported that aerobic exercise might increase plasma adiponectin levels in diabetes subjects when an intervention is accompanied by a reduction in weight or fat mass, but that study evaluated only 40 subjects and used only a three-week intervention [41]. The results of some studies with longer interventions suggested that regular physical activity and exercise in a weight-loss program increase adiponectin levels [22]. And previous epidemiological studies have shown that high levels of physical activity independently improved IL-6, C-reactive protein, leptin, TNF-α, and fibrinogen in healthy individuals [42,43]. We showed an independent association between exercise and adiponectin in general Japanese men. Our results suggested that regular exercise independently increased adiponectin levels.
Our study had some limitations. The cross-sectional design limited causal inferences. Although 71.2% of the hypoadiponectinemia subjects in our study did not change their diet habits and exercise within the last six months (chi-square test X2 = 4.36; df = 1 and p = 0.037), many cohort studies in the future should explain the association of adiponectin and lifestyle in Japanese men and women. Our assessment of dietary factors was based on self-reported dietary intake and questionnaires that asked about some dietary habits and the simple frequency of food factors. Although we thus did not evaluate dietary factors in detail, our results linking dietary factors and adiponectin levels were similar to those of previous studies. Moderate alcohol intake was not independently associated with higher adiponectin levels in our study. No report has explained the mechanisms that associate moderate alcohol intake and adiponectin expression, but Pischon et al. showed moderate alcohol consumption was independently associated with higher adiponectin levels in men living in the United States [25,26]. It is necessary to further study the association of adiponectin levels with alcohol consumption in the Japanese population.
In conclusion, our findings showed that adiponectin levels in general Japanese men are independently modulated smoking, dietary factors, and physical exercise. We have suggested that lifestyle habits might independently modulate adiponectin levels and that adiponectin might be the useful biomarker helping people prevent type 2 diabetes and cardiovascular disease by modifying their lifestyles.
Acknowledgements
We thank the participants in our study, are grateful to Sachiyo Tanaka, Toru Funahashi, MD, PhD, and Iichiro Shimomura, MD, PhD, for teaching us their method of adiponectin measurement, and thank our staff for helping collect data.
==== Refs
Maeda K Okubo K Shimomura I Funahashi T Matsuzawa Y Matsubara K cDNA cloning and expression of a novel adipose specific collagen-like factor, apM1 (AdiPose Most abundant Gene transcript 1) Biochem Biophys Res Commun 221 286 9 1996 Apr 16 8619847 10.1006/bbrc.1996.0587
Arita Y Kihara S Ouchi N Takahashi M Maeda K Miyagawa J Hotta K Shimomura I Nakamura T Miyaoka K Kuriyama H Nishida M Yamashita S Okubo K Matsubara K Muraguchi M Ohmoto Y Funahashi T Matsuzawa Y Paradoxical decrease of an adipose-specific protein, adiponectin, in obesity Biochem Biophys Res Commun 257 79 83 1999 Apr 2 10092513 10.1006/bbrc.1999.0255
Maeda N Takahashi M Funahashi T Kihara S Nishizawa H Kishida K Nagaretani H Matsuda M Komuro R Ouchi N Kuriyama H Hotta K Nakamura T Shimomura I Matsuzawa Y PPARgamma ligands increase expression and plasma concentrations of adiponectin, an adipose-derived protein Diabetes 2001 50 2094 9 11522676
Combs TP Wagner JA Berger J Doebber T Wang WJ Zhang BB Tanen M Berg AH O'Rahilly S Savage DB Chatterjee K Weiss S Larson PJ Gottesdiener KM Gertz BJ Charron MJ Scherer PE Moller DE Induction of adipocyte complement-related protein of 30 kilodaltons by PPARgamma agonists: a potential mechanism of insulin sensitization Endocrinology 2002 143 998 1007 11861525 10.1210/en.143.3.998
Maeda N Shimomura I Kishida K Nishizawa H Matsuda M Nagaretani H Furuyama N Kondo H Takahashi M Arita Y Komuro R Ouchi N Kihara S Tochino Y Okutomi K Horie M Takeda S Aoyama T Funahashi T Matsuzawa Y Diet-induced insulin resistance in mice lacking adiponectin/ACRP30 Nat Med 2002 8 731 7 12068289 10.1038/nm724
Matsubara M Maruoka S Katayose S Decreased plasma adiponectin concentrations in women with dyslipidemia J Clin Endocrinol Metab 2002 87 2764 9 12050247 10.1210/jc.87.6.2764
Hotta K Funahashi T Arita Y Takahashi M Matsuda M Okamoto Y Iwahashi H Kuriyama H Ouchi N Maeda K Nishida M Kihara S Sakai N Nakajima T Hasegawa K Muraguchi M Ohmoto Y Nakamura T Yamashita S Hanafusa T Matsuzawa Y Plasma concentrations of a novel, adipose-specific protein, adiponectin, in type 2 diabetic patients Arterioscler Thromb Vasc Biol 2000 20 1595 9 10845877
Cnop M Havel PJ Utzschneider KM Carr DB Sinha MK Boyko EJ Retzlaff BM Knopp RH Brunzell JD Kahn SE Relationship of adiponectin to body fat distribution, insulin sensitivity and plasma lipoproteins: evidence for independent roles of age and sex Diabetologia 2003 46 459 69 12687327
Weyer C Funahashi T Tanaka S Hotta K Matsuzawa Y Pratley RE Tataranni PA Hypoadiponectinemia in obesity and type 2 diabetes: close association with insulin resistance and hyperinsulinemia J Clin Endocrinol Metab 2001 86 1930 5 11344187 10.1210/jc.86.5.1930
Yamauchi T Kamon J Waki H Terauchi Y Kubota N Hara K Mori Y Ide T Murakami K Tsuboyama-Kasaoka N Ezaki O Akanuma Y Gavrilova O Vinson C Reitman ML Kagechika H Shudo K Yoda M Nakano Y Tobe K Nagai R Kimura S Tomita M Froguel P Kadowaki T The fat-derived hormone adiponectin reverses insulin resistance associated with both lipoatrophy and obesity Nat Med 2001 7 941 6 11479627 10.1038/90984
Hotta K Funahashi T Bodkin NL Ortmeyer HK Arita Y Hansen BC Matsuzawa Y Circulating concentrations of the adipocyte protein adiponectin are decreased in parallel with reduced insulin sensitivity during the progression to type 2 diabetes in rhesus monkeys Diabetes 2001 50 1126 33 11334417
Spranger J Kroke A Mohlig M Bergmann MM Ristow M Boeing H Pfeiffer AF Adiponectin and protection against type 2 diabetes mellitus Lancet 361 226 8 2003 Jan 18 12547549 10.1016/S0140-6736(03)12255-6
Chandran M Phillips SA Ciaraldi T Henry RR Adiponectin: more than just another fat cell hormone? Diabetes Care 2003 26 2442 50 12882876
Kumada M Kihara S Sumitsuji S Kawamoto T Matsumoto S Ouchi N Arita Y Okamoto Y Shimomura I Hiraoka H Nakamura T Funahashi T Matsuzawa Y Osaka CAD Study Group Coronary artery disease. Association of hypoadiponectinemia with coronary artery disease in men Arterioscler Thromb Vasc Biol 23 85 9 2003 Jan 1 12524229 10.1161/01.ATV.0000048856.22331.50
Ouchi N Kihara S Arita Y Maeda K Kuriyama H Okamoto Y Hotta K Nishida M Takahashi M Nakamura T Yamashita S Funahashi T Matsuzawa Y Novel modulator for endothelial adhesion molecules: adipocyte-derived plasma protein adiponectin Circulation 100 2473 6 1999 Dec 21–28 10604883
Okamoto Y Kihara S Ouchi N Nishida M Arita Y Kumada M Ohashi K Sakai N Shimomura I Kobayashi H Terasaka N Inaba T Funahashi T Matsuzawa Y Adiponectin reduces atherosclerosis in apolipoprotein E-deficient mice Circulation 106 2767 70 2002 Nov 26 12451000 10.1161/01.CIR.0000042707.50032.19
Pischon T Girman CJ Hotamisligil GS Rifai N Hu FB Rimm EB Plasma adiponectin levels and risk of myocardial infarction in men JAMA 291 1730 7 2004 Apr 14 15082700 10.1001/jama.291.14.1730
Ouchi N Kihara S Funahashi T Nakamura T Nishida M Kumada M Okamoto Y Ohashi K Nagaretani H Kishida K Nishizawa H Maeda N Kobayashi H Hiraoka H Matsuzawa Y Reciprocal association of C-reactive protein with adiponectin in blood stream and adipose tissue Circulation 107 671 4 2003 Feb 11 12578865 10.1161/01.CIR.0000055188.83694.B3
Krakoff J Funahashi T Stehouwer CD Schalkwijk CG Tanaka S Matsuzawa Y Kobes S Tataranni PA Hanson RL Knowler WC Lindsay RS Inflammatory markers, adiponectin, and risk of type 2 diabetes in the Pima Indian Diabetes Care 2003 26 1745 51 12766104
Shetty GK Economides PA Horton ES Mantzoros CS Veves A Circulating adiponectin and resistin levels in relation to metabolic factors, inflammatory markers, and vascular reactivity in diabetic patients and subjects at risk for diabetes Diabetes Care 2004 27 2450 7 15451915
Ryo M Nakamura T Kihara S Kumada M Shibazaki S Takahashi M Nagai M Matsuzawa Y Funahashi T Adiponectin as a biomarker of the metabolic syndrome Circ J 2004 68 975 81 15502375 10.1253/circj.68.975
Esposito K Pontillo A Di Palo C Giugliano G Masella M Marfella R Giugliano D Effect of weight loss and lifestyle changes on vascular inflammatory markers in obese women: a randomized trial JAMA 289 1799 804 2003 Apr 9 12684358 10.1001/jama.289.14.1799
Ziccardi P Nappo F Giugliano G Esposito K Marfella R Cioffi M D'Andrea F Molinari AM Giugliano D Reduction of inflammatory cytokine concentrations and improvement of endothelial functions in obese women after weight loss over one year Circulation 105 804 9 2002 Feb 19 11854119 10.1161/hc0702.104279
Qi L Rimm E Liu S Rifai N Hu FB Dietary glycemic index, glycemic load, cereal fiber, and plasma adiponectin concentration in diabetic men Diabetes Care 2005 28 1022 8 15855561
Pischon T Girman CJ Rifai N Hotamisligil GS Rimm EB Association between dietary factors and plasma adiponectin concentrations in men Am J Clin Nutr 2005 81 780 6 15817852
Shai I Rimm EB Schulze MB Rifai N Stampfer MJ Hu FB Moderate alcohol intake and markers of inflammation and endothelial dysfunction among diabetic men Diabetologia 2004 47 1760 7 15502925 10.1007/s00125-004-1526-0
Iwashima Y Katsuya T Ishikawa K Kida I Ohishi M Horio T Ouchi N Ohashi K Kihara S Funahashi T Rakugi H Ogihara T Association of hypoadiponectinemia with smoking habit in men Hypertension 2005 45 1094 100 15897361 10.1161/01.HYP.0000169444.05588.4c
Myers GL Kimberly MM Waymack PP Smith SJ Cooper GR Sampson EJ A reference method laboratory network for cholesterol: a model for standardization and improvement of clinical laboratory measurements Clin Chem 2000 46 1762 72 11067811
SPSS,Inc. SPSSx user's guide 1986 2 Chicago:SPSS
de Maat MP Pietersma A Kofflard M Sluiter W Kluft C Association of plasma fibrinogen levels with coronary artery disease, smoking and inflammatory markers Atherosclerosis 121 185 91 1996 Apr 5 9125293 10.1016/0021-9150(95)05716-1
Frohlich M Sund M Lowel H Imhof A Hoffmeister A Koenig W Independent association of various smoking characteristics with markers of systemic inflammation in men. Results from a representative sample of the general population (MONICA Augsburg Survey 1994/95) Eur Heart J 2003 24 1365 72 12871694 10.1016/S0195-668X(03)00260-4
Ouchi N Kihara S Arita Y Nishida M Matsuyama A Okamoto Y Ishigami M Kuriyama H Kishida K Nishizawa H Hotta K Muraguchi M Ohmoto Y Yamashita S Funahashi T Matsuzawa Y Adipocyte-derived plasma protein, adiponectin, suppresses lipid accumulation and class A scavenger receptor expression in human monocyte-derived macrophages Circulation 103 1057 63 2001 Feb 27 11222466
Kant AK Graubard BI Eating out in America, 1987–2000: trends and nutritional correlates Prev Med 2004 38 243 9 14715218 10.1016/j.ypmed.2003.10.004
Tomas E Tsao TS Saha AK Murrey HE Zhang CC Itani SI Lodish HF Ruderman NB Enhanced muscle fat oxidation and glucose transport by ACRP30 globular domain: acetyl-CoA carboxylase inhibition and AMP-activated protein kinase activation Proc Natl Acad Sci U S A 2002 99 16309 16313 12456889 10.1073/pnas.222657499
Yamauchi T Kamon J Minokoshi Y Ito Y Waki H Uchida S Yamashita S Noda M Kita S Ueki K Eto K Akanuma Y Froguel P Foufelle F Ferre P Carling D Kimura S Nagai R Kahn BB Kadowaki T Adiponectin stimulates glucose utilization and fatty-acid oxidation by activating AMP-activated protein kinase Nat Med 2002 8 1288 1295 12368907 10.1038/nm788
Nielsen JN Jorgensen SB Frosig C Viollet B Andreelli F Vaulont S Kiens B Richter EA Wojtaszewski JF A possible role for AMP-activated protein kinase in exercise-induced glucose utilization: insights from humans and transgenic animals Biochem Soc Trans 2003 31 186 90 12546682
Hara T Fujiwara H Nakao H Mimura T Yoshikawa T Fujimoto S Body composition is related to increase in plasma adiponectin levels rather than training in young obese men Eur J Appl Physiol 2005 94 520 6 15942768 10.1007/s00421-005-1374-8
Yatagai T Nishida Y Nagasaka S Nakamura T Tokuyama K Shindo M Tanaka H Ishibashi S Relationship between exercise training-induced increase in insulin sensitivity and adiponectinemia in healthy men Endocr J 2003 50 233 8 12803245 10.1507/endocrj.50.233
Hulver MW Zheng D Tanner CJ Houmard JA Kraus WE Slentz CA Sinha MK Pories WJ MacDonald KG Dohm GL Adiponectin is not altered with exercise training despite enhanced insulin action Am J Physiol Endocrinol Metab 2002 283 E861 5 12217905
Ferguson MA White LJ McCoy S Kim HW Petty T Wilsey J Plasma adiponectin response to acute exercise in healthy subjects Eur J Appl Physiol 2004 91 324 9 14586663 10.1007/s00421-003-0985-1
Yokoyama H Emoto M Araki T Fujiwara S Motoyama K Morioka T Koyama H Shoji T Okuno Y Nishizawa Y Effect of aerobic exercise on plasma adiponectin levels and insulin resistance in type 2 diabetes Diabetes Care 2004 27 1756 58 15220262
Panagiotakos DB Pitsavos C Chrysohoou C Kavouras S Stefanadis C ATTICA Study The associations between leisure-time physical activity and inflammatory and coagulation markers related to cardiovascular disease: the ATTICA Study Prev Med 2005 40 432 7 15530595 10.1016/j.ypmed.2004.07.010
Fung TT Hu FB Yu J Chu NF Spiegelman D Tofler GH Willett WC Rimm EB Leisure-time physical activity, television watching, and plasma biomarkers of obesity and cardiovascular disease risk Am J Epidemiol 152 1171 8 2000 Dec 15 11130623 10.1093/aje/152.12.1171
|
16262911
|
PMC1291392
|
CC BY
|
2021-01-04 16:39:19
|
no
|
Lipids Health Dis. 2005 Nov 2; 4:27
|
utf-8
|
Lipids Health Dis
| 2,005 |
10.1186/1476-511X-4-27
|
oa_comm
|
==== Front
Lipids Health DisLipids in Health and Disease1476-511XBioMed Central London 1476-511X-4-281627115210.1186/1476-511X-4-28ResearchHypercholesterolemia and apolipoprotein B expression: Regulation by selenium status Dhingra Sanjiv [email protected] Mohinder P [email protected] Department of Biophysics, Panjab University, Chandigarh-160014, India2005 5 11 2005 4 28 28 17 10 2005 5 11 2005 Copyright © 2005 Dhingra and Bansal; licensee BioMed Central Ltd.2005Dhingra and Bansal; licensee BioMed Central Ltd.This is an Open Access article distributed under the terms of the Creative Commons Attribution License (), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
Background
Apolipoprotein B (apoB) contains ligand-binding domain for the binding of LDL to LDL-R site, which enables the removal of LDL from circulation. Our recent data showed that selenium (Se) is involved in the lipid metabolism. The present study was aimed to understand the effect of Se deficiency (0.02 ppm) and selenium supplementation (1 ppm) on apoB expression in liver during hypercholesterolemia in male Sprague Dawley rats. Animals were fed with control and high cholesterol diet (2%) for 1 and 2 months. ApoB levels by ELISA and protein expression by western blot was done. Hepatic LDL receptor (LDL-R) activity (in vivo) and mRNA expression by RT-PCR was monitored.
Results
In selenium deficiency and on high cholesterol diet (HCD) feeding apoB levels increased and LDL-R expression decreased significantly after 2 months. On 1 ppm selenium supplementation apoB expression significantly decreased and LDL-R expression increased after 2 months. But after one month of treatment there was no significant change observed in apoB and LDL-R expression.
Conclusion
So the present study demonstrates that Se deficiency leads to up regulation of apoB expression during experimental hypercholesterolemia. Selenium supplementation upto 1 ppm leads to downregulation of apoB expression. Further, this study will highlight the nutritional value of Se supplementation in lipid metabolism.
==== Body
Background
The interaction between LDL and LDL receptor has a major role in determining plasma cholesterol levels [1,2] and apolipoprotein B (apoB) has the central role in this ligand-receptor interaction. Mutations in the apoB gene result in accumulation of LDL in circulation [3,4]. However, most of the studies suggested that one molecule of apoB exists per lipoproteinparticle, thus the quantity of apoB in fasting plasma predicts the number of LDL and VLDL particles [5,6]. Therefore, plasma apoB levels maybe a better assay of the concentration of atherogenic lipoproteinparticles than total or LDL cholesterol levels [7]. Furthermore, a cross-sectional study in patients who had coronary arterybypass graft surgery determined that apoB concentration wasa better discriminator than LDL cholesterol concentration inpredicting recurrent atherosclerotic disease in bypass grafts10 years after surgery [8].
Abnormalities in the apoB metabolism are responsible for the generation of hypercholesterolemia and increased risk of coronary heart disease [9]. Homma [10] reported a positive correlation between serum apoB levels and atherosclerotic conditions. Abraham et al. [11] have found a significantly increased production of apoB levels in cultured hepatocytes isolated from rats fed with atherogenic diet. Incubation of hepatocytes isolated from normal rats with added cholesterol resulted in an increased synthesis and secretion of apoB levels [12].
Several studies suggested that T3 is directly involved in the regulation of LDL-R and apoB expression [13,14]. Normally thyroid is the unique source of T4, but it secretes only 20% of the whole T3 in the body. Major amount of T3 is produced from T4 by 5'-deiodination in peripheral tissues [15]. This reaction is catalyzed by type-I 5'-iodothyronine deiodinase (5'-DI). Type-1 5'-iodothyronine deiodinase being a selenoprotein its activity decreases during selenium (Se) deficiency [16,17]. Hence during Se deficiency, T3 can not be produced in any quantity. This makes the role of Se important for lipid metabolism, since T3 is known to regulate the LDL-R level, which is further responsible for the maintenance of plasma cholesterol level. Wojcicki et al. [18] reported the protective role of Se against atherosclerosis.
Hence in view of all the above stated findings, present study is aimed to understand the effect of Se status on apoB levels under experimental hypercholesterolemic conditions in SD male rats. To the best of our knowledge, so far no other study has been reported linking Se status with apoB expression during hypercholesterolemia.
Results
Selenium levels
Se levels in the serum and liver decreased significantly (p < 0.001) in Se deficient groups (Ia and Ib) and increased in 1 ppm Se supplemented diet fed groups (IIIa and IIIb) in comparison to respective adequate groups (IIa and IIb). Significant decrease (p < 0.001) in the level was observed in HCD fed groups as compared to respective controls in all the three Se status i.e. deficient, adequate and excess groups. In deficient groups (Ia and Ib) and in HCD fed adequate group, Se level decreased significantly (p < 0.001), whereas in 1 ppm Se supplemented groups the level increased significantly (p < 0.001) after 2 months in comparison to 1-month data (Table 1).
Table 1 Selenium levels in liver (μg/g) and serum (μg/L), GSH-Px levels in liver (μmoles of NADPH oxidized/min/mg protein) after 1 and 2 months of control and high cholesterol diet (HCD) feeding.
Treatment period Se deficient group Se adequate group Se excess group
Control Ia HCD Ib Control IIa HCD IIb Control IIIa HCD IIIb
Se in liver
1 month 1.83 ± 0.26DDD 1.20 ± 0.14***### 3.85 ± 0.29 3.11 ± 0.39** 4.57 ± 0.43LL 2.51 ± 0.29***B
2 months 1.23 ± 0.22DDDAA 0.73 ± 0.05***AAA### 3.97 ± 0.36 2.54 ± 0.27***A 6.51 ± 0.65AAALLL 4.38 ± 0.44***AAABBB
Se in serum
1 month 1.54 ± 0.15DDD 0.76 ± 0.04***### 2.51 ± 0.24 2.11 ± 0.27* 3.84 ± 0.34LLL 2.62 ± 0.27***BB
2 month 0.98 ± 0.07DDDAAA 0.51 ± 0.02***AAA### 2.82 ± 0.29 1.62 ± 0.17***AA 4.73 ± 0.39AALLL 3.58 ± 0.31***AAABBB
GSH-Px
1 month 281.45 ± 34.89DDD 354.19 ± 32.79**### 387.42 ± 25.86 467.19 ± 27.66*** 568.82 ± 52.94LLL 686.19 ± 50.58**BBB
2 month 160.33 ± 22.78DDDAAA 463.73 ± 39.91***AAA### 393.16 ± 49.67 593.28 ± 35.38***AAA 672.91 ± 58.16AALLL 825.16 ± 63.85**AABBB
Data is represented as mean ± SD from 6 observations.
*p<0.05, **p<0.01, ***p<0.001 represent comparison between control and HCD groups; Ap<0.05, AAp<0.01, AAAp<0.001 comparison between 1 and 2 months; DDDp<0.001 comparison between group Ia and IIa; ###p<0.001 comparison between group Ib and IIb; LLp<0.01, LLLp<0.001 comparison between group IIa and IIIa; Bp<0.05, BBp<0.01, BBBp<0.001 comparison between group IIb and IIIb.
Glutathione peroxidase activity
Glutathione peroxidase (GSH-Px) activity in liver decreased significantly (p < 0.001) in Se deficiency (Ia and Ib) and it increased on 1 ppm Se supplementation (IIIa and IIIb) in comparison to respective adequate groups (IIa and IIb). On HCD feeding significant increase (p < 0.001) was observed in all the groups in comparison to respective controls. In Se deficient control group GSH-Px level decreased, whereas in HCD supplemented Se deficient and Se adequate as well as excess Se fed groups the level increased significantly (p < 0.001) after 2 months as compared to 1-month data (Table 1).
Total Cholesterol and LDL Level
In all the three Se status groups on HCD feeding significant increase (p < 0.001) in total cholesterol and LDL-cholesterol concentration was observed in comparison to respective control groups. In Se deficient groups (Ia and Ib) total cholesterol and LDL level increased and on 1 ppm Se supplementation it decreased significantly (p < 0.001) in comparison to respective adequate groups (IIa and IIb). In both the Se deficient groups (Ia and Ib) and in HCD fed adequate group lipid level increased significantly (p < 0.001), whereas it decreased on Se supplementation after 2 months in comparison to 1-month treatment period (Table 2).
Table 2 Total cholesterol (mg/dl), LDL-cholesterol (mg/dl) and apolipoprotein B (A405) levels in serum after 1 and 2 months of control and high cholesterol diet (HCD) feeding schedule.
Treatment period Se deficient group Se adequate group Se excess group
Control Ia HCD Ib Control IIa HCD IIb Control IIIa HCD IIIb
Cholesterol
1 month 107.87 ± 6.87DDD 246.54 ± 11.96##*** 83.62 ± 5.45 215.23 ± 13.45*** 74.43 ± 6.28L 184.29 ± 11.73***BB
2 months 120.23 ± 9.53DDDA 295.14 ± 13.43##***AAA 85.23 ± 5.24 269.82 ± 10.67***AAA 62.73 ± 5.38AALLL 165.34 ± 10.60***BBBA
LDL
1 month 41.47 ± 4.23DDD 94.82 ± 7.38#*** 32.16 ± 2.29 82.78 ± 7.17*** 28.62 ± 2.42L 70.88 ± 5.36***BB
2 months 48.35 ± 3.71ADDD 113.50 ± 9.16#***A 32.78 ± 2.75 99.73 ± 9.32***A 24.12 ± 2.06AALLL 60.59 ± 5.83***BBBAA
ApoB
1 month 0.195 ± 0.018 0.198 ± 0.019 0.191 ± 0.017 0.196 ± 0.016 0.189 ± 0.016 0.193 ± 0.018
2 months 0.287 ± 0.024DDDAAA 0.345 ± 0.028**AAA## 0.196 ± 0.016 0.275 ± 0.023***AAA 0.130 ± 0.011AAALLL 0.162 ± 0.015**AABBB
Data is represented as mean ± SD from 6 observations.
**p<0.01, ***p<0.001 represent comparison between control and HCD groups; Ap<0.05, AAp<0.01, AAAp<0.001 comparison between 1 and 2 months; DDp<0.01, DDDp<0.001 comparison between group Ia and IIa; ##p<0.01, ###p<0.001 comparison between group Ib and IIb; Lp<0.05, LLp<0.01, LLLp<0.001 comparison between group IIa and IIIa; BBp<0.01, BBBp<0.001 comparison between group IIb and IIIb.
T3 and T4 levels
Levels of T3 decreased and T4 increased significantly (p < 0.001) on HCD feeding in comparison to respective controls in all the three Se status groups. In Se deficiency (Ia and Ib) T3 decreased and T4 level increased in comparison to respective adequate groups, whereas on 1 ppm Se supplementation T3 level increased and T4 level decreased significantly (p < 0.001). In both the Se deficient groups and in HCD fed adequate group (IIb) T3 decreased and T4 increased significantly and in Se supplemented groups T3 increased and T4 decreased after 2 months in comparison to 1-month data (Fig. 1a &1b).
Figure 1 T3 (a), T4 (b) and 5'-DI (c) levels in serum in different groups: Ia-Se deficient control, Ib-Se deficient+HCD, IIa-Se adequate control, IIb-Se adequate+HCD, IIIa-Se excess control, IIIb-Se excess+HCD after 1 and 2 months. Data is represented as mean ± SD from 6 observations. **p<0.01, ***p<0.001 represent comparison between control and HCD groups; AAp<0.01, AAAp<0.001 comparison between 1 and 2 months; DDp<0.01, DDDp<0.001 comparison between group Ia and IIa; ##p<0.01, ###p<0.001 comparison between group Ib and IIb; Lp<0.05, LLp<0.01, LLLp<0.001 comparison between group IIa and IIIa; BBp<0.01, BBBp<0.001 comparison between group IIb and IIIb.
5'-DI activity and mRNA expression
5'-DI activity as well as mRNA expression in liver decreased significantly (p < 0.001) on HCD feeding and during Se deficiency in comparison to respective controls. On 1 ppm Se supplementation, significant increase (p < 0.001) in enzyme activity and mRNA expression was observed in comparison to adequate groups. In Se deficient groups (Ia and Ib) and in HCD fed adequate group the activity and mRNA expression decreased and it increased significantly in 1 ppm Se supplemented groups after 2 months in comparison to 1 month data (Fig. 1c and Fig. 2a &2b).
Figure 2 5'-DI mRNA analysis in liver in different groups: Ia-Se deficient control, IIa-Se adequate control, IIIa-Se excess control, Ib-Se deficient+HCD, IIb-Se adequate+HCD, IIIb-Se excess+HCD after 1 and 2 months. (a) mRNA expression by RT-PCR. (b) expression was quantified by densitometric analysis. β-actin was used as an internal control. Data is expressed as mean ± SD from 4 observations. *p<0.05, **p<0.01 represent comparison between control and HCD groups; Ap<0.05 comparison between 1 and 2 months; Dp<0.05, DDp<0.01 comparison between group Ia and IIa; ##p<0.01 comparison between group Ib and IIb; Lp<0.05, LLLp<0.001 comparison between group IIa and IIIa; Bp<0.05, BBBp<0.001 comparison between group IIb and IIIb.
LDL-R activity and mRNA expression
Percent remaining counts in blood after 120 h were almost same in all the groups after one month, hence no change was observed in LDL-R activity after one month of diet feeding schedule (Fig. 3a &3b). However after 2 months, the percent remaining counts were higher in Se deficient groups in comparison to adequate diet fed groups, so the LDL-R activity decreased significantly (p < 0.001) in Se deficiency. On 1 ppm selenium supplementation LDL-R activity increased significantly in comparison to adequate groups. On HCD feeding the receptor activity decreased significantly (p < 0.001) in all the three selenium status groups. After 2 months receptor activity decreased (p < 0.001) in Se deficient groups (Ia and Ib) and in HCD fed adequate group, whereas it increased significantly on 1 ppm selenium supplementation in comparison to 1 month treatment period (Fig. 4a &4b).
Figure 3 LDL-R activity in-vivo during selenium deficiency and on HCD feeding in different groups. (a) after 1 month of control diet feeding. (b) after 1 month of HCD feeding. Radiolabelled LDL was injected to the rats, % decrease in counts in blood with time was taken as a measure of clearance of LDL from animal blood and in turn the LDL-R activity.
Figure 4 LDL-R activity in-vivo during selenium deficiency and on HCD feeding in different groups. (a) after 2 months of control diet feeding. (b) after 2 months of HCD feeding. Radiolabelled LDL was injected to the rats, % decrease in counts in blood with time was taken as a measure of clearance of LDL from animal blood and in turn the LDL-R activity.
RT-PCR products of expected size i.e. 341 bp and 236 bp were obtained for LDL-R and β-actin. mRNA expression followed the same trend as it was observed for LDL-R activity i.e. no significant change was observed in expression after 1 month of treatment period. But after 2 months of diet feeding schedule mRNA expression decreased significantly (p < 0.001) in Se deficiency (Ia and Ib) in comparison to adequate groups i.e. 31% and 68% decrease was observed in groups Ia and Ib in comparison to IIa and IIb respectively. On 1 ppm Se supplementation, significant increase (p < 0.001) in RNA expression was observed in comparison to adequate groups. On HCD feeding in all the three selenium status groups significant decrease in the expression was observed. In Se deficient groups (Ia and Ib) and in HCD fed adequate group mRNA expression decreased and it increased significantly in 1 ppm Se supplemented groups after 2 months in comparison to 1 month data (Fig. 5a &5b).
Figure 5 Hepatic LDL-R mRNA analysis in different groups: Ia-Se deficient control, IIa-Se adequate control, IIIa-Se excess control, Ib-Se deficient+HCD, IIb-Se adequate+HCD, IIIb-Se excess+HCD after 1 and 2 months. (a) mRNA expression by RT-PCR. β-actin was also amplified (house keeping gene). (b) bands were quantified by densitometric analysis. Data is expressed as mean ± SD from 4 observations. *p<0.05, **p<0.01, ***p<0.001 represent comparison between control and HCD groups; Ap<0.05, AAp<0.01, AAAp<0.001 comparison between 1 and 2 months; DDDp<0.001 comparison between group Ia and IIa; ###p<0.001 comparison between group Ib and IIb; LLp<0.01, comparison between group IIa and IIIa; BBBp<0.001 comparison between group IIb and IIIb.
Apolipoprotein B levels
After one month of treatment, no significant change was observed in apoB levels by ELISA (Table 2). However after 2 months apoB levels in liver increased significantly (p < 0.001) in Se deficient control and high cholesterol diet (HCD) fed groups (Ia and Ib) in comparison to respective adequate groups (IIa and IIb). Whereas the level significantly (p < 0.001) decreased on 1 ppm Se supplementation in groups IIIa and IIIb. On HCD feeding a significant increase (p < 0.001) in apoB level was observed in comparison to respective control groups in all the three Se status groups after 2 months of diet feeding schedule. In selenium deficient groups (Ia and Ib), and in selenium adequate HCD fed group (IIb), the level increased and in 1 ppm selenium fed groups (IIIa and IIIb) it decreased significantly after 2 months of respective diet feeding in comparison to 1-month treatment period (Table 2).
Apolipoprotein B expression by western blot followed the similar trend as it was observed in ELISA. After one month of treatment, no significant change in band intensity was observed. So no change was there in apoB expression on high cholesterol diet feeding as well as in selenium deficiency and on 1 ppm selenium supplementation (Fig. 6).
Figure 6 Western Blot analysis of apolipoprotein B. Protein samples (30 μg) were resolved on 7.5% SDS-polyacrylamide electrophoresis and then electrophoretically transferred to PVDF membrane. Membrane was probed with antibody against apolipoprotein B. Lane1-Se deficient control; Lane2-Se adequate control; Lane3-Se excess control; Lane4-Se deficient+HCD; Lane5-Se adequate+HCD; Lane 6-Se excess+HCD. β-actin was used as an internal control.
After 2 months, immunoblot band intensity was apparently higher in Se deficient control and HCD fed groups (Ia and Ib) in comparison to respective adequate diet fed groups (IIa and IIb). Whereas the intensity was lower in 1 ppm Se supplemented groups IIIa and IIIb in comparison to adequate groups IIa and IIb respectively. On cholesterol supplemented diet feeding, apoB protein expression was found to be increased in comparison to respective control groups in all the three Se status groups after 2 months. In selenium deficient groups (Ia and Ib), and in selenium adequate cholesterol fed group (IIb), apoB expression increased and in 1 ppm selenium fed groups (IIIa and IIIb), it decreased after 2 months of respective diet feeding in comparison to 1-month treatment period (Fig. 6)
Discussion
Various studies suggested the prospective role of selenium in cardiovascular disorders [18]. Our results demonstrate that in selenium deficient animals significant increase in total cholesterol and LDL-cholesterol level was observed in comparison to adequate selenium fed animals [19]. Increased LDL is accumulatedin the intima, where it is oxidized. This oxidized LDL is associated with increased risk of coronary heart disease [20]. On 1 ppm Se supplementation lipid levels decreased significantly [21], Se supplementation might be protecting the LDL from oxidative modifications and further atherogenic changes [22,23]. Further, selenium supplementation leads to an increase in HDL cholesterol fraction [18]. HDL fraction increases the cholesterol elimination from tissues including smooth muscle cells in the aorta wall and facilitate the cholesterol transport to the liver, thus preventing its deposition and formation of atheromatous plaque [24].
In the present results in Se deficient groups, hepatic glutathione peroxidase (GSH-Px) activity decreased significantly in comparison to adequate groups (Table 1), so these observations confirm the Se deficiency, which is associated with decreased GSH-Px levels [25]. On high cholesterol diet feeding GSH-Px activity increased in all the groups. This increase in Se dependent GSH-Px on HCD feeding is attributed to the increased lipoperoxidative stress associated with cholesterol feeding as previously established in our lab [21]. Animals fed with the Se deficient diet had decreased hepatic 5'-DI activity as well as mRNA expression in comparison to the Se adequate groups [16,17]. During selenium deficiency hepatic stores of Se might be insufficient to allow the synthesis of 5'-DI. On 1 ppm Se supplementation GSH-Px and 5'-DI levels increased in control as well as HCD fed groups, this could be due to the fact that being selenoproteins the level of these two enzymes increased on Se supplementation.
Decreased level of T3 in selenium deficiency as well as on HCD feeding might be owing to the decreased conversion of T4 to T3 in the liver and other parts due to decreased 5'-DI (selenoprotein) expression during Se depletion [17]. Present studies revealed that in Se deficiency the LDL-R activity as well as mRNA expression is down regulated in comparison to adequate Se diet fed animals after 2 months of diet feeding (Fig. 3 &4). This could be due to the decreased T3 level during Se deficiency. T3 is directly involved in the regulation of LDL-R expression via modulation of SREBP-2 (sterol regulatory element-binding protein-2) gene expression. SREBP-2 is a major transcriptional regulator of cholesterol uptake through LDL-R [26]. In the present study we have observed that in Se deficiency, cholesterol level increased, this increased intracellular cholesterol level might be another reason to down regulate the LDL-R expression through feedback signaling pathway [27]. On feeding high cholesterol diet to the animals, LDL-R activity and mRNA expression decreased in all the three selenium status groups, so exogenous cholesterol given through diet is being used in the signaling pathway and probably it is suppressing the transcription of LDL-R through feedback mechanism. In case of 1 ppm Se supplemented groups increased T3 level might be upregulating the receptor expression.
In the current studies, on high cholesterol diet feeding for 2 months, apoB expression increased in all the three groups (Table 2; Fig. 6). This increased level of apoB on high cholesterol feeding is due to decreased expression of LDL-R during hypercholesterolemia as observed in the present studies. Decreased level of LDL-R is responsible for decreased clearance of apoB along with LDL, so these apolipoproteins are accumulated in the body [28,29]. During Se deficiency apoB levels in liver increased significantly in the present studies [30]. This could be due to the reason that selenium deficiency leads to decreased T3 levels and inturn hypothyroid state through decreased expression of 5'-DI enzyme [17,31], further hypothyroidism has been associated with increased level of apoB [14]. As T3 is involved in LDL-R expression, so reduced T3 levels during selenium deficiency is responsible for increase in apoB levels. Staels et al. [32] demonstrated that thyroid hormones activate the LDL-R, leading to an increased fractional catabolic rate of apoB without influencing its synthesis rate. Davidson et al. [33] reported that T3 administered to hypothyroid animals reduced the plasma apoB concentrations.
In the present studies on 1 ppm selenium supplementation for 2 months decreased expression of apoB was observed (Table 2; Fig. 6). This could be due to the increased level of T3 observed in the present study on Se supplementation through increased 5'-DI expression. So it results in increased catabolic rate of apoB through increased LDL-R expression. Davidson et al. [34] reported the suppressed synthesis of apoB during T3 supplementation. Walton et al. [35] suggested the increased catabolism of apoB through LDL receptors during hyperthyroidism. It is probably a combination of suppressed apoB synthesis and its increased elimination via LDL receptors during selenium supplementation here, which is regulating the apoB expression through T3 levels.
After one month of treatment, no significant change in the LDL-R and apoB expression was observed, this might be due to the reason that after 1 month the cholesterol accumulation might not be up to the extent that it could stimulate the feedback signaling pathway at translational as well as at transcriptional level to regulate LDL-R expression. So apoB catabolism through LDL receptors was not affected in different groups.
Conclusion
So, these results form the basis for a model that selenium status in the body regulates apolipoprotein B expression through selenoenzyme, 5'-DI. Selenium deficiency leads to decreased expression of 5'-DI, decreased T3 levels and decrease in LDL-cholesterol removal from blood through downregulation of LDL-R mRNA expression, ultimately decreased apoB catabolism through LDL receptors. Whereas Se supplementation upto 1 ppm leads to decreased apoB expression through increase in the LDL-R mRNA expression again via modulation of 5' -DI expression and in turn has the protective role against hypercholesterolemia. However, this interrelationship between selenium status and apoB expression warrants further investigation to decide the precise mechanism of lipid metabolism through the effect of Se status on the apolipoprotein B expression. Further studies must be undertaken to explore the therapeutic role of selenium supplementation in hypercholesterolemia.
Methods
Experimental Animals
Young male Sprauge-Dawley rats (100 g-body weight) were used in the present study. Animals were obtained from the Central Animal House, Panjab University, Chandigarh.
Treatment Protocol
Animals were acclimatized to the laboratory animal room and divided into three groups initially, group I (Se deficient diet fed), group II (Se adequate diet fed) and group III (Se excess diet fed). Feed and water were given ad libitum. This Se diet was given to the animals initially for 10 days so as to achieve the required Se status. The animals in these three groups were further divided into two each viz.: Group Ia (Se deficient control), Group Ib (Se deficient + high cholesterol diet fed); Group IIa (Se adequate control), Group IIb (Se adequate + high cholesterol diet fed); Group IIIa (Se excess control), Group IIIb (Se excess + high cholesterol diet fed). Treatment protocol was for 1 and 2 months.
Diet preparation
Yeast based synthetic Se deficient diet (supposed to contain 0.02 ppm Se) was prepared in the laboratory itself according to the composition given by Burk [36]. It contained torula yeast (inactivated) 30%, sucrose 56.99%, corn oil 6.67%, mineral mix 5%, vitamin mix 1%, dl-methionine 0.3% and vitamin E 0.04%. Se adequate and excess diet was prepared from Se deficient diet by supplementing it with 0.2 ppm and 1 ppm of Se as sodium selenite (Sigma Chemicals). 2% of cholesterol (Loba-Chemie, India) was added to the respective high cholesterol diet (HCD) groups.
After completion of diet feeding schedule, rats were kept on fasting for 10 hrs, anesthetized and exsanguinated. Serum and tissue (liver) samples were collected from each animal. Tissues were snap frozen in liquid nitrogen. Serum total cholesterol and LDL-cholesterol levels were estimated by enzymatic colorimetric kits obtained from E. MERCK diagnostic (Germany). Various parameters were carried out as detailed below
Selenium estimation
Selenium level was estimated by fluorimetric method [37], based on the principle that Se content in serum or tissue on acid digestion is converted to selenous acid. The reaction between selenous acid and aromatic-o-diamines such as 2,3-diamino-naphthalene (DAN) leads to the formation of 4,5-benzopiazselenol, which displays brilliant lime-green fluorescence when excited at 366 nm in cyclohexane. Fluorescence emission in extracted cyclohexane was read on fluorescence spectrophotometer using 366 nm as excitation wavelength and 520 nm as emission wavelength.
Se-dependent GSH-Px activity
Glutathione peroxidase (GSH-Px) activity was assayed using H2O2 as substrate [38]. The assay was carried out in the post-mitochondrial fraction (PMF) of liver as already published by us [39], the activity was expressed as μmoles of NADPH oxidized/min/mg protein. Total protein was done in all the samples [40].
T3 and T4 levels
Serum T3 and T4 estimation was done by radioimmunoassay (RIA) kits procured from BARC, Mumbai (Cat. No. RIAK-4/4A and RIAK-5/5A for T3 and T4 respectively).
Type-I 5'-iodothyronine deiodinase activity
Type-I iodothyronine deiodinase (5'-DI) activity in liver was estimated by following the method of Behne et al. [41].
LDL-R activity
LDL-R activity was estimated in vivo by following the method of Brown and Goldstein [42] and as per methodology already used by us [43]. Briefly, LDL isolated from overnight fasting human plasma using a single vertical spin density gradient ultra centrifugation [44], was radiolabeled with Na [131I] using chloramine-T [45]. Separation of labeled protein and unreacted iodide was done by gel filtration through sephadex G-25 column. The sterilized radiolabeled LDL was injected to rats of different groups. One ml of blood was taken 2 hrs after injection to measure the counts of radiolabeled LDL, and considered as counts at zero time interval or initial counts. Subsequently counts were also taken at 24, 48, 72, 96 and 120 hrs after injection. Percent decrease in counts at increasing time interval was taken as a measure of clearance of LDL from animal blood and in turn as indirect measurement of the LDL-R activity.
Apolipoprotein B levels by ELISA
Apolipoprotein B concentration was estimated in liver by ELISA using species specific (polyclonal anti rat apoB) antibody (Santa Cruz Biotechnology, Inc. Santa Cruz, CA). Optical density of each well was measured at 405 nm in ELISA reader (Stat Fax; Awareness Technology Inc., USA).
mRNA analysis
The mRNA expression for 5'-DI andLDL-R was done in liver using RT-PCR kit from QIAGEN.
RNA isolation
Total RNA from liver was extracted using TRI REAGENT (Molecular Research Centre, Inc. Ohio). The integrity and size distribution (quality) of RNA was examined by formaldehyde agarose gel electrophoresis.
RT-PCR
2 μg of total RNA template from different groups after treatment with DNase I (Ambion) was used in RT-PCR reaction. To the reaction mixture added 10 μl of 5X QIAGEN OneStep RT-PCR buffer (2.5 mM MgCl2 as final concentration), 2 μl of dNTP mix (10 mM of each dNTP), 5 μl of each forward and reverse gene specific primers (from 10 μM stock), 2 μl QIAGEN One Step RT-PCR Enzyme Mix, 1 μl RNase inhibitor (1 U/μl) and finally 25 μl of PCR grade RNase-free water (provided in the kit) to make total volume 50 μl. Mixed it gently by vortex and centrifuged it to collect all the components at the bottom of the PCR tubes. The PCR reaction was performed in the thermal cycler (Techne Ltd. England) using following conditions: the RT reaction was performed at 50°C for 50 min, initial PCR activation was done at 95°C for 15 min, followed by 35cycles of 94°C (denaturation) for 45 sec, 58.8°C (annealing) for 45 sec and 72°C (extension) for 1 min. Finally, incubated at 72°C for 10 min to extend any incomplete single strands.
Optimal oligonucleotide primer pairs for RT-PCR were selected with the aid of the software Gene Runner. The primer sequence (5' to 3') for rat 5'-DI gene coding (+) strand was TCTGGGATTTCATTCAAGGC, noncoding strand was TAGAGCCTCTCAGGCAGAGC, LDL-R gene coding (+) strand was ACCGCCATGAGGTACGTAAG, noncoding (-) strand was GGGTCTGGACCCTTTCTCTC and for rat β-actin gene coding (+) strand was AGAGCTATGAGCTGCCTGAC, and the noncoding (-) strand was CTGCATCCTGTCAGCCTACG. The length of RT-PCR products for 5'-DI, LDL-R and β-actin were 346, 341 bp and 236 bp respectively.
Final PCR products were analyzed on 1.5% agarose gel electrophoresis using 10 mM TE buffer. 5 μl of PCR product was used from each tube. Densitometric analysis of the bands was done by UviBandMap software (Uvitech, England).
Western Immunoblot Analysis
Western immunoblot analysis for apoB was done in liver. Tissue homogenate (10% w/v) from each group prepared in 20 mM tris-HCl buffer (pH 7.4) at 4°C was centrifuged at 3000 g for 10 min and the supernatant was used in the assay. Protein sample (30 μg) from each treatment group was separated by 7.5% SDS-polyacrylamide gel electrophoresis using minigel apparatus (BIORAD, UK). The separated proteins were electrophoretically transferred to PVDF membrane (Immobilon-P, Millipore, USA). Membrane was probed with antibody against apoB (Santa Cruz Biotechnology Inc. Santa Cruz, CA). β-actin was used as an internal control. Immunoblot analysis for β-actin was also done by using antibody against β-actin (Oncogene, USA).
Statistical analysis
Data is expressed as mean ± SD. Difference between different groups was tested using student's t test for unpaired values.
Authors' contributions
Both the authors SD and MPB have made substantive contributions to this study. Each author has participated in the work to take public responsibility for appropriate portions of the content. Both the authors have read and approved the final manuscript
Acknowledgements
Authors acknowledge the financial support given by Department of Atomic Energy, Government of India, Mumbai (India).
==== Refs
Lieu HD Withycombe SK Walker Q Rong JX Walzem RL Wong JS Hamilton RL Fisher EA Young SG Eliminating atherogenesis in mice by switching off hepatic lipoprotein Circulation 2003 107 1315 1322 12628954 10.1161/01.CIR.0000054781.50889.0C
Patalay M Lofgren IE Freake HC Koo SI Fernandez ML The lowering of plasma lipids following a weight reduction program is related to increased expression of the LDL receptor and lipoprotein lipase J Nutr 2005 135 735 739 15795426
Innerarity TL Mahley RW Weisgraber KH Bersot TP Krauss RM Vega GL Grundy SM Friedl W Davignon J McCarthy BJ Familial Defective Apolipoprotein B100: a mutation of apolipoprotein B that causes hypercholesterolemia J Lipid Res 1990 31 1337 1349 2280177
Kaiser M Kurktschiev TT Hanefeld M Intima-media thickness and atherosclerotic plaques in Familial Defective Apolipoprotein B-100 and Familial Hypercholesterolemia Annals N Y Acad Sci 2002 967 528 534
Vega GL Grundy SM Does measurement of apolipoprotein B have a place in cholesterol management? Arteriosclerosis 1990 10 668 671 2206113
Levinson SS Wagner SG Measurement of apolipoprotein B-containing lipoproteins for routine clinical laboratory use in cardiovascular disease Arch Pathol Lab Med 1992 116 1350 1354 1456883
Sniderman AD Silberberg J Is it time to measure apolipoprotein B? Arteriosclerosis 1990 10 665 677 2206112
Campeau L Enjalbert M Lesperance J Bourassa MG Kwiterovich PJ JrWacholder S Sniderman A The relation of risk factors to the development of atherosclerosis in saphenous-vein bypass grafts and the progression of disease in the native circulation. A study 10 years after aortocoronary bypass surgery N Engl J Med 1982 311 1329 1332 6333635
Whitfield AJ Barrett HR Van-bockxmeer FM Burnett JR Lipid disorders and mutations in the apoB gene Clin Chem 2004 50 1725 1732 15308601 10.1373/clinchem.2004.038026
Homma Y Predictors of atherosclerosis J Atheroscler Thromb 2004 11 265 70 15557708
Abraham R Kumar NS Kumar GS Sudhakaran PR Kurup PA Synthesis and secretion of apoB containing lipoproteins by primary cultures of hepatocytes isolated from rats fed atherogenic diet Atherosclerosis 1993 100 75 83 8318065 10.1016/0021-9150(93)90069-7
Kumar NS Abraham R Kumar GS Sudhakaran PR Kurup PA Synthesis and secretion of very low density lipoproteins-Modulation by cholesterol and phospholipids Ind J Biochem Biophys 1993 29 438 445
Ness GC Lopez D Chambers CM Newsome WP Cornelius P Long CA Harwood HJ Effects of l-triiodothyronine and the thyromimetic L-94901 on serum lipoprotein levels and hepatic low-density lipoprotein receptor, 3-hydroxy-3-methylglutaryl coenzyme A reductase and apo A-I gene expression Biochem Pharmacol 1998 56 121 129 9698096 10.1016/S0006-2952(98)00119-1
Mukhopadhyay D Plateroti M Anant S Nassir F Samarut J Davidson NO Thyroid hormone regulates hepatic triglyceride mobilization and apolipoprotein B messenger ribonucleic acid editing in a murine model of congenital hypothyroidism Endocrinology 2003 144 711 719 12538634 10.1210/en.2002-220741
Vander Geyten S Byamungu N Reyns GE Kuhn ER Darras VM Iodothyronine deiodinases and the control of plasma and tissue thyroid hormone levels in hyperthyroid tilapia (Oreochromis niloticus) J Endocrinol 2005 184 467 479 15749806 10.1677/joe.1.05986
Beckett GJ MacDougal DA Nicol F Arthur JR Inhibition of types I and II iodothyronine deiodinase activity in rat liver, kidney and brain produced by selenium deficiency Biochem J 1989 259 887 892 2730591
Dhingra S Singh U Bansal MP Effect of selenium depletion and supplementation on the kinetics of Type-I 5'- iodothyronine deiodinase and T3/T4 in rats Biol Trace Element Res 2004 97 95 104 10.1385/BTER:97:1:95
Wojcicki J Rozewicka L Wiszniewska BB Samochowiec L Juwiak S Kadlubowska D Tustanowski S Juzyszyn Z Effect of selenium and vitamin E on the development of experimental atherosclerosis in rabbits Atherosclerosis 1991 87 9 16 1670289 10.1016/0021-9150(91)90227-T
Huang K Liu H Chen Z Xu H Role of selenium in cytoprotection against cholesterol oxide-induced vascular damage in rats Atherosclerosis 2002 162 137 144 11947907 10.1016/S0021-9150(01)00707-9
Takahashi Y Zhu H Yoshimoto T Essential roles of lipoxygenases in LDL oxidation and development of atherosclerosis Antioxid Redox Signal 2005 7 425 431 15706089 10.1089/ars.2005.7.425
Kang BPS Bansal MP Mehta U Selenium supplementation and diet induced hypercholesterolemia in the rat: Changes in lipid levels, malonyldialdehyde production and the nitric oxide synthase activity Gen Physiol Biophys 1998 17 71 78 9675557
Hussein O Rosenblat M Refael G Aviram M Dietary selenium increases cellular glutathione peroxidase activity and reduces the enhanced susceptibility to lipid peroxidation of plasma and low-density lipoprotein in kidney transplant recipients Transplantation 1997 63 679 685 9075838 10.1097/00007890-199703150-00012
Gonca S Ceylan S Yardimoglu M Dalcik H Yumbul Z Kokturk S Filiz S Protective effects of vitamin E and selenium on the renal morphology in rats fed high-cholesterol diets Pathobiology 2000 68 258 263 11493758 10.1159/000055935
Pelton PD Patel M Demarest KT Nuclear receptors as potential targets for modulating reverse cholesterol transport Curr Top Med Chem 2005 5 265 282 15857310 10.2174/1568026053544588
Arthur JR Nicol F Backett GJ Selenium deficiency, thyroid hormone metabolism, and thyroid hormone deiodinases Am J Clin Nutr 1993 57 236 239
Shin DJ Osborne TF Thyroid hormone regulation and cholesterol metabolism are connected through sterol regulatory element-binding protein-2 (SREBP-2) J Biol Chem 2003 278 34114 34118 12829694 10.1074/jbc.M305417200
Goldstein JL Brown MS Regulation of the mevalonate pathway Nature 1990 343 425 430 1967820 10.1038/343425a0
Ouguerram K Chetiveaux M Zair Y Costet P Abifadel M Varret M Boileau C Magot T Krempf M Apolipoprotein B100 metabolism in autosomal-dominant hypercholesterolemia related to mutations in PCSK9 Arterioscler Thromb Vasc Biol 2004 24 1448 1453 15166014 10.1161/01.ATV.0000133684.77013.88
Brown MS Goldstein JL A receptor-mediated pathway for cholesterol homeostasis Science 1996 232 34 47 3513311
Mazur A Nassir F Gueux E Moundras C Bellanger J Grolier P Rock E Rayssiguier Y Diets deficient in selenium and vitamin E affect plasma lipoprotein and apolipoprotein concentration in the rat Br J Nutr 1996 76 899 907 9014658 10.1079/BJN19960096
Verhoelst CH Darras VM Doulabi BZ Reyns G Kuhn ER Vander Geyten S Type I iodothyronine deiodinase in euthyroid and hypothyroid chicken cerebellum Mol Cell Endocrinol 2004 214 97 105 15062548 10.1016/j.mce.2003.10.074
Staels B Van Tol A Chan L Will H Verhoeven G Auwerx J Alterations in thyroid status modulate apolipoprotein, hepatic triglyceride lipase, and low-density lipoprotein receptor in rats Endocrinology 1990 127 1144 1152 2387252
Davidson NO Carlos RC Lukaszewicz AM Apolipoprotein B mRNA editing is modulated by thyroid hormone but not growth hormone administration in the rat Mol Endocrinol 1990 4 779 785 2274057
Davidson NO Powell LM Wallis SC Scott J Thyroid hormone modulates the introduction of a stop codon in rat liver apolipoprotein B messenger RNA J Biol Chem 1988 263 13482 13494 3417667
Walton KW Scott PJ Dykes PW Davies JW The significance of alterations in serum lipids in thyroid dysfunction. II. Alterations of the metabolism and turnover of 131I-low-density lipoproteins in hypothyroidism and thyrotoxicosis Clin Sci 1965 29 217 238 5840973
Burk RF Production of selenium deficiency in the rat Methods in Enzymol 1987 143 307 313 3657552
Hasunuma R Ogawi T Kawaniska Y Fluorimetric determination of selenium in nanogram amounts in biological materials using 2,3-diaminonapthalene Anal Biochem 1982 26 242 245 6297332 10.1016/0003-2697(82)90510-3
Paglia DE Valentine WN Studies on the quantitative and qualitative characterization of erythrocyte glutathione peroxidase J Lab Clin Med 1967 70 158 168 6066618
Dhingra S Singh U Bansal MP Protective role of selenium status on T3/T4 kinetics in rats under hyperlipidemia Ind J Biochem Biophys 2003 40 260 264
Lowry OH Rosebrough NJ Farr AL Randall RJ Protein measurement with Folin-phenol reagent J Biol Chem 1951 193 265 275 14907713
Behne D Kyriakopoulos A Meinhold H Kohrle J Identification of type-1 iodothyronine 5'-deiodinase as a selenoenzyme Biochem Biophys Res Commun 1990 173 1143 1149 2268318 10.1016/S0006-291X(05)80905-2
Brown MS Goldstein JL How LDL-receptor influence cholesterol and atherosclerosis Sci Am 1984 251 52 60
Kang BPS Mehta U Bansal MP Hyperlipidemia and Type-I 5'-monodeiodinase activity: regulation by selenium supplementation Ind J Biochem Biophys 2000 7 183 187 10.1006/abbi.2000.1873
Chung BH Segrest JP Ray M Brunzell JD Hokanson JE Krauss RM Beaudrie K Cone JT Single vertical spin density gradient ultra centrifugation Methods in Enzymol 1986 128 181 209 3724502
Salahuddin H Singh O Talwar GP Radioiodination of antigens A handbook of practical immunology 1983 India: Vikas Publishing House 92 99
|
16271152
|
PMC1291393
|
CC BY
|
2021-01-04 16:39:19
|
no
|
Lipids Health Dis. 2005 Nov 5; 4:28
|
utf-8
|
Lipids Health Dis
| 2,005 |
10.1186/1476-511X-4-28
|
oa_comm
|
==== Front
Malar JMalaria Journal1475-2875BioMed Central London 1475-2875-4-521626908810.1186/1475-2875-4-52ResearchEvaluation of KO-Tab 1-2-3®: a wash-resistant 'dip-it-yourself' insecticide formulation for long-lasting treatment of mosquito nets Yates Alison [email protected]'Guessan Raphael 12Raphael.N'[email protected] Harparkash [email protected]éto Martin [email protected] Mark [email protected] London School of Hygiene & Tropical Medicine, London, WC1E 7HT, UK2 Centre de Recherche Entomologique, Cotonou, Benin2005 3 11 2005 4 52 52 15 8 2005 3 11 2005 Copyright © 2005 Yates et al; licensee BioMed Central Ltd.2005Yates et al; licensee BioMed Central Ltd.This is an Open Access article distributed under the terms of the Creative Commons Attribution License (), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
Introduction
Insecticide-treated nets (ITN) are an important method of preventing malaria. To remain effective, they need to be re-treated with pyrethroid insecticide at approximately yearly intervals. Systems for re-treating nets in Africa are limited, and the vast majority of nets in use have never been treated or were treated only once. Bayer Environmental Science (BES) has developed a long-lasting formulation 'KO-Tab 1-2-3®' which can be applied to the net post-manufacture, under field conditions, and renders the insecticide wash-resistant.
Methods
The performance of polyester nets treated with three kinds of BES long-lasting formulations, a conventional ITN (treated with standard KO-Tab) and PermaNet 2.0 were evaluated after washing samples of treated netting up to 30 times using standard WHO procedures. Performance was measured using 'three-minute exposure' and 'median time to knockdown' bioassay tests and by measuring the levels of deltamethrin using high-pressure liquid chromatography.
Results
The conventional ITN was largely stripped of deltamethrin within 5–10 washes and insecticidal efficacy in bioassay declined to suboptimal levels. With PermaNet and KO-Tab 1-2-3 the loss of deltamethrin was much slower: insecticide content halved within 20 washes and there was no loss of biological efficacy in three-minute exposure bioassays in WHO cylinders even after 30 washes. After 30 washes there remained on the netting 16% (4.4 mg/m2) of the loading dose of KO-Tab 1-2-3 and 28% (18.8 mg/m2) of the loading dose of PermaNet.
Conclusion
KO-Tab 1-2-3 was confirmed to be a long-lasting insecticide formulation. This finding raises the prospect of conventional polyester nets being converted into long-lasting insecticidal nets through simple dipping in the community or at home. This single development, if widely adopted, could transform the malaria control landscape in Africa and have a major impact on malaria.
==== Body
Introduction
Insecticide-treated nets (ITN) are an effective malaria control tool but to remain effective need to be retreated with pyrethroid insecticide about once a year [1]. This is a major constraint in developing countries where the infrastructure to provide repeated treatment of nets is inadequate or absent. The advent of long-lasting insecticidal nets (LLIN) has provided a technical solution to this problem [2-7]. With this technology, surface insecticide is replenished either from a resin matrix that coats the surface of fibres or by diffusion through the fibres as outer insecticide is removed during washing and normal use [8-10]. International donors and institutional buyers are increasingly opting for LLIN as their preferred choice of net. However, the majority of ITN in current use, particularly those available through the commercial retail sector, are not LLIN. Indeed the majority of nets in current use have either never been treated [11] or were treated only on purchase or distribution, and when non-insecticidal nets develop holes, which they inevitably do, they offer little or no protection against malaria [12].
Current LLIN are treated with insecticide during net manufacture. Bayer Environmental Sciences (BES) has developed a formulation technology 'KO-Tab 1-2-3' which offers the prospect of conventional nets being converted into LLIN through a dipping process that can be done post-manufacture under field conditions or in the home. A dip-it-yourself long-lasting treatment could solve the problem of having to regularly retreat conventional ITN.
This report describes the evaluation of polyester nets treated with 3 types of long-lasting formulation produced by BES, also a commercially available LLIN 'PermaNet 2.0®' and a conventionally treated net. Nets were washed up to 30 times and evaluated using two types of WHO bioassay test and by chemical residue assay using high-pressure liquid chromatography assay of residual deltamethrin.
Materials and methods
Preparation and washing procedure
Three treated nets were prepared by BES: KO-Tab 1-2-3 in which a conventional KO Tab (suspension concentrate formulation) is mixed in aqueous solution with a special binder to make a long-lasting wash-resistant treatment; BES902 in which a conventional KO-Tab is mixed with a different type of binder; and BES903 which has double the quantity of deltamethrin and binder. A long-lasting insecticidal net (PermaNet®) served as a positive control. One conventional KO-Tab treated net was prepared at LSHTM for comparison. The nets were laid flat for drying in a room kept at 25°C, and turned at 10 minutes intervals.
Fifteen samples of netting, each measuring 120 × 30 cm, were cut from each net. The samples were kept in triplets (one sample from each side and top of the net) and washed 0, 5, 10, 20 or 30 times. This scheme was expected to adjust for any variation in deposition due to uneven treatment or drying.
The netting was washed according to standard WHO procedures [10] in 2 g/litre soap solution (Le Chat Paillettes® Savon de Marseille) in deionised water at pH10-11. Net samples were washed individually for 10 minutes in 1 litre bottles containing 750 ml soap solution kept at 30°C in a motorized water bath adjusted to make 155 back-and-forth movements per minutes. The samples were then rinsed twice with clean deionised water at 30°C using the same procedure. The net samples were hung to dry at 24°C. There was a 24 h interval between washings and the 30 washings were carried out over six weeks.
Each 120 × 30 cm net sample was then sectioned into 3 pieces. The first piece was used for three minutes exposure bioassays, the second for median time to knock down tests, and the third for chemical assay.
Median time to knock down bioassay tests
Median time to knock down (MTKD) bioassays were performed at LSHTM using the wire ball frame technique [13]. In MTKD bioassays 11 female Anopheles stephensi (BEECH, insecticide susceptible strain), 2–5 days old, were exposed to test netting in a metal frame and the time taken for the median (6th) mosquito to be knocked down was recorded. Knockdown was defined as collapsed against the netting or fallen to the base of the ball frame and not moving. As each mosquito was knocked down it was removed using a mouth aspirator so if it did recover it would not be counted again. The time for the median mosquito to be knocked down was the end point of the test. Three replicate tests were carried out on each piece of net, giving a total of 9 replicates for each treatment/wash combination.
Three-minute exposure bioassay tests
Three-minute exposure bioassays were carried out by LSHTM staff based at the Centre de Recherche Entomologique de Cotonou in Benin. This type of test can be done using WHO plastic cones, wire ball frames, or WHO susceptibility test kits [13]. WHO cones are rather awkward and time-consuming for conducting tests of three minutes duration and present a large, non-insecticidal surface area if the test insecticide happens to be repellent. Instead, WHO susceptibility test kits were chosen as the exposure chamber for netting because use of these involves less mosquito handling, are easier to do and enable more precise exposures. For ease of handling the test netting was taped to a backing paper (same size as a WHO insecticide paper) before being inserted into the exposure tube and constrained with two metal rings. The kits' mesh lids were replaced with two layers of test netting to further reduce the ratio of untreated to treated surface inside the tube. Batches of 10 females were placed in the recovery tube of the test kit (which had been lined with plain paper). The mosquitoes were then blown into the exposure tube, kept there for three minutes, blown back into the recovery tube, and then held for 24 h together with a cotton wool pad soaked in sugar solution. Mortality was scored after 24 h. Tests were carried out using sugar-fed females of Anopheles gambiae Kisumu (insecticide susceptible), 2–5 days of age.
Seven replicates were carried out on each piece of netting, giving a total of 21 replicates (210 mosquitoes) for each treatment/wash combination.
Chemical analysis
Netting squares measuring 5 cm × 5 cm were cut after the washing cycles had been completed. Deltamethrin was extracted using acetonitrile and the extract was injected (20 μl per sample of net) onto HPLC. The analyses were carried out using a Dionex Summit range of equipment and software (Camberly, Surrey, UK). Samples were separated on an AcclaimR C18 120Å column (250 × 4.6 mm), eluted with water/acetonitrile (90:10%; v/v) at a flow rate of 2 ml/minutes and passed through a photodiode array detector (PDA-100) set at 275 nm. The authenticity of the detected peaks was determined by comparison of retention time, spectral extraction at 275 nm and spiking the sample with commercially available standard of deltamethrin. A calibration curve of deltamethrin was generated by Chromeleon (Dionex software) using known amounts of the standard (0–0.4 μg/ml) in acetonitrile injected onto the column. From this curve the amount of deltamethrin on the 25 cm2 pieces was estimated and the dosage per m2 calculated.
Statistical analysis
The three-minute mortality tests were analysed using blocked logistic regression. The median time to knockdown tests and chemical assays were analysed using analysis of variance and t-tests. Bioassay and chemical assay results were compared using linear regression.
Results
Chemical analysis
The results of HPLC analyses are presented in Figure 1. The loading dosage of deltamethrin on the KO-Tab treated net was close to the target of 25 mg/m2. The loading dosages of deltamethrin on the KO-Tab 1-2-3 treated net and on BES902 were within the same range of 25 mg/m2. The loading dosage on net BES903 was twice this amount (55.9 mg/m2), as this treatment was designed to resemble a PermaNet loading dosage. PermaNet itself showed a dosage of 66.7 mg/m2, somewhat higher than the manufacturer's target dosage of 55 mg/m2.
Figure 1 Mean (± CI) dosage of deltamethrin (mg/m2) before and after washing as measured using high-pressure liquid chromatography.
Washing mostly stripped the deltamethrin from the KO-Tab treated net. After 5 washes only 2.7% of the initial load remained, and after 10 washes only 1.7% (0.5 mg/m2) was present. After further washing residues were reduced to <0.2 mg/m2. With the KO-Tab 1-2-3 treated net, 73% of the initial load was still remaining after 5 washes and 62% was remaining at 10 washes. Only after 10 washes was there a more marked decline, with 25.6% remaining at 20 washes and 16% remaining at 30 washes.
With the net treated with formulation BES902, there was 83% still remaining at five washes and 81% at 10 washes. After 20 washes 38.5% remained.
By doubling the loading dose on net BES903 there was, at each wash interval examined, a comparatively higher dosage of deltamethrin still remaining compared to the KO-Tab 1-2-3 treated net and to net BES902. However, it is not clear why, after 30 washes, there was four times rather than two times as much active ingredient still remaining on net BES903 than on the KO-Tab 1-2-3 net or on net BES902.
With PermaNet the percentage of deltamethrin remaining was 80% of loading dose at five washes, 68% at 10 washes, 41% at 20 washes and 28% at 30 washes. At each of these wash points there was, relative to loading dose, proportionately more insecticide remaining on PermaNet than on the KO-Tab 1-2-3 treated net. With net BES903, on the other hand, the proportion of loading dose remaining at each wash point showed no clear difference from PermaNet. A notable difference between net BES903 and PermaNet was the narrower confidence interval on the PermaNet at each wash interval. This presumably represented a more homogeneous distribution of deltamethrin on the factory impregnated net than on the hand treated net.
Three-minute exposure bioassay tests
Prior to washing the baseline mortality for all formulations was 100% (Figure 2). At 5 wash cycles mortality remained at 100%. At 10 wash cycles mortality on the KO-Tab treated net decreased to 39.4% and at 30 washes it decreased to 16.1%. On the nets treated with the long-lasting formulations (KO-Tab 1-2-3, BES902, BES903) 100% or near 100% mortality was observed after 30 wash cycles. PermaNet showed 100% mortality after 30 wash cycles.
Figure 2 Three-minute exposure bioassay tests: proportion dead after 24 h. Seven replicate tests (70 mosquitoes exposed) were carried out on each formulation and wash interval.
HPLC analyses indicate that over 97% of deltamethrin was removed from the KO-Tab treated net after five washes (0.8 mg/m2 remaining), whereas according to three-minute bioassay tests there was still sufficient active ingredient remaining to kill 100% after five washes.
Median time to knock down tests
The results of median time to knockdown tests are shown in Figure 2. The time to knockdown of the median mosquito was quite consistent between replicates, as shown by narrow confidence intervals around the means of the medians. This enables small differences between treatments or washes to be detected using this technique.
The time to knockdown was fastest on unwashed nets, being inversely related to the amount of bioavailable deltamethrin. On unwashed netting the MTKD ranged from 461s on conventional KO-Tab to 502s on PermaNet and 560s on KO-Tab 1-2-3. Over the course of 30 washes the MTKD of conventional KO-Tab increased 2.4 fold to 1089s, whereas that of KO-Tab 1-2-3 only increased 1.3 fold to 742s and that of PermaNet only increased 1.1 fold to 554s. The increase in MTKD over 30 washes was significant for each of the treatments except for PermaNet which was borderline. For conventional KO-Tab the difference in MTKD over 30 washes (mean ± CI), 628 ± 99s (P < 0.0001), was greater than that of the long-lasting treatments; for KO-Tab 1-2-3 the difference was 181 ± 71s (P = 0.0001), for BES903 the difference was 136 ± 88s (P = 0.005) and for Permanet the difference was 52 ± 55s (P = 0.065). The difference in MTKD over 30 washes was therefore significantly less for PermaNet than for KO-Tab 1-2-3 or other Bayer long-lasting formulations.
Figure 4 shows that the MTKD test data are strongly correlated with the amount of deltamethrin detected on the nets using HPLC (P < 0.0001). The three-minute exposure bioassay results were also significantly correlated with deltamethrin content (P < 0.0001) but the relationship was less clear (Figure 5) owing to the consistently high mortalities observed with long lasting formulations over the 0–30 range of washes.
Figure 4 The relationship between median time to knockdown and dosage of deltamethrin.
Figure 5 The relationship between % mortality (from three minutes bioassays) and dosage of deltamethrin.
Discussion
A long-lasting insecticidal net is defined by WHO as a net which gives greater than 80% mortality in three-minute exposure bioassays after 20 standardized washes [10]. Strictly speaking this should be after three minutes exposure in WHO cones rather than WHO cylinders. Cones were originally developed for assessing indoor residual spraying, in which 30 minutes exposure bioassays are conducted on firm flat surfaces such as sprayed walls [11]. For testing pyrethroid treated netting, cones have two disadvantages: pyrethroids are repellent and there is a large surface area within the cone that is non-insecticidal so exposure may be shorter than the intended three minutes; secondly, because insertion and removal of mosquitoes involves use of an aspirator it is harder to expose all mosquitoes for a precise three minutes period within the cone. For this reason some research groups are switching from cones to 'ball/wire frames' [11] or WHO cylinders as these techniques have neither of these disadvantages. WHO is currently sponsoring multicentre trials to assess the comparability of cones and cylinders for net bioassays. The work presented here pre-dates the WHO initiated trials. Comparing our results with cylinders with earlier studies with cones, it would seem that mortality in cylinders is indeed higher than in cones: for example, after subjecting PermaNet 2.0 to 20 standardized washes, Graham et al. [7] observed 81.8% mortality and Gonzales et al.[9] observed 87.1% mortality in cones whereas we observed 100% mortality in cylinders. More important are that tests are controlled, and the candidate LLIN is compared with a standard LLIN both of which are subjected to the same standardized washing procedures. Our evidence using cylinder tests is that KO-Tab 1-2-3 is a long-lasting wash-resistant treatment on a par with PermaNet in terms of mortality over 30 washes. But until we have a calibration curve for cylinders versus cones it cannot be said that a KO-Tab 1-2-3 treated polyester net is a long-lasting insecticidal net by the strict WHO definition.
Whereas the three minutes exposure bioassays indicated that KO-Tab 1-2-3 retains an insecticidal efficacy equivalent to PermaNet's over the course of 30 washes, HPLC analyses indicated that the wash resistance of KO-Tab 1-2-3 may decline at a faster rate between 10 and 30 washes. By contrast, PermaNet and BES903 were equivalent in wash fastness both in three minutes bioassay and chemical assay. It can be inferred that a one-day interval between washes was sufficient time for regeneration (if any regeneration was necessary) because there was sufficient insecticide available after each washing to still kill 100% of mosquitoes even though insecticide was being steadily removed. These are important findings that offer the hope of conventional ITN being made long lasting through simple dipping procedures that do not require sophisticated industrial processes or equipment.
The sensitive median time to knockdown bioassays supported the findings of the HPLC analyses and revealed incremental differences in insecticide performance after washing that three minutes exposure tests did not pick up. Comparing median time to knockdown over the course of 0-30 washes, KO-Tab 1-2-3 showed an additional 129 second increase in MTKD over PermaNet. At this stage of evaluation it is not clear whether the increased MTKD or greater loss of active ingredient in KO Tab 1-2-3 over 30 washes would translate to a poorer performance compared to PermaNet under field conditions.
The biological efficacy of the conventional KO-Tab treated net continued long after most of the insecticide had been removed by washing. The apparent discrepancy between biological and chemical assay results has been observed before in other comparative trials of long-lasting nets and conventionally treated nets [7]. It would seem that a small amount of bioavailable pyrethroid remains quite firmly bound to the net after most of the insecticide has been washed off. This residue is clearly sufficient to kill all mosquitoes that come into tarsal contact with it. One way to understand how this could be is to consider that 'feet get equally wet whether one goes paddling in the sea or step in a puddle of rainwater'.
Two further challenges await KO-Tab 1-2-3. Can this formulation render other types of polymer or fabric long-lasting? Will the long lasting treatment stand up to wear and tear and routine washing over the lifetime of a net in normal everyday use? There is great diversity in the fabrics and materials used for making mosquito nets, particularly in traditional net using societies that were using nets before the ITN revolution of the last two decades. The utility and wash resistance of KO-Tab 1-2-3 needs to be confirmed on nets made of cotton, nylon, polyethylene and other synthetic materials before this product can have the widest possible application or impact against malaria. There might also be potential in treating non-net substrates such as curtains, canvas tents or blankets.
The next phase of testing of KO-Tab 1-2-3 should be an evaluation in experimental huts of treated polyester nets washed up to 20 times, as recommended in recent WHO guidelines[10]. The ultimate demonstration of the persistence and effectiveness of this product would be during normal household use over at least 3 years in a variety of cultural, environmental and epidemiological situations. In view of the time-scales required for evaluating LLIN, such studies should run concurrently with phase 2 experimental hut trials.
Authors' contributions
AY prepared the netting samples at LSHTM, undertook the median time to knockdown and chemical assays, summarised the data. RN supervised the mortality bioassays in Benin. HK carried out the HPLC chemical analysis. MR designed the study, analysed the data and prepared the manuscript.
Figure 3 Median time to knockdown tests: means and confidence intervals. Nine replicate tests were carried out on each formulation and wash interval.
Acknowledgements
This project was supported by Bayer Environmental Science and the Gates Malaria Partnership. We thank Dr Karin Horn of BES for advice and encouragement. We are grateful to Abibatou Oljo, Boco Pelagie and Hounkarin Lazare for technical assistance.
==== Refs
Lengeler C Insecticide-treated bednets and curtains for preventing malaria Cochrane Database Syst Rev 2004 CD000363 15106149
Itoh T Okuno T Evaluation of the polyethylene net incorporated with permethrin during manufacture of thread on efficacy against Aedes aegypti (Linnaeus) Med Entomol Zool 1996 47 171 174
N'Guessan R Darriet F Doannio JMC Chandre F Carnevale P Olyset net efficacy against pyrethroid-resistant Anopheles and Culex after 3 years' field use in Côte D'Ivoire Med Vet Entomol 2001 15 97 104 11297108 10.1046/j.1365-2915.2001.00284.x
Muller O Ido K Traore C Evaluation of a prototype long-lasting insecticide-treated mosquito net under field conditions in rural Burkina Faso Trans R Soc Trop Med Hyg 2002 96 483 484 12474472 10.1016/S0035-9203(02)90411-6
Kroeger A Skovmand O Phan QC Boewono DT Combined field and laboratory evaluation of a long-term impregnated bednet, PermaNet® Trans R Soc Trop Med Hyg 2004 98 152 155 15024924 10.1016/S0035-9203(03)00038-5
Tami A Mubyazi G Talbert A Mshinda H Duchon S Lengeler C Evaluation of Olyset insecticide-treated nets distributed seven years previously in Tanzania Malaria J 2004 3 1 9 10.1186/1475-2875-3-1
Graham KL Kayedi MH Maxwell C Kaur H Rehman H Malima R Curtis CF Lines JL Rowland MW Multi-country field trials comparing wash-resistance of PermaNet™ and conventional insecticide-treated nets against anopheline and culicine mosquitoes Med Vet Entomol 2005 19 72 83 15752180 10.1111/j.0269-283X.2005.00543.x
WHO Review of: Olyset net, Bifenthrin 10%WP Report of the 5th WHOPES Working Group Meeting WHO/HQ, Geneva, 30–31 October 2001, Document WHO/CDS/WHOPES/20014 World Health Organization, Geneva
WHO Review of: Vectobac WG, PermaNet, Gokilaht-S 5EC Report of the 7th WHOPES Working Group Meeting WHO/HQ, Geneva, 2–4 December 2003, Document WHO/CDS/WHOPES/20048 2004 World Health Organisation, Geneva
WHO Guidelines for laboratory and field testing of long-lasting insecticidal mosquito nets Document WHO/CDS/WHOPES/GCDPP/200511 2005 World Health Organisation, Geneva
WHO World Malaria Report 2005 World Health Organisation, Geneva
Mwangi TW Ross A Marsh K Snow RW The effects of untreated bednets on malaria infection and morbidity on the Kenyan coast Trans R Soc Trop Med Hyg 2003 97 369 372 15259458 10.1016/S0035-9203(03)90056-3
WHO Test procedures for insecticide resistance, bioefficacy and persistence of insecticides on treated surfaces Report of WHO Informal Consultation 1998: WHO/CDS/CPC/MAL/9812 World Health Organization, Geneva
|
16269088
|
PMC1291394
|
CC BY
|
2021-01-04 16:37:31
|
no
|
Malar J. 2005 Nov 3; 4:52
|
utf-8
|
Malar J
| 2,005 |
10.1186/1475-2875-4-52
|
oa_comm
|
==== Front
Microb Cell FactMicrobial Cell Factories1475-2859BioMed Central London 1475-2859-4-301626908610.1186/1475-2859-4-30ResearchCharacterization of the metabolic shift between oxidative and fermentative growth in Saccharomyces cerevisiae by comparative 13C flux analysis Frick Oliver [email protected] Christoph [email protected] Biochemical Engineering Institute, Saarland University, POB 151150, 66123 Saarbrücken, Germany2005 3 11 2005 4 30 30 29 9 2005 3 11 2005 Copyright © 2005 Frick and Wittmann; licensee BioMed Central Ltd.This is an Open Access article distributed under the terms of the Creative Commons Attribution License (), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
Background
One of the most fascinating properties of the biotechnologically important organism Saccharomyces cerevisiae is its ability to perform simultaneous respiration and fermentation at high growth rate even under fully aerobic conditions. In the present work, this Crabtree effect called phenomenon was investigated in detail by comparative 13C metabolic flux analysis of S. cerevisiae growing under purely oxidative, respiro-fermentative and predominantly fermentative conditions.
Results
The metabolic shift from oxidative to fermentative growth was accompanied by complex changes of carbon flux throughout the whole central metabolism. This involved a flux redirection from the pentose phosphate pathway (PPP) towards glycolysis, an increased flux through pyruvate carboxylase, the fermentative pathways and malic enzyme, a flux decrease through the TCA cycle, and a partial relocation of alanine biosynthesis from the mitochondrion to the cytosol. S. cerevisiae exhibited a by-pass of pyruvate dehydrogenase in all physiological regimes. During oxidative growth this by-pass was mainly provided via pyruvate decarboxylase, acetaldehyde dehydrogenase, acetyl-CoA synthase and transport of acetyl-CoA into the mitochondrion. During fermentative growth this route, however, was saturated due to limited enzyme capacity. Under these conditions the cells exhibited high carbon flux through a chain of reactions involving pyruvate carboxylase, the oxaloacetate transporter and malic enzyme. During purely oxidative growth the PPP alone was sufficient to completely supply NADPH for anabolism. During fermentation, it provided only 60 % of the required NADPH.
Conclusion
We conclude that, in order to overcome the limited capacity of pyruvate dehydrogenase, S. cerevisiae possesses different metabolic by-passes to channel carbon into the mitochondrion. This involves the conversion of cytosolic pyruvate either into acetyl CoA or oxaloacetate followed by intercompartmental transport of these metabolites. During oxidative growth mainly the NAD specific isoforms of acetaldehyde dehydrogenase and isocitrate dehydrogenase catalyze the corresponding reactions in S. cerevisiae, whereas NADPH supply under fermentative conditions involves significant contribution of sources other than the PPP such as e. g. NADPH specific acetaldehyde dehydrogenase or isocitrate dehydrogenase.
==== Body
Background
The budding yeast S. cerevisiae is an important biotechnological organism e.g. for production of ethanol [1], recombinant proteins [2], antibiotics [3] or biomass [4]. Additionally it is relevant as a model system to study metabolism due to its high homology with other eukaryotic organisms [5,6]. This explains the high interest in understanding metabolic function and regulation in this organism. S. cerevisiae is able to perform simultaneous respiration and fermentation at high growth rates even under fully aerobic conditions [7,8]. This Crabtree effect called phenomenon is visualized in chemostat culture of S. cerevisiae by a shift from biomass formation towards fermentative products at increased dilution rates [8]. Also other yeast species show this behavior [9]. Despite many studies in the past [10-12], a clear and univocal explanation of this phenomenon has not yet been provided [13,14]. The Crabtree effect has important consequences for industrial processes aiming at the production of yeast biomass, where the formation of fermentative by-products is undesired [15,16]. A powerful approach to investigate metabolic systems is 13C metabolic flux analysis which allows the quantification of in vivo activity of pathways and reactions [17-19]. In recent studies, which took the spatial arrangement of metabolic reactions in the compartments cytosol and mitochondrion into account, metabolic flux analysis was successfully applied to S. cerevisiae and related yeast strains [20-27]. In the present work we investigated the metabolic shift from oxidative to fermentative growth in S. cerevisiae by quantifying metabolic fluxes through the underlying reactions in the central metabolism at these different physiological states. For this purpose chemostat cultures with [1-13C] glucose as substrate were grown at different growth rates under aerobic glucose-limited conditions to metabolic and isotopic steady-state. The use of a continuous culture hereby allowed the precise adjustment of the physiological state of the cells, i.e. the relative activity of the fermentative pathways.
Results
Growth and product formation of S. cerevisiae
In order to identify the exact experimental conditions required to establish the physiological states of interest, chemostat cultures were carried out at different dilution rates between 0.10 h-1 and 0.45 h-1 (Figure 2). S. cerevisiae exhibited the typical growth behavior for this organism in continuous culture, characterized by a transition from purely oxidative metabolism at low growth rate to fermentative metabolism at high growth rate. At a growth rate below μ = 0.20 h-1 the metabolism of S. cerevisiae was purely oxidative, indicated by the absence of fermentation products and a high biomass yield. At higher growth rates the Crabtree effect was induced, which is shown by enhanced formation of ethanol and acetate. Hereby, the relative contribution of fermentation to the overall metabolic activity increased with increasing growth rate. At μ = 0.30 h-1 about 25 % of the utilized glucose was directed towards fermentation, indicating a still mainly respirative metabolism. Towards higher dilution rates the fraction of glucose channeled into fermentative pathways increased up to a value of more than 50 % at of μ = 0.40 h-1.
Tracer experiments with [1-13C] glucose as sole carbon source
Selected chemostat cultivations with [1-13C] glucose as substrate were carried out for the elucidation of metabolic fluxes. Hereby, three distinct dilution rates, representing purely-oxidative (0.15 h-1), mixed respiro-fermentative (0.30 h-1), and mainly fermentative metabolism (0.40 h-1) were selected for the flux studies. The resulting ethanol yield under these conditions was 0.0, 0.5, and 1.1 mol ethanol (mol glucose)-1, respectively. The stoichiometric data obtained from the 13C cultures agreed well with the corresponding values from the initially performed cultivations (Table 3). The major by-products were ethanol, acetate and glycerol. In all cases the carbon balance was almost closed underlining the high consistency of the data. Labeling patterns of amino acids from cell protein, analyzed after 5 residence times, remained constant over 4 samples taken in 1 h intervals from the culture, i. e. the relative error for the mass isotopomer fractions was typically below 1 %. This indicated that the cells were in isotopic steady-state. To quantify metabolic fluxes a 13C flux model was applied that involves metabolite and isotopomer balancing. The used model fully considers specific features of the underlying metabolism, e. g. bidirectional reactions, label scrambling in symmetric molecules, 13C incorporation from CO2 or naturally occurring isotopes [28]. For each scenario a comprehensive data set of 13C labeling data (Table 4) and stoichiometric data (Tables 2, 3) was used to calculate the fluxes.
Metabolic fluxes
The obtained flux distributions for S. cerevisiae growing in different physiological states are shown in Figure 3. The fluxes given are relative values normalized to the corresponding specific glucose uptake rate. The examined S. cerevisiae strain revealed the typical behavior linked to the switch from respiration to fermentation with increasing growth rate. This involved a redirection of flux from the TCA cycle towards the fermentative pathways. In addition to these previously described, well-known changes, a number of additional pathway fluxes were, however, affected by the change of the physiological regime.
PPP and glycolysis
The metabolic shift of S. cerevisiae was accompanied by a substantial decrease of the relative flux through the PPP (Figure 3). The decreasing contribution of the PPP with increasing growth rate has direct consequences for the supply of NADPH, required in various anabolic reactions, and certain anabolic precursors. Taking into account that two molecules NADPH are generated in the oxidative part of the PPP, the relative supply of NADPH is 110 % (normalized to the glucose uptake rate) at μ = 0.15 h-1. In comparison only 60 % and 30 % NADPH, respectively, are formed at μ = 0.30 h-1 and μ = 0.4 h-1. A similar trend can be seen for erythrose 4-phosphate and ribose 5-phosphate. The supply of these anabolic precursors was substantially lower under fermentative conditions.
Fluxes around the pyruvate node
Distinct flux changes were observed for the pathways utilizing cytosolic pyruvate, i. e. pyruvate carboxylase, pyruvate decarboxylase, and the pyruvate transporter into the mitochondrion. A surprising result was the increase of flux through pyruvate carboxylase with increasing growth rate. This reaction was thought to mainly serve for the supply of cytosolic oxaloacetate for anabolism. Under fermentative conditions at higher growth rate, however, the anabolic demand for this precursor was much lower as compared to purely oxidative growth (Table 2). Accordingly the increase of flux through pyruvate carboxylase was not driven by an increased anabolic demand. It was in fact linked to increased transport of oxaloacetate into the mitochondrion. Interestingly, pyruvate decarboxylase, the entry step towards the fermentative metabolism, exhibited in vivo activity not only during the formation of fermentative by-products. Significant flux through this pathway was also observed under conditions of purely oxidative growth, i.e. in the absence of overflow metabolism. Accordingly a high production of acetaldehyde resulted for all physiological states. Under purely oxidative growth, the formed acetaldehyde was completely utilized for the synthesis of cytosolic acetyl CoA. This was then either transported into the mitochondrion or directly channeled into anabolic pathways located in the cytoplasm. During the switch to fermentation the acetaldehyde formed was mainly channeled into the fermentative by-products. The highest flux through pyruvate decarboxylase was observed at maximum production of ethanol and acetate. The ingoing flux through the lower glycolytic chain was higher during fermentative growth as compared to purely oxidative growth (Figure 3).
Intercompartmental transport
A substantial transport of carbon from the cytosol into the mitochondrion resulted for all growth rates (Figure 3). Hereby the relative contribution of the different transporters depended on the physiological state. Under purely oxidative growth acetyl CoA, formed in high amounts via the fermentative routes, was the dominating metabolite transported into the mitochondrion. The acetyl CoA transport was strongly decreased under conditions of respiro-fermentative growth. During mainly fermentative metabolism the transport net flux practically decreased to zero. The flux of oxaloacetate transport increased with increasing growth rate. This is linked, as described above, with the increased flux through pyruvate carboxylase. Interestingly the activity of the oxaloacetate transporter correlated with the flux through malic enzyme. This might suggest that sufficient intercompartmental oxaloacetate transport is important for the activity of malic enzyme. The picture observed for the pyruvate transport was more complex. The relative transport flux increased from purely oxidative to respiro-fermentative growth. With further increase of the specific growth rate it substantially decreased. The total flux of carbon into the mitochondrion, as sum of the single transport reactions, was highest during purely oxidative growth.
TCA cycle
Overall, the metabolic shift from pure oxidation to fermentation was linked to a strong decrease of the TCA cycle flux (Figure 3). Under purely oxidative growth citrate synthase, the entry enzyme of this pathway, exhibited a relative flux of 73.1 %. The relative activity of this enzyme was 54 % at μ = 0.30 h-1 which indicates that the TCA cycle still contributed to a large extent to the metabolic activity. At μ = 0.40 h-1 the TCA cycle still operated as a cycle, but only at a relatively low activity. The observed flux changes include a shift of the relative contribution of fumarase and the oxaloacetate transporter to the supply of the mitochondrial malate/oxaloacetate pool.
Anabolism
Under purely oxidative growth both pathways for glycine biosynthesis were active, whereby the major fraction of glycine was formed via the serine pathway (Figure 3). The simultaneous in vivo activity of both pathways was maintained under fermentative conditions. An interesting finding resulted for the biosynthesis of alanine for biomass production. Whereas alanine was exclusively synthesized via the mitochondrial route at purely oxidative growth, cells partly switched to cytosolic alanine formation during the metabolic shift towards fermentation (Figure 3). Under mainly fermentative conditions the cytosolic pathway even provided the major fraction of proteinogenic alanine.
Goodness of fit and statistical flux evaluation
For the flux distributions, which are shown in Figure 3, an excellent agreement between experimentally determined and calculated labeling patterns was achieved. The deviation between measured and calculated mass isotopomers was typically below 3 % and thus rather small (Table 4). Generally the intracellular fluxes could be determined with high precision as indicated by the small 90 % confidence intervals (Table 5). This allows an unambiguous differentiation between the different physiological states of the cells. In the present work, the PPP reactions of transaldolase and transketolase were regarded reversible. Comparative flux calculations, setting these reactions as irreversible did lead to a significantly worse fit of the data, which was most pronounced for phenylalanine and tyrosine formed from precursors of the PPP. A model with irreversible PPP reactions thus could not describe the experimental data properly. Moreover this was also related to different results for several flux parameters such as the flux partitioning between glycolysis and PPP. The same was observed for glucose 6-phosphate isomerase. It seems therefore important to fully consider the reversible nature of the PPP reactions in metabolic flux analysis with S. cerevisiae as previously proposed by metabolic simulation studies [28,29].
Discussion
In the present work comparative 13C metabolic flux analysis of S. cerevisiae at different defined physiological states revealed that the transition from oxidative to fermentative utilization of glucose is reflected by substantial flux changes throughout the whole central carbon metabolism. The different physiological conditions, i. e. purely oxidative, mixed respiro-fermentative and mainly fermentative growth were established by variation of the dilution rate in chemostat culture. Hereby, the extent of the fermentation could be precisely adjusted and run into metabolic and isotopic steady-state.
We observed the previously described transition from respiration to fermentation as indicated by the redirection of flux from the TCA cycle towards the fermentation pathways [8]. It becomes, however, clear from the present study, that the regulatory changes linked to the metabolic transition were not restricted to these reactions, but affected a number of additional pathways.
(i) The metabolic shift had a strong effect on the flux partitioning between glycolysis and PPP. Whereas, at purely oxidative metabolism, the major carbon flux was channeled into the PPP, the opposite was observed for fermentative metabolism. The data of the present work together with previous flux studies of S. cerevisiae under various conditions clearly show that the flux partitioning between glycolysis and PPP is growth rate dependent (Figure 4A). A similar relation is found for the flux partitioning and the biomass yield. The reduced biomass yield at higher growth rates results in a reduced demand for erythrose 4-phosphate, ribose 5-phosphate and NADPH, respectively, which are all supplied by the PPP [45]. In the present work a significant amount of carbon was recycled from the PPP into glycolysis via fructose 6-phosphate and glyceraldehyde 3-phosphate by the reversible reactions of the PPP. This indicates that the demand for NADPH, but not for carbon precursors, determined the flux towards the PPP. A detailed discussion of the NADPH metabolism is given below. Previous enzymatic studies of S. cerevisiae in chemostat culture revealed that the in vitro activity, i.e. the capacity, of glucose 6-phosphate dehydrogenase slightly increased with the dilution rate between 0.1 h-1 and 0.4 h-1 [9]. The reduced PPP flux is therefore not due to a lack of capacity of glucose 6-phosphate dehydrogenase. In comparison to other studies the relative PPP flux observed under purely oxidative growth was slightly higher [27]. Due to the fact that this flux parameter could be determined with relatively high precision (Table 5) it seems likely that differences in cultivation conditions or the used strains are responsible for this.
(ii) Pyruvate decarboxylase, the key enzyme of fermentative metabolism, was found to already operate at a high activity under conditions in which alcoholic fermentation was absent (Figure 3). The substantial amount of acetaldehyde formed under these conditions was completely converted into cytosolic acetyl CoA which is either used for anabolism or transported into the mitochondrion. These findings disprove the common perception that only a small fraction of pyruvate is channeled through the fermentative by-pass at low glycolytic flux [30], but rather support previous speculations on fluxes around the pyruvate node linked to the Crabtree effect in S. cerevisiae [7]. Based on enzyme profiles, substrate affinities and intracellular pyruvate level studied in chemostat cultivation of S. cerevisiae at varied dilution rate it was previously supposed that pyruvate decarboxylase might be active even during purely oxidative metabolism leading to the formation of cytosolic acetyl CoA and its transport into the mitochondrion. The increased pyruvate level observed during fermentative growth [7] could also be an explanation for the observed flux redirection from mitochondrial to cytosolic alanine biosynthesis.
(iii) We found evidence for substantial in vivo activity of malic enzyme, as was previously reported for S. cerevisiae [29,35]. Additionally we could show a strong flux increase for malic enzyme, when the cells switched from pure oxidation to fermentative metabolism. The flux through malic enzyme increased almost linearly with increasing growth rate (Figure 4C). Although the present work and previous studies differ, concerning e.g. the investigated strain, the pH value or the nutrient status, the flux through malic enzyme is generally increased at high growth rate. As exception only a minor flux of 7 % was observed for malic enzyme in a batch culture of S. cerevisiae grown on glucose [27]. This might arise from the above mentioned differences in the experimental setup or the generally high uncertainty for the quantification of this flux parameter (Table 5). The exact physiological role of malic enzyme in S. cerevisiae so far remained quite intriguing [27]. Also this study does not really provide a satisfying insight into the function of malic enzyme. What at least can be stated is that the flux through malic enzyme seems to be related to the flux through the cytosolic pyruvate carboxylase and the oxaloacetate transporter. One might conclude that the increased activity of malic enzyme results indirectly from the flux redirection at the cytosolic pyruvate node.
(iv) The obtained fluxes allow a detailed inspection of the NADPH metabolism during the metabolic shift. Hereby, it should be noted that the performed flux analysis did not include a balance for NADPH. The fluxes reactions through NADPH supplying reactions were determined only from the fit of the labeling data and the metabolite balances given below. Generally S. cerevisiae can recruit different enzymes for the formation of NADPH involving the glucose 6-phosphate dehydrogenase and 6-phosphogluconate dehydrogenase in the PPP [31], NADPH specific acetaldehyde dehydrogenase [32], and NADPH specific isocitrate dehydrogenase [33]. Also malic enzyme is potentially capable to supply NADPH [34]. Clear differences resulted for the different physiological conditions (Figure 5). Linked to the strong decrease of the biomass yield, the anabolic NADPH demand decreased with increasing growth rate. To some extent this is attenuated by the higher protein content of S. cerevisiae at high growth rate and the correspondingly increased stoichiometric NADPH demand (Tables 1, 2). We additionally found that sources of NADPH varied with the physiological state.
During purely oxidative growth the PPP alone can account for the total anabolic NADPH demand. Under these conditions an additional NADPH supply by other reactions seems not necessary. Because no intercompartmental transport system for NADPH has been described for S. cerevisiae, it is likely that also mitochondrial enzymes are involved in the supply of NADPH. A potential candidate is mitochondrial NADPH specific isocitrate dehydrogenase [33]. This enzyme can, however, account only for a certain fraction of the total flux through the TCA cycle. An exclusive activity of NADPH specific isocitrate dehydrogenase can be excluded. Assuming this the total amount of NADPH supplied would be almost twice as high as the anabolic demand. Thus we conclude that to large extent mitochondrial NADH specific isoenzymes of isocitrate dehydrogenase contribute to the TCA cycle flux during oxidative growth of S. cerevisiae.
Under predominantly fermentative utilization of glucose only about 60 % of the totally required NADPH was provided by the PPP. This is an indication for an additional cytosolic NADPH source required during fermentative metabolism in S. cerevisiae. S. cerevisiae possesses several isoenzymes for acetaldehyde dehydrogenase with different specificity for NADH and NADPH [35]. Possibly the relative contribution of the different isoenzymes changes with the on-set of fermentation, so that instead of mainly NADH under oxidative growth, also NADPH is formed. Recently it was found that especially the NADPH dependent acetaldehyde dehydrogenase isoforms are involved in anaerobic fermentation of S. cerevisiae [35]. The increased NADPH formation by acetaldehyde dehydrogenase would probably lead to an increased level of intracellular NADPH, known to inhibit the PPP enzymes glucose 6-phosphate dehydrogenase and 6-phosphogluconate dehydrogenase [31]. The reduction of the PPP flux might therefore not only be caused by the reduced anabolic demand, but also by the activation of alternative cytosolic pathways for NADPH supply.
(v) An insight into the limiting reactions, involved in the complex flux redirection during the metabolic shift, can be derived from absolute carbon fluxes. As can be calculated from the relative fluxes and the specific glucose uptake rate (Figure 3) the transition from oxidative to fermentative metabolism in S. cerevisiae was accompanied by a more than 5-fold increase of the specific glucose uptake flux. It is interesting to see, how the increased influx of carbon is distributed throughout the metabolic network. The absolute carbon flux of enzymes involved in fermentation and the transport of cytosolic acetyl CoA into the mitochondrion are shown in Figure 6A. Pyruvate decarboxylase exhibits a tremendous flux increase with increasing growth rate. A completely different picture results for the enzymes catalyzing the subsequent metabolic steps, i.e. acetaldehyde dehydrogenase, acetyl CoA synthase and the transporter. The absolute carbon flux through acetyl CoA synthase remained constant for all physiological states investigated. Obviously this enzyme was already working at maximum capacity during purely oxidative growth. The saturation of acetyl CoA synthase obviously initiated the secretion of acetate into the medium with increasing growth rate. Additionally a strong decrease of the intracellular acetyl CoA level appears very likely. The insufficient activity of this enzyme had also consequences for the intercompartmental acetyl CoA transport. As shown, the absolute carbon flux through the transporter decreased to practically zero (Figure 6A). The cytosolic acetyl CoA was in fact completely channeled into anabolic reactions. Probably the reactions withdrawing cytosolic acetyl CoA into anabolism are favored by a higher affinity of the catalyzing enzymes. The insufficient capacity of acetaldehyde dehydrogenase and acetyl CoA synthase could even limit the supply of cytosolic acetyl CoA for anabolism, i.e. the formation of lysine and lipids. In this regard a significant decrease of the in vitro activity, i. e. the capacity, of acetaldehyde dehydrogenase and acetyl CoA synthase was previously observed during the metabolic shift from oxidative to fermentative metabolism [9]. It was concluded that probably these enzymes display the limiting steps that cause the secretion of ethanol and acetate at high growth rate. The present study provides further evidence that indeed these reactions are key points with respect to increased production of ethanol and acetate. The same principle can be applied to study the reactions downstream of the pyruvate pool (Figure 6B). The absolute carbon flux of pyruvate dehydrogenase was low under purely oxidative growth, but showed no further increase towards higher growth rate, despite the carbon flux entering the cytosolic pyruvate pool, and thus potentially available, was much higher. This indicates that pyruvate dehydrogenase already operates at maximum capacity during respiro-fermentative growth at μ = 0.3 h-1. The excess carbon entering the cytosolic pyruvate pool is obviously directed towards pyruvate carboxylase and fermentation. Probably also the activation of cytosolic alanine synthesis is related to this phenomenon. Pyruvate dehydrogenase thus displays an additional bottleneck. The insufficient capacity of pyruvate dehydrogenase seems to have an influence on the transport of pyruvate into the mitochondrion. Due to the fact that, during fermentative growth, a substantial fraction of mitochondrial pyruvate is already supplied via malic enzyme, the absolute carbon flux of the pyruvate transport is lower (Figure 6B). The absolute carbon flux through glucose 6-phosphate dehydrogenase increased about 1.5-fold when the dilution rate was increased from 0.15 h-1 to 0.4 h-1. In a previous study it was shown that, linked to the same shift in dilution rate, the intracellular level of this enzyme was increased about 1.6-fold [9]. Obviously, both glucose 6-phosphate dehydrogenase level and NADPH level are involved in the regulation of carbon flux into the PPP. In case of acetaldehyde dehydrogenase and acetyl CoA synthase the limited carbon flux at high dilution rate seems to be related to a decrease in enzyme level [9]. Additionally inhibitory effects, e.g. of acetaldehyde on acetaldehyde dehydrogenase [9], might be involved. The present study suggests that the flux redirection, caused by the regulatory network linked to the Crabtree effect, leads to a saturation of the capacity, i.e. the absolute flux, of acetaldehyde dehydrogenase, acetyl CoA synthase, and pyruvate dehydrogenase. It is the limitation of these enzymes that is then mainly responsible for the onset of the production of fermentative by-products during growth of S. cerevisiae at high growth rate.
Conclusion
Summarizing, the regulatory changes linked to the metabolic transition affect a number of pathways throughout the whole central metabolism of S. cerevisiae. It is known that yeast undergoes various changes involving e. g. change of enzyme levels, tolerance to stress, or morphology, when the metabolism switches from respirative to fermentative growth. These changes are controlled complex regulatory systems of transcriptional control as well as post-transcriptional control [36]. One of the important primary signals for the regulation of yeast metabolism under different nutrient conditions is the concentration of glucose [37]. As shown the glucose concentration in the medium indeed increased with increasing growth rate (Figure 2). The on-set of the glucose repression system was probably the major signal that triggered the observed changes in the intracellular carbon fluxes. This includes the observed decrease of flux through the TCA cycle with increasing growth rate, which probably results from direct repression of TCA cycle genes on the transcriptional level by the glucose repression system [38]. Similarly the strong increase of the specific glucose uptake rate and the glycolytic flux is probably a direct result of transcriptional induction of the corresponding genes [39]. Recent studies on fluxes and transcriptional levels revealed that a number of intracellular pathways in S. cerevisiae is probably regulated on the post-transcriptional level [36]. Metabolic flux analysis alone on itself as performed in the present study cannot distinguish between flux changes directly controlled or indirectly affected e.g. by regulation of interconnected reactions. However, it appears as a valuable tool to quantitatively assess metabolic changes in great detail and to provide important data leading to a deeper understanding of the metabolism of S. cerevisiae.
Methods
Organism and cultivation
S. cerevisiae ATCC 32167 was purchased from the American Type Strain and Culture Collection (Manassas, USA). The mineral medium used for cultivation contained 2 g/L glucose, 0.5 g/L (NH4)2HPO4, 1.0 g/L (NH4)2SO4, 0.05 g/L MgSO4 ·7 H2O, 0.025 g/L citric acid, 0.5 g/L KCl, 0.03 g/L CaCl2·2 H2O, 3 mg/L FeCl3·6 H2O, 2.1 mg/L MnSO4·H2O, 1.8 mg/L ZnSO4·7 H2O, 0.5 mg/L CuSO4·5 H2O, 60.3 mg/L myo-inositol, 30 mg/L Ca-panthotenate, 6 mg/L thiamin·HCl, 1.5 mg/L pyridoxine·HCl, 0.03 mg/L biotin, and 50 mmol/L phosphate buffer (pH 6.2). All chemicals were of analytical grade and purchased from Fluka (Buchs, Switzerland), Merck (Darmstadt, Germany), or Sigma (St. Louis, USA). In tracer studies normal glucose was replaced by an equal molar concentration of 99% [1-13C] glucose (Campro, Veenendal, Netherlands). Cultivations were performed in a continuously operated bioreactor (Meredos, Bovenden, Germany) with a culture volume of 100 ml at 30°C and 500 rpm. The aeration rate was kept at 100 mL/min by a mass flow controller (WMR Compact 4, Brooks Instruments, Veenendal, Netherlands). Composition of inlet and exhaust gas was determined on-line by mass spectrometry as described previously [40]. For flux quantification the culture was operated with the tracer substrate for five residence times, including on-line verification by a constant CO2 concentration in the exhaust gas, so that metabolic and isotopic steady-state could be assumed. Samples were then taken continuously from the outlet of the reactor and harvested at 4°C. Supernatant, obtained by centrifugation (16000 g, 5 min, Biofuge Pico, Heraeus, Hanau, Germany), was used for analysis of extracellular metabolites. The cell pellet was used for 13C labeling analysis by GC-MS.
Analysis of biomass and extracellular metabolites
Optical density (OD660) was measured at 660 nm (Novaspec II, Pharmacia, Freiburg, Germany). Cell dry mass (CDM) was determined gravimetrically after two-fold washing, centrifugation (10 min, 4°C, 3940 g; Labofuge 400 R, Heraeus, Hanau, Germany), and drying at 80°C until constant weight. A linear relationship of OD660 = 0.52 [g CDM /L] was obtained. Enzymatic test-kits were used for quantification of glucose, glycerol, acetate (Boehringer Mannheim, Mannheim, Germany) and ethanol (Sigma Diagnostics, St. Louis, USA) in the supernatant. Concentrations of 2-oxoglutarate, pyruvate, succinate, and fumarate in the supernatant were analyzed by HPLC [40]. Amino acid composition of the cell protein of S. cerevisiae was quantified by HPLC [41] after hydrolysis with 6 M HCl for 24 h at 105°C. The mean relative error of the stoichiometric measurements, i.e. the quantification of the yield coefficients, was 3.7 % (biomass yield), 4.0 % (ethanol yield), 9.3 % (acetate yield) and 11.3 % (glycerol yield).
GC-MS analysis of proteinogenic amino acids
Measurement of amino acid labeling patterns from cell hydrolysate was carried out by GC-MS in selective ion monitoring mode [42]. For a number of ion clusters all mass isotopomer fractions from the non-labeled (x0) to the fully 13C labeled (xn) variant could be measured with good signal intensity and accuracy (Table 4). Accordingly, a comprehensive data set was available for calculation of the metabolic fluxes. The relative measurement error for a mass isotopomer fraction from the four replicates at each dilution rate was typically below 1 %. It must be mentioned that some amino acids could not be considered for the flux calculation. These were tryptophan, methionine, cysteine and histidine, which did not yield suitable GC/MS signals, proline, which exhibited isobaric overlay, and leucine and isoleucine with ambiguous fragmentation pathways due to similarity of the amino acid side chain and the derivatization agent.
Metabolic network of S. cerevisiae
The metabolic network of S. cerevisiae comprising the major routes of the central carbon metabolism (PPP, glycolysis, fermentative pathways, TCA cycle, and anaplerosis) is shown in Figure 1. As shown, cytosol and mitochondrion are regarded as separate compartments. The model comprises separate pools for pyruvate, oxaloacetate and acetyl CoA in each of the two compartments, respectively. Anaplerotic carboxylation of pyruvate is located in the cytosol [43]. The malic enzyme is considered as mitochondrial reaction [34]. Additionally transport reactions for pyruvate, oxaloacetate and acetyl CoA are considered for intercompartmental carbon exchange [21,27,44]. The transport of pyruvate was assumed to be unidirectional due to its coupling to the proton motive force, whereas the transport reactions for oxaloacetate and acetyl CoA were considered reversible [26]. The reaction catalyzed by glucose 6-phosphate isomerase and the TCA cycle reactions interconverting oxaloacetate and fumarate were regarded as reversible steps, whereby the reversibility of these reactions was taken from the literature [27]. In addition to the central metabolic routes, biosynthetic pathways towards the different amino acids are implemented. In most of the cases amino acid formation is carried out by a single pathway in either one of the two compartments. Exceptions are the biosynthesis of alanine and glycine. For alanine biosynthesis a cytosolic and a mitochondrial pathway was implemented in the model. In addition to the known mitochondrial route, a cytoplasmic alanine amino transferase, generating alanine from cytosolic pyruvate, was recently identified in S. cerevisiae [25]. For glycine the alternative route via threonine aldolase [45,46] was taken into account in addition to the synthesis from serine. The following balance equations of intracellular metabolites were formulated for the examined network using the numbering of the fluxes of Figure 1.
Cytosolic Metabolite Pools
Glucose 6-phosphate: ν1 - ν2 - ν3 + ν4 - ν5 = 0 (1)
Fructose 6-phosphate: ν3 - ν4 - ν6 + ν8 - ν9 + ν10 - ν11 = 0 (2)
Pentose phosphate: ν2 - ν7 - ν8 + ν9 - 2ν12 + 2ν13 = 0 (3)
Erythrose 4-phosphate: - ν8 + ν9 + ν10 - ν11 - ν14 = 0 (4)
Sedoheptulose 7-phosphate: ν12 - ν13 - ν10 + ν11 - ν5 = 0 (5)
Glyceraldehyde 3-phosphate: ν6 - ν10 + ν11 + ν8 - ν9 + ν12 - ν13 + ν16 - ν17 - ν18 = 0 (6)
Dihydroxyacetone-phosphate: ν6 - ν15 - ν16 = 0 (7)
3-Phosphoglycerate: ν17 - ν19 - ν20 - ν24 = 0 (8)
Serine: ν20 - ν21 - ν22 = 0 (9)
Glycine: ν22 - ν23 + ν29 = 0 (10)
Threonine: ν28 - ν29 - ν30 = 0 (11)
Phosphoenolpyruvate: ν24 - ν25 - ν26 = 0 (12)
Pyruvatecyt: ν26 - ν27 - ν32 - ν33 - ν34 = 0 (13)
Acetaldehyde: ν32 - ν37 - ν38 = 0 (14)
Acetate: ν38 - ν39 - ν40 = 0 (15)
Acetyl CoAcyt: ν40 - ν41 - ν47 + ν48 = 0 (16)
Alaninecyt: ν33 - ν42 = 0 (17)
Oxaloacetatecyt: ν27 - ν28 - ν31 - ν35 + ν36 = 0 (18)
Mitochondrial Metabolite Pools
Pyruvatemit: ν34 - ν43 - ν44 + ν50 - ν51 = 0 (19)
Alaninemit: ν43 - ν45 = 0 (20)
Acetyl CoAmit: ν44 - ν46 + ν47 - ν48 - ν49 = 0 (21)
2-Oxoglutarate: ν49 - ν52 - ν53 = 0 (22)
Succinate: ν53 - ν54 + ν55 = 0 (23)
Oxaloacetate/Malatemit: ν35 - ν36 - ν49 - ν50 + ν54 - ν55 = 0 (24)
For each examined physiological state a different, growth rate dependent, cellular composition of S. cerevisiae was considered (Table 1). The macromolecular composition of a yeast cell in relation to the growth rate was calculated from literature data [47-49]. The composition of each macromolecule, however, was assumed to be identical for the different growth conditions [27]. The content and composition of lipids, nucleic acids, and carbohydrates was taken from the literature [48,49]. The amino acid composition of the cell protein was experimentally determined in this work. Based on these data and the stoichiometry of the anabolic pathways in yeast the demand for the different precursor metabolites could be calculated (Table 2) and used as measured fluxes in the flux estimation routine as described previously [41].
Mathematical modeling and flux analysis
Flux calculations were carried out with Matlab 6.1 and Simulink 3.0 (Mathworks Inc., Nattick, USA) combining metabolite balancing and isotopomer modeling. The performed approach utilized metabolite balancing (see mass balances given above) during each optimization step considering stoichiometric data on anabolic demand for biomass precursors (Table 2) and product secretion (Table 3). The set of fluxes that gave minimum deviation between experimental and simulated mass isotopomer fractions was taken as best estimate for the intracellular flux distribution. As error criterion a weighted sum of least squares was used [41]. Since the non-linear structure of isotopomer models may potentially lead to local minima, multiple parameter initializations were used to assure that the obtained flux distributions represented a global optimum. Statistical analysis of the obtained flux parameters was carried out by Monte-Carlo analysis, whereby random, normal-distributed noise was added to the measurement data [41,50].
Authors' contributions
OF carried out the experimental work including the chemostat cultivations, the 13C tracer studies and all analytics involved.
CW designed the study and carried out all computational work involving the construction of the metabolic network model of S. cerevisiae, the software implementation, the calculation of carbon fluxes, and the statistical analysis. He drafted and composed the manuscript.
Both authors read and approved the final manuscript.
Acknowledgements
We acknowledge the advice of Michael Hans in the performed tracer studies.
Figures and Tables
Figure 1 Metabolic network of the central cytosolic and mitochondrial metabolism of S. cerevisiae investigated in the present work. The network comprises glycolysis, pentose phosphate pathway, anaplerotic carboxylation, fermentative pathways, inter-compartmental transport of acetyl-CoA, pyruvate, and oxaloacetate, respectively, TCA cycle, malic enzyme and anabolic reactions from intermediary metabolites into anabolism.
Figure 2 Cultivation profile of S. cerevisiae in chemostat under aerobic, glucose-limited conditions at different growth rates: Concentration of cell dry mass (CDM), ethanol, acetate, and glucose.
Figure 3 Intracellular carbon flux distribution of S. cerevisiae cultivated in chemostat on [1-13C] glucose under aerobic glucose-limited conditions at different growth rates. All fluxes are given as relative fluxes normalized to the specific glucose uptake rate. For each reaction the fluxes corresponding to purely oxidative (μ = 0.15 h-1, qglc = 1.56 mmol g-1 h-1), respiro-fermentative (μ = 0.30 h-1, qglc = 4.90 mmol g-1 h-1), and mainly fermentative growth (μ = 0.40 h-1, qglc = 8.23 mmol g-1 h-1), respectively, are shown from top to bottom. For reversible reactions an additional arrow indicates the direction of the net flux and the values in the squared brackets are the obtained reversibilities of the corresponding enzymes. The fluxes correspond to the optimal fit between experimentally determined steady-state 13C labeling patterns of amino acids of the cell protein and 13C labeling patterns simulated via isotopomer modelling.
Figure 4 Correlation of specific growth rate and relative carbon fluxes in the central metabolism of S. cerevisiae: glucose 6-phosphate dehydrogenase (A), citrate synthase (B), and malic enzyme (C). Data shown comprise fluxes of S. cerevisiae ATCC 31267 determined at different growth rates in glucose-limited chemostat (this work, open circle), of S. cerevisiae CEN.PK 113.7D at μ = 0.1 h-1 in glucose-limited chemostat ([27],open triangle), at μ = 0.37 h-1 in batch culture under glucose excess ([27], closed triangle), of S. cerevisiae CEN.PK 113.7D at μ = 0.1 h-1 in glucose-limited chemostat ([26], open square), at μ = 0.4 h-1 in batch culture under glucose excess ([26], closed square). Note that Fiaux et al. [26] determined flux intervals instead of exact values, whereby the values given here represent the corresponding average of each interval.
Figure 5 NADPH metabolism of S. cerevisiae cultivated in chemostat under aerobic glucose-limited conditions at different growth rates. The fluxes correspond to purely oxidative (μ = 0.15 h-1), respiro-fermentative (μ = 0.30 h-1), and mainly fermentative growth (μ = 0.40 h-1).
Figure 6 Absolute carbon fluxes of S. cerevisiae cultivated in chemostat under aerobic glucose-limited conditions at different growth rates: Fluxes of pyruvate decarboxylase, acetaldehyde dehydrogenase and acetyl CoA synthase (A), and of intercompartmental pyruvate transport, pyruvate dehydrogenase and citrate synthase (B). For each enzyme the flux at purely oxidative (μ = 0.15 h-1), respiro-fermentative (μ = 0.30 h-1), and mainly fermentative growth (μ = 0.40 h-1) is shown from left to right. Additionally the 90 % confidence intervals for the fluxes are given.
Table 1 Macromolecular composition of Saccharomyces cerevisiae in glucose-limited chemostat culture at different dilution rates. The data given in g (g cell dry mass)-1 are calculated from literature data [27, 47-49].
Compound D = μ = 0.15 h-1 D = μ = 0.30 h-1 D = μ = 0.40 h-1
Protein 0.415 0.480 0.524
Carbohydrate 0.372 0.301 0.254
Lipids 0.070 0.070 0.070
RNA+DNA 0.069 0.096 0.113
Table 2 Anabolic demand of Saccharomyces cerevisiae during aerobic glucose-limited chemostat culture at different dilution rates. The data are given in μmol (g cell dry mass)-1 and are calculated from the cellular composition.
Compound D = μ = 0.15 h-1 D = μ = 0.30 h-1 D = μ = 0.40 h-1
Glucose 6-phosphate 2089 1694 1431
Erythrose 4-phosphate 243 281 307
Ribose 5-phosphate 115 127 141
Glyceraldehyde 3-phosphate 77 77 77
Phosphoglycerate (for lipids, nucleotides) 44 44 44
Phosphoglycerate (for serine + cysteine) 343 397 433
Phosphoglycerate/Oxaloacetate (for glycine)a 236 273 298
Oxaloacetate (for threonine, methionine, isoleucine) 393 455 496
Oxaloacetate (for others) 290 332 364
Phosphoenolpyruvate 457 529 577
Pyruvate (for alanine)b 304 352 384
Pyruvate (mitochondrial, for others) 981 1135 1237
Acetyl-CoA (cytosolic) 2108 2142 2164
Acetyl-CoA (mitochondrial) 216 250 273
2-Oxoglutarate 1008 1166 1272
NADPH 10088 11206 11956
a The actual precursor demand for glycine biosynthesis depends on the contribution of the serine pathway (from phosphoglycerate) and threonine aldolase (from oxaloacetate). Based on the relative activity of both pathways, the corresponding demand has to be added appropriately to phosphoglycerate and oxaloacetate, respectively.
b The demand of cytosolic and mitochondrial pyruvate for alanine synthesis depends on the relative contribution of cytosolic and mitochondrial route to alanine synthesis.
Table 3 Specific growth, substrate consumption, product formation rates and carbon balance of Saccharomyces cerevisiae cultured on [1-13C] glucose in chemostat at different growth rates for metabolic flux analysis.
Specific rate (C mmol h-1) Purely Oxidative Growth (μ = 0.15 h-1) Respiro-Fermentative Growth (μ = 0.30 h-1) Mainly Fermentative Growth (μ = 0.40 h-1)
Glucose - 0.96 - 1.88 - 2.30
Organic acids 0.01 0.01 0.01
Ethanol 0.00 0.31 0.82
Glycerol 0.01 0.05 0.08
Acetate 0.00 0.09 0.04
CDM 0.55 0.69 0.67
CO2 0.37 0.66 0.64
sum of substrates - 0.96 - 1.88 - 2.30
sum of products 0.94 1.81 2.26
carbon recovery [%] 97.9 96.3 98.3
Table 4 13C labeling patterns of amino acids in cell hydrolysates of Saccharomyces cerevisiae cultivatedon [1-13C] glucose in chemostat under aerobic glucose-limited conditions at different growth rates.
Measured Metabolite Purely Oxidative Growth (μ = 0.15 h-1) Respiro-Fermentative Growth (μ = 0.30 h-1) Mainly Fermentative Growth (μ = 0.40 h-1)
M0 M1 M2 M0 M1 M2 M0 M1 M2
Alanine (m/z 260, C1–3)
calca 0.647 0.328 0.596 0.380 0.539 0.435
expb 0.652 0.331 0.597 0.381 0.543 0.442
Alanine (m/z 232, C2–3)
calc 0.671 0.316 0.618 0.372 0.561 0.427
exp 0.666 0.325 0.615 0.374 0.549 0.442
Glycine (m/z 246, C2)
calc 0.922 0.074 0.932 0.065 0.964 0.035
exp 0.922 0.074 0.931 0.064 0.963 0.036
Valine (m/z 186, C2–5)
calc 0.450 0.424 0.117 0.365 0.457 0.165 0.308 0.471 0.203
exp 0.442 0.424 0.116 0.361 0.451 0.165 0.301 0.480 0.203
Valine (m/z 260, C2–5)
calc 0.450 0.424 0.117 0.365 0.457 0.165 0.308 0.471 0.203
exp 0.442 0.428 0.120 0.364 0.457 0.170 0.300 0.475 0.206
Valine (m/z 288, C2–5)
calc 0.434 0.425 0.128 0.346 0.451 0.179 0.288 0.464 0.220
exp 0.435 0.429 0.125 0.355 0.456 0.174 0.294 0.475 0.210
Serine (m/z 288, C2–3)
calc 0.694 0.303 0.628 0.369 0.566 0.430
exp 0.693 0.300 0.624 0.366 0.569 0.422
Serine (m/z 362, C2–3)
calc 0.694 0.303 0.628 0.369 0.566 0.430
exp 0.693 0.302 0.623 0.369 0.567 0.424
Serine (m/z 390, C1–3)
calc 0.682 0.311 0.611 0.375 0.560 0.431
exp 0.682 0.308 0.617 0.370 0.564 0.428
Threonine (m/z 376, C2–4)
calc 0.541 0.381 0.526 0.409 0.515 0.440
exp 0.540 0.387 0.521 0.409 0.504 0.455
Threonine (m/z 404, C1–4)
calc 0.496 0.396 0.504 0.413 0.488 0.444
exp 0.500 0.398 0.505 0.411 0.496 0.455
Aspartate (m/z 418, C1–4)
calc 0.496 0.396 0.504 0.413 0.488 0.444
exp 0.502 0.399 0.506 0.415 0.499 0.454
Arginine (m/z 414, C1–5)
calc 0.302 0.417 0.219 0.239 0.410 0.261 0.263 0.450 0.241
exp 0.309 0.415 0.218 0.245 0.406 0.260 0.263 0.450 0.244
Phenylalanine (m/z 336, C1–9)
calc 0.379 0.429 0.166 0.304 0.437 0.212 0.240 0.442 0.262
exp 0.371 0.417 0.166 0.305 0.431 0.212 0.238 0.437 0.263
Phenylalanine (m/z 302, C1–2)
calc 0.973 0.961 0.979
exp 0.957 0.952 0.975
Tyrosine (m/z 466, C1–9)
calc 0.379 0.429 0.166 0.304 0.437 0.212 0.240 0.442 0.262
exp 0.379 0.421 0.165 0.304 0.432 0.213 0.242 0.440 0.250
Tyrosine (m/z 302, C1–2)
calc 0.973 0.961 0.979
exp 0.972 0.964 0.966
Glutamate (m/z 432, C1–5)
calc 0.352 0.427 0.184 0.278 0.431 0.233 0.276 0.458 0.230
exp 0.353 0.429 0.184 0.278 0.432 0.233 0.279 0.467 0.228
a Calculated values predicted by the solution of the mathematical model corresponding to the optimized set of fluxes
b Experimental values obtained by GC-MS analysis of TBDMS-derivatized amino acids
Table 5 Statistical evaluation of metabolic fluxes of Saccharomyces cerevisiae cultured in chemostat under aerobic, glucose-limited conditions at μ = 0.15 h-1, μ = 0.30 h-1, and μ = 0.40 h -1.
Flux Parameter Interval for 90 % Confidencea (μ = 0.15 h-1) Interval for 90 % Confidence (μ = 0.30 h-1) Interval for 90 % Confidence (μ = 0.40 h-1)
Net Flux
glucose 6-phosphate isomerase [24.7 26.1] [58.5 61.3] [77.3 78.9]
fructose 1,6-bisphosphate aldolase [59.5 61.0] [77.6 79.6] [86.3 88.0]
triosephosphate isomerase [57.6 59.1] [72.1 74.1] [79.6 81.6]
glyceraldehyde 3-phosphate dehydrogenase [132.5 135.9] [157.7 162.0] [169.2 173.2]
enolase [126.9 130.7] [154.1 158.8] [165.2 169.9]
pyruvate kinase [122.4 126.5] [150.6 155.8] [162.2 167.4]
glucose 6-phosphate dehydrogenase [53.7 55.4] [28.1 31.2] [14.5 15.5]
transaldolase [18.3 18.9] [9.7 10.7] [5.1 5.4]
transketolase 1 [18.3 18.9] [9.7 10.7] [5.1 5.4]
transketolase 2 [15.9 16.6] [7.9 9.1] [3.6 4.0]
pyruvate carboxylase [21.6 27.0] [24.5 30.3] [25.5 37.3]
pyruvate decarboxylase [60.5 83.8] [74.4 92.6] [120.9 126.4]
acetyl CoA synthase [60.9 84.1] [10.9 29.3] [8.5 14.2]
intercompartmental pyruvate transport [17.8 38.1] [33.7 51.0] [7.5 11.2]
intercompartmental oxaloacetate transport [14.4 19.9] [18.3 24.9] [20.8 33.5]
intercompartmental acetyl CoA transport [41.0 63.6] [0.0 15.4] [0.0 2.8]
overall intercompartmental transport [94.3 100.1] [65.6 75.7] [33.2 41.2]
pyruvate dehydrogenase [9.8 36.5] [36.2 61.9] [16.3 31.2]
citrate synthase [69.3 77.4] [48.8 60.2] [17.2 29.4]
oxoglutarate dehydrogenase [59.2 68.1] [41.0 53.7] [10.4 23.9]
malic enzyme [4.5 10.4] [10.7 18.3] [14.0 28.0]
cytosolic alanine synthesis [0.0 0.0] [1.0 1.4] [0.6 1.2]
mitochondrial alanine synthesis [2.7 3.0] [0.7 1.1] [0.5 1.4]
glycine synthesis via serine [1.5 1.8] [0.2 1.2] [1.1 1.5]
glycine synthesis via threonine [0.5 0.7] [0.4 1.5] [0.0 0.4]
NADPH supply by PPP [107.4 110.8] [56.2 62.4] [29.0 31.0]
NADPH demand for anabolism [92.8 100.8] [61.7 72.8] [51.4 64.6]
Flux Reversibility b
glucose 6-phosphate isomerase [19.4 25.5] [21.2 31.9] [17.6 30.7]
transaldolase [0.0 0.3] [0.0 0.3] [0.0 0.3]
transketolase 1 [2.1 4.0] [20.7 36.4] [2.1 2.9]
transketolase 2 [0.0 0.8] [1.9 4.1] [0.0 0.2]
intercompartmental oxaloacetate transport [2.1 3.0] [0.1 0.4] [0.9 1.9]
intercompartmental acetyl CoA transport [17.4 28.6] [14.8 32.0] [17.3 30.1]
a The 90 % confidence intervals were obtained by a Monte-Carlo approach including 100 independent parameter estimation runs for each scenario with statistically varied experimental data assuming a normal distribution of the experimental error.
b Flux reversibility is defined as ratio of back flux to net flux.
==== Refs
Nissen TL Kielland-Brandt MC Nielsen J Villadsen J Optimization of ethanol production in Saccharomyces cerevisiae by metabolic engineering of the ammonium assimilation Metab Eng 2000 2 69 77 10935936 10.1006/mben.1999.0140
Kato A Nakamura S Ibrahim H Matsumi T Tsumiyama C Kato M Production of genetically modified lysozymes having extreme heat stability and antimicrobial activity against gram negative bacteria in yeast and in plant Nahrung 1998 42 128 130 9739552 10.1002/(SICI)1521-3803(199808)42:03/04<128::AID-FOOD128>3.3.CO;2-Q
Kim MD Lee TH Lim HK Seo JH Production of antithrombotic hirudin in GAL1-disrupted Saccharomyces cerevisiae Appl Microbiol Biotechnol 2004 65 259 262 15048590
Koh JH Yu KW Suh HJ Biological activities of Saccharomyces cerevisiae and fermented rice bran as feed additives Lett Appl Microbiol 2002 35 47 51 12081549 10.1046/j.1472-765X.2002.01131.x
Oliver SG Functional genomics: lessons from yeast Philos Trans R Soc Lond B Biol Sci 2002 357 17 23 11839178 10.1098/rstb.2001.1049
Dujon B European Functional Analysis Network (EUROFAN) and the functional analysis of the Saccharomyces cerevisiae genome Electrophoresis 1998 19 617 624 9588813 10.1002/elps.1150190427
Postma E Verduyn C Scheffers WA Van Dijken JP Enzymic analysis of the crabtree effect in glucose-limited chemostat cultures of Saccharomyces cerevisiae Appl Environ Microbiol 1989 55 468 477 2566299
Kaspar von Meyenburg H Energetics of the budding cycle of Saccharomyces cerevisiae during glucose limited aerobic growth Arch Mikrobiol 1969 66 289 303 5384632 10.1007/BF00414585
Serra A Strehaiano P Taillandier P Characterization of the metabolic shift of Saccharomyces bayanus var. uvarum by continuous aerobic culture Appl Microbiol Biotechnol 2003 62 564 568 12679850 10.1007/s00253-003-1288-5
Barford J Hall RJ An examination of the Crabtree effect in Saccharomyces cerevisiae: The role of respiratory adaptation J Gen Microbiol 1979 114 267 275
Sonnleitner B Käppeli O Growth of Saccharomyces cerevisiae is controlled by its limited respiratory capacity: formulation and verification of a hypothesis Biotechnol Bioeng 1986 28 927 937 10.1002/bit.260280620
Rieger M Käppeli O Fiechter A The role of limited respiration in the incomplete oxidation of glucose by Saccharomyces cerevisiae J Gen Microbiol 1083 129 653 661
Thierie J Modeling threshold phenomena, metabolic pathways switches and signals in chemostat-cultivated cells: the Crabtree effect in Saccharomyces cerevisiae J Theor Biol 2004 226 483 501 14759654 10.1016/j.jtbi.2003.10.017
Bellgardt KH Schügerl KBKH Baker's yeast production Bioreaction Engineering 2000 Berlin, Heidelberg, Springer Verlag 227 230
Lo Curto RB Tripodo MM Yeast production from virgin grape marc Bioresour Technol 2001 78 5 9 11265788 10.1016/S0960-8524(00)00175-9
Beudeker RF van Dam HW van der Plaat JB Vellenga K Verachtert HMR Developments in baker's yeast production Yeast biotechnology and biocatalysis 1990 New York, Marcel Dekker 103 146
Kelleher JK Flux estimation using isotopic tracers: common ground for metabolic physiology and metabolic engineering Metab Eng 2001 3 100 110 11289786 10.1006/mben.2001.0185
Sauer U High-throughput phenomics: experimental methods for mapping fluxomes Curr Opin Biotechnol 2004 15 58 63 15102468 10.1016/j.copbio.2003.11.001
Wittmann C Metabolic flux analysis using mass spectrometry Adv Biochem Eng Biotechnol 2002 74 39 64 11991183
Wittmann C Hans M Bluemke W Metabolic physiology of aroma-producing Kluyveromyces marxianus Yeast 2002 19 1351 1363 12402244 10.1002/yea.920
Maaheimo H Fiaux J Cakar ZP Bailey JE Sauer U Szyperski T Central carbon metabolism of Saccharomyces cerevisiae explored by biosynthetic fractional (13)C labeling of common amino acids Eur J Biochem 2001 268 2464 2479 11298766 10.1046/j.1432-1327.2001.02126.x
Blank LM Sauer U TCA cycle activity in Saccharomyces cerevisiae is a function of the environmentally determined specific growth and glucose uptake rates Microbiology 2004 150 1085 1093 15073318 10.1099/mic.0.26845-0
Blank LM Kuepfer L Sauer U Large-scale 13C-flux analysis reveals mechanistic principles of metabolic network robustness to null mutations in yeast Genome Biol 2005 6 R49 15960801 10.1186/gb-2005-6-6-r49
Blank LM Lehmbeck F Sauer U Metabolic-flux and network analysis in fourteen hemiascomycetous yeasts FEMS Yeast Res 2005 5 545 558 15780654 10.1016/j.femsyr.2004.09.008
dos Santos MM Gombert AK Christensen B Olsson L Nielsen J Identification of in vivo enzyme activities in the cometabolism of glucose and acetate by Saccharomyces cerevisiae by using 13C-labeled substrates Eukaryot Cell 2003 2 599 608 12796305 10.1128/EC.2.3.599-608.2003
Fiaux J Cakar ZP Sonderegger M Wuthrich K Szyperski T Sauer U Metabolic-flux profiling of the yeasts Saccharomyces cerevisiae and Pichia stipitis Eukaryot Cell 2003 2 170 180 12582134 10.1128/EC.2.1.170-180.2003
Gombert AK Moreira dos Santos M Christensen B Nielsen J Network identification and flux quantification in the central metabolism of Saccharomyces cerevisiae under different conditions of glucose repression J Bacteriol 2001 183 1441 1451 11157958 10.1128/JB.183.4.1441-1451.2001
Wittmann C Heinzle E Mass spectrometry for metabolic flux analysis Biotechnol Bioeng 1999 62 739 750 10099575 10.1002/(SICI)1097-0290(19990320)62:6<739::AID-BIT13>3.0.CO;2-E
Follstad BD Stephanopoulos G Effect of reversible reactions on isotope label redistribution--analysis of the pentose phosphate pathway Eur J Biochem 1998 252 360 371 9546650 10.1046/j.1432-1327.1998.2520360.x
Lei F Rotboll M Jorgensen SB A biochemically structured model for Saccharomyces cerevisiae J Biotechnol 2001 88 205 221 11434967 10.1016/S0168-1656(01)00269-3
Vaseghi S Baumeister A Rizzi M Reuss M In vivo dynamics of the pentose phosphate pathway in Saccharomyces cerevisiae Metab Eng 1999 1 128 140 10935926 10.1006/mben.1998.0110
Grabowska D Chelstowska A The ALD6 gene product is indispensable for providing NADPH in yeast cells lacking glucose-6-phosphate dehydrogenase activity J Biol Chem 2003 278 13984 13988 12584194 10.1074/jbc.M210076200
Minard KI McAlister-Henn L Sources of NADPH in yeast vary with carbon source J Biol Chem 2005 16179340
Boles E de Jong-Gubbels P Pronk JT Identification and characterization of MAE1, the Saccharomyces cerevisiae structural gene encoding mitochondrial malic enzyme J Bacteriol 1998 180 2875 2882 9603875
Saint-Prix F Bonquist L Dequin S Functional analysis of the ALD gene family of Saccharomyces cerevisiae during anaerobic growth on glucose: the NADP+-dependent Ald6p and Ald5p isoforms play a major role in acetate formation Microbiology 2004 150 2209 2220 15256563 10.1099/mic.0.26999-0
Daran-Lapujade P Jansen ML Daran JM van Gulik W de Winde JH Pronk JT Role of transcriptional regulation in controlling fluxes in central carbon metabolism of Saccharomyces cerevisiae. A chemostat culture study J Biol Chem 2004 279 9125 9138 14630934 10.1074/jbc.M309578200
Rolland F Winderickx J Thevelein JM Glucose-sensing and -signalling mechanisms in yeast FEMS Yeast Res 2002 2 183 201 12702307 10.1016/S1567-1356(02)00046-6
Schuller HJ Transcriptional control of nonfermentative metabolism in the yeast Saccharomyces cerevisiae Curr Genet 2003 43 139 160 12715202
Carlson M Glucose repression in yeast Curr Opin Microbiol 1999 2 202 207 10322167 10.1016/S1369-5274(99)80035-6
Hans MA Heinzle E Wittmann C Free intracellular amino acid pools during autonomous oscillations in Saccharomyces cerevisiae Biotechnol Bioeng 2003 82 143 151 12584756 10.1002/bit.10553
Wittmann C Heinzle E Genealogy profiling through strain improvement by using metabolic network analysis: metabolic flux genealogy of several generations of lysine-producing corynebacteria Appl Environ Microbiol 2002 68 5843 5859 12450803 10.1128/AEM.68.12.5843-5859.2002
Becker J Klopprogge C Zelder O Heinzle E Wittmann C Amplified Expression of Fructose 1,6-bisphosphatase in Corynebacterium glutamicum increases in vivo flux through the pentose phosphate pathway and lysine production on different carbon sources Appl Environ Microbiol 2005 in press
Walker ME Val DL Rohde M Devenish RJ Wallace JC Yeast pyruvate carboxylase: identification of two genes encoding isoenzymes Biochem Biophys Res Commun 1991 176 1210 1217 2039506 10.1016/0006-291X(91)90414-3
Nalecz MJ Nalecz KA Azzi A Purification and functional characterisation of the pyruvate (monocarboxylate) carrier from baker's yeast mitochondria (Saccharomyces cerevisiae) Biochim Biophys Acta 1991 1079 87 95 1888767
Liu JQ Nagata S Dairi T Misono H Shimizu S Yamada H The GLY1 gene of Saccharomyces cerevisiae encodes a low-specific L-threonine aldolase that catalyzes cleavage of L-allo-threonine and L-threonine to glycine--expression of the gene in Escherichia coli and purification and characterization of the enzyme Eur J Biochem 1997 245 289 293 9151955 10.1111/j.1432-1033.1997.00289.x
Monschau N Stahmann KP Sahm H McNeil JB Bognar AL Identification of Saccharomyces cerevisiae GLY1 as a threonine aldolase: a key enzyme in glycine biosynthesis FEMS Microbiol Lett 1997 150 55 60 9163906 10.1016/S0378-1097(97)00096-7
Lange HC Heijnen JJ Statistical reconciliation of the elemental and molecular biomass composition of Saccharomyces cerevisiae Biotechnol Bioeng 2001 75 334 344 11590606 10.1002/bit.10054
Oura E Dellweg H Biomass from carbohydrates Biotechnology 1983 3 Weinheim, VCH Verlag 3 42 10.1002/abio.370030102
Verduyn C Physiology of yeasts in relation to biomass yields Antonie Van Leeuwenhoek 1991 60 325 353 1807201 10.1007/BF00430373
Kiefer P Heinzle E Zelder O Wittmann C Comparative metabolic flux analysis of lysine-producing Corynebacterium glutamicum cultured on glucose or fructose Appl Environ Microbiol 2004 70 229 239 14711646 10.1128/AEM.70.1.229-239.2004
|
16269086
|
PMC1291395
|
CC BY
|
2021-01-04 16:05:47
|
no
|
Microb Cell Fact. 2005 Nov 3; 4:30
|
utf-8
|
Microb Cell Fact
| 2,005 |
10.1186/1475-2859-4-30
|
oa_comm
|
==== Front
Mol CancerMolecular Cancer1476-4598BioMed Central London 1476-4598-4-381624889910.1186/1476-4598-4-38ReviewEpigenetics of cervical cancer. An overview and therapeutic perspectives Dueñas-González Alfonso [email protected] Marcela [email protected] Myrna [email protected] Lucely [email protected] Claudia [email protected] Eduardo [email protected] Unidad de Investigación Biomédica en Cáncer, Instituto Nacional de Cancerología/Instituto de Investigaciones Biomédicas (INCan/IIB), Universidad Nacional Autónoma de Mexico (UNAM), Mexico City. Mexico2 Division of Clinical Research, Instituto Nacional de Cancerología (INCan), Mexico City, Mexico2005 25 10 2005 4 38 38 26 7 2005 25 10 2005 Copyright © 2005 Dueñas-González et al; licensee BioMed Central Ltd.2005Dueñas-González et al; licensee BioMed Central Ltd.This is an Open Access article distributed under the terms of the Creative Commons Attribution License (), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
Cervical cancer remains one of the greatest killers of women worldwide. It is difficult to foresee a dramatic increase in cure rate even with the most optimal combination of cytotoxic drugs, surgery, and radiation; therefore, testing of molecular targeted therapies against this malignancy is highly desirable. A number of epigenetic alterations occur during all stages of cervical carcinogenesis in both human papillomavirus and host cellular genomes, which include global DNA hypomethylation, hypermetylation of key tumor suppressor genes, and histone modifications. The reversible nature of epigenetic changes constitutes a target for transcriptional therapies, namely DNA methylation and histone deacetylase inhibitors. To date, studies in patients with cervical cancer have demonstrated the feasibility of reactivating the expression of hypermethylated and silenced tumor suppressor genes as well as the hyperacetylating and inhibitory effect upon histone deacetylase activity in tumor tissues after treatment with demethylating and histone deacetylase inhibitors. In addition, detection of epigenetic changes in cytological smears, serum DNA, and peripheral blood are of potential interest for development of novel biomolecular markers for early detection, prediction of response, and prognosis.
==== Body
Overview of cervical cancer
Epidemiology and treatment
Cervical cancer remains one of the greatest killers of women worldwide. According to Globocan 2000, it is estimated that in 2000 the numbers of patients diagnosed with and those who died from this disease were 470,606 and 233,372, respectively [1]. It is remarkable that these rates occur despite the fact that cervical cancer is a model for early detection due to its long and relatively well-known natural history, which offers an excellent opportunity for its detection before lesions become invasive [2].
Cervical cancer is currently staged clinically according the International Federation of Gynecology and Obststrics (FIGO) guidelines. In terms of treatment, invasive disease can be divided into three main groups: 1) early stage going from microinvasive disease IA1, IA2 to macroscopic disease confined to cervix and measuring <4 cm, IB1; 2) locally advanced FIGO stages IB2-IVA, and 3) IVB and recurrent disease [3].
Treatment of early stages
The recommended treatment for IA1 patients is either a local procedure such as conization or total hysterectomy depending on the patient's desire to remain fertile, whereas for IA2 patients the recommended procedure is a radical one including pelvic lymphadenectomy. On average, 8% of cases shows positive pelvic lymph nodes. As many women at this disease stage deserve to preserve fertility, radical trachelectomy is becoming an option for these patients. The same can apply for IB1 patients. In early cases that are surgically treated, the presence in the surgical specimen of a combination of intermediate-risk factors (vascular and lymphatic permeation, tumor size >2 cm, and deep cervical stroma invasion) or high-risk factors (positive pelvic lymph nodes, parametrial infiltration, and positive surgical margins) dictates use of adjuvant radiation or chemoradiation respectively. As a group, the prognosis of early-stage cases is fairly good with 5-year survival exceeding 90% [4,5]
Treatment of locally advanced stages
Results of treatment for these patients are far from optimal. In this regard, treatment of locally advanced cervical cancer has experienced no major changes for nearly 80 years during which exclusive radiation was considered the standard of care; thus, 5-year survival for stages IB2, IIB, IIIB, and IVA are 72.2, 63.7, 41.7, and 16.4%, respectively, according the 1998 Annual Report on the Results of Treatment in Gynaecological Cancer [6]. The lengthy permanence of this unimodal treatment was due, on the one hand, to the classical concept that cervical cancer is a disease that progresses in an orderly fashion (local, then regional, and at the very last, systemic); therefore, it could be effectively treated with a local modality such as radiation instead of a systemic modality such as chemotherapy. On the other hand, the role of surgery for locally advanced cases failed to treat the disease successfully by radical surgical procedures [7]. Over the last 20 years, however, an increasingly number of trials that incorporate either chemotherapy and/or surgery with radiation (neoadjuvant chemotherapy followed by radiation, neoadjuvant chemotherapy followed by surgery plus minus adjuvant radiation, and concurrent chemoradiation) have been performed in an attempt to improve treatment results. Radiation concomitant with cisplatin-based chemotherapy is considered the current standard of treatment. This combined modality produces an absolute increase in 5-year survival of 12% as compared with radiation alone. On the other hand, neoadjuvant chemotherapy when followed by surgery – but not when followed by radiation – yields a 15% increase in absolute 5-year survival. These data emerged from three meta-analyses of the literature based on individual patient analysis [8,9].
Treatment of IVB and recurrent disease
Patients with cervical cancer may present at diagnosis with distant metastases (stage IVB) or have, after primary treatment, pelvic recurrence, distant metastases, or a combination of both. Recurrence rates vary from up to 20% to 70% in early stages and locally advanced disease, respectively [10,11] and the majority of recurrences occur within 2 years of diagnosis. At this stage, cervical cancer is even more difficult to treat because this clinical situation is the result of a more malignant phenotype resulting from accumulation of genetic defects during tumor progression and previous therapies; thus, the tumor at this stage is commonly resistant to chemotherapy. Moreover, these patients frequently have a poor performance status that limits use of agressive chemotherapy and the majority of patients die as a result of uncontrolled disease. In this setting, the few patients who recur after initial surgical treatment can be salvaged with concurrent chemoradiation if the disease is local or locoregional [12]. Those who receive primary radiation or chemoradiation and have pelvic disease can be offered an ultraradical procedure such as pelvic exenteration; nonetheless, this procedure is currently limited to patients with small and central tumors that in these situations, pelvic exenteration may offer 5-year survival for up to 50% of patients [13]. Although some efforts have been devoted to extending the exenterative procedures to patients with higher disease burdens by use of intraoperative radiation [14], laterally extended pelvic exenteration [15], or pre-exenterative chemotherapy [16] none of these options are widely used. Unfortunately, patients with IVB and those with distant metastases – with or without pelvic relapse – have no option other than systemic chemotherapy that in this setting has limited value; cisplatin is the most active single agent [17] and more recently in combination with topotecan has shown a modest increase in time to progression and median survival as compared with single agent cisplatin. In any case, median survival remains between 6 and 12 months [18].
Molecular pathogenesis of cervical cancer
Human papillomavirus
Current experimental and epidemiologic information undoubtedly points to the human papillomavirus (HPV) as the primary causal agent in development of cervical carcinoma. Therefore, the study of its carcinogenic role continues to represent the mainstream research on the molecular biology of cervical cancer, with the idea that prophylactic and therapeutic applications of knowledge from this field could benefit millions of women afflicted, or at risk to be afflicted, with HPV-induced cervical cancer.
HPV classification is predicted on DNA sequence homology. At least 200 types have been identified and these have been classified into 16 groups [19]. Genital HPVs are classified according to their potential to induce malignant transformation as follows: high-risk types (16, 18, 31, 33, 35, 39, 45, 51, 52, 56, 58, 59, 68, 73, and 82); probable high-risk types (26, 53, and 66), and low-risk types (6, 11, 40, 42, 43, 44, 54, 61, 70, 72, 81, and CP6108) [20]. Among high-risk strains, HPV 16 and 18 are those most closely associated with cervical carcinoma and are found in >50% and 20% of squamous cell carcinomas, respectively [21].
A large body of knowledge supports the view that high-risk HPV types (HR-HPV) have the ability to transform cells into a malignant phenotype. Nevertheless, only a minority of cervical lesions infected with HR-HPV inevitably progress to cervical carcinoma, as indicated by frequent spontaneous clearance of HPV infection and the long delay between onset of persistent infection and emergence of the malignancy. For that reason, studies have been focused on analyzing the participation of possible viral and cellular factors governing HPV-induced malignancy.
HPV structure
Human papillomaviruses (HPV) belong to the Papovaviridae family. They consist of a 72-capsomere capside containing the viral genome. Capsomers are composed of two structural proteins: the 57 kD late protein L1, which accounts for 80% of the viral particle, and the 43–53 kD minor capside protein L2. The HPV genome consists of eight kilobasepairs (Kbp) and is a double-stranded DNA molecule. Arrangement of the 8–10 open reading frames (ORFs) within the genome is similar in all papillomavirus types, and partly overlapping ORFs are arranged on a sole DNA strand. The genome can be divided into three regions: the long control region (LCR) without coding potential; the region of early proteins (E1–E8), and the region of late proteins (L1 and L2) [22].
Early gene products
E1 and E2 encode proteins that are vital for extrachromosomal DNA replication and completion of the viral life cycle [23]. A hallmark of HPV-associated cervical carcinoma is loss of the expression of viral E2 protein [24]. A fusion product consisting of the small E8 ORF with part of the E2 protein has been described. This fusion protein is able to repress viral DNA replication as well as transcription, and is therefore believed to play a major role in the maintenance of viral latency observed in the basal cells of infected epithelium [25,26]. The E4 protein is expressed in the later stages of infection when complete virions are being assembled, and is not known to have transforming properties; however, it is considered to play an important role in the maturation and replication of the virus [27]. The E5 in open reading frame is often deleted in cervical carcinoma cells, indicating that it might not be essential in maintaining the malignant transformation of the host cell. Nevertheless, it has been reported that E5 protein possesses a weak transforming activity [28].
E6 and E7 proteins
E6 and E7 are the most important oncogenic proteins. Transcription of E6 and E7 genes has been observed to occur always in cervical carcinomas, being the first indication of a main role of these viral genes in HPV-associated tumorigenesis. The immortalization and transforming potential of E6 and E7 proteins have been demonstrated in tissue culture and in experimental animal models [29]. From the studies of E6-p53 and E7-pRb models, numerous actions have been identified of viral gene products on cellular proteins. Therefore, several findings hint at possible ways by which HPV-infected cells may escape controls governing cell growth and proliferation.
The E6 protein of high-risk HPV anogenital types shows weak oncogenic potential in the majority of established cell lines, and cooperation with E7 protein is required for full transforming capacity. Discovery of the inactivation of the tumor suppressor genes p53 and pRB by E6 and E7 oncoproteins provided a basic explanation of how high-risk HPV types induce their oncogenic effects on cervical cells [30]. E6 has many interactions with cellular proteins; nevertheless, its key action is inhibition of the function of tumor suppressor protein p53 by enhancing its degradation through the ubiquitin pathway [31,32]. To inhibit p53 function, E6 requires a cellular protein called E6-associated protein (E6AP). In non-infected cells, ubiquitin-mediated degradation of p53 is triggered by the mdm-2 protein, while in HR HPV-infected cells the E6-E6AP complex replaces mdm-2 in control of cellular p53 levels. This shift shortens the p53 half-life and reduces its levels in cervical carcinoma cells to less than one half of the level found in normal ephithelial cells [33]. It is known that increase in p53 levels plays a critical role in the induction of genes that results in cell cycle arrest [34], allowing repair of damaged DNA or activation of apoptotic pathways [35]. Therefore, cells expressing E6 maintain low levels of functional p53, altering normal response to DNA damage and favoring accumulation of genomic mutations. Binding of the E7 oncoprotein on pRB provides a complementary function. Binding releases transcription factor E2F that activates expression of genes that stimulate DNA synthesis in the cell. If earlier E6 action had freed the same cell from p53 control, that cell survives into the S phase with damaged DNA and, through E7 action, is able to replicate the HPV DNA [36]. Oncogenic properties of E6 and E7, as well as their effects on p53 and pRB, have provided the general basis for further investigations of the role of HPV in carcinogenesis in the HPV-infected cervix. Research in the action of the two oncoproteins have shown how they subvert key cell cycle and regulatory processes such as cyclins, cyclin-dependant kinases (CDKs), and cyclin-dependant kinase inhibitors (CDIs), among other interactions, to transform and immortalize the host cell [37].
DNA integration
HVP DNA is usually extrachromosomal or episomal in benign cervical precursor lesions. Cancer tissues may contain both episomal and integrated HPV DNAs at the same time, although integration appears to occur more frequently in HPV 18- than in HPV 16-associated cervical cancer [28]. During HPV DNA integration, the viral genome usually breaks in the E1/E2 region. The break generally leads to loss of the E1 and E2 regions. Loss of E2 results in uncontrolled and increased expression of E6 and E7 oncogenic proteins. Increased expression of E6 and E7, meanwhile, has been observed to lead to malignant transformation of host cells and to tumor formation [38]. HPV viral integration into host genome DNA is associated with progression from polyclonal to monoclonal status in cervical intraepithelial neoplasia (CIN), and these events play a fundamental role in the progression from low- to high-grade cervical neoplasia [39].
Epigenetics
Epigenetics can be defined as the study of genoma function that is contained outside of DNA itself and by means of which stable alterations in gene expression are set. Epigenetics is a well-established phenomenon that plays a major role in a diversity of biological processes such as embryonic development, cancer biology, and immune system response, among many others. The two most widely studied epigenetic changes are DNA methylation and histone acetylation; however, the picture is much more complicated than this, with new players coming onto the scene such as the RNA interference phenomenon, which has proven to be implicated in transcriptional silencing through small duplex RNA molecules that recruit silencing complexes to the chromatin [40,41,22].
DNA methylation
DNA methylation is a covalent chemical modification that occurs at the cytosine ring, resulting in the addition of a methyl (CH3) group at the carbon 5 position. According to the fact that DNA is made up of four bases and that therefore 16 possible dinucleotide combinations can occur, the CpG dinucleotide should have a frequency of 6%. However, the actual presence is only 5–10% of its predicted frequency. The human genome is not methylated uniformly and contains regions of unmethylated segments interspersed with methylated region. In contrast to the remainder of the genome, smaller regions of DNA, called CpG islands – and ranging from 0.5–5 kb and occurring on average every 100 kb – have distinctive properties. These regions are unmethylated, GC-rich (60–70%), have a ratio of CpG to GpC of at least 0.6, and thus do show no suppression of dinucleotide CpG frequency. Approximately one half of all genes in humans have CpG islands, and these are present in both housekeeping genes and genes with tissue-specific patterns of expression [43-46]. At least three functional DNA methyltransferases (DNMTs) have been identified; the most abundant is DNMT1, which preferentially methylates hemi-methylated DNA. DNTM1 localizes to replication foci and is responsible for maintaining proper methylation levels during replication and possibly DNA repair [47,48]. Other known functional methyltransferases are DNMT3a and DNMT3b, which are responsible for de novo methylation during embryogenesis [49]. DNMT3a and DNMT3b have equal preferences for hemi-methylated and non-methylated DNA, and thus have been classified as de novo methyltransferases[50]. In addition to DNMTs, the machinery of methylation includes demethylases, methylation centers triggering DNA methylation, and methylation protection centers [51]. The effect of DNA methylation on gene transcription can only be seen in the context of chromatin remodeling players. DNA methylation can directly interfere with transcriptional factor binding and thus inhibit replication [52], in addition to the ability of DNA methyltransferases DNMT1, DNMT3a, and DNMT3b to repress transcription in a methylation-independent manner [53]. Methyl-CpG binding proteins, which can recognize methylated DNA, have been shown to associate with large protein complexes containing HDACs and chromatin-remodelling activities, and it has also been suggested that DNA methylation could produce gene silencing by methyl binding domain proteins that recruit histone methyltransferases, which methylate lysine 9 in histone H3 and subsequently repress gene transcription [54]; as a result, histones are deacetylated and gene transcription is most often repressed.
Histones and post-translational modifications
How double-strand DNA is packaged into the dynamic structure of chromatin is crucial for the process of transcriptional control by regulating transcription factor accessibility to DNA regulatory sequences. Chromatin is constituted of nucleosomes, which are comprised of 146 base pairs of DNA wrapped around a core of two copies each of H2A, H2B, H3, and H4 histone proteins. These proteins suffer post-translational modifications that play a prominent role in gene expression regulation and signal transduction pathways such as methylation, acetylation, ubiquitination, phosphorylation, and sumoylation, which determine chromatin architecture and ultimately gene transcription [57].
The most widely studied modification is acetylation. Addition of charge-neutralizing acetyl groups to lysine residues on histones disrupts interactions with DNA, resulting in chromatin decompactation, greater access of DNA to transcription factors, and the presence of a transcriptionally active genomic locus. This post-translational modification depends on the net local balance between activities of histone acetyltransferase (HAT) and histone deacetylase (HDAC). In general, increased levels of histone acetylation (hyperacetylation) are found in more decondensed euchromatin, whereas decreased levels of acetylation (hypoacetylation) are characteristic of more condensed heterochromatin [58]. However, this mechanical model is an oversimplification of how gene transcription is regulated as additional histone modifications influence transcription [59].
Histone methylation can occur on lysine and arginine residues, giving the cell another layer of regulatory options, for example, lysine 9 in histone H3. It is currently known that histone arginine methylation is more dynamic, correlating well with gene activation and its loss from target arginines in H3 and H4 with gene inactivation. In contrast, lysine methylation appears to be a more stable mark. In this sense, methylation of lysine 4 in histone H3 correlates with gene activation, whereas methylation of lysines 9 and 27 in histone H3 correlates with repression [61-63]. Phosphorylation is another important and long-appreciated histone modification often associated with chromosome/chromatin condensation that includes mitosis, meiosis, apoptosis, and DNA damage, events regulated by different histone kinases (for example, members of the Aurora/AIK family [64-66]. Along with this post-translational modification of histone proteins, sequence-specific DNA binding by transcription factors and other protein remodeling factors determine a histone code for gene-specific transcriptional control that may dictate which modification or specific combinations of histone modifications can affect distinct downstream events by altering chromatin structure and/or generating a binding platform for protein effectors that can specifically recognize the modification and initiate gene transcription or repression [67].
Epigenetic alterations in cancer
Because of the close interplay between DNA methylation and histone modifications, it is expected that both mechanisms are operating in disease processes such as cancer; nonetheless, for the majority of tumor types the epigenetic defects could be just one of the many molecular cell alterations that lead to the malignant phenotype.
DNA methylation and cancer
Abnormalities in DNA methylation have long been associated with cancer. Both hypo- and hypermethylation play a prominent role in carcinogenesis, and their contribution shows scarcely defined boundaries. It has long been known that in cancer cells both alterations coexist: malignant tumors show global hypomethylation and regional hypermethylation. Whether one must precede the other or whether both should start at the same time remains to be elucidated. In terms of carcinogenesis, the first observations in fact were done on hypomethylation [69]; later, the discovery of regional hypermethylation as a means to silence the tumor suppressor genes expression gained the most attention [70].
Hypermethylation and gene silencing
Observations that tumor suppressor genes can be inactivated not only through structural changes (mutation, deletion) but also by lack of expression due to promoter hypermethylation positioned tumor suppressor gene epigenetic silencing as a well-established oncogenic process [71]. The first suppressor gene known to be hypermethylated and silenced was RB [72], which was followed by multiple publications describing similar findings for a variety of tumor suppressor genes, among them p16, MLH1, VHL, and E-cadherin [73].
Whether gene promoter hypermetylation is the cause or consequence for the tumor suppressor gene silencing is still a matter of controversy; nevertheless, these views are not mutually exclusive. That DNA methylation is causal has been shown by the ability of diverse pharmacologic compounds and molecular techniques to reactivate gene expression upon inhibition of DNA methylation in cancer cells [74].
On the other hand, other findings suggest that hypermethylation-induced gene silencing could be secondary to changes that determine gene expression, such as chromatin modification, so that methylation helps to maintain the silenced status of the gene. Strong support for the second view came from experiments showing that methylation of histone H3 lysine 9 – that is, chromatin modification – occurred, along with re-silencing of p16 in absence of DNA methylation in cells in which p16 had previously been activated by knocked out of DNA methyltransferase [75] and by data demonstrating p16 silencing in mammary epithelial cells that had escaped senescense and had demethylated the promoter [76].
Hypomethylation and gene activation
It is known that tumor cells have global DNA hypomethylation that can be as high as 60% less than their normal counterparts [77]. This hypomethylation occurs mainly in the body of genes (coding regions and introns), as well as in pericentromeric regions of chromosomes rich in repetitive DNA sequences [79]. Interestingly, hypomethylation is progressive from premalignant conditions to fully developed malignancies [80]. The main mechanisms put forward in attempting to explain cancer causation by hypomethylation include chromosome instability and reactivation of transposable elements and/or inappropriate gene activation [81]
There are two pieces of convincing evidence linking hypomethylation with chromosomal instability. The congenital disorder ICFs syndrome (immunodeficiency, chromosomal instability, and facial anomalies) caused by mutations at DNMT3b demonstrates loss of methylation in classical satellite DNA and mitogen-inducible formation of bizarre multiradial chromosomes that contain arms from chromosomes 1 and 16 [82]. This disorder, however, is not associated with cancer, but common somatic tumors such as breast, ovarian, and other epithelial tumors commonly have unbalanced chromosomal translocations with breakpoints in the pericentromeric DNA of chromosomes 1 and 16 [83]. In mouse models with an inactivated allele of NF1 and p53 genes, introduction of a hypomorphic DNMT1 allele caused a 2.2-fold increase in LOH frequency [84].
Finally, some reports have stressed the fact that many CpG islands are normally methylated in somatic tissues [85], and that demethylation could lead to activation of nearby genes such as HRAS. Indeed, experimental demonstration exists that hypomethylation leads to activation of genes important for cancer development, including promoter CpG demethylation and overexpression of 14-3-3sigma, maspin, heparanase, and S100A4 in several tumor types [86-88]. The question here is whether over-expression was indeed caused by hypomethylation or whether promoters are hypomethylated secondary to its high transcriptional activity. There are data showing that the sole hypomethylation as achieved by pharmacologic means is not sufficient to activate gene expression. In this context, some genes are not permisive for expression; this means that despite the fact that methylation is relieved the necessary ancillary factors to activate transcription are not present. Others are permissive and therefore reactivated by demetylation, whereas for others hypomethylation does not affect their levels of expression but can be over-expressed due to activation of signalling pathways known to activate them [89].
Chromatin and cancer
All classical genetic alterations – for instance, mutations in oncogenes or tumor suppressor genes of malignant cells – eventually affect gene transcription (mutant Ras, HER2 amplifications), or are transcription factors in themselves (c-myc, p53). It is therefore not surprising that the machinery of transcription control could be directly involved in the carcinogenesis process. Although the complex nature of the regulation of transcription is clear, certainly a disruption in the balance of activities of enzymes in charge of maintaining histone acetylation status is expected to occur in cancer. Among histone acetylases, the coding genes of p300/CBP have been found altered in some neoplasms. Mutations have been observed in epithelial tumors such as lung, esophageal, ovarian, and gastric tumors [90-93]. Chromosomal translocations involving CBP or p300 that in turn disrupt transcription by its fusion with partern genes are well-described molecular defects leading to hematologic malignancies such as some forms of acute myeloid leukemias [94,95]. Histone deacetylase activity leading to aberrantly repressed transcription was one of the first described leukemogenesis events. In acute promyelocytic leukemia, PML-RARα translocation generates a fusion gene product that represses transcription by associating with a co-repressor complex that contains HDAC activity [96]. Similar mechanisms account for other types of leukemia and lymphomas such as AML1-ETO and LAZ3/BCL6, respectively [97,98].
Despite the fact that participation of DNA methylation and chromatin in the carcinogenic process is unquestionable, it must be borne in mind that the split of epigenetics and genetics as separate types of defects in cancer is very artificial. In fact, according to the definition of epigenetics as genetic information not contained in the DNA sequence itself, current evidence demonstrates that primary genetic defects (mutations in genes with no known primarily methylating or chromatin-modifying activity such as growth factor receptors, adhesion molecules, etc,, or mutations in genes that in themselves affect DNA methylation or chromatin such as DNMTs or HAT/HDAC) are those leading to altered DNA methylation and chromatin changes. Demonstration that exogenous or endogenous carcinogens without causing primarily gene mutations lead to epigenetic abnormalities should prove that epigenetics is by itself one of the carcinogenic steps.
Epigenetic alterations in cervical cancer
Because infection with high-risk types of human papillomavirus is needed for cervical cancer development, it is important to consider the epigenetic changes occurring in the viral genome that can influence the virus-driven carcinogenic process as well as epigenetic changes in the host genome. Table 1 summarizes epigenetic alterations found in cervical cancer.
Table 1 Main epigenetic alterations in cervical cancer
Alteration Meaning
HPV-related
Methylation of HPV-E2 binding sites De-repression of E6 and E7 HPV oncoproeteins?
Methyation at HPV-E6 and E7 LCR Cause or consequence of E6/E7 over-expression?
E6 and/or E7 interaction with DNMTs? Silencing of cellular tumor suppressor genes?
Interaction between E7 with HDACs Aid for cell transformation
Interaction between E6 with HATs Aid for cell transformation
Host cell-related
Regional DNA hypermethylation Silencing of tumor suppressor genes
Global DNA hypomethylation Genomic instability?, oncogen over-expression?
Abnormal pattern of chromatin Unknown
Loss of imprinting at H19/IGF2 loci Tumor progression?
H3 hyper-phosphorylation & acetylation Associated with carcinogenesis Progression
HPV and methylation
The realization that viral infections, by insertion of viral genes into host genomes, can trigger host defense mechanisms such methylation machinery activation has aroused interest in the study of epigenetic events occurring in both virus and host genomes [99]. Human genomes harbor DNA sequences resembling retroviral long terminal repeats and the transposable elements, and indeed there are indications that under some situations inappropiate "activation" of these normally silenced sequences could play a role in the carcinogenic process [100]. In addition, it is also established that some viruses can find ways to adapt different tactics to regulate expression of their genes through modulation of DNA methylation; thus, a virus may silence activation of its genes in a manner that favors establishment of persistent infection and evades the host immune defense [101]. In addition to this, viral oncoproteins can possess the ability to modulate directly or indirectly the methylation machinery in order to silence cellular genes that could interfere with its tumor promoting actions. A very illustrative example of this is how the Epstein-Barr virus oncogene product, latent membrane protein 1, induces downregulation of E-cadherin gene expression via activation of DNA methyltransferases [102].
The role of HPV genome DNA hypermetylation has of late been the subject of study. One of the first indications of the importance of DNA methylation and viral gene expression came from studies of cell transfection with HPV-16 in-vitro methylated genomes, demonstrating that under these circumstances DNA is transcriptionally repressed [103]. In SiHa and CasKy cell lines that harbor HPV-16 and have a couple of and multiple viral genome copies, respectively, a conserved profile of CpG hyper and hypomethylation was found by using scanning with the restriction enzyme McfBC. Hypermethylation was found in genomic segments overlying late genes, while the long control region and the E6 and E7 oncogenes were unmethylated in SiHa cells. Interestingly, evaluation of smears of normal, precursor, and invasive lesion smears of 81 patients showed that as lesion severity increases, there is progressive hypomethylation on these LCR and E6 gene regions; thus, hypermethylation was found in 52% of smears from asymptomatic women, in 21.7% of preinvasive lesions, and only in 6.1% of invasive-case smears. These findings lead the authors to postulate that neoplastic transformation can be suppressed by gene hypermethylation, whereas hypometylation accompanies or causes cancer progression [104]. These findings however, were not totally coincident in another study that studied L1 and LCR regions by bisulfite modification in 115 clinical samples. First, high heterogeneity on methylation status was noted among patients and even in different samples of the same patient. As expected, methylation frequency was ca. 30% in the L1 region and lower in other positions, particularly at a CpG site located in the linker between two nucleosomes positioned over HPV-16 enhancer and promoter. However, methylation at most sites was consistently higher in carcinomas as compared with dysplasia, possibly related to the tandem repetition and chromosomal integration that occurs in invasive lesions [105]. In another study performed in two HPV-18 cervical cancer cell lines, HeLa and C4-1, and clinical samples, there was also clonal heterogeneity in the methylation status of the different regions analyzed. When it came to clinical sample analysis, there was complete or partial HPV-enhancer methylation in three of six tumors and complete demethylation in eight smears from asymptomatic patients. Likewise, promoter methylation was found in three of six cancers and in four of six smears [106]. The latter two studies suggest that methylation status of viral oncogenes in lesions perhaps is perhaps solely the result of their transcriptional activity level and not a causal event for neoplastic progression.
Further data on the influence of DNA methylation in the HPV life cycle comes from another study that focused on methylation of E2, the early gene that contributes to multiple biological processes including viral transcription and viral DNA replication. It has been shown that the capacity of E2 protein to bind E2BS in vitro is inhibited by methylation of these cytosines [107]. Kim et al. performed a methylation analysis by bisulfite modification of E2 binding site within LCR in DNA isolated from an immortalized epithelial cell line isolated from an HPV 16-infected patient and demonstrated that this region is selectively hypomethylated in the highly differentiated cell populations, whereas poorly differentiated basal-like cells were heavily methylated particularly in E2 binding sites. These observations may indicate that the methylation state of the viral genome, and particular that of E2BSs, may vary during the viral life cycle, providing a novel means for modulating E2 function as infection progresses [108]. It will be of major interest to analyze human papillomavirus oncogene expression in cervical tumors before and after treatment of patients with DNA methylation inhibitors.
Hypermethylated genes in cervical cancer
There are numerous reports demonstrating that tumor suppressor genes belonging to nearly every cancer pathway or function category have silenced or diminished their expression due to abnormal promoter hypermethylation in cervical carcinoma (Table 2).
Table 2 Tumor suppressor genes hypermethylated in invasive cervical cancer
Gene Rate Function Reference
DcR1/DcR2 100% Apoptosis 113
hTERT 57% Apoptosis 122
p73 39% Apoptosis 129
p16 8–42% Cell-cycle 130–136
PTEN 58% WNT-pathway 142
E-cadherin 28–80.5% WNT-pathway 143–145
APC 11–94% WNT-pathway 133,135,136
MGMT 5–81% DNA repair 133,134,136,144
FANCF 30% FA-BRAC pathway 161
BRAC1 6.1% FA-BRAC pathway 133
hMLH1 5% Mismatch repair 134
RASSF1A 0–45% Negative ras-effector 144,172–174
DAPK 45–100% Metastasis/cell death 133,135,136,144
TSLC1 58–65% Tumor suppressor 179,180
FHIT 11–88% DNA repair?/cell death? 133,134,135,136,144
HIC1 18–45% Transcription factor 133,135
RARβ 33–66% Cell differentiation 133,136,200,201
TIMP2/TIMP3 47%/1–10% Tissue inhibitor MTs 144,202,203
Caveolin-1 6% Caveolae membrane 205
ER α 25% Steroid hormone receptor 136
Apoptosis-related genes
It is now established that failure of cells to undergo apoptosis is crucial for cancer development and progression, but most importantly this phenomenon participates in intrinsic or acquired resistance of cancer cells to chemotherapy and radiation. Identification of points in the apoptotic pathway at which dysregulation occurs would potentially open up new therapeutic opportunities in situations where conventional cancer treatments fail. One of the first indications of the role of methylation for inactivation of key apoptotic genes came from the study showing that Apaf-1 was silenced in melanoma instead of being lost or mutated [109].
Studies analyzing apoptosis-related genes that can be inactivated by methylation in cervical cancer are limited. One study has shown that decoy receptors DcR1 and DcR2 can be the target for abnormal methylation that leads to their silencing [110]. These molecules are members of the tumor necrosis factor receptor superfamily, which includes TNFR1, Fas, and the decoy receptors for TRAIL. Upon engagement by their respective ligands, TNFR1 and FAS recruit adaptor molecules and activate a cascade of caspases. Death-inducing decoy receptors DR4 and DR5 and DcR1 and DcR2 are structurally related; nonetheless, DcR1 completely lacks the intra-cellular death domain and DcR2 contains a truncated, nonfunctional death domain and appears unable to induce apoptosis. Hence, DcR1 or DcR2 have been postulated to function as oncogenes because of their postulated anti-apoptotic effects [111,112]. In cervical carcinoma, a study has found that all 50 cases analyzed had methylation of either DcR1 and/or DcR2 [113], suggesting that cervical cancer cells, by downregulating decoy receptor expression, obtain a growth advantage.
Telomerase activation is a critical element in cellular immortalization and cancer. The end of the chromosome, the telomere, plays a critical role in chromosome structure and function. A certain length of the telomere is important for cell division, and the telomere may serve as a "mitotic clock" for cell proliferation. Normal human somatic cells express low or undetectable telomerase activity, whereas in immortal eukaryotic cells as well as in cancer cells telomerase activity increase is apparently necessary to ensure proliferation. Telomerase is a ribonucleoprotein comprising an RNA template, the telomerase-associated protein, and the catalytic subunit (hTERT) [114,115]. Telomerase activity has been demonstrated in various types of gynecologic cancers [116]. Data on hTERT expression in cervical cancer has revealed that 0–33% of normal cervices exhibited hTERT mRNA expression, whereas 80–100% of cervical cancers showed hTERT expression [117-120] The fact that the hTERT gene promoter has a CpG island and high overall GC content suggests a possible role for methylation in regulation of hTERT gene expression; however, the relationship between gene promoter and expression is unclear for this gene. Despite it is expected that hypermethylation decreases gene expression, a study has found a correlation between reduced expression and catalytic subunit activity with demethylation [121]. This may explain what was found with regard to better prognosis of patients with cervical cancer whose tumors lack hTERT methylation [122].
The p53 pathway responds to stresses that can disrupt the fidelity of DNA replication and cell division, resulting in activation of the p53 protein as a transcription factor that initiates either growth arrest or apoptosis. This apoptosis pathway is disrupted in the majority of human cancers by downregulation or loss of p14ARF, upregulation of MDM2, or mutation of p53 [123,124]. However, this pathway, by virtue of its multiple positive and negative feedback loops, can be the target of aberrant methylation in some of their components. p73 is a member of the p53 family that encodes two different proteins expressed under the control of two independent promoters and that have opposite activities: the transcriptionally active full-length TAp73 and the NH2-terminally truncated dominant-negative Np73 [125]. TAp73 has been reported as involved in cellular response to DNA damage induced by radiation and chemotherapeutic agents and when it is overexpressed in cells, it activates transcription of p53-responsive genes such as p21, Bax, Mdm2, and GADD45 and inhibits cell growth in a p53-like manner by inducing apoptosis [126,127]. It has been reported that p73 transcription can be regulated by the promoter and exon 1, which is rich in CpG dinucleotides [128], and its transcriptional silencing through methylation is a common event in some leukemias, lymphomas, and brain tumors, as well as in ovarian cell lines but not in breast, renal, and colon cancers. A recent study found that epigenetic modification of p73 via CpG-island hypermethylation represents a critical alternative mechanism for inactivation of this gene in cervical cancer and high incidence of p73 hypermethylation (38.8%) in cervical cancer but not in controls; in addition, its methylation was correlated with loss of its p73 expression. Importantly, radioresistant cancer samples had significantly higher methylation rate than radiosensitive cancer samples, and in vitro demethylation successfully restored p73 expression in cervical cancer cell lines previously found to have methylated p73 and lack of p73 mRNA and protein expression [129].
Cell cycle-related genes
It is well-established that cancer cells evolve in part by overriding normal cell-cycle regulation. Normal cell cycle progression relies on the cell's ability to translate extracellular signals such as those produced by growth factor receptor stimulation and extracellular matrices to efficiently replicate DNA and divide. Proper cell- cycle regulation is essential for cells and requires a number of players, among them cyclin-dependent kinases and their binding partners along with natural inhibitory molecules such as p16, Rb, and p15 that play an essential role. Within this class, the p16 gene has been one of the most studied in cervical cancer. Aberrant methylation of the p16 gene occurs early within tumor cell populations in both in situ and invasive tumors at frequencies that vary from 10 up to 100% [130-136]. As the cell cycle is primarily deregulated by the HPV, the molecular contribution of p16 inactivation is unclear; however, the fact that this not only is a very early event in cervical carcinogenesis but is more frequently methylated in advanced tumors [132] suggests that its reactivation could have therapeutic value. Despite the fact that Rb and p15 are known to be inactivated by methylation in other tumors, no reports exist on cervical tumors.
WNT pathway
The Wnt signaling pathway, named for its most upstream ligands, the Wnts, is involved in various differentiation events during embryonic development and leads to tumor formation when aberrantly activated. Within this pathway, there are a number of participating molecules, and the pathway is regulated by a multiprotein complex consisting of, among others, members of β-catenin, the key component, adenomatous polyposis coli APC, Axin, and GSK-3β. [137]. In the absence of Wnt stimulation, β-catenin accumulates in cytosol to then be translocated to nucleus, leading to transcription of target genes. This pathway is also involved in calcium-dependent cell adhesion by virtue of the interaction between β-catenin and cadherin [138]. There are mutations in APC, another key regulator of the pathway that promotes β-catenin proteolysis and reduces its transcriptional activity. PTEN is a lipid and protein phosphatase that is a negative regulator of phosphatidylinositol 3 (PI-3) kinase-dependent signaling and influences the WNT pathway by hindering activation of integrin-linked kinase (ILK), which inhibits GSK-3 β and thereby causes accumulation of β-catenin [139]. The WNT signaling pathway is the most frequently altered pathwayin the majority of cancers; for instance, it has been demonstrated that nearly all colorectal cancers have at least one activating mutation in this pathway [140]. As such, individual components of the pathway can be targeted by epigenetic inactivation. A recent study analyzing 310 colorectal carcinomas for eight members of the signaling cascade, including APC, β-catenin, AXIN2, TCF4, WISP3, E-cadherin, and PTEN. Hypermethylation on E-cadherin and APC were found at frequencies between 36 and 42% [141].
Studies on cervical cancer have uncovered that hypermethylation of these genes is not uncommon. A study in 62 cases of squamous cell carcinomas showed that while PTEN mutations were absent, promoter methylation was found in 58% of cases. Interestingly, patients with persistent disease or patients who died of the disease had a significantly higher percentage of PTEN methylation than those without evidence of recurrence. Multivariate Cox regression models confirmed PTEN was an important significant predictor for both total and disease-free survival after controlling age, pathologic grade, and clinical stage [142]. Inactivation of the cadherin-mediated cell adhesion system caused by aberrant methylation is a common finding in human cancers. Methylation frequency of E-cadherin in cervical cancer varies from between 28 and 80.5% [143-145] and appears to have prognostic significance, cases with no promoter methylation having a better outcome in univariate and multivariate analyses [146]. Mutations are the main mechanism of inactivation for APC, particularly for colon and other tumors from the gastrointestinal tract. However, APC promoter hypermethylation occurs in other cancers. Frequencies of methylation in 208 primary human tumors of multiple origins were as follows: stomach (13 of 38, 34%); pancreas (6 of 18, 33%); liver (6 of 18, 33%), and esophagus (4 of 27, 15%; it was less common in tumors of bladder (2 of 19, 10%), kidney (1 of 12, 8%), or breast (1 of 19, 5%), or was not observed at all in brain (0 of 10), lung (0 of 17), head and neck (0 of 10), or ovary (0 of 20) [147]. In endometrial cancer, hypermethylation occurs at an increased frequency, particularly in MSI+ endometrial tumors [148] as well as in cervical cancer, with rates varying from 11 to 94% [133,135,136].
DNA repair
Alkyating agents induce O6-alkylguanines that can lead to mutations and to cell death unless repaired. The major pathway of repair involves transfer of the alkyl group from DNA to a cysteine acceptor site in the protein O6-alkylguanine-DNA alkyltransferase. Alkyltransferase brings about this transfer without The need for cofactors and DNA is restored completely by the action of a single protein, but the cysteine acceptor site is not regenerated and the number of O6-alkylguanines that can be repaired is equal to the number of active alkyltransferase molecules. A significant fraction of human tumor cell lines do not express the alkyltransferase gene; thus, they are much more sensitive to mutagenesis and killing by alkylating agents [149]. The MGMT gene product removes mutagenic and cytotoxic adducts from O(6)-guanine in DNA, the preferred point of attack of many carcinogens (i.e., methylnitrosourea) and alkylating chemotherapeutic agents (i.e., BCNU, temozolamide, etc.). As a consequence, its lack of expression produces opposite effects for cancer development and progression: First, tumors acquire a mutator phenotype characterized by generation of transition point mutations in key genes such as p53 and K-ras, but at the same time lack of enzymatic activity renders tumors more sensitive to the killing effects of alkylating drugs [150]. While these observations bear clinical and practical implications as predictive or prognostic markers for response in CNS tumors [151], its silencing by hypermethylation can be associated with higher stages, worse survival, or mutations in secondary genes that adversely affect the prognosis of patients with tumors such as gastric, colorectal, head, and neck carcinomas, [152-154] and even in low-grade astrocytomas [155]. There is scarce information concerning the role of MGMT gene in cervical cancer; a number of studies have analyzed the frequency of MGMT promoter hypermethylation, which varies from 5–81% [133,134,136,144]. Interestingly, the five cases with MGMT or BRCA1 methylation did not respond to chemoradiation [133].
FA-BRCA pathway
Fanconi anemia (FA) is an autosomal recessive chromosomal instability syndrome characterized by hypersensitivity to DNA cross-linking agents and predisposition to cancer, especially leukemia [156]. FA patients are also prone to various solid malignancies, including squamous cell carcinoma. FA is a genetically heterogeneous disease with genes for seven FA complementation (FANC) groups identified [157]. FANC genes are essential in DNA repair pathways in normal cellular response to cisplatin and other DNA cross-linking agents. FANC proteins interact with BRCA genes in a pathway that involves a number of other genes [158]. Recently, it has been shown that promoter hypermethylation of FANCF gene disrupts the FA-BRCA pathway, resulting in cisplatin resistance in ovarian cancer [159]. FANCF promoter hypermethylation has also been identified in squamous cell carcinomas of lung and oral cavity [160]. In cervical cancer, a study has shown methylation of BRCA1 in 6.1% [133] of invasive tumors, whereas FANCF hypermethylation rate was 30% [161]. Interestingly, hypermetylation of these genes was mutually exlusive in the analyzed cases [161], suggesting the important role of disruption of this pathway for cancer. This abnormality seems to be a late event in cervical carcinogenesis, as no hypermetylation was observed in any case of preinvasive disease [161].
Mismatch repair
Cells with dysfunction of mismatch repair genes hMLH1 and hMSH2, as well as hMSH3, hMSH6, and hPMS2, show mutation rates up to 1,000-fold greater than those observed in normal cells [162]. The mutator phenotype, which can be measured by microsatellite instability analysis, has been detected in tumors of patients with hereditary nonpolyposis colorectal sporadic and other types of cancers [163]. Mutations and loss of expression due to gene promoter hypermetylation are the main mechanisms of inactivation of members of this gene family [164]. Hypermethylation and loss of hMLH1 protein expression has been associated with chemotherapy resistance in ovarian and other tumors [165]. The relevancy of this phenomenon has been recently demonstrated by acquisition of hypermethylation of the gene in relapsed ovarian cancer after being treated with chemotherapy, which predicts poor overall survival [166]. Existing data on cervical cancer with regard to hMHL1 expression status and methylation is limited. While some studies have found protein expression loss in invasive lesions [167], others have found the opposite [168], while presence of microsatellite instability appears to correlate with a worse prognosis [169] but not with response to cisplatin in a neoadjuvant setting [170]. Regarding gene promoter methylation, its presence is rare in cervical cancers [134].
Miscellaneous
RASSF1A
The Ras Association Domain family 1 (RASSF1A) gene consists of two main variants (RASSF1A and RASSF1C), which are transcribed from distinct CpG-island promoters. Aberrant methylation of the RASSF1A promoter region is one of the most frequent epigenetic inactivation events detected in human cancer and leads to silencing of RASSF1A. Hypermethylation of RASSF1A has been observed in a variety of primary tumors including cervix. The product of this gene is involved in several important functions including apoptotic signaling, microtubule stabilization, and mitotic progression and may act as a negative Ras effector inhibiting cell growth and inducing cell death. Its loss of expression in several tumor types is related with worse prognosis [171]. Studies in patients with cervical cancer have demonstrated its silencing by methylation in up to 30% of tumors [172-174,144].
DAPK
DAP-kinase (DAPk) is a Ca(2+)/calmodulin (CaM)-regulated Ser/Thr kinase that functions as a positive mediator of programmed cell death. It associates with actin microfilament and has a unique multidomain structure [175,176]. In cervical cancer, it is methylated in up to 100% of cases [133,135,136,144], which suggests that its loss of expression is needed for cervical cancer progression as seen in an experimental model, in which loss of DAP-Kinase expression aids in metastatic potential of lung cancer cells [177].
TSLC1
The recently identified IGSF4/TSLC1 gene codes for an immunoglobulin Ig-like intercellular adhesion molecule that mediates calcium-independent homophilic or heterophilic interactions independently of Ca2. This gene was first identified as a tumor suppressor gene in lung cancer and silencing can derive from loss of heterozygosity or promoter hypermethylation [178]. This gene also has tumor suppressor effects on cervical cancer, as demonstrated by transfection studies in which IGSF4 cDNA was introduced into SiHa cells. Transfectants displayed a marked reduction in anchorage-independent growth and when injected into nude mice, these were less able to generate tumors. Progression of the cervical lesion is accompanied by loss of expression of IGSF4. In two studies, it was found that normal epithelium and CIN-1 lesions are free of methylation at IGSF4, whereas methylation rate in CIN-III is 35%, which increases to 58 and 65% in invasive tumors [179,180]. These data demonstrate that IGSF4 may have a role in cervical cancer development.
FHIT
The FHIT gene is a tumor suppressor gene located on chromosome 3p14.2 and LOH on the short arm of chromosome 3 and has been detected in up to 75% of cervical carcinomas [181,182]. Many studies have reported altered FHIT expression in a variety of carcinomas including head and neck, lung, kidney, gastrointestinal, and breast cancer and in 68% of cervical carcinoma cell lines [183]. Additional studies have confirmed that the FHIT gene is abnormally expressed in 30–78% of cervical dysplasia, carcinoma cell lines, and primary tumors [184-188,133,134,136,144]. The antitumorigenic mechanism of FHIT is not yet clear; however, for cervical and lung cancer cell lines its reintroduction by use of an adenoviral vector induces strong suppressive effects and reduces tumorigenicity due to an apoptotic effect associated with caspase-8 cleavage and activation of the effector caspase-3 [189]. Hypermethylation of this gene has been associated with loss of expression and advanced stages of cervical carcinoma, suggesting its participation in disease progression [190].
HIC1
HIC1 (hypermethylated in cancer) is a tumor suppressor gene unique in the sense that it has never been found mutated but is found silenced by hypermethylation. This gene encodes a zinc-finger transcription factor that belongs to a group of proteins known as the POZ family. HIC1 is hypermethylated and transcriptionally silent in several types of human cancer. Homozygous disruption of Hic1 impairs development and results in embryonic and perinatal lethality in mice, while heterozygous mice develop many different spontaneous malignant tumors including a predominance of epithelial cancers in males and lymphomas and sarcomas in females. Complete loss of Hic1 function in heterozygous mice appears to involve dense methylation of the remaining wild-type allele promoter [191]. It has recently been shown that its loss of function accentuates the tumorigenic effect of loss of p53 [192]. It has been found that the HIC1 gene is down regulated in many cervical cancer cell lines and re-expressed upon treatment with a demethylating drug [133]. In primary cervical tumors, its methylation rate varies between 18 and 45% [133,135]. These results support the tumor suppressor role of HIC1 and its inactivation by promoter methylation in cervical cancer.
RARβ2
Retinoic acid is essential for regulation of epithelial cell differentiation. The intracellular effects of RA are mediated by RA-binding nuclear receptors, including RA receptors alpha, beta, and gamma. Ligand-activated receptors induce transcription of target genes by binding to RA-responsive elements in promoter regions. One target gene is the RAR beta gene, which encodes a potential tumor suppressor. Complete or partial inhibition of gene expression has been observed in many tumor cell lines and in primary human tumors [193]. A well-known mechanism of RAR-β2 inhibition demonstrated in breast and colon carcinomas is hypermethylation of its promoter [194]. In cervical cancer, the RARβ2 gene is of particular interest because retinoic acid inhibits transformation of human keratinocytes by HPV-16 [196] and leads to regression of moderate cervical dysplasia [197]. In addition, RA in combination with interferon alfa is a highly effective antitumor therapy for patients with cervical cancer [198]. However, the existence of intrinsic or acquired resistance to retinoic acid is well-established [199]. Rate of RAR-β2 methylation progressively increases from 11% in low-grade to 29% in high-grade lesions and from 33–63% in invasive cancers [133,136,200,201], suggesting that this abnormality is an early event in multistage cervical carcinogenesis.
Tissue inhibitors of metalloproteinases
TIMP-2 and TIMP-3 are endogenous inhibitors of matrix metalloproteinases that possess growth promoting effect in cell culture, and anti-tumoral, anti-apoptotic, and anti-angiogenic effects in animal model systems in vivo. It has been shown that these endogenous inhibitors are downregulated by methylation in cervical carcinomas in a variable proportion of patients [144,202,203] and may contribute to progression of cervical cancer.
Caveloin
Caveolin-1 is commonly downreglulated in cervical cancer cells and its introduction via gene transfer restores caveolin-1 protein expression and impairs growth in SiHa cells, which supports its role as a tumor suppressor gene [204]. Although it is known that HPV-E6 oncoprotein reduces caveolin expression, in a small percentage of cervical cancer tumors silencing occurs via promoter methylation [205].
Global hypomethylation
As stated previously, tumor cells may have up to 60% less global DNA hypomethylation than their normal counterparts [77,78] and interestingly, DNA hypomethylation is progressive from premalignant conditions to fully developed malignancies [80]. The main mechanisms set forward in attempting to explain cancer causation by hypomethylation are chromosome instability and reactivation of transposable elements and/or inappropriate gene activation [81]. Analyses of global DNA methylation have been performed in samples covering the full spectrum of cervical lesions. Kim et al., analyzed 41 samples from colposcopically identifiable lesions for methylation by incorporation of [3H]-S-adenosylmethionine. As expected, the extent of 3H-methyl group incorporation was increased three- and seven-fold in DNA from cervical dysplasia and cancer as compared with DNA of normal cervices and within dysplasias as long as they progress from normal to low- and high-grade [206]. Subsequently, this group performed a similar analysis in 83 cases and obtained essentially the same results with regard to DNA hypomethylation, highlighting the fact that lower methylated DNA correlated wtih serum folate levels [207]. Using a computer-assisted assay based on quantitative analysis of DNA methylation in individual interphase nuclei by immunolabelling with anti-5-methylcytosine antibodies, progressive hypomethylation was observed in dysplastic and cancer cells as compared to normal controls [208]. These data, along with observations of gene promoter hypermethylation of a number of genes during pre-invasive to invasive stages of cervical cancer, demonstrate that both phenomena are coincident during carcinogenesis of the cervix uteri.
Histone alterations in cervical cancer
At the chromatin level, there are some indications that the pattern of chromatin distribution in smears may aid in diagnosis of cervical neoplasia, particularly for glandular lesions [209], but the molecular bases for these chromatin alterations have yet to be determined. Contrariwise, histone changes at a global level in cancer and normal cells have only recently been studied [210]. A recent report showed that during the tumorigenic process, cancer cells had a loss of monoacetylated and trimethylated forms of histone H4, predominantly at acetylated Lys16 and trimethylated Lys20 residues of histone H4, which were associated with hypomethylation of DNA repetitive sequences, a hallmark of cancer cells [211]. In this line, tumor cell acetylation level and methylation of histones in prostate carcinoma cells identified two disease subtypes with distinct biological behaviors in patients with prostatic carcinoma [212]. In cervical cancer, it has been reported that phosphorylated and acetylated forms of histone H3 in cytologic smears shows a marked association of histone H3 modifications with progression of the disease from CIN I to CIN II and CIN III [213].
The balance between histone deacetylases and histone acetyl transferase activities is a major player in regulation of gene transcription [214]; hence, this balance must be mantained in normal cells, or otherwise unchecked cell proliferation and cell death would occur. E6 and E7 oncoproteins of HPV target numerous cellular proteins to disturb cell growth and proliferation [30] including HDACs and HATs. The HPV-E7 protein from high-risk types [215] binds to HDACs, and this interaction occurs through an intermediary protein called Mi2β, a member of the nucleosome remodeling and histone deacetylation (NURD) complex that possesses the ability to modify chromatin structure through both deacetylation of histones and ATP-dependent nucleosome repositioning. Binding of HDACs to E7 is independent of binding to Rb, and mutations on E7 abolishes its binding to HDAC1 and results in a loss of the ability of E7 to efficiently transform rodent fibroblasts [216]. E6 protein of HPV high-risk types also shares, with other small DNA tumor viruses, the capacity for targeting CBP/p300 in an interaction involving the C-terminal zinc finger of E6 and CBP residues 1808–1826 to downregulate p53 transcriptional activity independently of removal of cellular p53 protein through the proteosome degradation pathway [217]. As for E7, binding of E6 to the transcriptional co-activator p300/CBP is essential for cell transformation [218].
Loss of imprinting
Imprinting refers to the condition established during gametogenesis, dictating that a specific parental allele is or is not expressed in the offspring. Loss of imprinting is implicated not only in developmental disorders such as Prader-Willi syndrome (PWS) and Angelman syndrome (AS), but also in cancer [219]. In Wilms tumors, loss of imprinting leads to biallelic expression of IGF2 and reciprocal loss of expression of H19, a non-transcribed RNA with tumor suppressor activities, while IGF2 promotes tumor formation by inhibiting apoptosis [220,221]. In cervical cancer, a sole study has found a high frequency of both loss of heterozygosity and loss of imprinting for H19 and IGF2 genes, suggesting that they participate in the molecular pathogenesis of cervical cancer [222].
Translational opportunities of epigentic alterations in cervical cancer
Identification of numerous epigenetic alterations of cervical cancer in all stages of the disease process offers new possibilities of diagnosis and treatment of cervical cancer (Table 3).
Table 3 Translational opportunities from epigenetic alterations in cervical cancer
Early detection Identification of a set of hypermethylated genes from cytological smears
Prognostic/predictive Determination of global methylation or histone modifications in tumor or peripheral blood cells Determination of hypermethylated gene promoters from serum/plasma DNA
Therapy DNA methylation inhibitors and/or HDACs inhibitors alone or as chemo- or radio-sensitizers
Early detection
Cervical cancer remains a model disease for screening due to its long natural history and ease of sampling and reading cytologic abnormalities; dramatic reductions in mortality from this neoplasia achieved in countries with well-organized detection programs validate this fact [223]. Nevertheless, the test as such has low sensitivity, though high specificity for detecting CIN-3 [224]. More recently, testing for high-risk types of HPV infection are in use to aid in the triage of women with atypical squamous cells of undertermined significance; nonetheless, the large number of women requiring additional testing and the inability of cytology or HPV testing to identify women at higher risk for disease progression impose greater efforts on health systems. Thus, novel screening alternatives are needed. The realization that genetic and epigenetic alterations are present at the earliest steps of the malignant progression of cervix uteri has led to testing the presence of these abnormalities, such as p16 expression [225,226]. A large number of studies looking at the methylation status of tumor suppressor genes have uncovered that some genes are found hypermethylated in preinvasive lesions, raising the possibility that testing for methylation of either of these or a set of these may prove to be a useful screening tool [133-136]. However, there is limited information with respect to the sensitivity and specificity of methylated genes for identification of women with cervical dysplasia and cancer as well as comparisons of results using different sources of samples, either exfoliated cells or paraffin-embedded fresh biopsy samples. In this regard, a very comprehensive study investigated the methylation profile of 20 genes (p16, p15, CCND2, RASSF1, RARb, TWIST1, SYK, HIC1, VHL, PRDM2, SFN, MLH1, MGMT, APC, CDH1, and CDH13) in exfoliates and biopsies of 319 women that participated in a cytology screening study. By logistic regression, the authors determined the best set of candidate genes for employment as a disease markers. First, they found similar detection rate of methylation regardless of sample source, and second, they found that CDH13, DAPK1, RARb, and TWIST1 were the genes showing statistically signficant increase with lesion severity, and DAPK1, RARb, and TWIST1, the best panel of hypermethylated genes. At least one of the three genes was hypermethylated in 57% of samples with CIN-3/CIS and in 74% with invasive cancer, but in only 5% of samples with CIN-1. Estimated specificity of the panel was 95% with sensitivity of 74% (95% confidence interval [CI 95% ], 73–75%) for invasive carcinoma and 52% (95%,49–55%) for CIN-3/CIS. These findings provide preliminary evidence on the potential usefulness of a panel of genes to be tested for hypermethylation in cytology samples; however, additional studies are needed before this epigenetic-based screening test could be adopted [227].
Methylated genes in serum of patients with cervical cancer as biomarkers
Circulating nucleic acids represent a biomarker that might be used in early detection of cancer, in the follow-up of patients with cancer or as a prognostic factor. Presence of nucleic acids in plasma or serum of patients with cancer has been recognized since the 1970s, but it was not until 1989 that the neoplastic characteristics of plasma DNA in patients with cancer were recognized [228], and this was followed by detection by PCR-based techniques of specific gene mutations present in the primary tumor, such as Ras family members in tumors such as pancreas and myelodysplastic syndromes [229,230]. Within this field of study, cell-free circulating RNA and nucleosomal DNA have also been studied in patients with cancer [231,232]. The feasibility of applying a PCR-based technique such as the Methylaton-Specific PCR (MSP) [233] for amplification of gene promoters from DNA circulating in serum or plasma paved the way for the study of epigenetic alterations not only in serum plasma of patients with cancer, but also in other biologic fluids, such as urine and sputum [234,235].
The methylation status of several genes present in the serum or plasma of patients with cervical cancer has been studied with regard to their prognostic signficance. In a study analyzing serum samples of 93 patients with the methyLight technique for gene promoter methylation of CALCA, hTERT, MYOD1, PR, and TIMP3 genes, methylation rates were 62, 0, 25, 79, and 4% respectively, and in all but one case corresponding primary tumors also had these genes methylated. When authors looked at survival, methylation of MYOD1 was strongly associated with shorter, disease-free and overall survival with median survival of 1.9 and 6.1 years for methylated and unmethylated cases, respectively [203]. The same group of researchers reported on the prognostic significance of CDH1 (E-cadherin) and CDH13 (H-cadherin) using the same technique and the same number of patients. The main finding was that aberrant methylation of either CDH1 or CDH13 was found in 40 (43%) patients, and median survival for these patients with methylated genes was 1.2 vs. 4.3 years in CDH1 or CDH13 methylation-free patients [143]. Data suggesting that methylation of gene promoters in patients with cervical cancer is a common phenomenon were reported by another group that analyzed DAPK, p16, and MGMT genes; these authors also found a strong correspondence between methylation in serum and in primary tumors with methylation frequencies in serum of 40, 10, and 7.5% of these genes, respectively [236]. Together, these data encourage further studies to find a set of methylated genes that would have prognostic significance but that would also serve as surrogate markers of efficacy of epigenetic therapies.
Treatment
Demethylating agents
Reactivation of tumor suppressor genes that have been silenced by a epigenetic mechanism such as gene promoter methylation is a very attractive molecular target for cance therapy. Inhibitors of DNA methylation have demonstrated the ability to inhibit hypermethylation, restore suppressor gene expression, and exert antitumor effects in in vitro and in vivo laboratory models. There are several demethylating agents currently being evaluated in preclinical and clinical studies (Table 4). The classical demethylating agents comprise the analogs of deoxycytidine:5-azacytidine, 5-aza-2-deoxycytidine, 1-β-D-arabinofuranosil-5-azacytosine, and dihydro-5-azacytidine. 5-azacytidine and its analog are the most studied and were developed over 30 years ago as classical cytotoxic agents, but were subsequently discovered to be effective DNA methylation inhibitors [237]; these were tested as such in several phase II studies against solid tumors demonstrating very modest activity [238]. To the contrary, their antileukemic activity was very promising and both are being revived as a consequence of their demonstrated inhibitory activity upon DNA methylation and gene-reactivating function. Currently, 5-azacytidine is Federal Drug Administration (FDA)-approved for use against myelodysplastic syndrome, and the hydrosoluble analog 5-aza-2-deoxycytidine is being tested in a variety of solid tumors as DNA demethylating agent [239]. Despite the poor activity against solid tumors of these nucleoside analogs, it is remarkable that the proof of the concept iwith regard to the ability of demethylating compounds to demethylate and reactivate tumor suppresor gene expression has been demonstrated in solid tumors [136,240]. Whether or not the reactivating effect can translate into a clinical response on its own or in combination with classical cytotoxic therapies remains to be demonstrated.
Table 4 Epigenetic therapy agents
DNA methylation inhibitors
deoxycytidine analogs:
5-azacytidine, 5-aza-2-deoxycytidine, 1-β-D-arabinofuranosil 5-azacytosine, dihydro- 5-azacytidine
nucleic acid-based:
MG98 antisense oligonucleotide
cytidine deaminase analogs:
zebularine
non-nucleoside analogs:
(-)-epigallocatechin-3-gallate, procaine, procainamide, hydralazine
Histone deacetylase inhibitors
Small molecular weight carboxilates:
sodium butyrate, valproic acid, sodium phenylbutyrate and pivaloyloxymethyl butyrate
Hydroxamic acids:
SAHA, trichostatin A, SBHA
Benzamides:
CI-994, MS-275
Epoxyketones:
trapoxin B, 2-amino-8-oxo-9,10-epoxydecanoic acid
Cyclic peptides:
apicidin, depsipeptide
Hybrid molecules:
CHAP 31, CHAP50
As a second category of demethylating agents, we note the antisense oligonucleotide MG98 against the 3' untranslated region of DNMT1 mRNA, which codes for the enzyme DNA methyltransferase 1 that is responsible for maintenance of DNA methylation [241]. This agent has shown an ability to inhibit DNMT1 expression without affecting other DNMTs, and to cause demethylation with re-expression of p16 in bladder and colon cancer cell lines as well as to produce tumor growth inhibition in nude mice bearing human lung and colon xenografts [242]. MG98 has been evaluated in a phase I trial in patients with advanced or refractory solid tumors and has demostrated its tolerability despite the fact that dose-limiting toxicities of transaminitis, thrombocytopenia, and fatigue prohibited higher doses. Its molecular efficacy was demonstrated by producing a decrease in DNMT mRNA levels in six of 10 patients [243].
The fact that deoxycytidine analogs such as current cytotoxic agents are not only carcinogenic but also exhibit neutropenia as their dose-limiting toxicity even when used at doses required for demethylation [244] has renewed interest in finding effective and less toxic demethylating agents. Zebularine is a new oral cytidine analog originally synthesized as a cytidine deaminase inhibitor that has been shown to cause demethylation and reactivation of a silenced and hypermethylated p16 gene in human bladder tumor cells grown in nude mice. Zebularine was also demonstrated as minimally cytotoxic in vitro and in vivo and can be given continuously at a lower dose to maintain demethylation for a prolonged period, only possible because of its low-toxicity profile; to date, there are no results of clinical trials with this agent [245]. Within this class of so-called "non-toxic and orally administered agents" there is the green tea major polyphenol (-)-epigallocatechin-3-gallate (EGCG), which resulted as an effective inhibitor of DNMT activity at micromolar concentrations and that was able to demethylate and reactivate expression of several tumor suppressor genes such as p16, RAR-β2, and MGMT in cancer cell lines [246]. There is another class of so-called "old drugs" whose demethylating activity upon gene promoters of tumor suppressor genes was recently highlighted. Procainamide, a nonnucleoside inhibitor of DNA methyltransferases approved for treatment of cardiac arrhythmias, can demethylate the GSTP1 promoter, a common somatic genome change in human prostate cancer and reactivates in vitro and in nude mice the expression of the gene [247]. A related drug, procaine, has also the ability of demethylating and reactivating tumor suppressor gene expression, such as the RARβ2 gene in a breast cancer cell line effect that is accompanied by growth-inhibitory actions [248]. Our group has recently shown in vitro and in vivo promoter demethylation and tumor suppressor gene transcriptional reactivation mediated by the antihypertensive compound hydralazine, a well-tolerated drug devoid of the common side effects of cytotoxic chemotherapy agents [249]. Its DNA demethylating activity can be explained by the interaction between its nitrogen atoms with residues Lys162 and Arg240 of the DNA methyltransferase active site, as shown in a silico modeling [250].
Histone deacetylase inhibitors
HDACs are seen as a potential target for cancer treatment. HDAC inhibition has been reported to induce tumor cell differentiation, apoptosis, or growth arrest, depending on the experimental system [251,252], and to sensitizer cells to chemotherapy [253] or radiation therapy [254] However, the HDAC-dependent mechanisms accounting for the observed and rather selective modulation of gene expression, as well as specific patterns of antitumor activity, remain poorly understood.
Currently, there are six structurally distinct classes of HDAC inhibitors at diverse stages of preclinical and clinical development (Table 4); these act by binding to various portions of catalytic domains within class I and II HDACs: 1) Small molecular weight carboxilates (sodium butyrate, valproic acid, sodium phenylbutyrate and pivaloyloxymethyl butyrate); 2) hydroxamic acids (suberoylanilide hydroxamic acid -SAHA-, trichostatin A, suberic bishydroxamic acid -SBHA-, and others); 3) benzamides (CI-994 and MS-275); 4) epoxyketones (trapoxin B, HC-toxin, or 2-amino-8-oxo-9,10-epoxydecanoic acid); 5) cyclic peptides (apicidin, depsipeptide), and 6) hybrid molecules CHAP 31 and CHAP50) [255].
In general, HDAC inhibitors are at most at the early stages of clinical development. Among them, SAHA has shown in a recently reported phase I study of 73 patients with advanced cancer a complete response in a patient with transformed diffuse large B-cell lymphoma for 17 months, three partial responses in B-cell lymphoma, laryngeal cancer, and papillary thyroid cancer, and prolonged stabilization in patients with renal carcinoma [256]. In another study of MS-275 in 31 heavily pre-treated patients despite the fact that no objective responses were observed, 15 cases achieved disease stabilization for 62–309 days [257]. In another phase I study with CI-994, one of 42 patients with a pre-treated lung adenocarcinoma showed a partial response over 2 years, whereas three additional patients with small-cell lung cancer, colorectal, and renal carcinoma had stable disease [258]. In a phase II study of pivaloyloxymethyl butyrate in pre-treated non small-cell lung cancer, three (6.4%) of 47 patients achieved partial responses and 14 (30%) patients achieved stable disease [259].
The strong interplay between DNA hypermethylation and histone deacetylation for silencing and modulating the expression of a number of cancer-related genes predicts not only a synergy in gene expression at global [260] and individual gene levels [261] but also in antitumor activity. For instance, combinations of decitabine with trichostatin A or depsipeptide synergistically reactivate silenced tumor suppressor genes including MLH1, TIMP3, CDKN2B, CDKN2A, ARHI, gelsolin, and maspin [262-264] and increase the level of tumor cell apoptosis [265]. Thus, a logical step forward is to combine a demethylating with a histone deacetylase inhibitor for cancer treatment.
Epigenetic therapy in cervical cancer
Despite ample experimental evidence supporting the development of drugs that target the epigenoma via inhibition of DNA methylation or histone modification as cancer therapy, clinical results are pending to date. At present, the majority of the information has been generated in hematologic neoplasms and advanced or refractory solid tumors. The main findings of this form of therapy in cervical cancer are depicted in Table 5.
Table 5 Main findings of epigenetic therapy in cervical cancer
Lack of response to DNA methylation inhibitor 5-aza-2-deoxycytidine as a single agent in advanced or recurrent cervical cancer*
Response rate of 38.1% with the DNA methylation inhibitor 5-aza-2-deoxycytidine plus cisplatin in advanced or recurrent cervical cancer*
Demethylation and reactivation of the expression of several tumor suppressor genes in the tumors of cervical cancer patients in a phase I trial of the DNA methylation inhibitor hydralazine
Sustained stabilization of disease in a patient with cervical cancer treated within a phase I trial of the HDAC inhibitor MS-275
Major response in a patient with cervical cancer being treated with the HDAC inhibitor valproic acid followed by a single dose of epirubicin within a phase I trial
H3 and H4 hyperacetylation as well as inhibition of deacetylase activity in the tumors of cervical cancer patients with cervical cancer in a phase I trial of the HDAC inhibitor magnesium valproate
Ongoing phase II trial of the combination of hydralazine and magnesium valproate added to cisplatin chemoradiation in FIGO stage IIIB patients
*DNA methylation was not analyzed in these trials.
Decitabine was used against advanced or recurrent cervical carcinoma in a small phase II study at a time when it was already known that this drug was a demethylating agent. The schedule used was 75 mg/m2 per hour 3 times on day 1 at 7-hours intervals and repeated every 5 weeks. None of the 14 patients evaluated for response responded or even had disease stabilization, and one patient died from neutropenic sepsis [266]. Aparicio et al. reported a phase I study using 5-aza-2'-deoxycytidine in patients with advanced solid tumors using escalating doses of 20, 30, and 40 mg/m2 employing a 72-hour continuous infusion every 28 days. Quantitative Methyl-Light reaction was used to evaluate changes in promoter methylation in 19 genes, but no consistent evidence of gene demethylation was documented despite the fact that grade 4 neutropenia was found in nearly one third of the patients. This latter finding argues against its use as a DNA demethylating agent in solid tumors because despite such toxicity, steady-state levels reached during the study (0.1–0.2 μM) are below levels needed in vitro to demethylate gene promoters [244]. On the other hand, a number of genes showed increased methylation, which could be derived from the cytotoxicity of this nucleoside analog. It is well-known that the majority of cytotoxic agents lead to increase in DNA methylation in vitro and in cancer patients [267]. In an attempt to exploit the synergism observed between decitabine and cisplatin, a phase II study with this combination was performed in advanced or recurrent cervical cancer. Initial dose of cisplatin was 40 mg/m2, whereas decitabine dosage was 50 mg/m2 for 3 consecutive days every 21 days. Due to toxicity, the cisplatin dose was reduced to 30 mg/m2. Twenty five patients were included in the study; 24 were eligible for evaluation of toxicity, whereas 21 were eligible for evaluation of tumor responses. A total of 75 cycles of chemotherapy were administered to patients, with a median of three cycles (range: 1–8 cycles per patient). The most frequently observed side effect was grade III and IV neutropenia in 68.0% of cases; one patient died of complications caused by drug-related neutropenic sepsis. Of a total of 21 patients evaluable for tumor response, eight (38.1%) achieved a partial response, whereas stable disease was documented in five cases (23.8%); median survival was 5 months [268]. DNA methylation at global or individual gene was not evaluated; thus, the merit of this combination as DNA methylation-targeted therapy remains to be established. Among other DNA demethylating agents, in a phase I study hydralazine was administered to cohorts of four patients at the following dose levels: 1) 50 mg/day; 11) 75 mg/day; III) 100 mg/day, and IV), 150 mg/day. Tumor biopsies and peripheral blood samples were taken the day before and after treatment to analyze by MSP and RT-PCR the genes APC, MGMT; ER, GSTP1, DAPK, RARβ, FHIT, and p16 pre- and post-treatment for DNA promoter methylation and gene expression. The drug was well-tolerated and rates of demethylation at the different dose levels were as follows: 50 mg/day, 40%; 75 mg/day, 52%; 100 mg/day, 43%, and 150 mg/day, 32%; nine of 12 informative cases (75%) re-expressed the gene. Interestingly, there was neither change in methylation status of H19 and clone 1.2 nor changes in global DNA methylation in peripheral blood mononuclear cells [136].
With regard to histone deacetylase inhibitors for cervical cancer, limited clinical information obtains from a phase I study of MS-275; a patient with cervical cancer had a sustained period of 10 months of stable disease [257], supporting the potential activity of this class of drugs for this tumor type. Also encouraging are recent data from a phase I study for metastatic solid tumors in which valproic acid was administered as an IV loading dose followed by five oral doses were administered every 12 hours followed by a dose of epirubicin at day 3. At the time of reporting, 16 patients have been treated at four dose levels: VPA 15, 30, 45, and 60 mg/kg, with epirubicin at 75 mg/m2. The maximum tolerated dose has not been reached and dose escalation is continuing. Major responses were observed in all tumor types including in anthracycline failures and in anthracycline-resistant cancers such as melanoma and cervical carcinoma, suggesting that inhibition of HDAC activity may chemosensitize tumor cells [269]. Use of valproic acid in the form of magnesium valproate was recently reported in a phase I study where 12 newly diagnosed patients with cervical cancer were treated after a baseline tumor biopsy and blood sampling at the following dose levels (four patients each): 20 mg/kg; 30 mg/kg, or 40 mg/kg for 5 days via oral route; at day 6, when steady state of valproic acid was achieved, tumor and blood sampling were repeated. All patients completed the study medication and mean daily dose for all patients was 1,890 mg with depressed level of consciousness grade 2 as most frequent toxicity. After treatment, there was hyperacetylation of H3 and H4 in tumors of nine and seven patients, respectively, whereas six patients demonstrated hyperacetylation of both histones, and tumor deacetylase activity decreased in eight patients (80%). These data demonstrate for the first time that magnesium valproate at a dose between 20 and 40 mg/kg inhibits deacetylase activity and hyperacetylates histones in malignant solid tumors [270]
It is remarkable that DNA methylation inhibitors such as decitabine [271] and zebularine [272] as well as HDAC inhibitors including SAHA, M344, depsipeptide [273], and valproic acid [274] are radiosensitizers. In addition, it is very noteworthy that HDAC inhibitors such as phenylbutyrate, trichostatin A, and valproic acid are able to reduce cutaneous radiation toxicity following radiotherapy [275]. Therefore, radiation or chemoradiation in combination with DNA demethylating agents and/or HDAC inhibitors is a research avenue to be explored. Based on these data, a phase II study of transcriptional therapy with the demethylating hydralazine plus the histone deacetylase inhibitor magnesium valproate added to chemoradiation with cisplatin is ongoing in FIGO stage IIIB cervical carcinoma.
Conclusion
Epigenetic alterations are at least if not more important than genetic defects for the development and progression of malignant diseases. In cervical cancer, a number of epigenetic alterations occurring during all stages of cervical carcinogenesis in both human papillomavirus and host cellular genomes have been identified. These include global DNA hypomethylation, hypermetylation of key tumor suppressor genes, and histone modifications. From the diagnostic point of view, because epigenetic abnormalities occur very early in the carcinogenic process they can potentially be exploited as molecular markers for early detection. In this sense, identification of a set of genes hypermetylated in cytologic smears could offer novel means for screenning. Assessment of hypermethylated genes in primary tumor or in serum DNA may serve as a prognostic factor or as a means of predicting response to radiation, chemotherapy, and transcriptional agents.
In the therapeutic field, transcriptional therapy is a very promising form of cancer treatment that is being extensively evaluated. It is too early to evaluate usefulness. However, it has now been demonstrated that inhibitors of DNA methylation and histone deacetylases can reactivate expression of tumor suppressor genes and induce hystone hyperacetylation in the tumors of patients with cervical cancer after treatment with these agents.
Authors' contributions
AD-G conceived and wrote the manuscript; ML contributed with the HPV section; MC, LC, CA and EC participated in the discussion, analysis of the content.
Acknowledgements
This work was supported by CONACyT grants SALUD-2002-C01-6579 and AVANCE C01-294, and by Psicofarma, S.A. de C.V., Mexico.
==== Refs
Parkin DM Bray F Ferlay J Pisani P Estimating the world cancer burden: Globocan 2000 Int J Cancer 2001 94 153 156 11668491 10.1002/ijc.1440
Richart RM Barron BA A follow-up study of patients with cervical displasia Am J Obstet Gynecol 1969 105 386 371 5810787
Dueñas-Gonzalez A Cetina L Mariscal I de la Garza J Modern management of locally advanced cervical carcinoma Cancer Treat Rev 2003 29 389 399 12972357 10.1016/S0305-7372(03)00068-9
Whitney CW Sause W Bundy BN Malfetano JH Hannigan EV Fowler WC JrClarke-Pearson DL Liao SY A randomized comparison of fluorouracil plus cisplatin versus hydroxyurea as an adjunct to radiation therapy in stages IIB-IVA carcinoma of the cervix with negative para-aortic lymph nodes. A Gynecologic Oncology Group and Southwest Oncology Group Study J Clin Oncol 1999 17 1339 1348 10334517
Sedlis A Bundy BN Rotman MZ Lentz SS Muderspach LI Zaino RJ A randomized trial of pelvic radiation therapy versus no further therapy in selected patients with stage IB carcinoma of the cervix after radical hysterectomy and pelvic lymphadenectomy: A Gynecologic Oncology Group Study Gynecol Oncol 1999 73 177 183 10329031 10.1006/gyno.1999.5387
Annual Report on the Results of Treatment in Gynaecological Cancer J Epidemiol Statist 1998 3 5 34
Brunschwig A Daniel W Evaluation of pelvic exenteration for advanced cancer of the cervix Surg Gynecol Obstet 1956 103 337 341 13360640
Green JA Kirwan JM Tierney JF Survival and recurrence after concomitant chemotherapy and radiotherapy for cancer of the uterine cervix: a systematic review and meta-analysis Lancet 2001 358 781 786 11564482 10.1016/S0140-6736(01)05965-7
Neoadjuvant Chemotherapy for Locally Advanced Cervical Cancer Meta-analysis Collaboration Neoadjuvant chemotherapy for locally advanced cervical cancer: a systematic review and meta-analysis of individual patient data from 21 randomised trials Eur J Cancer 2003 39 2470 2486 14602133 10.1016/S0959-8049(03)00425-8
López MJ Spratt JS Exenterative pelvic surgery J Surg Oncol 1999 72 102 114 10518108 10.1002/(SICI)1096-9098(199910)72:2<102::AID-JSO12>3.0.CO;2-1
Hockel M Knapstein P The combined operative and radiotherapeutic treatment of recurrent tumors infiltrating the pelvic wall Gynecol Oncol 1992 46 20 28 1634136 10.1016/0090-8258(92)90189-P
Maneo A Landoni F Cormio G Colombo A Placa F Pellegrino A Mangioni C Concurrent carboplatin/5-fluorouracil and radiotherapy for recurrent cervical carcinoma Ann Oncol 1999 10 803 807 10470427 10.1023/A:1008356010556
Friedlander M Guidelines for the treatment of recurrent and metastatic cervical cancer Oncologist 2002 7 342 347 12185296
Gemignani ML Alektiar KM Leitao M Mychalczak B Chi D Venkatraman E Barakat RR Curtin JP Radical surgical resection and high-dose intraoperative radiation therapy (HDR-IORT) in patients with recurrent gynecologic cancers Int J Radiat Oncol Biol Phys 2001 50 687 694 11395237 10.1016/S0360-3016(01)01507-3
Hockel M Laterally extended endopelvic resection: surgical treatment of infrailiac pelvic wall recurrences of gynecological malignancies Am J Obstet Gynecol 1999 180 306 312 9988791
López-Graniel C Dolores-Velázquez R Cetina L González-Enciso A Chanona J Uribe-Domínguez J Candelaria M de la Garza J Dueñas-González A Pre-exenterative chemotherapy for patients with recurrent cervical cancer patients who do not met the classical criteria for pelvic exenteration. A novel therapeutic approach BMC Cancer
Bonomi P Blessing JA Stehman FB DiSaia PJ Walton L Major FJ Randomized trial of three cisplatin dose schedules in squamous-cell carcinoma of the cervix: a Gynecologic Oncology Group study J Clin Oncol 1985 3 1079 1085 3894589
Long HJ 3rdBundy BN Grendys EC JrBenda JA McMeekin DS Sorosky J Miller DS Eaton LA Fiorica JV Randomized phase III trial of cisplatin with or without topotecan in carcinoma of the uterine cervix: A Gynecologic Oncology Group study J Clin Oncol 2005 21 4626 4633 15911865 10.1200/JCO.2005.10.021
zur Hausen Papillomaviruses in human cancers Proc Assoc Am Physicians 1999 111 581 587 10591087 10.1046/j.1525-1381.1999.99723.x
Muñoz N Bosch FX de Sanjosé S Herrero R Castellsague X Shah KV Snijders PJ Meijer CJ International Agency for Research on Cancer Multicenter Cervical Cancer Study Group. Epidemiologic classification of human papillomavirus types associated with cervical cancer N Engl J Med 2003 348 518 527 12571259 10.1056/NEJMoa021641
Walboomers JM Jacobs MV Manos MM Bosch FX Kummer JA Shah KV Snijders PJ Peto J Meijer CJ Muñoz N Human papillomavirus is a necessary cause of invasive cervical cancer worldwide J Pathol 1999 189 12 19 10451482 10.1002/(SICI)1096-9896(199909)189:1<12::AID-PATH431>3.0.CO;2-F
Baker CC Phelps WC Lindgren V Braun MJ Gonda MA Howley PM Structural and transcriptional analysis of human papillomavirus type 16 sequences in cervical carcinoma cell lines J Virol 1987 61 962 967 3029430
Ward P Coleman DV Malcolm DB Regulatory mechanisms of the papillomavirus Trends Genet 1989 5 97 99 2543107 10.1016/0168-9525(89)90037-1
Thierry F Benotmane MA Demeret C Mori M Teissier S Desaintes O A genomic approach reveals a novel mitotic pathway in papillomavirus carcinogenesis Cancer Res 2004 64 895 903 14871818 10.1158/0008-5472.CAN-03-2349
Stubenrauch F Laimins LA Human papillomavirus life cycle: active and latent phases Semin Cancer Biol 1999 9 379 386 10712884 10.1006/scbi.1999.0141
Stubenrauch F Zoble T Iftner T The E8 domain confers a novel long-distance transcriptional repression activity on the E8E2Cprotein of high-risk human papillomavirus type 31 J Virol 2001 75 4139 4149 11287563 10.1128/JVI.75.9.4139-4149.2001
Brown DF Fan L Jones J Bryan J Colocalization of human papillomavirus type 11 E1 E4 and L1 proteins in human foreskin implants grown in athymic mice Virology 1994 201 46 54 8178488 10.1006/viro.1994.1264
Crucius K Auvinen E Alonso A Enhancement of EGF and PMA-mediated MAP kinase activation in cells expressing the human papillomavirus type 16 E5 protein Oncogene 1997 15 1437 1444 9333019 10.1038/sj.onc.1201312
Munger K Howley PM Human papillomavirus immortalization and transformation functions Virus Res 2002 89 213 228 12445661 10.1016/S0168-1702(02)00190-9
Howley PM Scheffner M Munger K Oncoproteins encoded by the cancer-associated papillomavirus target the products of retinoblastoma and p53 tumor suppressor genes Quant Biol 1991 56 159 155
Henstermann A Linares LK Ciechanover A Complete switch form Mdm2 to human papillomavirus E6-mediated degradation of p53 in cervical cancer cells Proc Natl Acad Sci USA 2001 98 1218 1223 11158620 10.1073/pnas.031470698
Mantovani F Banks L The human papillomavirus E6 protein and its contribution to malignant progression Oncogene 2001 20 7874 7887 11753670 10.1038/sj.onc.1204869
Werness BA Levine AJ Howley PM Association of human papillomavirus types 16 and 18 E6 proteins with p53 Science 1990 248 76 79 2157286
Di Leonardo A Linke SP Clarkin K Wahl GM DNA damage triggers a prolonged p53-dependent G1 arrest and long-term induction of Cip1 in normal human fibroblasts Genes Dev 1994 8 2540 2551 7958916
Smith ML Chen IT Zhan Q O'Connor PM Fornace AJ Jr Involvement of the p53 tumor suppressor in repair of u.v.-type DNA damage Oncogene 1995 10 1053 1059 7700629
Hwang ES Riese DJ 2ndSettleman J Nilson LA Honig J Flynn S DiMaio D Inhibition of cervical carcinoma cell line proliferation by the introduction of a bovine papillomavirus regulatory gene Virology 1993 67 3720 3729
Southern SA Herrington CS Disruption of cell cycle control by human papillomavirus with special reference to cervical carcinoma Int J Gynecol Cancer 2000 10 263 274 11240685 10.1046/j.1525-1438.2000.010004263.x
Doorbar J Ely S Sterling J McLean C Crawford L Specific interaction between HPV16 E1–E4 and cytokeratins results in collapse of the epithelial cell intermediate filament network Nature 1991 352 824 827 1715519 10.1038/352824a0
Ueda Y Enomoto T Miyatake T Ozaki K Yoshizaki T Kanao H Ueno Y Nakashima R Shroyer KR Murata Y Monoclonal expansion with integration of high-risk type human papillomavirus is an essential step for cervical carcinogenesis: association of clonal status and human papillomavirus infection with clinical outcome in cervical intraepithelial neoplasia Lab Invest 2003 83 1517 1527 14563953 10.1097/01.LAB.0000092234.68751.83
Jablonka E Lamb MJ The changing concept of epigenetics Ann N Y Acad Sci 2002 981 82 96 12547675
Egger G Liang G Aparicio A Jones PA Epigenetics in human disease and prospects for epigenetic therapy Nature 2004 429 457 463 15164071 10.1038/nature02625
Kawasaki H Taira K Transcriptional gene silencing by short interfering RNAs Curr Opin Mol Ther 2005 7 125 131 15844619
Colot V Rossignol JL Eukaryotic DNA methylation as an evolutionary device Bioessays 1999 21 402 411 10376011 10.1002/(SICI)1521-1878(199905)21:5<402::AID-BIES7>3.0.CO;2-B
Antequera F Bird A CpG islands EXS 1993 64 169 185 8418949
Bird AP CpG-rich islands and the function of DNA methylation Nature 1986 321 209 213 2423876 10.1038/321209a0
Antequera F Bird A Number of CpG islands and genes in human and mouse Proc Natl Acad Sci U S A 1993 90 11995 11999 7505451
Robert MF Morin S Beaulieu N Gauthier F Chute IC Barsalou A MacLeod AR DNMT1 is required to maintain CpG methylation and aberrant gene silencing in human cancer cells Nat Genet 2003 33 61 65 12496760 10.1038/ng1068
Wang KY James-Shen CK DNA methyltransferase Dnmt1 and mismatch repair Oncogene 2004 23 7898 7902 15378011 10.1038/sj.onc.1208111
Okano M Bell DW Haber DA Li E DNA methyltransferases Dnmt3a and Dnmt3b are essential for de novo methylation and mammalian development Cell 1999 99 247 257 10555141 10.1016/S0092-8674(00)81656-6
Liu K Wang YF Cantemir C Muller MT Endogenous assays of DNA methyltransferases: evidence for differential activities of DNMT1, DNMT2, and DNMT3 in mammalian cells in vivo Mol Cell Biol 2003 23 2709 2719 12665573 10.1128/MCB.23.8.2709-2719.2003
Reik W Dean W Walter J Epigenetic reprogramming in mammalian development Science 2001 293 1089 1093 11498579 10.1126/science.1063443
Zhu WG Srinivasan K Dai Z Duan W Druhan LJ Ding H Yee L Villalona-Calero MA Plass C Otterson GA Methylation of adjacent CpG sites affects Sp1/Sp3 binding and activity in the p21(Cip1) promoter Mol Cell Biol 2003 23 4056 4065 12773551 10.1128/MCB.23.12.4056-4065.2003
Nguyen CT Gonzales FA Jones PA Altered chromatin structure associated with methylation-induced gene silencing in cancer cells: correlation of accessibility, methylation, MeCP2 binding and acetylation Nucleic Acids Res 2001 29 4598 4606 11713309 10.1093/nar/29.22.4598
Geiman TM Robertson KD Chromatin remodeling, histone modifications, and DNA methylation- how does it all fit together? Cell Biochem 2002 87 117 125 10.1002/jcb.10286
Baylin SB Reversal of gene silencing as a therapeutic target for cancer- roles for DNA methylation and its interdigitation with chromatin Novartis Found Symp 2004 259 226 233 15171257
Bird A 2002 DNA methylation patterns and epigenetic memory Genes Dev 2000 16 6 21 11782440 10.1101/gad.947102
Chakravarthy S Park YJ Chodaparambil J Edayathumangalam RS Luger K Structure and dynamic properties of nucleosome core particles FEBS Lett 2005 579 895 898 15680970 10.1016/j.febslet.2004.11.030
Kuo MH Allis CD Roles of histone acetyltransferases and deacetylases in gene regulation Bioessays 1998 20 615 626 9780836 10.1002/(SICI)1521-1878(199808)20:8<615::AID-BIES4>3.0.CO;2-H
Wang C Fu M Mani S Wadler S Senderowicz AM Pestell RG Histone acetylation and the cell-cycle in cancer Front Biosci 2001 6 D610 D629 11282573
Davie JR Covalent modifications of histones: expression from chromatin templates Curr Opin Genet Dev 1998 8 173 178 9610407 10.1016/S0959-437X(98)80138-X
Trievel RC Structure and function of histone methyltransferases Crit Rev Eukaryot Gene Expr 2004 14 147 169 15248813 10.1615/CritRevEukaryotGeneExpr.v14.i3.10
Daniel JA Pray-Grant MG Grant PA Effector proteins for methylated histones: an expanding family Cell Cycle 2005
Lachner M O'Carroll D Rea S Mechtler K Jenuwein T Methylation of histone H3 lysine 9 creates a binding site for HP1 proteins Nature 2001 410 116 120 11242053 10.1038/35065132
Mahadevan LC Willis AC Barratt MJ Rapid histone H3 phosphorylation in response to growth factors, phorbol esters, okadaic acid, and protein synthesis inhibitors Cell 1991 65 775 783 2040014 10.1016/0092-8674(91)90385-C
Nowak SJ Corces VG Phosphorylation of histone H3 correlates with transcriptionally active loci Genes Dev 2000 14 3003 3013 11114889 10.1101/gad.848800
Pascreau G Arlot-Bonnemains Y Prigent C Phosphorylation of histone and histone-like proteins by aurora kinases during mitosis Prog Cell Cycle Res 2003 5 369 374 14593731
Agalioti T Chen G Thanos D Deciphering the transcriptional histone acetylation code for a human gene Cell 2002 111 381 392 12419248 10.1016/S0092-8674(02)01077-2
de la Cruz X Lois S Sánchez-Molina S Martínez-Balbas MA Do protein motifs read the histone code? Bioessays 2005 27 164 175 15666348 10.1002/bies.20176
Feinberg AP Vogelstein B Hypomethylation distinguishes genes of some human cancers from their normal counterparts Nature 1983 301 89 92 6185846 10.1038/301089a0
de Bustros A Nelkin BD Silverman A Ehrlich G Poiesz B Baylin SB The short arm of chromosome 11 is a "hot spot" for hypermethylation in human neoplasia Proc Natl Acad Sci U S A 1988 85 5693 5697 2840671
Laird PW Jaenisch R DNA methylation and cancer Hum Molec Genet 1994 3 1487 1495 7849743
Greger V Passarge E Hopping W Messmer E Horsthemke B Epigenetic changes may contribute to the formation and spontaneous regression of retinoblastoma Hum Genet 1989 83 155 158 2550354 10.1007/BF00286709
Santini V Kantarjian HM Issa JP Changes in DNA methylation in neoplasia: pathophysiology and therapeutic implications Ann Intern Med 2001 134 573 586 11281740
Szyf M Targeti NG DNA methylation in cancer Ageing Res Rev 2003 2 299 328 12726777 10.1016/S1568-1637(03)00012-6
Bachman KE Histone modifications and silencing prior to DNA methylation of a tumor suppressor gene Cancer Cell 2003 3 89 95 12559178 10.1016/S1535-6108(02)00234-9
Clark SJ Melki J DNA methylation and gene silencing in cancer: which is the guilty part? Oncogene 2002 21 5380 5387 12154400 10.1038/sj.onc.1205598
Goelz SE Vogelstein B Hamilton SR Feinberg AP Hypomethylation of DNA from benign and malignant human colon neoplasms Science 1985 228 187 190 2579435
Feinberg AP Gehrke CW Kuo KC Ehrlich M Reduced genomic 5-methylcytosine content in human colonic neoplasia Cancer Res 1988 48 1159 1161 3342396
Ehrlich M DNA hypomethylation, cancer, the immunodeficiency, centromeric region instability, facial anomalies syndrome and chromosomal rearrangements J Nutr 2002 132 2424S 2429S 12163705
Dunn BK Hypomethylation: one side of a larger picture Ann N Y Acad Sci 2003 983 28 42 12724210
Gamma-Sosa MA Slagel VA Trewyn RW Oxenhandler R Kuo KC Gehrke CW Ehrlich M The 5-methylcytosine content of DNA from human tumors Nucleic Acids Res 1983 11 6883 6894 6314264
Eden A Gaudet F Waghmare A Jaenisch R Chromosomal instability and tumors promoted by DNA hypomethylation Science 2003 300 455 12702868 10.1126/science.1083557
Hansen RS Wijmenga C Luo P Stanek AM Canfield TK Weemaes CM Gartler SM The DNMT3B DNA methyltransferase gene is mutated in the ICF immunodeficiency syndrome Proc Natl Acad Sci U S A 1999 96 14412 14417 10588719 10.1073/pnas.96.25.14412
Narayan A Ji W Zhang XY Marrogi A Graff JR Baylin SB Ehrlich M Hypomethylation of pericentromeric DNA in breast adenocarcinomas Int J Cancer 1998 77 833 838 9714050 10.1002/(SICI)1097-0215(19980911)77:6<833::AID-IJC6>3.0.CO;2-V
Strichman-Almashanu LZ A genome-wide screen for normally methylated human CgG islands that can identify novel imprinted genes Genome Res 2002 12 543 554 11932239 10.1101/gr.224102. Article published online before print in March 2002
Ogishima T Shiina H Breault JE Tabatabai L Bassett WW Enokida H Li LC Kawakami T Urakami S Ribeiro-Filho LA Terashima M Fujime M Igawa M Dahiya R Increased heparanase expression is caused by promoter hypomethylation and up-regulation of transcriptional factor early growth response-1 in human prostate cancer Clin Cancer Res 2005 11 1028 1036 15709168
Sato N Maitra A Fukushima N Van Heek NT Matsubayashi H Iacobuzio-Donahue CA Rosty C Goggins M Frequent hypomethylation of multiple genes overexpressed in pancreatic ductal adenocarcinoma Cancer Res 2003 63 4158 4166 12874021
Akiyama Y Maesawa C Ogasawara S Terashima M Masuda T Cell-type-specific repression of the maspin gene is disrupted frequently by demethylation at the promoter region in gastric intestinal metaplasia and cancer cells Am J Pathol 2003 163 1911 1919 14578190
Karpf AR Lasek AW Ririe TO Hanks AN Grossman D Jones DA Limited gene activation in tumor and normal epithelial cells treated with the DNA methyltransferase inhibitor 5-aza-2'-deoxycytidine Mol Pharmacol 2004 65 18 27 14722233 10.1124/mol.65.1.18
Koshiishi N Chong JM Fukasawa T Ikeno R Hayashi Y Funata N Nagai H Miyaki M Matsumoto Y Fukayama M p300 Gene alterations in intestinal and diffuse types of gastric carcinoma Gastric Cancer 2004 7 85 90 15224194 10.1007/s10120-004-0273-8
Ward R Johnson M Shridhar V van Deursen J Couch FJ CBP truncating mutations in ovarian cancer J Med Genet 2005 42 514 518 15937088 10.1136/jmg.2004.025080
So CK Nie Y Song Y Loss of heterozygosity and internal tandem duplication mutations of the CBP gene are frequent events in human esophageal squamous cell carcinoma Clin Cancer Res 2004 10 19 27 14734447 10.1158/1078-0432.CCR-03-0160
Kishimoto M Kohno T Okudela K Otsuka A Sasaki H Tanabe C Sakiyama T Hirama C Kitabayashi I Minna JD Takenoshita S Yokota J Mutations and deletions of the CBP gene in human lung cancer Clin Cancer Res 2005 11 512 519 15701835
Sobulo OM Borrow J Tomek R Reshmi S Harden A Schlegelberger B Housman D Doggett NA Rowley JD Zeleznik-Le NJ MLL is fused to CBP, a histone acetyltransferase, in therapy-related acute myeloid leukemia with a t(11;16)(q23;p13.3) Proc Natl Acad Sci U S A 1997 94 8732 8777 9238046 10.1073/pnas.94.16.8732
Chaffanet M Gressin L Preudhomme C Soenen-Cornu V Birnbaum D Pebusque MJ MOZ is fused to p300 in an acute monocytic leukemia with t(8;22) Genes Chromosomes Cancer 2000 28 138 144 10824998 10.1002/(SICI)1098-2264(200006)28:2<138::AID-GCC2>3.0.CO;2-2
Lin RJ Nagy L Inoue S Shao W Miller WH JrEvans RM Role of the histone deacetylase complex in acute promyelocytic leukaemia Nature 1998 391 811 814 9486654 10.1038/35895
Licht JD AML1 and the AML1-ETO fusion protein in the pathogenesis of t(8;21) AML Oncogene 2001 20 5660 5679 11607817 10.1038/sj.onc.1204593
Melnick A Carlile G Ahmad KF Kiang CL Corcoran C Bardwell V Prive GG Licht JD Critical residues within the BTB domain of PLZF and Bcl-6 modulate interaction with corepressors Mol Cell Biol 2002 22 1804 1818 11865059 10.1128/MCB.22.6.1804-1818.2002
Verma M Viral genes and methylation Ann N Y Acad Sci 2003 983 170 180 12724222
Tao Q Huang H Geiman TM Lim CY Fu L Qiu GH Robertson KD Defective de novo methylation of viral and cellular DNA sequences in ICF syndrome cells Hum Mol Genet 2002 11 2091 2102 12189161 10.1093/hmg/11.18.2091
Tao Q Robertson KD Stealth technology: how Epstein-Barr virus utilizes DNA methylation to cloak itself from immune detection Clin Immunol 2003 109 53 63 14585276 10.1016/S1521-6616(03)00198-0
Tsai CN Tsai CL Tse KP Chang HY Chang YS The Epstein-Barr virus oncogene product, latent membrane protein 1, induces the downregulation of E-cadherin gene expression via activation of DNA methyltransferases Proc Natl Acad Sci U S A 2002 99 10084 10089 12110730 10.1073/pnas.152059399
Rosl F Arab A Klevenz B zur Hausen H The effect of DNA methylation on gene regulation of human papillomaviruses J Gen Virol 1993 74 791 801 8388016
Badal V Chuang LS Tan EH Badal S Villa LL Wheeler CM Li BF Bernard HU CpG methylation of human papillomavirus type 16 DNA in cervical cancer cell lines and in clinical specimens: genomic hypomethylation correlates with carcinogenic progression Virology 2003 77 6227 6234 10.1128/JVI.77.11.6227-6234.2003
Kalantari M Calleja-Macías IE Tewari D Hagmar B Lie K Barrera-Saldaña HA Wiley DJ Bernard HU Conserved methylation patterns of human papillomavirus type 16 DNA in asymptomatic infection and cervical neoplasia J Virol 2004 78 12762 12772 15542628 10.1128/JVI.78.23.12762-12772.2004
Badal S Badal V Calleja-Macías IE Kalantari M Chuang LS Li BF Bernard HU The human papillomavirus-18 genome is efficiently targeted by cellular DNA methylation Virology 2004 324 483 492 15207633 10.1016/j.virol.2004.04.002
Thain A Jenkins O Clarke AR Gaston K CpG methylation directly inhibits binding of the human papillomavirus type 16 E2 protein to specific DNA sequences J Virol 1996 70 7233 7235 8794373
Kim K Garner-Hamrick PA Fisher C Lee D Lambert PF Methylation patterns of papillomavirus DNA, its influence on E2 function, and implications in viral infection J Virol 2003 77 12450 12459 14610169 10.1128/JVI.77.23.12450-12459.2003
Soengas MS Capodieci P Polsky D Mora J Esteller M Opitz-Araya X McCombie R Herman JG Gerald WL Lazebnik YA Cordon-Cardo C Lowe SW Inactivation of the apoptosis effector Apaf-1 in malignant melanoma Nature 2001 409 207 211 11196646 10.1038/35051606
van Noesel MM van Bezouw S Salomons GS Voute PA Pieters R Baylin SB Herman JG Versteeg R Tumor-specific down-regulation of the tumor necrosis factor-related apoptosis-inducing ligand decoy receptors DcR1 and DcR2 is associated with dense promoter hypermethylation Cancer Res 2002 62 2157 2161 11929838
Ashkenazi A Dixit VM Apoptosis control by death and decoy receptors Curr Opin Cell Biol 1999 11 255 260 10209153 10.1016/S0955-0674(99)80034-9
Ozoren N El-Deiry WS Cell surface death receptor signaling in normal and cancer cells Semin Cancer Biol 2003 13 135 147 12654257 10.1016/S1044-579X(02)00131-1
Shivapurkar N Toyooka S Toyooka KO Reddy J Miyajima K Suzuki M Shigematsu H Takahashi T Parikh G Pass HI Chaudhary PM Gazdar AF Aberrant methylation of trail decoy receptor genes is frequent in multiple tumor types Int J Cancer 2004 109 786 792 14999791 10.1002/ijc.20041
Weinrich SL Pruzan R Ma L Ouellette M Tesmer VM Holt SE Reconstitution of human telomerase with the template RNA component hTR and the catalytic protein subunit hTRT Nature Genet 1997 17 498 502 9398860 10.1038/ng1297-498
Lange T Activation of telomerase in a human tumor Proc Natl Acad Sci U S A 1994 91 2882 2885 8159672
Yokoyama Y Takahashi Y Shinohara A Lian Z Tamaya T Telomerase activity in the female reproductive tract and neoplasms Gynecol Oncol 1998 68 145 149 9514809 10.1006/gyno.1997.4921
Nakano N Watney E McDougall JK Telomerase activity and expression of telomerase RNA component and telomerase catalytic subunit gene in cervical cancer Am J Pathol 1998 153 857 864 9736034
Snijders PJ van Duin M Walboomers JM Steenbergen RD Risse EK Helmerhorst TJ Telomerase activity exclusively in cervical carcinomas and a subset of cervical intraepithelial neoplasia grade III lesions: strong association with elevated messenger RNA levels of its catalytic subunit and high-risk human papillomavirus DNA Cancer Res 1998 58 3812 3818 9731489
Takakura M Kyo S Kanaya T Tanaka M Inoue M Expression of human telomerase subunits and correlation with telomerase activity in cervical cancer Cancer Res 1998 58 1558 1561 9537264
Wisman GB Knol AJ Helder MN Krans M Vries EG Hollema H Telomerase in relation to clinicopathologic prognostic factors and survival in cervical cancer Int J Cancer 2001 91 658 664 11267977 10.1002/1097-0215(200002)9999:9999<::AID-IJC1099>3.0.CO;2-7
Guilleret I Benhattar J Demethylation of the human telomerase catalytic subunit (hTERT) gene promoter reduced hTERT expression and telomerase activity and shortened telomeres Exp Cell Res 2003 289 326 334 14499633 10.1016/S0014-4827(03)00281-7
Widschwendter A Muller HM Hubalek MM Wiedemair A Fiegl H Goebel G Mueller-Holzner E Marth C Widschwendter M Methylation status and expression of human telomerase reverse transcriptase in ovarian and cervical cancer Gynecol Oncol 2004 93 407 416 15099954 10.1016/j.ygyno.2004.01.036
Vousden KH p53: death star Cell 2000 103 691 694 11114324 10.1016/S0092-8674(00)00171-9
Harris SL Levine AJ The p53 pathway: positive and negative feedback loops Oncogene 2005 24 2899 2908 15838523 10.1038/sj.onc.1208615
Melino G De Laurenzi V Vousden KH p73: friend or foe in tumorigenesis Nature Rev Cancer 2002 2 605 615 12154353 10.1038/nrc861
Zhu J Jiang J Zhou W Chen X The potential tumor suppressor p73 differentially regulates cellular p53 target genes Cancer Res 1998 58 5061 5065 9823311
Yang A Kaghad M Wang Y p63, A p53 homolog at 3q27-29, encodes multiple products with transactivating, death-inducing, and dominant-negative activities Molec Cell 1998 2 305 316 9774969 10.1016/S1097-2765(00)80275-0
Banelli B Casciano I Romani M Methylation-independent silencing of the p73 gene in neuroblastoma Oncogene 2000 19 4553 4556 11002429 10.1038/sj.onc.1203807
Liu SS Leung RC Chan KY Chiu PM Cheung AN Tam KF Ng TY Wong LC Ngan HY p73 Expression is associated with the cellular radiosensitivity in cervical cancer after radiotherapy Clin Cancer Res 2004 10 3309 3316 15161684 10.1158/1078-0432.CCR-03-0119
Nakashima R Fujita M Enomoto T Haba T Yoshino K Wada H Kurachi H Sasaki M Wakasa K Inoue M Buzard G Murata Y Alteration of p16 and p15 genes in human uterine tumours Br J Cancer 1999 80 458 467 10408854 10.1038/sj.bjc.6690379
Nuovo GJ Plaia TW Belinsky SA Baylin SB Herman JG In situ detection of the hypermethylation-induced inactivation of the p16 gene as an early event in oncogenesis Proc Natl Acad Sci U S A 1999 96 12754 12759 10535995 10.1073/pnas.96.22.12754
Wong YF Chung TK Cheung TH Nobori T Yu AL Yu J Batova A Lai KW Chang AM Methylation of p16INK4A in primary gynecologic malignancy Cancer Lett 1999 136 231 235 10355753 10.1016/S0304-3835(98)00327-9
Narayan G Arias-Pulido H Koul S Vargas H Zhang FF Villella J Schneider A Terry MB Mansukhani M Murty VV Frequent promoter methylation of CDH1, DAPK, RARB, and HIC1 genes in carcinoma of cervix uteri: its relationship to clinical outcome Molec Cancer 2003 2 24 12773202 10.1186/1476-4598-2-24
Virmani AK Muller C Rathi A Zoechbauer-Mueller S Mathis M Gazdar AF Aberrant methylation during cervical carcinogenesis Clin Cancer Res 2001 7 584 589 11297252
Dong SM Kim HS Rha SH Sidransky D Promoter hypermethylation of multiple genes in carcinoma of the uterine cervix Clin Cancer Res 2001 7 1982 1986 11448914
Zambrano P Segura-Pacheco B Pérez-Cárdenas E Cetina L Revilla-Vázquez A Taja-Chayeb L Chávez-Blanco A Angeles E Cabrera G Sandoval K Trejo-Becerril C Chanona-Vilchis J Dueñas-González A A phase I study of hydralazine to demethylate and reactivate the expression of tumor suppressor genes BMC Cancer 2005 5 44 15862127 10.1186/1471-2407-5-44
Uthoff SM Eichenberger MR McAuliffe TL Hamilton CJ Galandiuk S Wingless-type frizzled protein receptor signaling and its putative role in human colon cancer Molec Carcinogen 2001 31 56 62 10.1002/mc.1039
Hajra KM Fearon ER Cadherin and catenin alterations in human cancer Genes Chromosomes Cancer 2002 34 255 268 12007186 10.1002/gcc.10083
Persad S Troussard AA McPhee TR Mulholland DJ Dedhar S Tumor suppressor PTEN inhibits nuclear accumulation of beta-catenin and T cell/lymphoid enhancer factor 1-mediated transcriptional activation J Cell Biol 2001 153 1161 1174 11402061 10.1083/jcb.153.6.1161
Giles RH van Es JH Clevers H Caught up in a Wnt storm: Wnt signaling in cancer Biochim Biophys Acta 2003 1653 1 24 12781368
Thorstensen L Lind GE Lovig T Diep CB Meling GI Rognum TO Lothe RA Genetic and epigenetic changes of components affecting the WNT pathway in colorectal carcinomas stratified by microsatellite instability Neoplasia 2005 7 99 108 15802015
Cheung TH Lo KW Yim SF Chan LK Heung MS Chan CS Cheung AY Chung TK Wong YF Epigenetic and genetic alternation of PTEN in cervical neoplasm Gynecol Oncol 2004 93 621 627 15196854 10.1016/j.ygyno.2004.03.013
Widschwendter A Ivarsson L Blassnig A Muller HM Fiegl H Wiedemair A Muller-Holzner E Goebel G Marth C Widschwendter M CDH1 and CDH13 methylation in serum is an independent prognostic marker in cervical cancer patients Int J Cancer 2004 109 163 166 14750164 10.1002/ijc.11706
Kang S Kim JW Kang GH Park NH Song YS Kang SB Lee HP Polymorphism in folate- and methionine-metabolizing enzyme and aberrant CpG island hypermethylation in uterine cervical cancer Gynecol Oncol 2005 96 173 180 15589597 10.1016/j.ygyno.2004.09.031
Chen CL Liu SS Ip SM Wong LC Ng TY Ngan HY E-cadherin expression is silenced by DNA methylation in cervical cancer cell lines and tumours Eur J Cancer 2003 39 517 523 12751384 10.1016/S0959-8049(02)00175-2
Muller HM Widschwendter A Fiegl H Goebel G Wiedemair A Muller-Holzner E Marth C Widschwendter M A DNA methylation pattern similar to normal tissue is associated with better prognosis in human cervical cancer Cancer Lett 2004 209 231 236 15159026 10.1016/j.canlet.2003.12.016
Esteller M Sparks A Toyota M Sánchez-Cespedes M Capella G Peinado MA González S Tarafa G Sidransky D Meltzer SJ Baylin SB Herman JG Analysis of adenomatous polyposis coli promoter hypermethylation in human cancer Cancer Res 2000 60 4366 4371 10969779
Zysman M Saka A Millar A Knight J Chapman W Bapat B Methylation of adenomatous polyposis coli in endometrial cancer occurs more frequently in tumors with microsatellite instability phenotype Cancer Res 2002 62 3663 12097272
Pegg AE Byers TL Repair of DNA containing O6-alkylguanine FASEB J 1992 6 2302 2310 1544541
Esteller M Herman JG Generating mutations but providing chemosensitivity: the role of O6-methylguanine DNA methyltransferase in human cancer Oncogene 2004 23 1 8 14712205 10.1038/sj.onc.1207316
Hegi ME Diserens AC Gorlia T Hamou MF de Tribolet N Weller M Kros JM Hainfellner JA Mason W Mariani L Bromberg JE Hau P Mirimanoff RO Cairncross JG Janzer RC Stupp R MGMT gene silencing and benefit from temozolomide in glioblastoma N Engl J Med 2005 352 997 1003 15758010 10.1056/NEJMoa043331
Park TJ Han SU Cho YK Paik WK Kim YB Lim IK Methylation of O(6)-methylguanine-DNA methyltransferase gene is associated significantly with K-ras mutation, lymph node invasion, tumor staging, and disease free survival in patients with gastric carcinoma Cancer 2001 92 2760 2768 11753949 10.1002/1097-0142(20011201)92:11<2760::AID-CNCR10123>3.0.CO;2-8
Qi J Zhu YQ Huang MF Yang D Hypermethylation of CpG island in O6-methylguanine-DNA methyltransferase gene was associated with K-ras G to A mutation in colorectal tumor World J Gastroenterol 2005 11 2022 2025 15800999
Zuo C Ai L Ratliff P Suen JY Hanna E Brent TP Fan CY O6-methylguanine-DNA methyltransferase gene: epigenetic silencing and prognostic value in head and neck squamous cell carcinoma Cancer Epidemiol Biomarkers Prev 2004 13 967 975 15184253
Komine C Watanabe T Katayama Y Yoshino A Yokoyama T Fukushima T Promoter hypermethylation of the DNA repair gene O6-methylguanine-DNA methyltransferase is an independent predictor of shortened progression free survival in patients with low-grade diffuse astrocytomas Brain Pathol 2003 13 176 184 12744471
Kutler DI Singh B Satagopan J A 20-year perspective on the International Fanconi Anemia Registry (IFAR) Blood 2003 101 1249 1256 12393516 10.1182/blood-2002-07-2170
Joenje H Patel KJ The emerging genetic and molecular basis of Fanconi anaemia Nature Rev Genet 2001 2 446 457 11389461 10.1038/35076590
D'Andrea AD Grompe M The Fanconi anaemia/BRCA pathway Nat Rev Cancer 2003 3 23 34 12509764 10.1038/nrc970
Taniguchi T Tischkowitz M Ameziane N Disruption of the Fanconi anemia-BRCA pathway in cisplatin-sensitive ovarian tumors Nature Med 2003 9 568 574 12692539 10.1038/nm852
Marsit CJ Liu M Nelson HH Posner M Suzuki M Kelsey KT Inactivation of the Fanconi anemia/BRCA pathway in lung and oral cancers: implications for treatment and survival Oncogene 2004 23 1000 1004 14647419 10.1038/sj.onc.1207256
Narayan G Arias-Pulido H Nandula SV Basso K Sugirtharaj DD Vargas H Mansukhani M Villella J Meyer L Schneider A Gissmann L Durst M Pothuri B Murty VV Promoter hypermethylation of FANCF: disruption of Fanconi Anemia-BRCA pathway in cervical cancer Cancer Res 2004 64 2994 2997 15126331 10.1158/0008-5472.CAN-04-0245
Jass JR HNPCC and sporadic MSI-H colorectal cancer: a review of the morphological similarities and differences Fam Cancer 2004 3 93 100 15340259 10.1023/B:FAME.0000039849.86008.b7
Aaltonen LA Peltomaki P Mecklin JP Jarvinen H Jass JR Green JS Lynch HT Watson P Tallqvist G Juhola M Sistonen P Hamilton SR Kinzler KW Vogelstein B de la Chapelle A Replication errors in benign and malignant tumors from hereditary non-polyposis colorectal cancer patients Cancer Res 1994 54 1645 1648 8137274
Deng G Chen A Hong J Chae HS Kim YS Methylation of CpG in a small region of the hMLH1 promoter invariably correlates with the absence of gene expression Cancer Res 1999 59 2029 2033 10232580
de las Alas MM Aebi S Fink D Howell SB Los G Loss of DNA mismatch repair: effects on the rate of mutation to drug resistance Natl Cancer Inst 1997 89 1537 1541 10.1093/jnci/89.20.1537
Gifford G Paul J Vasey PA Kaye SB Brown R The acquisition of hMLH1 methylation in plasma DNA after chemotherapy predicts poor survival for ovarian cancer patients Clin Cancer Res 2004 10 4420 4426 15240532 10.1158/1078-0432.CCR-03-0732
Ciavattini A Piccioni M Tranquilli AL Filosa A Pieramici T Goteri G Immunohistochemical expression of DNA mismatch repair (MMR) system proteins (hMLH1, hMSH2) in cervical preinvasive and invasive lesions Pathol Res Pract 2005 201 21 25 15807307 10.1016/j.prp.2004.09.012
Kwasniewska A Gozdzicka-Jozefiak A Postawski K Miturski R Evaluation of DNA mismatch repair system in cervical dysplasias and invasive carcinomas related to HPV infection Eur J Gynaecol Oncol 2002 23 231 235 12094961
Chung TK Cheung TH Wang VW Yu MY Wong YF Microsatellite instability, expression of hMSH2 and hMLH1 and HPV infection in cervical cancer and their clinico-pathological association Gynecol Obstet Invest 2001 52 98 103 11586036 10.1159/000052951
Ercoli A Ferrandina G Genuardi M Zannoni GF Cicchillitti L Raspaglio G Carrara S Mancuso S Scambia G Microsatellite instability is not related to response to cisplatin-based chemotherapy in cervical cancer Int J Gynecol Cancer 2005 15 308 311 15823117 10.1111/j.1525-1438.2005.15221.x
Dammann R Schagdarsurengin U Seidel C Strunnikova M Rastetter M Baier K Pfeifer GP The tumor suppressor RASSF1A in human carcinogenesis: an update Histol Histopathol 2005 20 645 663 15736067
Cohen Y Singer G Lavie O Dong SM Beller U Sidransky D The RASSF1A tumor suppressor gene is commonly inactivated in adenocarcinoma of the uterine cervix Clin Cancer Res 2003 9 2981 2984 12912945
Kuzmin I Liu L Dammann R Geil L Stanbridge EJ Wilczynski SP Lerman MI Pfeifer GP Inactivation of RAS association domain family 1A gene in cervical carcinomas and the role of human papillomavirus infection Cancer Res 2003 63 1888 1893 12702579
Yu MY Tong JH Chan PK Lee TL Chan MW Chan AW Lo KW To KF Hypermethylation of the tumor suppressor gene RASSFIA and frequent concomitant loss of heterozygosity at 3p21 in cervical cancers Int J Cancer 2003 105 204 209 12673680 10.1002/ijc.11051
Deiss LP Feinstein E Berissi H Cohen O Kimchi A Identification of a novel serine/threonine kinase and a novel 15-kD protein as potential mediators of the gamma interferon-induced cell death Genes Dev 1995 9 15 30 7828849
Shohat G Shani G Eisenstein M Kimchi A The DAP-kinase family of proteins: study of a novel group of calcium-regulated death-promoting kinases Biochem Biophys Acta 2002 1600 45 50 12445458
Inbal B Cohen O Polak-Charcon S Kopolovic J Vadai E Eisenbach L Kimchi A DAP kinase links the control of apoptosis to metastasis Nature 1997 390 180 184 9367156 10.1038/36599
Watabe K Ito A Koma YI Kitamura Y IGSF4: a new intercellular adhesion molecule that is called by three names, TSLC1, SgIGSF and SynCAM, by virtue of its diverse function Histol Histopathol 2003 18 1321 1329 12973698
Steenbergen RD Kramer D Braakhuis BJ Stern PL Verheijen RH Meijer CJ Snijders PJ TSLC1 gene silencing in cervical cancer cell lines and cervical neoplasia J Natl Cancer Inst 2004 96 294 305 14970278
Li J Zhang Z Bidder M Funk MC Nguyen L Goodfellow PJ Rader JS IGSF4 promoter methylation and expression silencing in human cervical cancer Gynecol Oncol 2005 96 150 158 15589594 10.1016/j.ygyno.2004.08.050
Wistuba II Montellano FD Milchgrub S Virmani AK Behrens C Chen H Ahmadian M Nowak JA Muller C Minna JD Gazdar AF Deletions of chromosome 3p are frequent and early events in the pathogenesis of uterine cervical carcinoma Cancer Res 1997 57 3154 3158 9242443
Larson AA Kern S Curtiss S Gordon R Cavenee WK Hampton GM High resolution analysis of chromosome 3p alterations in cervical carcinoma Cancer Res 1997 57 4082 4090 9307297
Hendricks DT Taylor R Reed M Birrer MJ FHIT gene expression in human ovarian, endometrial, and cervical cancer cell lines Cancer Res 1997 57 2112 2115 9187105
Birrer MJ Hendricks D Farley J Sundborg MJ Bonome T Walts MJ Geradts J Abnormal Fhit expression in malignant and premalignant lesions of the cervix Cancer Res 1999 59 5270 5274 10537308
Connolly DC Greenspan DL Wu R Ren X Dunn RL Shah KV Jones RW Bosch FX Muñoz N Cho KR Loss of fhit expression in invasive cervical carcinomas and intraepithelial lesions associated with invasive disease Clin Cancer Res 2000 6 3505 3510 10999736
Greenspan DL Connolly DC Wu R Lei RY Vogelstein JT Kim YT Mok JE Muñoz N Bosch FX Shah K Cho KR Loss of FHIT expression in cervical carcinoma cell lines and primary tumors Cancer Res 1997 57 4692 4698 9354423
Nakagawa S Yoshikawa H Kimura M Kawana K Matsumoto K Onda T Kino N Yamada M Yasugi T Taketani Y A possible involvement of aberrant expression of the FHIT gene in the carcinogenesis of squamous cell carcinoma of the uterine cervix Br J Cancer 1999 79 589 594 10027335 10.1038/sj.bjc.6690093
Yoshino K Enomoto T Nakamura T Sun H Ozaki K Nakashima R Wada H Saitoh J Watanabe Y Noda K Murata Y FHIT alterations in cancerous and non-cancerous cervical epithelium Int J Cancer 2000 85 6 13 10585575 10.1002/(SICI)1097-0215(20000101)85:1<6::AID-IJC2>3.0.CO;2-6
Roz L Gramegna M Ishii H Croce CM Sozzi G Restoration of fragile histidine triad (FHIT) expression induces apoptosis and suppresses tumorigenicity in lung and cervical cancer cell lines Proc Natl Acad Sci U S A 2002 99 3615 3620 11891319 10.1073/pnas.062030799
Wu Q Shi H Suo Z Nesland JM 5'-CpG island methylation of the FHIT gene is associated with reduced protein expression and higher clinical stage in cervical carcinomas Ultrastruct Pathol 2003 27 417 422 14660280
Chen WY Zeng X Carter MG Morrell CN Chiu Yen RW Esteller M Watkins DN Herman JG Mankowski JL Baylin SB Heterozygous disruption of Hic1 predisposes mice to a gender-dependent spectrum of malignant tumors Nature Genet 2003 33 197 202 12539045 10.1038/ng1077
Chen W Cooper TK Zahnow CA Overholtzer M Zhao Z Ladanyi M Karp JE Gokgoz N Wunder JS Andrulis IL Levine AJ Mankowski JL Baylin SB Epigenetic and genetic loss of Hic1 function accentuates the role of p53 in tumorigenesis Cancer Cell 2004 6 387 398 15488761 10.1016/j.ccr.2004.08.030
Chambon P The retinoid signaling pathway: molecular and genetic analyses Semin Cell Biol 1994 5 115 125 8068884 10.1006/scel.1994.1015
Widschwendter M Berger J Hermann M Muller HM Amberger A Zeschnigk M Widschwendter A Abendstein B Zeimet AG Daxenbichler G Marth C Methylation and silencing of the retinoic acid receptor-beta2 gene in breast cancer J Natl Cancer Inst 2000 92 826 832 10814678 10.1093/jnci/92.10.826
Sun SY Retinoic acid receptor beta and colon cancer Cancer Biol Ther 2004 3 87 88 14726690
Khan MA Jenkins GR Tolleson WH Creek KE Pirisi L Retinoic acid inhibition of human papillomavirus mediated transformation of human keratinocytes Cancer Res 1993 53 905 909 8381332
Meyskens FL Surwit E Moon TE Childers JM Davis JR Dorr RT Jonson CS Alberts DS Enhancement of regression of cervical intraepithelial neoplasia II (moderate dysplasia) with topically applied all-trans retinoic acid: a randomized trial J Natl Cancer Inst 1994 86 539 543 8133537
Lippman SM Kavanagh JJ Paredes-Espinoza M Delgadillo-Madrueno F Paredes Casillas P Hong WK Massimini G Holdener EE Kazakoff IH 13-cis-Retinoic acid plus interferon-α2a in locally advanced squamous cell carcinoma of cervix J Natl Cancer Inst 1993 85 499 500 8445678
Geraddts J Chen JY Russell EA Yankaskas JR Nieves L Minna JD Human cancer cell lines exhibit resistance to retinoic acid treatment Cell Growth Differ 1993 4 799 809 8274449
Ivanova T Petrenko A Gritsko T Vinokourova S Eshilev E Kobzeva V Kisseljov F Kisseljova N Methylation and silencing of the retinoic acid receptor-beta 2 gene in cervical cancer BMC Cancer 2002 2 4 11945179 10.1186/1471-2407-2-4
Feng Q Balasubramanian A Hawes SE Toure P Sow PS Dem A Dembele B Critchlow CW Xi L Lu H McIntosh MW Young AM Kiviat NB Detection of hypermethylated genes in women with and without cervical neoplasia J Natl Cancer Inst 2005 97 273 282 15713962
Ivanova T Vinokurova S Petrenko A Eshilev E Solovyova N Kisseljov F Kisseljova N Frequent hypermethylation of 5' flanking region of TIMP-2 gene in cervical cancer Int J Cancer 2004 108 882 886 14712492 10.1002/ijc.11652
Widschwendter A Muller HM Fiegl H Ivarsson L Wiedemair A Muller-Holzner E Goebel G Marth C Widschwendter M DNA methylation in serum and tumors of cervical cancer patients Clin Cancer Res 2004 10 565 571 14760078 10.1158/1078-0432.CCR-0825-03
Razani B Altschuler Y Zhu L Pestell RG Mostov KE Lisanti MP Caveolin-1 expression is down-regulated in cells transformed by the human papilloma virus in a p53-dependent manner. Replacement of caveolin-1 expression suppresses HPV-mediated cell transformation Biochemistry 2000 39 13916 13924 11076533 10.1021/bi001489b
Chan TF Su TH Yeh KT Chang JY Lin TH Chen JC Yuang SS Chang JG Mutational, epigenetic and expressional analyses of caveolin-1 gene in cervical cancers Int J Oncol 2003 23 599 604 12888893
Kim YI Giuliano A Hatch KD Schneider A Nour MA Dallal GE Selhub J Mason JB Global DNA hypomethylation increases progressively in cervical dysplasia and carcinoma Cancer 1994 74 893 899 8039116
Fowler BM Giuliano AR Piyathilake C Nour M Hatch K Hypomethylation in cervical tissue: is there a correlation with folate status? Cancer Epidemiol Biomarkers Prev 1998 7 901 906 9796635
de Capoa A Musolino A Della Rosa S Caiafa P Mariani L Del Nonno F Vocaturo A Donnorso RP Niveleau A Grappelli C DNA demethylation is directly related to tumour progression: evidence in normal, pre-malignant and malignant cells from uterine cervix samples Oncol Rep 2003 10 545 549 12684621
Watanabe S Iwasaka T Yokoyama M Uchiyama M Kaku T Matsuyama T Analysis of nuclear chromatin distribution in cervical glandular abnormalities Acta Cytol 2004 48 505 513 15296341
Bernstein BE Humphrey EL Erlich RL Schneider R Bouman P Liu JS Kouzarides T Schreiber SL Methylation of histone H3 Lys 4 in coding regions of active genes Proc Natl Acad Sci U S A 2002 99 8695 8700 12060701 10.1073/pnas.112318199
Fraga MF Ballestar E Villar-Garea A Boix-Chornet M Espada J Schotta G Bonaldi T Haydon C Ropero S Petrie K Iyer NG Pérez-Rosado A Calvo E López JA Cano A Calasanz MJ Colomer D Piris MA Ahn N Imhof A Caldas C Jenuwein T Esteller M Loss of acetylation at Lys16 and trimethylation at Lys20 of histone H4 is a common hallmark of human cancer Nature Genet 2005 37 391 400 15765097 10.1038/ng1531
Seligson DB Horvath S Shi T Yu H Tze S Grunstein M Kurdistani SK Global histone modification patterns predict risk of prostate cancer recurrence Nature 2005 435 1262 1266 15988529 10.1038/nature03672
Anton M Horky M Kuchtickova S Vojtesek B Blaha O Immunohistochemical detection of acetylation and phosphorylation of histone H3 in cervical smears Ceska Gynekol 2004 69 3 6 15112379
Vigushin DM Coombes RC Targeted histone deacetylase inhibition for cancer therapy Curr Cancer Drug Targets 2004 4 205 218 15032670 10.2174/1568009043481560
Longworth MS Laimins LA The binding of histone deacetylases and the integrity of zinc finger-like motifs of the E7 protein are essential for the life cycle of human papillomavirus type 31 Virology 2004 78 3533 3541 10.1128/JVI.78.7.3533-3541.2004
Brehm A Nielsen SJ Miska EA McCance DJ Reid JL Bannister AJ Kouzarides T The E7 oncoprotein associates with Mi2 and histone deacetylase activity to promote cell growth EMBO J 1999 18 2449 2458 10228159 10.1093/emboj/18.9.2449
Zimmermann H Degenkolbe R Bernard HU O'Connor MJ The human papillomavirus type 16 E6 oncoprotein can down-regulate p53 activity by targeting the transcriptional coactivator CBP/p300 J Virol 1999 73 6209 6219 10400710
Bernat A Massimi P Banks L Complementation of a p300/CBP defective-binding mutant of adenovirus E1a by human papillomavirus E6 proteins J Genet Virol 2002 83 829 833
Rainier S Relaxation of imprinted genes in human cancer Nature 1993 362 747 749 8385745 10.1038/362747a0
Ehrlich M Hypomethylation and hypermethylation of DNA in Wilms tumors Oncogene 2002 21 6694 6702 12242669 10.1038/sj.onc.1205890
Brown KW Williams JC Maitland NJ Mott MG Genomic imprinting and the Beckwith-Wiedemann syndrome Am J Hum Genet 1990 46 1000 1001 2339685
Douc-Rasy S Barrois M Fogel S Ahomadegbe JC Sthelin D Coll Riou G High incidence of loss of heterozygosity and abnormal imprinting of H19 and IGF2 genes in invasive cervical carcinomas. Uncoupling of H19 and IGF2 expression and biallelic hypomethylation of H19 Oncogene 1996 12 423 430 8570220
Clarke EA Anderson TW Does screening by "Pap" smears help prevent cervical cancer? A case-control study Lancet 1979 2 1 4 87887 10.1016/S0140-6736(79)90172-7
Fahey MT Irwing L Macaskill P Meta-analysis of pap test accuracy Am J Epidemiol 1995 141 680 689 7702044
Nieh S Chen SF Chu TY Lai HC Lin YS Fu E Gau CH Is p16(INK4A) expression more useful than human papillomavirus test to determine the outcome of atypical squamous cells of undetermined significance-categorized Pap smear? A comparative analysis using abnormal cervical smears with follow-up biopsies Gynecol Oncol 2005 97 35 40 15790434 10.1016/j.ygyno.2004.11.034
Kalof AN Evans MF Simmons-Arnold L Beatty BG Cooper K p16INK4A immunoexpression and HPV in situ hybridization signal patterns: potential markers of high-grade cervical intraepithelial neoplasia Am J Surg Pathol 2005 29 674 679 15832093 10.1097/01.pas.0000155164.78785.c2
Feng Q Balasubramanian A Hawes SE Toure P Sow PS Dem A Dembele B Critchlow CW Xi L Lu H McIntosh MW Young AM Kiviat NB Detection of hypermethylated genes in women with and without cervical neoplasia Natl Cancer Inst 2005 97 273 282
Stroun M Anker P Maurice P Lyautey J Lederrey C Beljanski M Neoplastic characteristics of the DNA found in the plasma of cancer patients Oncology 1989 46 318 322 2779946
Sorenson GD Pribish DM Valone FH Memoli VA Bzik DJ Yao SL Soluble normal and mutated DNA sequences from single-copy genes in human blood Cancer Epidemiol Biomarkers Prev 1994 3 67 71 8118388
Vasioukhin V Anker P Maurice P Lyautey J Lederrey C Stroun M Point mutations of the N-ras gene in the blood plasma DNA of patients with myelodysplastic syndrome or acute myelogenous leukemia Br J Haematol 1994 86 774 779 7918071
Anker P Stroun M Progress in the knowledge of circulating nucleic acids: plasma RNA is particle-associated. Can it become a general detection marker for a cancer blood test? Clin Chem 2002 48 1210 1211 12142375
Trejo-Becerril C Pérez-Cárdenas E Treviño-Cuevas H Taja-Chayeb L García-López P Segura-Pacheco B Chávez-Blanco A Lizano-Soberón M González-Fierro A Mariscal I Wegman-Ostrosky T Dueñas-González A Circulating nucleosomes and response to chemotherapy: an in vitro, in vivo and clinical study on cervical cancer patients Int J Cancer 2003 104 663 668 12640671 10.1002/ijc.11003
Herman JG Circulating methylated DNA Ann N Y Acad Sci 2004 1022 33 39 15251936 10.1196/annals.1318.006
Dulaimi E Uzzo RG Greenberg RE Al-Saleem T Cairns P Detection of bladder cancer in urine by a tumor suppressor gene hypermethylation panel Clin Cancer Res 2004 10 1887 1893 15041703 10.1158/1078-0432.CCR-03-0127
Wang YC Lu YP Tseng RC Lin RK Chang JW Chen JT Shih CM Chen CY Inactivation of hMLH1 and hMSH2 by promoter methylation in primary non-small cell lung tumors and matched sputum samples J Clin Invest 2003 111 887 895 12639995 10.1172/JCI200315475
Yang HJ Liu VW Wang Y Chan KY Tsang PC Khoo US Cheung AN Ngan HY Detection of hypermethylated genes in tumor and plasma of cervical cancer patients Gynecol Oncol 2003 93 435 440 15099958 10.1016/j.ygyno.2004.01.039
Jones PA Taylor SM Cellular differentiation, cytidine analogs and DNA methylation Cell 1980 20 85 93 6156004 10.1016/0092-8674(80)90237-8
Abele R Clavel M Dodion P Bruntsch U Gundersen S Smyth J Renard J van Glabbeke M Pinedo HM The EORTC Early Clinical Trials Cooperative Group experience with 5-aza-2'-deoxycytidine (NSC 127716) in patients with colo-rectal, head and neck, renal carcinomas and malignant melanomas Eur J Cancer Clin Oncol 1987 23 1921 1924 2449354 10.1016/0277-5379(87)90060-5
Issa JP Decitabine Curr Opin Oncol 2003 15 446 451 14624227 10.1097/00001622-200311000-00007
Chan AT Tao Q Robertson KD Flinn IW Mann RB Klencke B Kwan WH Leung TW Johnson PJ Ambinder RF Azacytidine induces demethylation of the Epstein-Barr virus genome in tumors J Clin Oncol 2004 22 1373 1381 15007085 10.1200/JCO.2004.04.185
Goffin J Eisenhauer E DNA methyltransferase inhibitors- state of the art Ann Oncol 2002 13 1699 116 12419742 10.1093/annonc/mdf314
Beaulieu N Fournel M Macleod A Antitumor activity of MG98, an antisense oligonucleotide targeting DNA methyltransferase-1 (DNMT1). [Abstract] Clin Cancer Res 2001 7 3800S
Davis AJ Moore MJ Gelmon KA Phase 1 and pharmacodynamic study of human DNA methyltransferase (MeTase) antisense oligodeoxynucleotide (ODN), MG98, administered as 21-day infusion q4 weekly [Abstract] Clin Cancer Res 2000 6 4517S
Segura-Pacheco B Trejo-Becerril C Pérez-Cárdenas E Taja-Chayeb L Mariscal I Chávez A Acuña C Salazar AM Lizano M Dueñas-González A Reactivation of tumor suppressor genes by the cardiovascular drugs hydralazine and procainamide and their potential use in cancer therapy Clin Cancer Res 2003 9 1596 603 12738711
Angeles E Vázquez-Valadez VH Vázquez-Valadez O Velázquez-Sánchez AM Ramírez A Martínez L Díaz-Barriga S Romero-Rojas A Cabrera G López-Castañares R Dueñas-González A Computational studies of 1-Hydrazinophtalazine (Hydralazine) as antineoplastic agent. Docking studies on methyltransferase Lett Drug Design Discov 2005 4 282 286 10.2174/1570180054038413
Shao Y Gao Z Marks PA Jiang X Apoptotic and autophagic cell death induced by histone deacetylase inhibitors Proc Natl Acad Sci U S A 2004 101 18030 18035 15596714 10.1073/pnas.0408345102
Knupfer MM Pulzer F Schindler I Hernaiz Driever P Knupfer H Keller E Different effects of valproic acid on proliferation and migration of malignant glioma cells in vitro Anticancer Res 2001 21 347 351 11299760
Kim MS Blake M Baek JH Kohlhagen G Pommier Y Carrier F Inhibition of histone deacetylase increases cytotoxicity to anticancer drugs targeting DNA Cancer Res 2003 63 7291 7300 14612526
Zhang Y Jung M Dritschilo A Jung M Enhancement of radiation sensitivity of human squamous carcinoma cells by histone deacetylase inhibitors Radiat Res 2004 161 667 674 15161353
Drummond DC Noble CO Kirpotin DB Guo Z Scott GK Benz CC Clinical development of histone deacetylase inhibitors as anticancer agents Annu Rev Pharmacol Toxicol 2005 45 495 528 15822187 10.1146/annurev.pharmtox.45.120403.095825
Kelly WK O'Connor OA Krug LM Chiao JH Heaney M Curley T MacGregore-Cortelli B Tong W Secrist JP Schwartz L Richardson S Chu E Olgac S Marks PA Scher H Richon VM Phase I study of an oral histone deacetylase inhibitor, suberoylanilide hydroxamic acid, in patients with advanced cancer J Clin Oncol 2005 23 3923 3931 15897550 10.1200/JCO.2005.14.167
Ryan QC Headlee D Acharya M Sparreboom A Trepel JB Ye J Figg WD Hwang K Chung EJ Murgo A Melillo G Elsayed Y Monga M Kalnitskiy M Zwiebel J Sausville EA Phase I and pharmacokinetic study of MS-275, a histone deacetylase inhibitor, in patients with advanced and refractory solid tumors or lymphoma J Clin Oncol 2005 23 3912 3922 15851766 10.1200/JCO.2005.02.188
Prakash S Foster BJ Meyer M Wozniak A Heilbrun LK Flaherty L Zalupski M Radulovic L Valdivieso M LoRusso PM Chronic oral administration of CI-994: a phase 1 study Invest New Drugs 2001 19 1 11 11291827 10.1023/A:1006489328324
Reid T Valone F Lipera W Irwin D Paroly W Natale R Sreedharan S Keer H Lum B Scappaticci F Bhatnagar A Phase II trial of the histone deacetylase inhibitor pivaloyloxymethyl butyrate (Pivanex, AN-9) in advanced non-small cell lung cancer Lung Cancer 2004 45 381 386 15301879 10.1016/j.lungcan.2004.03.002
Shi H Wei SH Leu YW Rahmatpanah F Liu JC Yan PS Nephew KP Huang TH Triple analysis of the cancer epigenome: an integrated microarray system for assessing gene expression, DNA methylation, and histone acetylation Cancer Res 2003 63 2164 2171 12727835
Cameron EE Bachman KE Myohanen S Herman JG Baylin SB Synergy of demethylation and histone deacetylase inhibition in the re-expression of genes silenced in cancer Nature Genet 1999 21 103 107 9916800 10.1038/5047
Primeau M Gagnon J Momparler RL Synergistic antineoplastic action of DNA methylation inhibitor 5-AZA-2'-deoxycytidine and histone deacetylase inhibitor depsipeptide on human breast carcinoma cells Int J Cancer 2003 103 177 184 12455031 10.1002/ijc.10789
Bovenzi V Momparler RL Antineoplastic action of 5-aza-2'-deoxycytidine and histone deacetylase inhibitor and their effect on the expression of retinoic acid receptor beta and estrogen receptor alpha genes in breast carcinoma cells Cancer Chemother Pharmacol 2001 48 71 76 11488527 10.1007/s002800100294
Keen JC Yan L Mack KM Pettit C Smith D Sharma D Davidson NE A novel histone deacetylase inhibitor, scriptaid, enhances expression of functional estrogen receptor alpha (ER) in ER negative human breast cancer cells in combination with 5-aza 2'-deoxycytidine Breast Cancer Res Treat 2003 81 177 186 14620913 10.1023/A:1026146524737
Zhu WG Lakshmanan RR Beal MD Otterson GA DNA methyltransferase inhibition enhances apoptosis induced by histone deacetylase inhibitors Cancer Res 2001 61 1327 1333 11245429
Vermorken JB Tumolo S Roozendaal KJ Guastalla JP Splinter TA Renard J 5-aza-2'-deoxycytidine in advanced or recurrent cancer of the uterine cervix Eur J Cancer 1991 27 5216 217
Nyce JW Drug-induced DNA hypermethylation: a potential mediator of acquired drug resistance during cancer chemotherapy Mutat Res 1997 386 153 161 9113116 10.1016/S1383-5742(96)00051-8
Pohlmann P DiLeone LP Cancella AI Caldas AP Dal Lago L Campos O JrMonego E Rivoire W Schwartsmann G Phase II trial of cisplatin plus decitabine, a new DNA hypomethylating agent, in patients with advanced squamous cell carcinoma of the cervix Am J Clin Oncol 2002 25 496 501 12393992 10.1097/00000421-200210000-00015
Munster PN Marchion DC Bicaku E Sullivan P Beam C Mahany J Lush R Sullivan DM Daud A Phase I trial of the histone deacetylase inhibitor, valproic acid and the topoisomerase II inhibitor, epirubicin: a clinical and translational study [Abstract] Procc ASCO 2005 24
Chávez-Blanco A Segura-Pacheco B Pérez-Cárdenas E Taja-Chayeb L Cetina L Candelaria M Cantú D Pérez-Plasencia C Cabrera G Trejo-Becerril C Angeles E González-Fierro A García-López P Zambrano P Dueñas-González A Histone acetylation and histone deacetylase activity of magnesium valproate in tumor and peripheral blood of patients with cervical cancer. A phase I study Molec Cancer 2005 4 22 16001982 10.1186/1476-4598-4-22
Forrer C Hofstetter B Pruschy M Ciernik IF Treatment with demethylating agent 5-Aza-2'-Deoxy-Cytidine (DAC) leads to enhanced sensitivity to ionizing radiation (IR) in colorectal carcinoma [Abstract] Radiat Oncol 2004 73 S1 S497
Dote H Cerna D Burgan WE Carter DJ Cerra MA Hollingshead MG Camphausen K Tofilon PJ Enhancement of in vitro and in vivo tumor cell radiosensitivity by the DNA methylation inhibitor zebularine Clin Cancer Res 2005 11 4571 4579 15958643 10.1158/1078-0432.CCR-05-0050
Zhang Y Jung M Dritschilo A Jung M Enhancement of radiation sensitivity of human squamous carcinoma cells by histone deacetylase inhibitors Radiat Res 2004 161 667 674 15161353
Camphausen K Cerna D Scott T Sproull M Burgan WE Cerra MA Fine H Tofilon PJ Enhancement of in vitro and in vivo tumor cell radiosensitivity by valproic acid Int J Cancer 2005 114 380 386 15578701 10.1002/ijc.20774
Chung YL Wang AJ Yao LF Antitumor histone deacetylase inhibitors suppress cutaneous radiation syndrome: implications for increasing therapeutic gain in cancer radiotherapy Molec Cancer Ther 2004 3 317 325 15026552
|
16248899
|
PMC1291396
|
CC BY
|
2021-01-04 16:36:35
|
no
|
Mol Cancer. 2005 Oct 25; 4:38
|
utf-8
|
Mol Cancer
| 2,005 |
10.1186/1476-4598-4-38
|
oa_comm
|
==== Front
Mol PainMolecular Pain1744-8069BioMed Central London 1744-8069-1-321626642910.1186/1744-8069-1-32Short ReportSingle subject pharmacological-MRI (phMRI) study: Modulation of brain activity of psoriatic arthritis pain by cyclooxygenase-2 inhibitor Baliki M [email protected] J [email protected] DR [email protected] AV [email protected] Departments of Physiology, Northwestern University Feinberg School of Medicine, Chicago, Illinois, 60611, USA2 Anesthesiology, Northwestern University Feinberg School of Medicine, Chicago, Illinois, 60611, USA2005 2 11 2005 1 32 32 9 9 2005 2 11 2005 Copyright © 2005 Baliki et al; licensee BioMed Central Ltd.2005Baliki et al; licensee BioMed Central Ltd.This is an Open Access article distributed under the terms of the Creative Commons Attribution License (), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
We use fMRI to examine brain activity for pain elicited by palpating joints in a single patient suffering from psoriatic arthritis. Changes in these responses are documented when the patient ingested a single dose of a selective cyclooxygenase-2 inhibitor (COX-2i). We show that mechanical stimulation of the painful joints exhibited a cortical activity pattern similar to that reported for acute pain, with activity primarily localized to the thalamus, insular, primary and secondary somatosensory cortices and the mid anterior cingulum. COX-2i resulted in significant decreased in reported pain intensity and in brain activity after 1 hour of administration. The anterior insula and SII correlated with pain intensity, however no central activation site for the drug was detected. We demonstrate the similarity of the activation pattern for palpating painful joints to brain activity in normal subjects in response to thermal painful stimuli, by performing a spatial conjunction analysis between these maps, where overlap is observed in the insula, thalamus, secondary somatosensory cortex, and anterior cingulate. The results demonstrate that one can study effects of pharmacological manipulations in a single subject where the brain activity for a clinical condition is delineated and its modulation by COX-2i demonstrated. This approach may have diagnostic and prognostic utility.
==== Body
Introduction
Over the last fifteen years, functional MRI (fMRI) and positron emission tomography (PET) have been used to unravel brain circuitry underlying pain perception and study the properties of these areas in acute and chronic pain conditions (for a recent review see [1]). Recently, the utility of combining fMRI with pharmacology has been demonstrated by a number of groups [2-6]. Brain imaging studies in combination with various analgesics have also been described regarding the impact of examined chemicals on brain activity for pain [7-11]. These studies examine acute experimental pain conditions, and demonstrate the possibility of delineating brain regions modulated in normal subjects, for centrally acting drugs, such as opiates and ketamine. In the present study we show the potential of studying clinical pain conditions, and tracking the efficacy of pharmacological interventions in an individual patient, where multiple repeat scans are performed before and after administering a single dose of the analgesic that the patient was using to manage satisfactorily his arthritis.
We examine the analgesic efficacy of a selective cyclooxygenase-2 (COX-2) inhibitor on psoriatic arthritis pain. Traditional nonsteroidal anti-inflammatory drugs (NSAIDs) are nonselective inhibitors of COX-1 and COX-2, which catalyze transformation of arachadonic acid to prostaglandin. Substantial clinical evidence shows that COX-2 selective inhibitors are effective for treating osteoarthritis, rheumatoid arthritis, and other inflammatory pain conditions [12]. Differential elevation of COX-2 has been documented in synovial tissue of patients with inflammatory arthritis, including patients with psoriatic arthritis [13], and in animal studies of inflammatory pain COX-2 elevation is observed in the periphery and in the central nervous system [14]. Thus, there is good scientific evidence for management of inflammation and pain of psoriatic arthritis with selective COX-2 inhibitors (COX-2i). Here we examine the effects of a single dose of a COX-2i on brain activity for joint pressure allodynia associated with psoriatic pain.
Results
A single subject with psoriatic pain was studied. The subject skipped a single dose of his COX-2i 12 hours prior to the scanning session. Brain activity was performed for palpating the painful joints in 4 fMRI scans prior to administration of COX-2i, in 4 fMRI scans 1 hour post drug ingestion, and in 2 fMRI scans 3 hours post drug ingestion.
COX-2i treatment decreased pain
A single 200 mg dose of selective COX-2i reduced baseline pain, joint stimulation pain, and restored ability to ambulate. At the start of the study the patient was not able to stand on his legs due to severe ankle joint pain. One hour after ingesting the medicine he still had very limited mobility. After three hours, he was able to walk. Left panel of Figure 1A shows an example rating of pain when joints of the hand are palpated (collected during an fMRI session, prior to ingesting the COX-2i), where the average baseline pain is about 3 and stimulus-evoked pain is about 6 (on a 0–10 pain scale, 10 = maximum imaginable pain). The medication decreased baseline pain by 50% (over 4 hours post-treatment); stimulus evoked pain from 7.93 +/- 0.73 to 3.15 +/- 0.55 (1 hour post COX-2i, p < 0.001), to 2.2 +/- 0.12 (3 hours post COX-2i, p < 0.001 (right panel of figure 1A).
Figure 1 A. Example pain intensity rating for manually palpating painful arthritic joints (grey areas is duration of palpation). Right panel shows that the mean rating of pain intensity for palpating painful joints, which significantly decreased 1 and 3 hours after administration of COX-2i when compared to the baseline (p < 0.001). B. Brain activity in ten consecutive scans is presented for palpating painful joints (each column is a separate scan). The first 4 scans are prior to drug ingestion (pre COX-2i), the next four start 1 hour post drug ingestion, the last two are 3 hours post-drug ingestion. Four transverse slices are shown for each scan (top to bottom: z = -12, 10, 28 and 44). Note that brain activity decrease after cox-2i is evident in these individual scans. C. Average group activity minus visual controls using fixed effects analysis for palpating painful joints across all conditions (pre and post COX-2i). Activity is seen bilaterally in insular cortex and thalamus, in addition to right primary and secondary somatosensory areas and mid anterior cingulum. Activity maps are presented in MNI space, x y z coordinates in mm.
COX-2i treatment decreased pain related brain activity
We observe widespread brain activity for palpating painful joints prior to drug administration. This activity was significantly attenuated across all areas 1 hour after ingestion of COX-2i, and there was no significant brain activity after 3 hours of drug ingestion (figure 1B).
Overall brain activity for palpating painful joints was determined by averaging all 10 scans, prior and post-drug ingestion. The activity mainly included areas coding the sensory properties of the acute painful stimulus, including bilateral anterior and posterior insula together with secondary somatosensory cortex, multiple portions of anterior cingulate, primary somatosensory cortex, basal ganglia, and thalamus (Figure 1C, Table I).
Table 1 Brain regions activated for palpating painful joints in a psoriatic arthritis patient
Region Coordinates Z value Cluster Index Voxels P value
x y z
R sup temporal pole 46 6 -6 6.24 10 835 p < 10-14
R ant insula * 40 20 -6 3.98 10
R inf parietal, S2 * 42 -22 16 3.93 10
L ant insula * -42 22 -2 4.66 9 321 p < 10-06
L ant insula/inf frontal (45/47) -48 20 -6 4.52 9
R ant thalamus 10 -4 6 3.65 8 258 p < 10-05
R putamen 26 6 -4 2.70 8
R S1 hand (3) * 44 -20 48 5.43 7 210 p < 10-04
R S1/M1 hand (3/4) 38 -28 68 4.08 7
R mid temporal 68 -22 8 4.56 6 203 0.000117
bi lingual -4 -66 10 4.57 5 200 0.000136
R thalamus * 16 -22 8 4.56 4 187 0.000263
mid ACC (24) * 2 22 34 4.48 3 158 0.00122
R SMA (6) 10 -6 52 3.68 2 135 0.00439
L thalamus * -12 -24 6 4.01 1 123 0.0088
L = left; R = right; bi = bilateral; mid = middle; ant = anterior; inf = inferior; post = posterior; sup = superior ; S1 = primary somatosensory cortex; S2 = secondary somatosensory cortex; M1 = primary motor cortex; ACC = anterior cingulate; SMA = supplementary motor area; numbers in parenthesis = Brodmann areas. Activity for brain areas marked by * are shown in Figure 1C.
Brain areas that correlated with pain intensity ratings are shown in figure 2. The main areas include anterior insular cortex, and secondary sensory/posterior insular cortex. The average BOLD signal in these two areas shows the response profile of brain activity in each region and the decrease in magnitude of this response after drug ingestion. In both regions, magnitude of activity (Z-values) is strongly correlated with pain ratings (figure 2 last two panels).
Figure 2 Higher-level covariate analysis between pain intensity for palpating joints (mean rating of pain during scan) and brain activity across all 10 scans results in two main clusters: left insular cortex (red) R2 = 0.78 and right secondary somatosensory area (blue) R2 = 0.77. Middle panels show the average change in BOLD signal during stimulation (mean stimulus duration = 13.3 seconds) for these areas prior to, and 1 and 3 hours after the administration of COX-2i. Right panels show the relationship between activity (Z score) and pain for a 1 cc volume centered at the peak of the covariate analysis, for each region.
It is worthy to note that no brain areas showed significant increases in activity after the administration of drug (i.e. the contrast: post – pre drug ingestion scans yielded no positive results, and no brain areas were negatively correlated with pain). Thus, we could not identify any brain area responding to the drug.
Overlap between brain activity in the patient with psoriatic pain and acute thermal pain in normal subjects
The average brain activity pattern for palpating joints in the psoriatic arthritis patient closely resembles brain activity observed for acute pain, as reported by many groups in the past, see recent review [1]. To directly demonstrate this similarity we compared the spatial activity pattern in this patient to group-averaged brain activity in normal subjects following thermal painful stimulation. This similarity is illustrated in figure 3. The overlap between these activation maps included bilateral thalamus and insula, anterior cingulate, and secondary somatosensory cortex.
Figure 3 Similarity of brain activity for psoriatic pain in a single subject to thermal pain in a group of normal subjects. Top panel: The close proximity of brain activity is shown for the entire brain. Average group activity maps for thermal painful stimuli in 7 healthy control subjects (red) and for palpating painful joints in one patient (blue). Conjunction analysis between heat pain in healthy normal subjects and psoriatic arthritis pain indicates regions of overlap (outlined in black in lower panels), including: bilateral insula (z = -8 to +4) and thalamus (z = +4), ACC (z = 34), and right S2 region (z = +10 to +22\).
Discussion
The study demonstrates feasibility of studying effects of a single dose of an analgesic on brain activity for a clinical pain condition in an individual subject. We show that even in a single subject using repeated fMRI scans it is possible to demonstrate the efficacy of an analgesic in relation to brain activity, to identify the regions involved in the pain, and the regions specifically modulated by the analgesic.
Peripheral vs. central actions of COX-2 inhibitors remains unknown in man, although there is evidence for their central efficacy in rodents [14]. We could not identify brain regions responding to the drug, independent of pain modulation, implying that the drug-induced analgesia is probably not mediated through active inhibition at the level of the cortex. However, this negative finding may also be due to many other reasons (like insufficient power), especially since this is a single subject study, and needs to be determined in a larger population of patients.
The pattern of brain activity observed for the psoriatic arthritis joint pain is similar to brain activity observed by many groups for acute pain, and especially for acute thermal painful stimuli, see for example [15-18], and many others [1]. A recent meta-analysis of human brain imaging studies regarding regional activity for acute pain indicates that incidence of activity for anterior cingulate, primary somatosensory, secondary somatosensory, and insular cortices, as well as the thalamus ranges from 75% to 94% [1]. Consistent with these observations recent rodent studies indicate that manipulating activity in anterior cingulate and insula can modify pain-like behavior [19-21]. All of these regions were identified as activated in our psoriatic arthritis patient, and most of them (all but primary somatosensory cortex) showed overlap of activity between psoriatic pain and thermal pain in normal subjects. Anterior insula and secondary somatosensory cortex/posterior insula were the two brain regions that robustly reflected changes in pain perception as a consequence of ingesting the COX-2i. A number of groups have studied brain regions specifically coding painful stimuli for acute pain [16,22,23]. The study by Bornhovd et al. [22] examined laser stimulation evoked responses and differentiated between detection of painful stimuli vs. coding of painful stimuli, and showed that anterior insula and secondary somatosensory cortex are the ones that best code pain intensity. Thus, the psoriatic pain modulation by COX-2i closely agrees with the results of Bornhovd et al. [22]. In contrast, our recent study of brain activity in chronic back pain patients with radiculopathy, and using a data collection and analysis approach identical to the present study, indicated that back pain intensity was represented in medial prefrontal cortex activity, while duration of back pain was reflected by anterior insula activity (preliminary report by Baliki et al. Society Neuroscience Abstract 2003). Therefore, at least in the subject that we have studied, the psoriatic arthritis joint pain seems more like acute thermal pain and less like chronic neuropathic back pain. This result is consistent with the clinical assessment of psoriatic arthritis being more an inflammatory condition than neuropathic [13] and the patient's responsiveness to the COX-2i is in agreement with animal studies showing that COX-2 expression and responsiveness to COX-2i seems specific to inflammatory rather than neuropathic pain conditions [14,24], but of course this notion remains to be demonstrated in a population of psoriatic arthritis patients.
In conclusion, this case study has important clinical implications in making judgments as to the efficacy of pharmacotherapies on brain circuitry for clinical pain conditions that a patient may be suffering from, and demonstrates a method that can identify brain regions that may differentiate between patients with similar clinical conditions but with different responses to the same pharmacotherapy. In general, the methodology provides an objective approach that may be used for drug development and testing effects of drugs in individual cases.
Materials and methods
A single patient with psoriatic arthritis pain in multiple joints was recruited to the study. An additional 7 healthy subjects were studied for thermal pain. Subjects were instructed regarding the procedure and signed a consent form. Institutional Review Board of Northwestern University approved the procedures.
The patient agreed to skip one dose of the analgesic and anti-inflammatory selective COX-2i (200 mg Celebrex every 12 hours, Pfizer) 12 hours prior to the start of the study. The subject knew that this would dramatically limit his mobility and requested help in transportation from home to the scanner. The subject rated his pain on a verbal analog scale (0 = no pain, 10 = most imaginable pain) at the start of the study and at different time intervals throughout the course of the study. The participant's ability to ambulate was tested at the start and at various intervals during the study, as an alternate measure for knee joint pain. He underwent 10 joint palpation scans, 4 prior and 6 after ingesting the analgesic, 2 fMRI scans for visual-motor controls, and anatomic scans for proper registration of the fMRI scans to standard space.
Brief outline of case history
The subject was a 53 year old male with a past medical history significant for plantar-palmar psoriasis and mild gastric reflux from a small hiatal hernia. Past medical history was notable only for a lack of any orthopedic procedures on any joints. The psoriasis had been present since age 39, but 7 months prior to his participation in this investigation, the psoriasis became acutely more severe but remained purely a dermatologic problem. At 3 months prior to the investigation, the patient noted an acute onset of swelling in the left knee associated with significant pain on movement. This was clinically diagnosed as psoriatic arthritis, and within weeks involved all joints of the extremities with severe swelling and pain with movement. Further, pain along the entire spine was noted. The severity of the pain significantly interfered with sleep and with all activities of daily living. Daily doses of ibuprofen of 3.2 g and of naprosyn 2.4 g failed to control the pain. After 3 months of poorly controlled pain, he was started on celecoxib 200 mg bid. Within 1 hour of the first dose, profound relief was noted, and this relief was sustained throughout the use of the medication allowing him to sleep and restore full activities of daily living. However, despite the pain control with the celecoxib, joint swelling was unchanged.
Brain Imaging
Brain activity was studied with fMRI while the patient rated fluctuations of ongoing pain using a finger-spanning device to continuously rate and log subjective perception of pain during fMRI data collection [25]. The patient underwent an initial training phase prior to scanning, in which he learned to use the finger-span device. During brain imaging sessions, the finger span device is synchronized and time locked with fMRI image acquisition and connected to a computer providing visual feedback. This device was also used in performing visual control scans.
Rating pain during stimulation of arthritic joints
The patient rated the stimulus pain with the finger-span device. During a given functional imaging session, 10 stimuli ranging in duration from 10 to 15 seconds were applied to the arthritic limbs and digits. (see fig 1A). The patient rated pain on a scale of 0 to 10 and was provided with a visual feedback of the rating. The stimulus was delivered manually by an investigator who applied variable duration and intensity pressures on the painful joints. Stimulus intensity was non-painful when applied on the investigator but was always rated painful (of variable intensity) by the patient. In some scans stimuli were applied to the hand joints (left and right), in others to the kneecap (left and right). Given the small number of fMRI scans, we did not attempt to distinguish between stimulus sites (neither body side nor upper vs. lower limb distinctions were tested).
Rating pain during thermal stimulation in healthy normals
Healthy volunteers (n = 7; average age = 30.3 ± 3.2) were scanned during acute thermal stimulation where they rated the stimulus pain with the finger-span device. During a given functional imaging session, 9 noxious thermal stimuli ranging in duration from 10 to 30 seconds were applied to the healthy subjects. The stimuli were applied both on the lower back area and volar surface of the hands. Subjects were instructed to rate their pain on a scale of 0 to 10 and were provided with a visual feedback of their rating (average reported pain across all subjects and runs was 5.0 ± 1.6). A purpose built, fMRI compatible thermal stimulator delivered fast ramping (10°C per second) painful thermal stimuli (baseline 38°C, peak temperatures 46°C and 48°C) via a contact probe (1 × 1.5 cm peltier). Durations and intensities of thermal stimuli as well as inter-stimulus intervals were presented in a pseudorandom fashion. This variation in interval was adopted to decrease the regularity of stimulus presentation and thus reduce volunteers' ability to predict the arrival of the next stimulus. In addition, the relatively long duration inter-stimulus interval reduced sensitization.
Visual control
The patient and normal subjects were instructed to follow as closely as possible fluctuations of a bar projected on a screen in time. This visual tracking provides an adequate visual-motor control since it is similar to the pain rating finger-span task, with the important difference being that now the finger movement (i.e. variations in magnitude) is correlated with a visual input rather than pain.
Functional Magnetic Resonance Data Analysis
Functional MR data for the arthritic patient was acquired with a 1.5T Siemens (VISION) whole-body scanner with echo-planar imaging (EPI) capability using the standard radio-frequency head coil. Multi-slice T2*-weighted echo-planar images were obtained with the following parameters: repetition time TR = 3.5 s, echo time TE = 70 ms, flip angle = 90°, slice thickness = 3 mm, in-plane resolution = 3.475 × 3.475 mm2. The 36 slices covered the whole brain from the cerebellum through to the vertex. An average of 240 volumes were acquired per event per condition in all participants. A No-Flow T1-weighted anatomical MRI image was also using the following parameters: TR = 22 ms, TE = 5.6 ms, flip angle = 20°, matrix 256 × 256 and a FOV of 240 mm, with 160 mm coverage in the slice direction. Anatomic images were used to register functional images in standard space.
Functional MR data for thermal pain in normal healthy subjects were acquired with a 3T Siemens Trio whole-body scanner with echo-planar imaging (EPI) capability using the standard radio-frequency head coil. Multi-slice T2*-weighted echo-planar images were obtained with the following parameters: repetition time TR = 2.5 s, echo time TE = 70 ms, flip angle = 90°, slice thickness = 3 mm, in-plane resolution = 3.475 × 3.475 mm2. The 36 slices covered the whole brain from the cerebellum to the vertex. An average of 400 volumes were acquired per event per condition in all participants. A T1-weighted anatomical MRI image was also acquired for each subject using the following parameters: TR = 2.1 s, TE = 4.38 ms, flip angle = 8°, FOV = 220 mm, slice thickness = 1 mm, in-plane resolution = 0.86 × 0.86 mm2 and number of sagittal slices = 160.
Image analysis to reveal significant brain activity based on changes in BOLD signal was performed using FMRIB Expert Analysis Tool (FEAT, [26], ). The preprocessing of each subject's individual scan time-series of fMRI volumes encompassed: slice time correction; motion correction using MCFLIRT; spatial smoothing using a Gaussian kernel of full-width-half-maximum 5 mm; nonlinear high-pass temporal filtering (Gaussian-weighted least-squares straight line fitting, filter cutoff of 100 seconds) and subtraction of the mean of each voxel time-course from that time-course (i.e. intensity normalization). The fMRI signal was then linearly modeled on a voxel by voxel basis using FMRIB's Improved Linear Model (FILM) with local autocorrelation correction [27,28].
Each condition described above (thermal pain, arthritic pain and visual ratings) was considered to generate a hemodynamic response described by the convolution of the corresponding finger span rating with a generalized hemodynamic response function (gamma function: lag = 6 seconds, standard deviation = 3 seconds). Head motion vector (derived from motion correction) was used at this level as a covariate of no interest to further remove any residual variance due to head motion. The significance of the model fit to each voxel time-series was calculated, yielding statistical parametric maps for each scan. Average group activity maps were generated by subtracting the visual control scans from the pain activity maps using FEAT in a second level random and fixed effects group analysis, following the co-registration of individual scans to standard space (152 subject average Montreal Neurological Institute (MNI) space, ). This results in a Z-score map of statistically significant pain-related activity across different conditions. For random-effects, cluster-based correction of the Z-statistic images was performed. The raw Z-statistic images from the group analysis were thresholded at Z-scores > 2.3. For each resulting cluster of spatially connected voxels surviving the Z threshold, a cluster probability threshold of P = 0.01 was applied to the computed significance of that cluster, which corrects for multiple comparisons [29].
Stereotactic coordinates of local maxima within areas of significant activity change were determined for each condition. The localization of these local maxima and clusters was determined and assessed by reference to a standard stereotactic atlas [30] and reported in MNI coordinates.
Covariate analysis for arthritic pain was performed for all stimulation scans with mean pain rating of each scan. This is a higher level analysis where first level brain activity maps are used to determine brain regions that correlate across all scans with the corresponding pain intensity. The resultant map shows clusters of voxels that significantly covary with pain intensity across scans. Based on the results of this analysis, we perform a regional correlation analysis using the mean Z-value of a 27-voxel (1 cc) volume centered at the peak within a cluster of interest, relating this value to pain intensity across scans.
Central activation by COX 2i was investigated by a higher level analyses that was performed to determine brain areas that shows increased activity post COX-2i. A simple linear model was utilized in the set up of this covariate analyses: scans performed 3 hours post, 1 hour post and pre COX-2i were regressed with 1,0,-1 respectively. No brain areas exhibited significant corrolation (i.e. no voxels had a linear increase in activity across the 3 sessions).
Average time-course of BOLD response for arthritic pain was calculated for each fMRI scan by and then averaged across session. Pre-processed fMRI images are re-averaged relative to the start time of stimuli, 2 brain volumes (7 seconds) prior to the start time and 9 (31.5 seconds) post-start time. The average duration of stimuli across all scans is 13.3 seconds. Time-courses for coordinates of interest are then extracted from each scan, averaged over the 27-voxel neighborhood, and then averaged across sessions. Variability of these curves is expressed in S.E.M. over the scans. These BOLD responses show the time-course of activity in a given brain region relative to the stimulus, and indicate its changes at different time points from drug ingestion.
The thermal pain responses in normal subjects was used to indicate the spatial similarity between this map and the map obtained for psoriatic pain. Overlap between the two maps was identified using conjunction analysis, where a spatial mask is generated from the group-averaged thermal map, thresholded as Z-value > 2.3. The intersection between this mask and brain activity for psoriatic pain (averaged across all scans) determines brain regions commonly activated in both conditions. It is important to note that no statistical comparisons were made for intensity of activity between maps for thermal stimulation in normal healthy volunteers and arthritic pain since each data were collected on different strength magnets.
==== Refs
Apkarian AV Bushnell MC Treede RD Zubieta JK Human brain mechanisms of pain perception and regulation in health and disease Eur J Pain 2005 9 463 484 15979027 10.1016/j.ejpain.2004.11.001
Furey ML Pietrini P Haxby JV Cholinergic enhancement and increased selectivity of perceptual processing during working memory Science 2000 290 2315 2319 11125148 10.1126/science.290.5500.2315
Hariri AR Mattay VS Tessitore A Fera F Smith WG Weinberger DR Dextroamphetamine modulates the response of the human amygdala Neuropsychopharmacology 2002 27 1036 1040 12464460 10.1016/S0893-133X(02)00373-1
Honey GD Bullmore ET Soni W Varatheesan M Williams SC Sharma T Differences in frontal cortical activation by a working memory task after substitution of risperidone for typical antipsychotic drugs in patients with schizophrenia Proc Natl Acad Sci U S A 1999 96 13432 13437 10557338 10.1073/pnas.96.23.13432
Leslie RA James MF Pharmacological magnetic resonance imaging: a new application for functional MRI Trends Pharmacol Sci 2000 21 314 318 10918638 10.1016/S0165-6147(00)01507-8
Tracey I Prospects for human pharmacological functional magnetic resonance imaging (phMRI) J Clin Pharmacol 2001 Suppl 21S 28S 11452725
Casey KL Svensson P Morrow TJ Raz J Jone C Minoshima S Selective opiate modulation of nociceptive processing in the human brain J Neurophysiol 2000 84 525 533 10899224
Wise RG Rogers R Painter D Bantick S Ploghaus A Williams P Rapeport G Tracey I Combining fMRI with a pharmacokinetic model to determine which brain areas activated by painful stimulation are specifically modulated by remifentanil Neuroimage 2002 16 999 1014 12202088 10.1006/nimg.2002.1146
Wise RG Williams P Tracey I Using FMRI to quantify the time dependence of remifentanil analgesia in the human brain Neuropsychopharmacology 2004 29 626 635 14679387 10.1038/sj.npp.1300364
Zubieta JK Smith YR Bueller JA Xu Y Kilbourn MR Jewett DM Meyer CR Koeppe RA Stohler CS Regional mu opioid receptor regulation of sensory and affective dimensions of pain Science 2001 293 311 315 11452128 10.1126/science.1060952
Rogers R Wise RG Painter DJ Longe SE Tracey I An investigation to dissociate the analgesic and anesthetic properties of ketamine using functional magnetic resonance imaging Anesthesiology 2004 100 292 301 14739803 10.1097/00000542-200402000-00018
Chavez ML DeKorte CJ Valdecoxib: a review Clin Ther 2003 25 817 851 12852704 10.1016/S0149-2918(03)80110-8
Siegle I Klein T Backman JT Saal JG Nusing RM Fritz P Expression of cyclooxygenase 1 and cyclooxygenase 2 in human synovial tissue: differential elevation of cyclooxygenase 2 in inflammatory joint diseases Arthritis Rheum 1998 41 122 129 9433877 10.1002/1529-0131(199801)41:1<122::AID-ART15>3.0.CO;2-8
Samad TA Moore KA Sapirstein A Billet S Allchorne A Poole S Bonventre JV Woolf CJ Interleukin-1beta-mediated induction of Cox-2 in the CNS contributes to inflammatory pain hypersensitivity Nature 2001 410 471 475 11260714 10.1038/35068566
Casey KL Minoshima S Morrow TJ Koeppe RA Comparison of human cerebral activation pattern during cutaneous warmth, heat pain, and deep cold pain J Neurophysiol 1996 76 571 581 8836245
Coghill RC Sang CN Maisog JM Iadarola MJ Pain intensity processing within the human brain: a bilateral, distributed mechanism J Neurophysiol 1999 82 1934 1943 10515983
Gelnar PA Krauss BR Sheehe PR Szeverenyi NM Apkarian AV A comparative fMRI study of cortical representations for thermal painful, vibrotactile, and motor performance tasks Neuroimage 1999 10 460 482 10493903 10.1006/nimg.1999.0482
Tracey I Becerra L Chang I Breiter H Jenkins L Borsook D Gonzalez RG Noxious hot and cold stimulation produce common patterns of brain activation in humans: a functional magnetic resonance imaging study Neurosci Lett 2000 288 159 162 10876085 10.1016/S0304-3940(00)01224-6
Johansen JP Fields HL Glutamatergic activation of anterior cingulate cortex produces an aversive teaching signal Nat Neurosci 2004 7 398 403 15004562 10.1038/nn1207
Jasmin L Rabkin SD Granato A Boudah A Ohara PT Analgesia and hyperalgesia from GABA-mediated modulation of the cerebral cortex Nature 2003 424 316 320 12867983 10.1038/nature01808
Calejesan AA Kim SJ Zhuo M Descending facilitatory modulation of a behavioral nociceptive response by stimulation in the adult rat anterior cingulate cortex
1 Eur J Pain 2000 4 83 96 10833558 10.1053/eujp.1999.0158
Bornhovd K Quante M Glauche V Bromm B Weiller C Buchel C Painful stimuli evoke different stimulus-response functions in the amygdala, prefrontal, insula and somatosensory cortex: a single-trial fMRI study Brain 2002 125 1326 1336 12023321 10.1093/brain/awf137
Derbyshire SW Jones AK Gyulai F Clark S Townsend D Firestone LL Pain processing during three levels of noxious stimulation produces differential patterns of central activity Pain 1997 73 431 445 9469535 10.1016/S0304-3959(97)00138-3
Broom DC Samad TA Kohno T Tegeder I Geisslinger G Woolf CJ Cyclooxygenase 2 expression in the spared nerve injury model of neuropathic pain
1 Neuroscience 2004 124 891 900 15026129 10.1016/j.neuroscience.2004.01.003
Apkarian AV Krauss BR Fredrickson BE Szeverenyi NM Imaging the pain of low back pain: functional magnetic resonance imaging in combination with monitoring subjective pain perception allows the study of clinical pain states Neurosci Lett 2001 299 57 60 11166937 10.1016/S0304-3940(01)01504-X
Jezzard P Mathews P Smith SM Functional MRI: An introduction to methods 2001 Oxford, Oxford University Press
Woolrich MW Ripley BD Brady M Smith SM Temporal autocorrelation in univariate linear modeling of FMRI data Neuroimage 2001 14 1370 1386 11707093 10.1006/nimg.2001.0931
Woolrich MW Behrens TE Beckmann CF Jenkinson M Smith SM Multilevel linear modelling for FMRI group analysis using Bayesian inference Neuroimage 2004 21 1732 1747 15050594 10.1016/j.neuroimage.2003.12.023
Friston KJ Holmes AP Worsley KJ Poline JB Frith CD Frackowiak RS Statistic parametric maps in functional imaging: A general linear approach Hum Brain Mapp 1995 2 189 210 10.1002/hbm.460020402
Talairach J Tournoux P Co-planar stereotactic atlas of the human brain 1988 Stuttgart,, Thieme
|
16266429
|
PMC1291397
|
CC BY
|
2021-01-04 16:40:08
|
no
|
Mol Pain. 2005 Nov 2; 1:32
|
utf-8
|
Mol Pain
| 2,005 |
10.1186/1744-8069-1-32
|
oa_comm
|
==== Front
Nutr JNutrition Journal1475-2891BioMed Central London 1475-2891-4-281625313510.1186/1475-2891-4-28ResearchAge as a determinant of nutritional status: A cross sectional study Forster Sarah [email protected] Salah [email protected] Sheffield Institute for Studies on Ageing, The University of Sheffield & Barnsley Hospital, Northern General Hospital, Sheffield, S5 7AU, UK2005 27 10 2005 4 28 28 17 8 2005 27 10 2005 Copyright © 2005 Forster and Gariballa; licensee BioMed Central Ltd.2005Forster and Gariballa; licensee BioMed Central Ltd.This is an Open Access article distributed under the terms of the Creative Commons Attribution License (), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
Background
Undenutrition is known to be prevalent and largely unrecognised in older patients; however, aberrations in indicators of nutritional status may simply reflect effects of age and/or functional disability.
Objective
The aim of this study was to measure the effect, if any of age on nutritional status in older patients.
Design
445 randomly selected hospitalised patients consented to nutritional status assessment derived from anthropometric, haematological, and biochemical data within 72 hours of admission. Nutritional status was compared between those age < 75 years and those aged 75 years or more. Using multiple regression models, we measured the association between age and nutritional assessment variables after adjusting for disability, chronic illness, medications, smoking and tissue inflammation.
Results
Body weight, body mass index, mid-upper arm circumference, haemoglobin, serum albumin and plasma ascorbic acid were all significantly lower in people aged ≥ 75 years compared with those < 75 years of age. Although riboflavin (vitamin B2), 25OH VitD3, red-cell folate and vitamin B12 concentrations were lower in those aged ≥ 75 years, differences were not statistically significant. After adjusting for disability and co-morbidity in a multivariate analysis, age alone had a significant and independent effect on important anthropometric and biochemical nutritional assessment variables.
Conclusion
Increasing age is independently associated with poor nutritional status. This may partly explain the poor clinical outcome in older patients.
==== Body
Introduction
Societies worldwide have experienced a considerable increase in the number of older people [1]. There is a growing recognition that age-related physiological changes may predispose to protein-energy undernutrition, in the elderly, particularly in the presence of other factors associated with aging, including social and psychological, variables and the presence of disease [2,3]. Consequently, undenutrition is known to be prevalent and largely unrecognised in older patients. There is also evidence which links protein-energy undernutrition or its markers with clinical outcomes in acute and non-acute hospital settings, and additional evidence indicating that nutritional supplements can improve outcome in some of these settings [4-6].
Good nutrition may contribute significantly to the health and well being of older individuals, and to their ability to recover from illness. However there is no gold standard for diagnosing undernutrition, and most clinically available nutrition screening instruments lack sensitivity and specificity. In addition, abnormal nutritional indicators may simply reflect effects of age, functional disability, or severe underlying disease [7,8]. Furthermore, the difficulties in detecting early signs of undernutrition are similar to those encountered in the early recognition of many age-related diseases [7]. However, in the case of nutritional deficiency there are two further difficulties: for almost every nutrient there is a long latency period before a low intake leads to overt clinical manifestations, and early diagnosis must depend upon the findings of abnormalities of special tests, including biochemical and haematological investigations. Second, in the elderly the true significance of abnormal results of these tests is not fully understood [9]. The aim of this study was to measure the effects of aging on nutritional status in older patients.
Subjects & Methods
Four hundred and forty five acutely ill older patients who took part in a randomised, double blind, placebo-controlled single-centre trial of nutritional supplementation were included. The randomisation sequence was generated by the trial statistician, concealed in sequentially numbered, sealed opaque envelopes and kept in a clerical office at a different city. Admission diagnoses included ischaemic heart disease (myocardial infarction & angina), heart failure, atrial fibrillation, chronic obstructive pulmonary disease, chest and urinary tract infections, septicaemia, stroke, Parkinson's disease anaemia, diabetes, osteoarthritis, rheumatoid arthritis syncope, falls, fracture limbs, elective surgery knee and hip surgery. Inclusion criteria were: age ≥ 65 years; stable medical condition; able to swallow and able to sign an informed written consent form. Patients excluded from the study were those with severe medical or psychiatric illness including those with gastric surgery, malabsorption, or morbid obesity, in coma, diagnosed severe dementia (abbreviated mental test < 6), and malignancy, living in institution and patients already on supplements. The study was approved by Barnsley research ethics committee.
Clinical and nutritional assessment
Following informed written consent and recruitment to the study all patients had baseline assessment. The assessment included demographic and medical data including current diagnosis, history of chronic illnesses, smoking, alcohol and drug intake, nutritional status and disability (Barthel score).
Nutritional status was assessed from anthropometric, haematological and biochemical data within 72 hours of admission. All anthropometrics measurements were performed by a single observer (SF) using standard methods with intra observer's differences assessed prior to the commencement of the study. Mid-upper arm circumference (MUAC) and triceps skin folds (TSF) were measured by a flexible tape and Harpenden Skin fold callipers accurate to 0.2 mm (Practical Metrology Sussex UK) respectively and the mean of three measures was recorded. The local Pathology Laboratory performed routine tests including serum ferritin, albumin and transferrin measurements. C-reactive protein (CRP) concentration, a marker of tissue inflammation, was measured by a modified latex-enhanced immuno-turbidimetric assay (normal ≤10 mg/L). The inter-assay coefficient of variation (c.v.) was 3.9%. Red cell folate and plasma vitamin B12 were measured on the Architect (Abbott Laboratories) using chemiluminescent microparticle immunoassay technology. The inter-assay coefficients of variation were 12.6% and 8.4% respectively. Riboflavin status was assessed as the erythrocyte glutathione reductase activity coefficient (EGRAC), using the Cobas BioAutoanalyser (Roche Diagnostics), giving an inter-batch cv of 7.6%. Ratios above 1.4 were considered to reflect biochemical deficiency of riboflavin [10]. Plasma total ascorbic acid was measured by a fluorescence assay automated for the Cobas BioAutoanalyser [11], giving inter-batch cv's of 8.4%. 25-Hydroxy Vitamin D3 was measured by HPLC using Chromsystems (Germany). This method uses a Chromsystems reagent kit, which allows the safe chromatographic determination of 25-OH-vitamin D3 on a simple isocratic HPLC system with UV detection [12]. The inter-assay coefficient of variation was 4.6%.
Disability at baseline was assessed using the Barthel score on a 20-point scale [13]. The Barthel scores 10 functions on a scale 0 (fully dependent) to 20 (independent).
Statistical analyses
Statistical analyses were performed with SPSS software, version 11.0 (SPSS Inc., Chicago). Descriptive tests (median and inter-quartile range) were used to describe the baseline characteristics of the subjects. Mann Whitney U tests was used to test between-group differences where appropriate with a p-value of <0.05 regarded as statistically significant. A forward stepwise multiple regression analysis was performed to determine the influence of age on nutritional status as measured by body mass index (BMI), MUAC, TSF, haemoglobin, serum albumin and vitamin C after adjusting for a number of independent clinical variables including disability (Barthel score), co morbidity (previous illnesses, drugs and smoking) and tissue inflammation (CRP).
Results
Between July 2001, and May 2004, 445 patients were recruited. Table 1 shows baseline clinical characteristics including gender, smoking, alcohol, consumption, chronic illness, drugs, disability, mental status, tissue inflammation (CRP), haemoglobin, glucose and renal function.
Table 1 Baseline characteristics of subjects age < 75 years compared with those aged 75 years or more [n(%) unless stated otherwise]
Variable <75 years (n = 181) ≥75 years (n = 264)
Gender, female 78 (43)| 133 (50)
Smoking Never smoked 50 (28) 87 (33)
Ex smoker 80 (44) 136 (52)
Current smoker 44 (24) 37 (14)
Number of subjects with alcohol >14 units 26 (14) 20 (8)
Chronic disease IHD 58 (32) 104 (39)
Hypertension 63 (35) 80 (30)
Stroke 14 (8) 37 (14)
COPD 35 (19) 49 (19)
Medications + 0–2 66 (38) 77 (30)
3–6 97 (55) 166 (64)
>6 12 (7) 16 (6)
Barthel Score *, median (Q1-Q3) 19 (14–20) 18 (13–20)
C-reactive protein median (Q1-Q3) 17 (5–79) 19.7 (5–59)
Fasting blood glucose median (Q1-Q3) 5.6 (5–6.4) 5.6 (5–6.2)
* P < 0.05
+ Values do not always add up because of some missing data
Subjects aged ≥75 years were significantly more disabled and had non significantly higher levels of co-morbidity and CRP concentration compared with those less than 75 years of age. Age alone affected both anthropometric and nutritional biochemical measurements. BMI, MUAC, haemoglobin, serum albumin and plasma ascorbic were all significantly lower in persons > 75 years of age compared with those younger than 75 years (Table 2). Riboflavin (EGRAC), 25OH VitD3, red cell folate and vitamin B12 concentrations were nonsignificantly lower in those aged ≥75 years.
Table 2 Anthropometric & biochemical nutritional data of subjects age < 75 years compared with those aged 75 years or more [median(interquartile range)]
N <75 years old N ≥75 years old
Body weight (kg) 166 70.0 (60–79) 231 63.5 (56.5–72.5)**
BMI + 165 26 (23–28) 231 25 (22–27)*
MUAC (cm) + 175 29.5 (26.5–32.0) 259 27.5 (25.5–29.5)**
TSF (mm) + 174 16 (11–21) 259 15 (11–19)
Haemoglobin (g/dl) 181 13.3 (12–15) 259 12.8 (11–14) **
Albumin (g/L) 164 39 (36–42) 245 37 (34–40)**
Transferrin (μg/L) 143 2.15 (1.79–2.44) 199 2.10 (1.76–2.51)
Ferritin (pmol/L) 156 254 (103–497) 215 238 (124–521)
Red cell folate (nmol/L) 147 378 (437–888) 210 359 (269–504)
Vitamin B12 (pmol/L) 154 239 (170–349) 215 214 (154–337)
Vitamin B2 (EGRAC) 124 1.27 (1.18–1.40) 174 1.28 (1.18–1.38)
Plasma ascorbic acid (μmol/L) 129 21.3 (18.7–35.8) 193 15.2 (5.6–32.5)*
25OH Vit D3 (nmol/L) 107 36.0 (18.0–59.0) 145 33.0 (17.2–53.5)
** p value < 0.01, * p value < 0.05
+ body mass index (BMI), Mid-upper arm circumference (MUAC) and triceps skin folds (TSF)
Tables 3 &4 summarize results of the multiple regression analysis for the association between age and other clinical variable and nutritional assessment parameters. Among those included in the multiple regression models, age and smoking revealed a statistically significant association with BMI, MUAC and TSF (Table 3). However, haemoglobin and serum albumin were primarily associated with age, disability (Barthel score) and tissue inflammation. Vitamin C values were mainly predicted by smoking and tissue inflammation. For individuals, aged ≥ 75 years BMI, MUAC and TSF were 1.3, 1.95 cm and 0.02 mm less respectively compared with those patients aged less than 75 years (Table 3). Similarly for patients aged 75 years or more haemoglobin, serum albumin and vitamin C values were 0.4 g/dL, 1.5 g/L and 4.4 μmol/L less respectively, compared with patients aged less than 75 years. However the results for vitamin C did not reach statistical significance (Table 4).
Table 3 Multiple regression result of age, disability, co-morbidity, smoking and acute phase response on BMI, MUAC and TSF
Body mass index MUAC (cm) TSF (mm)
Regression coefficient (95% C.I) P value Regression coefficient (95% C.I) P value Regression coefficient (95% C.I) P value
Age (<75; ≥ 75 yrs) -.1.30 (-2.0 to -.27) 0.010 -1.95 (-2.7 to -1.2) 0.000 -0.02 (-0.09 to-0.01) 0.009
Barthel score (<11; ≥ 11) -.13. (-1.3 to 1.1) 0.832 0.43 (-0.5 to 1.3) 0.352 -0.03 (-0.5 to 1.3) 0.556
Chronic illnesses .036 (-.20 to .41) 0.508 -0.02 (-0.2 to 0.3) 0.618 -0.05 (-0.01 to 0.02) 0.310
All medications -.035 (-.32 to .16) 0.523 -0.14 (-0.3 to 0.1) 0.162 0.02 (-0.01 to 0.01) 0.772
Smoking (Never, Ex, Current) -1.16 (-1.78 to -.55) 0.000 -0.87 (-1.4 to -0.4) 0.001 -0.02 (-0.08 to -0.02) 0.000
CRP (≤10, >10 mg/L) .01 (-.94 to 1.7) 0.062 -0.02 (-0.7 to 0.8) 0.824 0.05 (-0.02 to 0.06) 0.305
Table 4 Multiple regression result of age, disability, co-morbidity, smoking and acute phase response on haemoglobin, Serum albumin and vitamin C
Haemoglobin (g/dl) Serum Albumin g/L Vitamin C μmol/L
Regression coefficient (95% C.I) P value Regression coefficient (95% C.I) P value Regression coefficient (95% C.I) P value
Age (<75; ≥ 75 yrs) -0.4 (-0.9 to -0.03) 0.035 -1.5(-2.3 to -0.7) 0.000 -4.4 (-9.2 to .41) 0.073
Barthel score (<11; ≥ 11) 1.0 (0.5 to 1.6) 0.000 2.5 (1.5 to 3.5) 0.000 5.6 (-0.4 to 11.5) 0.066
Chronic illnesses 0.02 (-0.11 to 0.18) 0.650 0.2 (-0.1 to 0.5) 0.190 0.3 (-1.3 to 2.0) 0.709
All medications -0.03 (-0.15 to 0.08) 0.575 0.04 (-0.1 to .3) 0.407 0.8 (-0.5 to 2.1) 0.225
Smoking (Never, Ex, Current.) 0.03 (-0.19 to 0.39) 0.514 -0.38 (-0.9 to 1.9) 0.191 -3.8 (-7.1 to -.53) 0.023
CRP (≤10, >10 mg/L) -0.9 (-1.3 to -0.5) 0.000 -3.7 (-4.6 to -2.9) 0.000 -6.5 (-11.5 to -1.5) 0.011
Discussion
The main findings of this study were that patients aged 75 years or more had poorer nutritional status compared with those younger than 75 years. After adjusting for nonnutritional clinical risk indicators, increasing age was strongly and independently correlated with poor nutritional status.
Ageing in man is accompanied by changes, which may impair food acquisition, digestion, and metabolism. Anorexia and weight loss are common and important clinical problems in older people, and the causes are multifactorial [2,3]. There is a growing recognition that age-related physiological anorexia may predispose to protein-energy undernutrition in older persons, particularly in the presence of other pathological factors associated with aging [2,3]. For example, alteration in smell and taste and poor dental health directly decrease food intake or influence food selection [7]. In general, physical activity and lean body mass decrease with aging, while body fat, increases. These factors may decrease energy requirements and intake [7]. Lower food intake may lead to lower intake of both macro- and macronutrients, and even in relatively healthy persons, mild, sub-clinical nutritional deficiencies are known to be common. For example, the most recent National Diet and Nutrition Survey people aged 65 years and older in the UK highlighted low dietary intakes and poor biochemical status for a range of micronutrients including vitamin C, riboflavin, thiamine and folate, and modest intakes of vitamin E [14].
In this study we also found a significant independent association between former and current smokers status and poor nutritional status. Smoking is known to be associated with oxidative stress, poor nutritional status, weight loss, and increased mortality [15,16]. Because smokers are known to have low intakes of fruits and vegetables that are rich in antioxidants, they are therefore more likely to be susceptible to oxidative damage caused by free radicals [15,16]. Palaniappan et al. have recently shown that smokers consume significantly fewer fruits and vegetables than non-smokers, leading to lower intake of folate and vitamin C [16]. Although mechanisms leading to poor nutritional in smokers are not known, possible contenders include poor taste and appetite and pro inflammatory effect of smoking or possible confounding effect.
Numerous studies in the past have demonstrated a strong association between serum albumin levels and clinical outcome [4,5,17]. A study of predictors of early non-elective hospital readmission in elderly patients has found that individuals with any amount of weight loss and no improvement in albumin concentrations during the first month after hospitalisation were at a much higher risk of readmission than were those who maintained or increased their post-discharge weight and had repleted their serum albumin concentrations [17].
Many conditions such as disability, acute and chronic diseases may influence nutritional status in ageing patients [4,18-20]. Because poor nutritional status may be partly related to clinical risk indicators other than age, such as underlying disease state, tissue inflammation, diet-medication interactions, or functional capacity, we made an attempt to control for these factors. By adjusting for the influence of these factors on nutritional status, it was possible to identify a potential independent effect of age on nutritional status. Because patients with diseases severe medical and psychiatric illnesses such as liver, gastrointestinal, kidney or neoplasm were excluded from this study, it is possible that increasing age alone is causally related to poor nutritional status.
Although in this cross-sectional study we have identified age as a potentially important risk factor for undernutrition in a hospital inpatient population, our results may not be extrapolated to healthy populations of older individuals.
It is also not clear whether nutritional deficiencies in individuals aged 75 years and older reflect poor nutrient intake, increased demand for nutrients, or simply represent underlying co-morbidities. Although this study was not designed to answer these questions, targeting this cohort with nutritional supplementation may help to overcome the potentially detrimental effects of nutritional deficits in this population.
Nutritional support studies to determine the optimal timing and composition of nutritional therapy relative to a patient's age are long overdue. This has the potential to improve nutritional status and lead to better rehabilitation outcome, decreased readmission rate, improved quality of life, and contribute to reducing health care costs.
==== Refs
Population: by gender and age 1901–2026: Social Trends 2001; 31 (ST 31103) 2001 Office of National Statistics, UK
Gariballa SE Nutrition and older people: special consideration for ageing Clinical Medicine 2004 4 411 413 15536868
MacIntosh CG Morley JE Horowitz M Chapman IM Anorexia of Ageing Nutrition 2000 16 983 995 11054605 10.1016/S0899-9007(00)00405-6
Gariballa SE Parker SG Taub N Castleden CM The influence of nutritional status on clinical outcome after acute ischaemic stroke Am J of Clin Nutr 1998 68 275 281 9701183
Sullivan DH Walls RC Impact of nutritional status on morbidity in a population of geriatric rehabilitation patients J Am Geriatr Soc 1994 42 471 77 8176139
Milne AC Potter J Avenell A Protein and energy supplementation in elderly people at risk from malnutrition Cochrane Library issue Oxford 2002 3
Gariballa SE Sinclair AJ Nutrition, Ageing and ill-health Br J Nutr 1998 80 7 21 9797639
Klein Kinney J Jeejeebhoy K Alpers D Hellerstein M Murray M Nutrition support in clinical practice: Review of published data Am J Clin Nutr 1997 66 683 706 9280194
Exton-Smith AN Exton-Smith AN, Caird FI Nutritional status: diagnosis and prevention of malnutrition Metabolic and nutritional disorders in the elderly 1980 John Wright & Sons, Bristo
Powers HJ Bates CJ Duerdin JM Effects of riboflavin deficiency in rats on some aspects of iron metabolism International Journal of Vitamin and Nutrition Research 1983 53 371 376
Veuillemier JP Keck E Fluorimetric assay of vitamin C in biological materials using a centrifugal analysis with fluorescence attachment Micronutrient Analysis 1989 5 25 34
Turpeinen U H U Stenman U "Determination of 25 Hydroxyvitamin D in Serum by HPLC and Immunoassay Clinical Chemistry 2003 49 1521 1523 12928235 10.1373/49.9.1521
The Royal College of Physicians of London and The British Geriatrics Society Standardised assessment scales for older people A report of joint workshop 1992 1 27
Finch S D W Lowe C Bates CJ Prendice A Smithers G Clarke PC National Diet and Nutrition Survey people aged 65 years and older 1998 1 London: The Stationary Office
Cochrane WJ Afolabi OA Investigation into nutritional status, dietary intake and smoking of patients with COPD J Hum Nutr Dietet 2004 17 3 11 10.1046/j.1365-277X.2003.00478.x
Palaniappan U Starkey LJ O'Loughlin J Gray-Donald K Fruits and vegetable consumption is lower and saturated fat intake is higher among Canadians reporting smoking J Nutr 2001 131 1952 1958 11435513
Friedmann JM Jensen GL Smiciklas-Wright McCamish MA Predicting early nonelective hospital readmission in nutritionally compromised older adults Am J Clin Nutr 1997 65 1714 20 9174465
Weinsier RL Heimburger DC Distinguishing malnutrition from disease: the search goes on Am J Clin Nutr 1997 66 1063 9356520
Gaggiotti G Ambrosi S Onorato G Piloni S Amadio L Spazzafumo L La Rocca R "The influence of age and sex on nutritional parameters in subjects aged 60 years and over Archives of Genernology and Geriatrics 1995 20 117 128 10.1016/0167-4943(94)00578-U
Louw JA Werbeck A Louw M Kotze T Cooper R Labadarios D "Blood vitamin concentrations during the acute phase response Critical Care Medicine 1992 20 934 941 1617986
|
16253135
|
PMC1291398
|
CC BY
|
2021-01-04 16:39:30
|
no
|
Nutr J. 2005 Oct 27; 4:28
|
utf-8
|
Nutr J
| 2,005 |
10.1186/1475-2891-4-28
|
oa_comm
|
==== Front
Reprod Biol EndocrinolReproductive biology and endocrinology : RB&E1477-7827BioMed Central London 1477-7827-3-591625092010.1186/1477-7827-3-59MethodologyEasy and rapid method for the determination of gene expression in cumulus cells incubated for oocyte maturation Kumamoto Kanako [email protected] Haifeng [email protected] Hideaki [email protected] Takato [email protected] Graduate School of Biosphere Science, Hiroshima University, Higashi-Hiroshima, Hiroshima 739-8528, Japan2005 27 10 2005 3 59 59 2 9 2005 27 10 2005 Copyright © 2005 Kumamoto et al; licensee BioMed Central Ltd.2005Kumamoto et al; licensee BioMed Central Ltd.This is an Open Access article distributed under the terms of the Creative Commons Attribution License (), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
Background
The objectives of this study were to develop an easy and rapid method for measuring gene expression in a small number of cells by real-time PCR without RNA extraction and purification, and to use this method to determine more precisely IGF-I gene expression in the cumulus cells surrounding oocytes.
Methods
First, after small numbers of cumulus cells were lysed in cell lysis buffer, they were digested with various concentrations of DNase I for different periods at 37°C to determine the optimal conditions for digestion of genomic DNA in the lysate. Since nonspecific amplification was liable to occur when the non-purified RT product of the cell lysate was used for real-time PCR with the given primers, the optimal conditions for Mg2+ and annealing temperature were well investigated. Further, to create the same conditions as in the actual sample reaction for measurement by real-time PCR, RT-minus product was added to the reaction mixture of the standard curve, and then the amplification efficiency was assessed. Next, IGF-I gene expression in cumulus cells collected from cumulus oocyte complexes (COCs) every 4 h during maturation was determined using the developed method.
Results
The optimal conditions for measuring gene expression using the cell lysate from a small number of cells were as follows: incubation of the cell lysate with 0.16 U/microL DNase I with 10 U/microL for 30 min, an Mg concentration of 1.5 mM for amplification of target gene by real-time PCR using RT-product of the cell lysate. When the RT-minus products added to the reaction mixture for the standard curve, which was prepared for purified 18SrRNA plasmid, the PCR efficiency was similar between the sample and the standard.
The IGF-I gene expression in the cumulus cells was elevated up through the first 8 h of the culture and then declined gradually by the end of maturation, with the maximal gene expression (778-fold) seen at 8 h.
Conclusion
It can be concluded that the method developed here, in which equivalent to cumulus cells collected from 0.03–0.075 COCs were employed per reaction, permits rapid and easy determination of target gene expression in a limited number of cells using real-time PCR without RNA extraction.
==== Body
Background
Isobe et al (2001) have shown that a loss of the gap junction within cumulus cells, i.e. an interruption in the transmission of signaling factors from the cumulus to the oocytes, triggers the resumption of meiosis in pig oocytes [1]. Eppig (2001) also provided biochemical evidence for the signaling factors exchanged between the oocyte and its surrounding cumulus cells, and concluded that the factors are essential for inducing and regulating the differentiation of follicular compartments from one specific developmental stage to the next, thus ensuring the development of the oocyte [2,3]. Furthermore, using specific inhibitors of RNA polymerase II, such as alpha – amanitin or 6-dichloro-1-b-D-ribofuranosyl-benzimidazole, it has been demonstrated that an initial transcriptional event in the cumulus cells that surround the oocyte is an absolute requirement for gonadotrophin-mediated resumption of meiosis in mammalian species including the pig [4], mouse [5] and cow [6,7]. In bovine cumulus-oocyte complexes (COCs), transcriptional events initiated by FSH did not occur when COCs resumed meiosis spontaneously or in the presence of hCG [8]. These findings strongly suggest that gene expressions in the surrounding cumulus cells are essential for the meiotic resumption of oocytes [[9] and [5]].
On the other hand, real-time polymerase chain reaction (PCR) offers substantial advantages over conventional reverse transcriptase – polymerase chain reaction (RT-PCR) for the molecular study of gene expression in any kind of cell, including cumulus cells. Some studies have investigated gene expression by means of real-time PCR using a large number of cumulus cells [10,11] collected from many COCs, since it is difficult to develop a template for investigating gene expression. Only a few gene expressions have been revealed in cumulus cells during oocyte maturation due to the very small number of cumulus cells surrounding an oocyte.
The objective of the present study was to develop a rapid and easy method for determining gene expression using small quantities of the target template in a small number of cells, such as cumulus cells surrounding an oocyte. This method was then used for analyzing IGF-I gene expression in the surrounding cumulus cells during in vitro maturation of bovine oocytes.
Materials and methods
Oocyte recovery
The ovaries were obtained at a slaughter house and transported to the laboratory in a saline solution at around 30°C. Oocytes were aspirated from follicles of 2 to 5 mm in diameter using a 10 ml syringe attached to a 21 gage needle. The recovered cumulus-oocyte complexes (COCs) were washed with phosphate-buffered saline (PBS) containing 0.05% polyvinylpyrrolidone (Sigma, USA) and 100 μg/ml kanamycin. The COCs, which were completely surrounded by at least 4–5 layers of non-expanded cumulus in combination with a homogeneous cytoplasm, were used for cumulus cell collection before and after incubation.
Preparation of the cumulus cell lysates
Three freshly isolated or incubated COCs were vortexed for 2 min in PBS supplemented with 0.002% hyaluronidase, and the oocytes were removed from the cumulus cell suspension. The cumulus cells were washed three times with the PBS provided in a cells-to-cDNA II kit (Ambion, Texas, USA) by centrifugation for 3 min at 4°C at 1200 g, then the pelleted cumulus cells were flash frozen in liquid nitrogen and stored at -80°C until use.
To make a cell lysate, 50 μl of the cell lysis buffer that was provided in the cells-to-cDNA II kit was added to 0.5 ml of small tube (AB gene, Surrey, UK) containing the pelleted cumulus cells and vortexed for 5 min in order to mix the cells. To lyse the cells, the suspension was heated using a block-type thermal cycler system according to the manufacturer's instructions.
DNase digestion and reverse transcript
Cell lysates were supplemented with the different concentrations (final concentrations of 0.04 U/μl and 0.08 U/μl) of DNase I provided in the cells-to-cDNA II kit or with two concentrations (final concentrations of 0.08 U/μl and 0.16 U/μl) of RNase-free certified DNase I with 10 U/μl (Roche, Penzberg, Germany). They were then incubated at 37°C for various periods. Immediately after the end of the incubation, each tube was heated for 15 min at 75°C to inactivate the DNase.
To synthesize the first strand of cDNA as per the cells-to-cDNA II kit protocol, 5 μl cell lysate, 4 μl dNTP Mix, 2 μl oligo dT and 5 μl RNase-free water were assembled in an RNase-free 0.5 ml tube, then heated for 3 min at 70°C. After this mixture was cooled on ice, 2 μl 10 × RT buffer, 1 μl M-MLV reverse transcriptase and 1 μl RNase inhibitor were added to the reaction tubes. Reverse transcription was carried out for 30 min at 42°C, followed by incubation at 95°C for 10 min. RT-minus products with all the reaction components except for the reverse transcriptase were produced for each sample, and were then employed for real-time PCR in order to demonstrate that the template for the PCR product was cDNA, not genomic DNA.
Real-time PCR
The primers for 18S rRNA and insulin-like growth factor I (Table 1) were designed using the Gen Bank sequences (accession numbers: AF176811 and X15726) within only one exon, and they had an 80–200 bp product. They were synthesized by Nihon Gene Research Lab Inc. (Sendai, Japan).
Table 1 Primer sequences of 18SrRNA and IGF-I for real-time PCR and plasmid construction (†).
Genes Accession No. Primer sequences
18SrRNA AF176811 5'-AACGGCTACCACATCCAA-3' 5'-GACTCATTCCAATTACAGGGC-3'
5' – TGACGGGGAATCAGGGTT-3' (†) 5' – CAAAGTAAACGCTTCGGGC-3' (†)
IGF – I X15726 5'-GGA GTG CAG GAA ACA AGA AC-3' 5'-TTT TGG TAG GTC TTC TGG TG-3'
To construct the standard curve, a DNA fragment amplified with the bovine 18SrRNA primers for plasmid construction was cloned into the PCR 2.1™ vector according to the manufacturer's instructions (Invitrogen Life Technologies, Carlsbad, CA). The plasmid DNA with the target sequence was purified with the help of the QIA Filter Plasmid Maxi Kit (Qiagen, Westburg, The Netherlands). The IGF-I fragments used to generate the standard curve were amplified with the primer for real-time PCR, collected from agarose gel, and then precipitated with ethanol. The concentrations of plasmid and target sequence DNA were evaluated by measuring the absorbance at 260 nm and expressed in a copy number of target fragments per milliliter.
The PCR was performed in a light cycler (Roche, Hercules, CA) using Light Cycler-Fast Start DNA Master SYBR Green I (Roche) as per the kit instructions with the following modifications: the optimal concentration of MgCl2 (1.5 mM) was used for both primers (18SrRNA and IGF-I), which had been explored in experiment 1 and 2 as described below; 200 nM of each primer; 2 μl SYBR Green I master mix; and either 2 or 5 μl of RT product in a final volume of 20 μl.
The basic protocol for real-time PCR was an initial incubation at 95°C for 10 min to activate modified Taq polymerase. Fifty-five cycles of PCR were carried out with denaturation at 95°C for 10 s, then at annealing temperature for 5 s, then extended at 72°C for 10 s. After amplification, a melting curve was generated by heating the samples to 95°C for 0s, followed by cooling to 65°C for 15s and heating to 95°C at a rate of 0.1°C/s at which stage SYBR Green I fluorescence was measured continuously. The melting temperatures and target gene expression were analyzed using Light Cycler Software (version 3.5: Roche). After analysis, the target gene expression was normalized to 18SrRNA.
The RT-minus control from the previous step and minus-template PCR control were included in each reaction of real-time PCR as negative controls. The minus-template PCR should have all the PCR components, with water substituted for the RT reaction aliquot, thus verifying that none of the PCR reagents are contaminated with DNA.
Experimental design
This study consisted of the two following experimental phases; 1) construction of a rapid and easy method for determining gene expression using RNA crude solution (cell lysate) as a template, and 2) analysis of IGF-I gene expression in cumulus cells using the method developed.
Experiment 1: Construction of a rapid and easy method for determining gene expression using RNA crude solution (cell lysate) as a template.
The aim of this experiment was to establish an easy and rapid method for determining gene expression using a small number of cells. The cumulus cells collected from COCs before maturation were lysed in the cell lysis buffer provided in the Cells-to-cDNA II kit. Cell lysate containing genomic DNA was digested for 15 to 60 min with various concentrations of ether the DNase I provided in the Cells-to-cDNA II kit or RNase-free DNase I. Next, the Mg concentration and annealing temperature for real-time PCR using the RT product of the cell lysate were well investigated, since the RT-product may contain the cations that originated in the cell lysis buffer and the RT buffer. Based on the results of this experiment, we determined the optimal conditions for the digestion of genomic DNA in cell lysate, as well as the concentration of Mg and the annealing temperature for real-time PCR when the RT-product of the cell lysate corresponding to the number of cells from 0.075 COC was amplified by real-time PCR with 18SrRNA primers. Further, to ensure that the PCR efficiency (E = 10-1/s - 1) was similar between the sample and the standard, we analyzed whether the addition of RT-minus products to the reaction mixture for the standard curve which was prepared for purified 18SrRNA plasmid, affected the PCR efficiency.
Experiment 2: Analysis of IGF-I gene expression in cumulus cells using the method developed.
This experiment was designed to test whether IGF-I gene expression in cumulus cells surrounding an oocyte during its maturation can actually be measured by the method developed in Experiment 1. Five COCs were cultured in 100 μl of TCM 199 supplemented with 10% fetal cow serum (FCS; Gibco BRL, NY, USA), 1.3 μg/ml LH (Sigma Chemical Co., St. Louis, MO), 0.6 μg/ml FSH (Sigma), 0.2 mM Na – pyruvate (Sigma), and 50 μg/ml gentamycin in a 96-well plate covered with a plate seal. The oocyte cultivation was carried out at 39°C in a humidified atmosphere of 5% CO2 in air. Cumulus cells were collected from COCs cultured for 0 h, 4 h, 8 h, 12 h, 16 h, 20 h, and 24 h, then lysed in the provided cell lysis buffer. The cell lysate was treated for 30 min with 0.16 U/μl of RNase-free DNase I, and then reverse-transcribed with oligo dT primer as described above. The RT-product of the cell lysate was amplified by real-time PCR with IGF-I primers, and the resultant gene expression was normalized to 18SrRNA.
Statistical analysis
Data are expressed as means ± SD, and statistical comparisons using Stat View were based on an analysis of variance (ANOVA) and on Fisher's PLSD test for post hoc comparisons. A p-value < 0.05 was considered significant. A log normal distribution was fitted to each sample set of data, since the logarithmic transformation of real-time PCR data was performed before the statistical analysis.
Results and Discussion
In investigating gene expression in cumulus cells, RNA extraction from a large number of the cumulus cells collected from many COCs [12] may allow us to determine a low copy-number target. However, it is difficult to determine the kinetics of gene expression in cumulus cells during oocyte maturation, when a large number of cumulus cells from COCs at the same stage could not be collected. Therefore, further studies are needed to develop the method for determining gene expression in a limited number of cells. The crude RNA solution (cell lysate) was used as a template for the reverse transcript.
Experiment 1: Construction of a rapid and easy method for determining gene expression using RNA crude solution (cell lysate) as a template.
The aim of this experiment was to develop a rapid and easy method for determining gene expression by using minute quantities of target template in a cell lysate prepared from a limited number of cells, such as cumulus cells surrounding an oocyte. The cell lysate contained genomic DNA, and the contaminating genomic DNA might be amplified in real-time PCR, leading to inaccurate quantification of the transcript of interest. In general, the confounding signal from the contaminating genomic DNA could be attenuated by using the primer that was designed to cross an exon/exon boundary and/or by using an RNA extraction method, such as the acid guanidinium thiocyanate-phenol-chloroform method. Besides, complete DNase I digestion of contaminating DNA could be used for real-time PCR. In this study, however, a crude RNA solution (cell lysate) that contained an excess of genomic DNA was subjected to RT as a template, because the present experiment focuses on the construction of a simpler method for determining the gene expression using cell lysate without RNA extraction. Furthermore, the primers employed in this experiment were designed for only single exon in order to develop a method in which even primers made from unknown genetic structures and cloned cDNA could be used. Thus, complete digestion of genomic DNA in the cell lysate was a prerequisite for real-time PCR in the present experiment.
The genomic DNA in cell lysate derived from cumulus cells that were collected from COCs aspirated from 2–7 mm follicles of slaughter house ovaries was digested for 15 to 60 min with 0.04 and 0.08 U/μl DNase I, which was provided in the cells-to-cDNA II kit, and the cell lysate was reverse-transcribed without RT enzyme. The RT-minus product of the lysate was then amplified by real-time PCR with 18SrRNA primers to verify the residual gene contamination (Fig 1). Two amplification curves in late cycles and two melting peaks at 85°C were noted in the data acquired from the amplification of the RT-minus product of the lysate for both periods of incubation with 0.08 U/μl of the provided DNase I. These results showed that the gene contaminants in the cell lysate treated for 15 to 60 min with 0.08 U/μl of the provided DNase I were not digested completely. Likewise, there were two reasonable amplification curves in earlier cycles and two larger melting peaks at 85°C in the data amplified from the RT-minus product of the cell lysate that was treated with the commercial-kit-recommended concentration (0.04 U/μl) of the provided DNase I. This result indicates that the contaminating DNA remained high in the cell lysate after digestion (data not shown).
Figure 1 Amplification curves (A) and melting curves (B) of 18SrRNA amplified by real-time PCR using the RT-minus product of the cell lysate treated for 15 min and 60 min with 0.08 U/μl of the DNase I provided in the cells-to-cDNA II kit.
Further, the cell lysate was treated for 15, 30 and 60 min with 0.08 and 0.16 U/μl RNase-free DNase I. The resultant amplification and melting curve analysis indicated that the cell lysates digested for various time periods with 0.08 U/μl RNase-free DNase I were also remained not to digest genomic DNA contaminants (Fig 2). In contrast, no amplification curve or melting peak of 18SrRNA-specific residual gene amplicons was observed in the RT – minus product of the sample treated for 30 and 60 min with 0.16 U/μl of RNase-free DNase I (Fig 3). Thus, it was concluded that digestion with 0.16 U/μl of RNase- free DNase I for 30 min is enough to eliminate genomic DNA in the celll lysate.
Figure 2 Amplification curves (A) and melting curves (B) of 18SrRNA amplified by real-time PCR using the RT-minus product of the cell lysate treated for 15, 30 and 60 min with 0.08 U/μl RNase-free DNase I.
Figure 3 Amplification curves (A) and melting curves (B) of 18SrRNA amplified by real-time PCR using the RT – minus product of the cell lysate treated for 15, 30 and 60 min with 0.16 U/μl RNase-free DNase I.
The primer pair of the target gene for real-time PCR is usually designed to be the optimal Mg concentration ranging from 2 to 5 mM [13-15]. It is known that the optimal concentration of Mg in real-time PCR is liable to be higher than that in block-type PCR[16]. Thus, determination of the optimal Mg concentration for 18SrRNA within the range of expected Mg concentration (2 mM to 5 mM) was carried out based on the melting and amplification curves of 18SrRNA by real-time PCR using the RT plus and minus products of the cell lysate, which was treated for 30 min with 0.16 U/μl RNase-free DNase I (Fig 4). Amplification curves were found to increase the fluorescence intensity in the product with or without RT when an Mg concentration of 2 mM and an annealing temperature of 60°C were used. Single melting peaks at 85 and 76°C were found in the RT-plus and minus products, respectively. When the PCR products resulting from real-time PCR of the RT-plus and minus products of the lysate were analysed by gel-electrophoresis, single bands of each PCR product were exhibited at 120 bp for the RT-plus product and at 60 bp for the RT-minus product. The real-time PCR of the RT-minus product using an Mg concentration of 3 mM–5 mM at the various annealing temperatures (59°C to 61°C) gave a melting peak at the same melting temperature (76°C) as that given when an Mg concentration of 2 mM was used, indicating that the Mg concentration (3 – 5 mM) for real-time PCR was beyond the optimal Mg concentration for the 18SrRNA primer pair (data not shown).
On the other hand, it has been reported that large molar excess cations, especially Mg2+, produce a nonspecific product [17,18]. Thus, it is possible that the nonspecific bands that were amplified by real-time PCR of the RT-minus product using 2–5 mM Mg concentrations as described above may be produced by a large molar excess of cations that may be contained in the cell lysis buffer and/or RT buffer.
The following studies were carried out to optimize the Mg2+ concentration and annealing temperature for the 18SrRNA primer pair of real-time PCR. When an Mg concentration of 1.5 mM and annealing temperature of 56°C were used in real-time PCR, two melting peaks were found at 85 and 77°C, indicating that PCR products amplified at annealing temperature of 56°C contained nonspecific products also (Fig 5). Comparing the specific melting peak for 18SrRNA at 85°C in melting curves amplified at annealing temperature of 60 and 61°C, the peak amplified at annealing temperature of 60°C was larger than that at 61°C. The size of the target gene amplified at 60°C was 120 bp when the size of the target amplicons was certified by 3% agarose gel electrophoresis. These findings showed that an Mg concentration of 1.5 mM, which was slightly low compared with the expected Mg concentration of the primer pair, and an annealing temperature of 60°C were optimal for the determination of 18SrRNA expression using the RT product of the cell lysate.
Figure 4 Amplification curves (A) and melting curves (B) of 18SrRNA amplified by real-time PCR using the RT-minus and plus products of the cell lysate (certified residual DNA contamination-free) at 2 mM MgCl2, and gel-electrophoresis pattern of PCR product (C). Lane 1: 100-bp Ladder size marker; Lane 2: amplification product of RT-minus (60 bp). Lane 3:. amplification product of RT-plus (120 bp).
Figure 5 Melting curve (A) amplified at various annealing temparature and gel-electrophoresis pattern (B) of 18SrRNA amplified at 60° of annealing temperature by real-time PCR using the RT product of the cell lysate treated with 0.16 U/μl RNase-free DNase I for 30 min. Lane 1: 100-bp Ladder size marker; Lane 2: amplification product of 18SrRNA (120 bp) in gel-electrophoresis pattern.
Figure 6 Amplification efficiency and slope of 18SrRNA were assessed by using a reaction mixture of standard curve supplemented with RT-minus control product.
Figure 7 IGF-I expression in cumulus cells surrounding an oocyte during in vitro maturation. Each value represents the mean ± SD. Bars with different letters are significantly different (P < 0.05).
However, the PCR efficiency was low in the standard curve of purified plasmid with the sequence of 18SrRNA using an Mg concentration of 1.5 mM (data not shown). This suggests that the 1.5 mM Mg concentration employed for real-time PCR using purified plasmid might be insufficient to produce the same PCR efficiency as was found with the RT product of the cell lysate. Therefore, the RT-minus product of the cell lysate, which was digested with RNase-free DNase I was added to the reaction mixture for the production of the standard curve to ensure an amplification efficiency similar to that observed in the RT product of the cell lysate (Fig 6). The amplification efficiency of the 18SrRNA gene was 0.97 (slope, -3.391), which was similar to the efficiency (0.99) of the standard curve generated from a dilution series of the RT product of the cell lysate (data not shown). Taken together, these findings suggested that the method involving real-time PCR using the cell lysate treated fo 30 min with 0.16 U/μl RNase-free DNase I allowed accurate quantification of gene expression in a small number of the cells.
Experiment 2: Analysis of IGF-I gene expression in cumulus cells using the method developed.
In this experiment, whether IGF-I gene expression in the cumulus cells surrounding an oocyte during its maturation can actually be measured by the method developed here using cell lysate was tested. An Mg concentration of 1.5 mM and an annealing temperature of 56°C for IGF-I primers proved to produce a high yield of amplicon when assessed in the same way as in Experiment 1. The resultant slope of the standard curve and PCR efficiency were -3.307 and 1.00, respectively.
Transcripts of the target (IGF-I) and reference (18SrRNA) genes were analyzed by real-time PCR using the cell lysate prepared from cultured COCs, and then IGF-I gene expression was normalized by the 18SrRNA gene. Cumulus cells collected from COCs before maturation had low expression ratios of IGF-I genes (Fig 7). IGF-I mRNA expression in the cumulus cells surrounding an oocyte cultured for 8 h was strongly up-regulated, compared to that in cumulus cells collected from COCs before maturation. IGF-I mRNA expression then gradually decreased from 8–24 h of incubation. There was no IGF-I gene expression in the cumulus cells from COCs incubated for 24 h. This is the first report showing that the expression levels of IGF-I in the cumulus cells were dramatically increased during oocyte maturation.
Several growth factors including EGF and IGF-I have been demonstrated to be synthesized in the ovary and to have some effects on cumulus cell functions such as expansion[19] and proliferation [20]. However, Nuttinck et al. recently found that no significant difference was observed in bovine IGF-I gene expression between cumulus cells collected from COCs before maturation and cultured for 24 h, and they then concluded that IGF-I gene expression was unaffected by the differentiation stage of the COCs [21]. In this study, IGF-I mRNA expression in cumulus cells increased abruptly at 8 h of incubation. It is well known that germinal vesicle breakdown (GVBD) occurs in bovine oocytes after 8 h of incubation [22]. This evidence suggests that IGF-I may be responsible for the meiotic resumption of bovine oocytes. This assumption is also supported by several reports showing that IGF-I stimulates the in vitro maturation of cattle [23,24] and rat [25] oocytes, and mediates the meiotic resumption of Xenopus[26].
Conclusion
In conclusion, since IGF-I gene expression in cumulus cells surrounding an oocyte during its maturation can actually be measured by the method that employs cell lysate derived from a limited number of cells, we can conclude that the method described in this study is a useful tool for the rapid and easy analysis of target gene expression.
Authors' contributions
KK was responsible for the design, coordination of the experiments, carried out the molecular genetic studies, and drafted the manuscript. HW carried out the oocyte maturation. HY collaborated in the oocyte maturation and molecular genetic studies. TT was responsible for design and coordination of the study. He analyzed the data and helped to drafted the manuscript. All authors read and approved the final manuscript.
Acknowledgements
We are grateful to the staff of the Meat Inspection Office of Hiroshima city for supplying the bovine ovaries. We also thank Dr. Yoshimura Y, Hiroshima Univ., for helpful suggestions. We thank Mr. Kumar AT and Dr. Aboagla EM for his English proofreading of the manuscript.
==== Refs
Isobe N Terada T Effect of the factor inhibiting germinal vesicle breakdown the disruption of gap junctions and cumulus expansion of cumulus-oocyte complexes cultured in vitro Reproduction 2001 121 249 257 11226049 10.1530/rep.0.1210249
Van den Hurk R Abir R Telfer EE Bevers MM Primate and bovine immature oocytes and follicles as sources of fertilizable oocytes Hum Reprod Update 2000 6 457 474 11045877 10.1093/humupd/6.5.457
Eppig JJ Oocyte control of ovarian follicular development and function in mammals Reproduction 2001 122 829 838 11732978 10.1530/rep.0.1220829
Meinecke B Meinecke-Tillmann S Effects of alpha-amanitin on nuclear maturation of porcine oocytes in vitro J Reprod Fertil 1993 98 195 201 8345464
Rodriguez KF Petters RM Crosier AE Farin CE Roles of gene transcription and PKA subtype activation in maturation of murine oocytes Reproduction 2002 123 799 806 12052234 10.1530/rep.0.1230799
Tatemoto H Terada T Time-dependent effects of cycloheximide and α-amanitin on meiotic resumption and progression in bovine follicular oocytes Theriogenology 1995 43 1107 1113 16727697 10.1016/0093-691X(95)00074-I
Sirard MA Richard F Mayes M Controlling meiotic resumption in bovine oocytes: a review Theriogenology 1998 49 483 497 10732029 10.1016/S0093-691X(97)00420-2
Wolf CJ Farin CE Effect of gonadotropins on the ability of 5,6-dichloro-1-β-D-ribofuranosylbenzimidazole (DRB) to inhibit germinal vesicle breakdown in bovine oocytes Theriogenology 1996 46 759 768 16727940 10.1016/S0093-691X(96)00234-8
Rodriguez KF Farin CE Developmental capacity of bovine cumulus oocyte complexes after transcriptional inhibition of germinal vesicle breakdown Theriogenology 2004 61 1499 1511 15036980 10.1016/j.theriogenology.2003.08.016
Dragovic RA Ritter LJ Schulz SJ Amato F Armstrong DT Gilchrist RB Role of oocyte-secreted growth differentiation factor 9 in the regulation of mouse cumulus expansion Endocrinology 2005 146 2798 2806 15761035 10.1210/en.2005-0098
Zhang XJ Moyse J Jarari N Confino E Kazer R Pentraxin-3 gene expression in cumulus cells may be a molecular marker for egg quality Fertility and Sterility 2003 80 81 10.1016/S0015-0282(03)02016-8
Otsuka Y Yanaihara A Iwasaki S Hasegawa J Yanaihara T Okai T Localization and gene expression of steroid sulfatase by RT-PCR in cumulus cells and relationship to serum FSH levels observed during in vitro fertilization J Exp Clin Assist Reprod 2005 2 6 15823201 10.1186/1743-1050-2-6
Broberg EK Nygårdas M Salmi AA Hukkanen V Low copy number detection of herpes simplex virus type 1 mRNA and mouse Th1 type cytokine mRNAs by Light Cycler quantitative real-time PCR J Virol Methods 2003 112 53 65 12951213 10.1016/S0166-0934(03)00191-5
Bischoff C Lüthy J Altwegg M Baggi F Rapid detection of diarrheagenic E. coli by real-time PCR J Microbiol Methods 2005 61 335 341 15767009 10.1016/j.mimet.2004.12.007
Cui XS Jeong YJ Jun JH Kim NH Insulin-like growth factor-I alters apoptosis related genes and reduces apoptosis in porcine parthenotes developing in vitro Theriogenology 2005 63 1070 1080 15710194 10.1016/j.theriogenology.2004.05.017
Teo I A Choi J W Morlese J Taylor G Shaunak S LightCycler qPCR optimisation for low copy number target DNA J Immunol Methods 2002 270 119 133 12379344 10.1016/S0022-1759(02)00218-1
Inns M A Gelfand D Sninsky J White T PCR Protocols: A guide to methods and applications 1990 Academic Press
Henegariu O Heerema NA Dlouhy SR Vance GH Vogt PH Multiplex PCR: Critical Parameters and Step-by-Step Protocol Bio Techniques 1997 23 504 511
Rieger D Luciano AM Modina S Pocar P Lauria A Gandolfi F The effects of epidermal growth factor and insulin-like growth factor I on the metabolic activity, nuclear maturation and subsequent development of cattle oocytes in vitro J Reprod Fertil 1998 112 123 130 9538337
Kölle S Stojkovic M Boie G Wolf E Sinowatz F Growth Hormone-Related Effects on Apoptosis, Mitosis, and Expression of Connexin 43 in Bovine In Vitro Maturation Cumulus-Oocyte Complexes Biol Reprod 2003 68 1584 1589 12606495 10.1095/biolreprod.102.010264
Nuttinck F Charpigny G Mermillod P Loosfelt H Meduri G Freret S Grimard B Heyman Y Expression of components of the insulin-like growth factor system and gonadotropin receptors in bovine cumulus-oocyte complexes during oocyte maturation Domest Anim Endocrinol 2004 27 179 195 15219936 10.1016/j.domaniend.2004.03.003
Sirard MA Florman HM Leibfried-Rutledge ML Barnes FL Sims ML First NL Timing of nuclear progression and protein synthesis necessary for meiotic maturation of bovine oocytes Biol Reprod 1989 40 1257 1263 2775818 10.1095/biolreprod40.6.1257
Sirotkin AV Taradajnik TE Makarevich AV Bulla J Effect of follicular cells, IGF-I and tyrosine kinase blockers on oocyte maturation Anim Reprod Sci 1998 51 333 344 9686315 10.1016/S0378-4320(98)00071-2
Lorenzo PL Illera MJ Illera JC Illera M Enhancement of cumulus expansion and nuclear maturation during bovine oocyte maturation in vitro by the addition of epidermal growth factor and insulin-like growth factor I J Reprod Fertility 1994 100 697 701
Feng P Catt KJ Knecht M Transforming growth factor-beta stimulates meiotic maturation of the rat oocyte Endocrinology 1988 122 181 186 3275534
Conti M Andersen CB Richard F Mehats C Chun SY Horner K Jin C Tsafriri A Role of cyclic nucleotide signaling in oocyte maturation Mol Cell Endocrinol 2002 187 153 159 11988323 10.1016/S0303-7207(01)00686-4
|
16250920
|
PMC1291399
|
CC BY
|
2021-01-04 16:37:13
|
no
|
Reprod Biol Endocrinol. 2005 Oct 27; 3:59
|
utf-8
|
Reprod Biol Endocrinol
| 2,005 |
10.1186/1477-7827-3-59
|
oa_comm
|
==== Front
Reprod Biol EndocrinolReproductive biology and endocrinology : RB&E1477-7827BioMed Central London 1477-7827-3-601625963810.1186/1477-7827-3-60ResearchImmunohistochemical evidence for an endocrine/paracrine role for ghrelin in the reproductive tissues of sheep Miller David W [email protected] Joanne L [email protected] Yvonne A [email protected] Una [email protected] Alanna [email protected] Clare L [email protected] Richard G [email protected] School of Veterinary and Biomedical Sciences, Murdoch University, South Street, Murdoch, WA, Australia2 Sustainable Livestock Systems Group, Scottish Agricultural College, Bucksburn, Aberdeen, UK3 School of Biological Sciences, University of Aberdeen, Aberdeen, UK4 Early Life Nutrition Group, Rowett Research Institute, Greenburn Rd, Bucksburn, Aberdeen, UK2005 31 10 2005 3 60 60 3 10 2005 31 10 2005 Copyright © 2005 Miller et al; licensee BioMed Central Ltd.2005Miller et al; licensee BioMed Central Ltd.This is an Open Access article distributed under the terms of the Creative Commons Attribution License (), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
Background
The gut hormone, ghrelin, is involved in the neuroendocrine and metabolic responses to hunger. In monogastric species, circulating ghrelin levels show clear meal-related and body weight-related changes. The pattern of secretion and its role in ruminant species is less clear. Ghrelin acts via growth hormone secretagogue receptors (GHSR-1a) to alter food intake, fat utilization, and cellular proliferation. There is also evidence that ghrelin is involved in reproductive function. In the present study we used immunohistochemistry to investigate the presence of ghrelin and GHSR-1a in sheep reproductive tissues. In addition, we examined whether ghrelin and GHSR-1a protein expression is developmentally regulated in the adult and fetal ovine testis, and whether there is an association with markers of cellular proliferation, i.e. stem cell factor (SCF) and proliferating cell nuclear antigen (PCNA).
Methods
Antibodies raised against ghrelin and its functional receptor, GHSR-type 1a, were used in standard immunohistochemical protocols on various reproductive tissues collected from adult and fetal sheep. GHSR-1a mRNA presence was also confirmed by in situ hybridisation. SCF and PCNA immunoexpression was investigated in fetal testicular samples. Adult and fetal testicular immunostaining for ghrelin, GHSR-1a, SCF and PCNA was analysed using computer-aided image analysis. Image analysis data were subjected to one-way ANOVA, with differences in immunostaining between time-points determined by Fisher's least significant difference.
Results
In adult sheep tissue, ghrelin and GHSR-1a immunostaining was detected in the stomach (abomasum), anterior pituitary gland, testis, ovary, and hypothalamic and hindbrain regions of the brain. In the adult testis, there was a significant effect of season (photoperiod) on the level of immunostaining for ghrelin (p < 0.01) and GHSR-1a (p < 0.05). In the fetal sheep testis, there was a significant effect of gestational age on the level of immunostaining for ghrelin (p < 0.001), GHSR-1a (p < 0.05), SCF (p < 0.05) and PCNA (p < 0.01).
Conclusion
Evidence is presented for the presence of ghrelin and its receptor in various reproductive tissues of the adult and fetal sheep. In addition, the data indicate that testicular expression of ghrelin and its receptor is physiologically regulated in the adult and developmentally regulated in the fetus. Therefore, the ghrelin ligand/receptor system may have a role (endocrine and/or paracrine) in the development (cellular proliferation) and function of the reproductive axis of the sheep.
==== Body
Background
Ghrelin is an acylated polypeptide hormone secreted predominantly by endocrine cells of the stomach [1,2]. Several lines of evidence implicate ghrelin in the regulation of growth hormone (GH) release, energy balance, food intake and body weight [3-6], with the effects mediated via a 7-transmembrane G protein-coupled receptor, the GH secretagogue receptor (GHSR) type 1a [2]. Evidence is also accumulating to suggest that ghrelin may play a role in the central regulation of reproduction. For example, exogenous ghrelin has been shown to inhibit luteinising hormone (LH) secretion in rats, both in vivo and in vitro [7,8], and immunoreactivity and gene expression for ghrelin and its functional receptor have been found in the hypothalamic region of the rat and human brain, an area known to be important in the control of reproduction [2,9-11]. There may also be peripheral (paracrine) effects of ghrelin on the reproductive axis, with reports of the ghrelin ligand/receptor system being present in rat and human gonads [12-15]. Although ruminant species also appear to utilise the ghrelin system to modulate neuroendocrine and metabolic responses to hunger [1,6,16-20], little is known of the tissue distribution of ghrelin and its receptor, nor of its link to the reproductive axis in these species.
Although ghrelin is predominantly produced by the stomach, gene expression for ghrelin and its receptor in human and rat tissues indicates a widespread distribution in various organs, including the intestine, heart, kidney, liver, lung, pancreas, placenta, pituitary, gonads and brain [21-23]. The peripheral role of ghrelin is unclear, although it has been indicated that it is involved in the control of cellular apoptosis and proliferation [24-29]. The expression of ghrelin receptor in the rat testis has also been shown to be up-regulated around the time of pubertal gonad development [30]. In sheep, the non-breeding season (which is primarily regulated by photoperiod) is associated with a marked reduction in testis weight, fewer germ cells maturing beyond the spermatocyte stages, and decreased steroidogenic activity of the Leydig cells [31]. The possible paracrine role of ghrelin in the testis in this seasonal process has not been investigated.
Development of the reproductive organs during fetal and post-natal life is essential for normal sexual function in adulthood. In rats, it is known that the fetus and placenta secrete ghrelin, and that the ghrelin receptor is present in fetal tissues [32-35]. It is speculated that alteration in maternal, placental or fetal ghrelin during pregnancy, such as that caused by maternal feed restriction [36-38], might contribute to programming of adult infertility via central or peripheral mechanisms. One such mechanism could involve a link between ghrelin and cell proliferation in the control of fetal testicular development. Ghrelin has been shown to specifically inhibit the proliferative activity of immature Leydig cells by down-regulating stem cell factor (SCF) gene expression [39]. It remains equivocal whether the ghrelin system is developmentally regulated in fetal reproductive organs, and whether there are interactions with other paracrine regulators of gonadal development such as SCF.
In the present study we used immunohistochemistry to test the hypothesis that ghrelin and its functional receptor, GHSR-1a, are present in tissues of the ovine reproductive axis, namely in adult hypothalamus, pituitary and gonads. In addition, we examined whether (a) ghrelin and GHSR-1a protein expression is developmentally or physiologically regulated in the adult and fetal ovine testis and (b) ghrelin expression is linked to proliferative activity in the developing fetal testis.
Methods
Animals and tissue collection
All experimental procedures involving animals were conducted under the authority of the Animals (Scientific Procedures) Act, UK, 1986 after Home Office and local ethical committee approval. Tissue samples including stomach (abomasum; i.e. control tissue), brain (hypothalamus & hindbrain), pituitary gland, testis and ovary from adult Suffolk-cross sheep collected from a local abattoir, were immersion-fixed in Bouin's solution for 6 h. Testicular tissue samples from 12 reproductively active (short-day photoperiod, 8 h light:16 h dark) and 10 reproductively regressed (long-day photoperiod, 16 h light:8 h dark) adult Soay rams (2 years old) housed under these artificial lighting conditions for 12 weeks were also collected [40] and immersion-fixed in Bouin's solution for 6 h. Testicular tissue samples for in situ hybridisation analysis of GHSR-1a mRNA expression were also collected, immediately frozen on dry ice and stored at -80C. Fetal testicular tissue samples were collected from Greyface ewes killed, as part of another study [41], on days 30, 40, 50, 70, 100 and 140 of gestation (term ≈ day 145). Fetuses (n = 7 male fetuses at each gestational age) were quickly recovered, immediately exsanguinated, and either the hind torsos were fixed intact (days 30 and 40 of gestation) or testes were fixed following dissection (on days 50, 70, 100 and 140 of gestation). Fetal testicular tissue samples were immersion-fixed in Bouin's solution for 6 h. After fixation, all tissue samples were rinsed and then stored in 70% alcohol before processing and embedding into paraffin wax using standard procedures. Sections (5 μm) were cut and mounted on poly-L-lysine coated glass slides and dried overnight at 42°C prior to immunohistochemical analysis.
Immunohistochemistry
Tissue sections were dewaxed in Histoclear, re-hydrated through a graded ethanol series (100%, 95% &, 70%) and washed in Tris-buffered saline for 2 × 5 min. Where antigen retrieval procedures were necessary (see below), this was achieved by microwaving (800 W) the tissue sections in 0.01 M citrate buffer on full power for 3 × 5 minutes, followed by a 20 minute rest period. Sections were placed in a DAKO autostainer and incubated at room temperature with the appropriate primary antibodies as follows: (a) anti-human ghrelin (Phoenix Europe GmbH, Karlsruhe, Germany) at a 1:600 dilution for 30 mins, (b) anti-human growth hormone secretagogue receptor-1a (Phoenix Europe GmbH, Karlsruhe, Germany) at a 1:400 dilution, (c) anti-ovine SCF (kindly supplied by Dr K McNatty, Wallaceville Animal Research Station, Upper Hutt, New Zealand, and characterised by Tisdall et al. [42]), at a 1:450 dilution and (d) anti-rat PCNA (Novacastra Laboratories Ltd., Newcastle, UK) at a 1:100 dilution. Ghrelin peptide sequences are highly conserved between species [43,44] and the cross-reactivity and specificity of the anti-human ghrelin and GHSR-1a antibodies were tested using control ovine stomach tissue.
Ghrelin and SCF immunoreactivity required antigen retrieval and used the DAKO ChemMate peroxidase/DAB detection system (DakoCytomation Ltd, Ely, UK). In brief, this comprised a peroxidase block step with 3% hydrogen peroxide for 5 mins, followed by a 30 minute incubation with secondary biotinylated link antibody (ChemMate A solution) and a 30 minute incubation with peroxidase substrate (ChemMate B solution). In between each of these steps the slides were rinsed with Tris-buffered saline (TBS). Diaminobenzidine tetrahydrochloride (DAB: DakoCytomation Ltd, Ely, UK) was then applied in two 5 minute incubations. Sections were counterstained with haematoxylin Z (Vector Laboratories Ltd, Peterborough, UK). The negative controls were produced by substituting the primary antibody with normal rabbit serum at the same dilution as the primary antibody.
Ghrelin receptor (GHSR-1a) and PCNA immunoreactivity were detected using the Vectastain Elite ABC kit (Vector Laboratories Ltd, Peterborough, UK) The protocol was as described above with exception of the following: the secondary antibody was a biotinylated rabbit anti-goat IgG (Vector Laboratories Ltd, Peterborough, UK) in a 1:200 dilution. The tissue sections were then incubated with the Vectastain Elite ABC reagent (Vector Laboratories Ltd, Peterborough, UK) for 30 minutes before undergoing the final steps (as above). Negative controls were produced by substituting the primary antibody with normal rabbit or horse serum (1:200).
To determine antibody specificity, all primary antibodies were incubated overnight with immunising peptide. Following incubation, antibody-antigen preparations were centrifuged and the supernatant applied to selected tissue sections.
Testicular GHSR-1a gene expression
Messenger RNA levels were quantified by in situ hybridisation using techniques previously described in detail by Mercer et al. [45]. A riboprobe complementary to GHSR-1a was generated from cloned cDNA from the hypothalamus of rat [46]. The sequence of this cDNA fragment used by Tups and colleagues [46] was compared to the same region of the ovine gene using the Multalin multiple sequence alignment website. Across this region of the GHSR-1a gene, there was 90% homology between rat and sheep.
Frozen adult sheep testicular sections (20 μm) were cut onto glass slides double-coated with gelatin and poly-L-lysine, with six or seven sections mounted on each slide. Briefly, slides were fixed in cold NBF, acetylated, and hybridised overnight at 58°C using [35S]-labelled cRNA probes (2 × 107 c.p.m./ml). Slides were treated with RNase A, desalted, with a final high stringency wash (30 min) in 0.1 × SSC at 60°C, dried and exposed for 4 weeks to Kodak Biomax MR Film (Kodak, Rochester, NY, USA). Autoradiographic images were captured using the Image-Pro Plus system (Media Cybernetics, Maryland, USA).
Image and statistical analysis
Immunostained area for ghrelin and GHSR-1a in adult testis tissue from the Soay rams, and ghrelin, GHSR-1a, SCF and PCNA in fetal testis tissue was analysed using computer aided image analysis. The system was composed of a Zeiss axioplan microscope (Zeiss, West Germany) and HV-C20 Hitachi camera (Hitachi, Japan) connected to a computer running Image-Pro Plus system. Four sections per testis were quantified for tissue section area stained (brown colour) over six randomly selected fields of view previously shown to be sufficient to stabilise the mean and standard error [47]. The total area of positively stained cells (brown) was selected and expressed as a sum of pixels. All nuclei were then selected (brown = immunopositive + blue = haematoxylin Z) and the positively stained cells were expressed as a percentage of the total.
Image analysis data were subjected to one-way ANOVA, with differences between testicular immunostaining levels of ghrelin, GHSR-1a, SCF and PCNA at different stages of gestation in the fetuses determined by Fisher's least significant difference.
Results
Antibody specificity
The specificity of the three antibodies (anti-ghrelin, anti-GHSR1a and anti-SCF) was confirmed by the absence of immunostaining when the primary antibody was replaced with serum from the species in which the antibody was raised. Additionally, immunostaining was abolished when the antisera were pre-incubated with the immunising peptide (Fig. 1). In all tissues investigated, the ghrelin immunopositive cells exhibited perinuclear staining and/or more widespread cytoplasmic staining. At high magnification it was confirmed that the nuclei were indeed negative (Fig. 1). The GHSR-1a and SCF positive cells both exhibited cytoplasmic staining.
Figure 1 Immunohistochemical localisation of ghrelin and GHSR-1a in the sheep stomach and pituitary gland. Photomicrographs of sheep stomach (abomasum) and anterior pituitary (AP) sections immunostained with antibodies against either ghrelin or GHSR-1a. (A) Stomach section showing positive staining for ghrelin (brown) in tunica mucosa (MUC), tunica submucosa (SUB) and tunica muscularis (MUSC). (B) Positive immunostaining in the stomach was abolished when the antiserum was pre-incubated with the immunising peptide (ghrelin). (C) High magnification micrograph of MUC showing the cytoplasmic and perinuclear (arrow) nature of the immunostaining for ghrelin. (D) AP section showing positive staining for ghrelin in most cells. (E) AP section showing positive staining for GHSR-1a in most cells. (F) Positive immunostaining for GHSR-1a in the AP was abolished when the antiserum was pre-incubated with the immunising peptide. (G) High magnification micrograph of the AP showing the cytoplasmic and perinuclear (arrows) nature of the immunostaining for GHSR-1a. The scale bar of A represents 100 μm, for B, D, E, F they represent 50 μm, and for C, G they represent 20 μm. The insert is the negative control.
Ghrelin and GHSR-1a in adult sheep tissues
There was general immunopositive staining for ghrelin and GHSR-1a throughout hypothalamic region of the brain, including the median eminence (ME), arcuate nucleus (ARC), ventromedial hypothalamus (VMH) and ependymal lining (EL) of the third cerebral ventricle (Fig. 2). Ghrelin and GHSR-1a were also present in the hindbrain where they were found in distinct neuronal bodies in the nucleus tract solitarus (NTS). The immunoreactivity of ghrelin in the NTS neuronal bodies showed general cytoplasmic staining with some nuclear/perinuclear staining also detectable. (Fig. 2). In the anterior pituitary gland, ghrelin and GHSR-1a immunoreactivity was also present, with GHSR-1a showing cytoplasmic and perinuclear staining in some cells (Fig. 1) In the adult ovary, ghrelin and GHSR-1a were immuno-localised to ovarian follicles at all developmental stages (primordial, primary, secondary, pre-antral and antral). Both proteins were also co-localised to the granulosa cells and some staining was observed in the thecal cells of the larger follicles (Fig. 3). Ghrelin and GHSR-1a immunostaining was also present in the luteal cells of the corpus luteum. In some of the sections their appeared to be some positive staining for ghrelin and GHSR-1a on the oocyte (especially in the larger oocytes). However, this finding was not consistent across all oocytes. In the adult testis, ghrelin immunostaining was predominant in the germ and Sertoli cells, with the germ cells showing intense perinuclear staining (Fig 3). Moreover, there appeared to be more intense ghrelin staining in the germ cells in all developmental stages prior to the first meiotic division. Lower level immunostaining was also observed in the interstitial tissue (where Leydig cells are localised: Fig 3g). GHSR-1a protein was also detected in cord Sertoli, and germ cells. (Fig 3h, Fig 4c,d). In contrast to ghrelin, staining for the receptor appeared to be more predominant in the interstitium (Fig 3h: GHSR-1a v Fig 3g: ghrelin). In addition, in-situ hybridisation for GHSR-1a mRNA showed clear rings of hybridisation within the testis, indicative of gene expression in the interstitial tissue around the seminiferous tubules (Fig. 3i).
Figure 2 Immunohistochemical of ghrelin and GHSR-1a in the sheep brain. Photomicrographs of sheep brain sections. (A) Hypothalamic section including the third ventricle (3V) showing positive staining for ghrelin in the median eminence (ME), arcuate nucleus (ARC) and ventromedial hypothalamus (VMH). (B) Hypothalamic section showing positive staining for GHSR-1a in the ME, ARC and VMH. (C) High magnification micrograph showing the staining for ghrelin in the VMH. (D) High magnification micrograph showing the staining for GHSR-1a in the VMH. (E) High magnification micrograph showing the staining for ghrelin in the ARC, ME and ependymal lining (EL) of the 3V. (F) High magnification micrograph showing the staining for GHSR-1a in the ARC, ME and EL. (G) Hindbrain section including the fourth ventricle (4V) showing positive staining for ghrelin in the nucleus of the tractus solitarus (NTS). (H) Hindbrain section showing positive staining for GHSR-1a in the NTS. (I) High magnification micrograph showing the staining for ghrelin in the NTS. (J) High magnification micrograph showing the staining for GHSR-1a in the NTS. The scale bars of A, B represent 150 μm, for C, D, E, F G, H they represent 50 μm, and for I, J they represent 20 μm. The insert is the negative control.
Figure 3 Immunohistochemical and in situ localisation of ghrelin and GHSR-1a in sheep gonads. Photomicrographs of sheep ovarian and testicular sections. (A) Ovarian section showing positive immunostaining for ghrelin in the stroma, primordial follicles (PF) and the ovarian surface epithelium (OSE). (B) Ovarian section showing positive immunostaining for ghrelin in secondary follicles (SF) and granulosa cells (GC). (C) High magnification micrograph showing positive immunostaining for ghrelin in the corpus luteum (CL). (D) Ovarian section showing positive immunostaining for GHSR-1a in the stroma, PF and OSE. (E) Ovarian section showing positive immunostaining for GHSR-1a in the SF and GC. (F) High magnification micrograph showing positive immunostaining for GHSR-1a in the CL. (G) Testicular section of seminiferous tubules showing positive immunostaining for ghrelin in the Sertoli cells (SC), pre-spermatogonia (PS), round spermatids (RS) and in the interstitium (INT). (H) Testicular section of seminiferous tubules showing positive immunostaining for GHSR-1a in the SC, PS, RS and INT. (I) Autoradiograph of adult testis sections following in situ hybridisation to an antisense 35S-labelled riboprobe to GHSR-1a mRNA showing hybridisation mainly in the interstitial areas between the seminiferous tubules. The scale bars of A, B, D, E, I represent 50 μm, and for C, F, G, H they represent 150 μm The insert is the negative control (in situ sense control for I).
Figure 4 Immunolocalisation of ghrelin and GHSR-1a in the testes from adult sheep maintained in either short- or long-day photoperiods. Photomicrographs of adult sheep testicular sections kept for 12 weeks in either a short day (SD) or long day photoperiod (LD). (A) Section from a SD sheep showing positive staining for ghrelin in the interstitium (INT), Sertoli cells (SC), pre-spermatogonia (PS) and round spermatids (RS). (B) Section from a LD sheep showing reduced staining for ghrelin in the testis compared to the SD sheep. (C) Section from a SD sheep showing positive staining for GHSR-1a. (D) Section from a LD sheep showing reduced staining for GHSR-1a in the testis compared to the SD sheep. (E) The scale bars of A, B, D, E represent 50 μm.
Effect of photoperiod on Ghrelin and GHSR-1a in the adult Soay testis
Testis size was reduced in the LD group (data not shown), as expected [40]. Ghrelin and GHSR-1a immunostaining was demonstrated in germ, Sertoli and interstitial cells of adult Soay testes collected in both the short-day (SD: reproductively active) and long-day (LD: reproductively regressed) photoperiods (Fig. 4). Comparison between the two groups indicated that the number of ghrelin and ghrelin receptor positive cells was reduced in the testes from the LD (reproductively regressed) Soay rams. This was verified after statistical analysis of the image analysis data indicating a significant increase in immunostaining levels in the SD photoperiod for both ghrelin (p < 0.01) and GHSR-1a (p < 0.05) (Fig. 5). More detailed histological examination indicated that the decreased ghrelin staining was predominantly associated with reduced staining of the testicular cords (Sertoli and germ cells). In contrast, photoperiod appeared to influence GHSR-1a staining in both the cords and interstitial areas.
Figure 5 Effect of photoperiod on the percentage of cells positively immunostained for ghrelin and GHSR-1a in adult sheep testes. Ghrelin (A) and GHSR-1a (B) in adult sheep testes (short days, SD versus long days, LD). Values with different alphabetical superscripts are significantly different to one another: a versus b = p < 0.05; a versus c = p < 0.01. Values are means ± S.E.M.
Ghrelin, GHSR-1a, SCF and PCNA in the fetal sheep testis
The ovine fetal testis is visible as an outgrowth of undifferentiated cells from the mesonephros at gestational day 30 (Fig. 6a–d). At this stage it was impossible to visually differentiate between somatic and germ cells. At day 50 (Fig. 6e–h), small testicular cords were distinguishable towards the periphery of the gonadal tissue, and germ cells were present surrounded by cells that will constitute the Sertoli cell population at a later gestational stage. At day 70 (Fig. 6i–l) the testicular cords were more pronounced. Germ and Sertoli cells were identifiable within the cords. Interstitial cells, probably Leydig cells and their precursors, were clearly identifiable.
Figure 6 Immunolocalisation of ghrelin and GHSR-1a in fetal sheep testes. Photomicrographs of fetal sheep testicular sections. (A, B, C, D) Sections from fetuses at day 30 of gestation showing positive immunostaining for ghrelin (A), GHSR-1a (B), SCF (C) and PCNA (D) in the fetal testis and mesonephros (meson.). (E, F, G, H) Testicular sections from fetuses at day 50 of gestation showing positive immunostaining for ghrelin (E), GHSR-1a (F), SCF (G) and PCNA (H). (I, J, K, L) Testicular sections from fetuses at day 70 of gestation showing positive immunostaining for ghrelin (I), GHSR-1a (J), SCF (K) and PCNA (L). (M, N, O, P) Testicular sections from fetuses at day 140 of gestation showing positive immunostaining for ghrelin (M), GHSR-1a (N), SCF (O) and PCNA (P) in the seminiferous cords (C) and interstitium (INT). Arrows depict germ cells. The scale bars of A, B, C, D represent 150 μm, and for the rest they represent 100 μm.
At day 30, low level immunostaining for ghrelin was observed in the gonadal/mesonephros complex (Fig. 6a). From day 70 (Fig. 6i) onwards ghrelin staining intensity increased, and at day 140 (Fig. 6m), ghrelin was clearly localised to the Sertoli, germ and some interstitial cells. Maximum GHSR-1a immunostaining intensity was evident at day 50 (Fig. 6b), although lower levels were present at all other gestational time points. From day 70 onwards, positive staining for GHSR-1a was restricted to the Sertoli and interstitial cells (Fig. 6j,n)
Cells stained for SCF were most prominent in the fetal testis at day 30 (Fig. 6c). The later gestational stages appeared to have fewer cells of lower staining intensity. From day 70 onwards SCF immunopositive cells were localised to the interstitium and to the endothelial cells of the blood vessels. PCNA-immunopositive cells were detected in the gonadal/mesonephros complex at day 30 (Fig. 6d) and was predominantly localised to the seminiferous cords at the later gestational ages examined. Positive PCNA staining was seen in the Sertoli and germ cells and staining was also evident in the Leydig cell containing interstitial area (Fig. 6h,l,p).
There was a significant overall effect of gestational age on the immunostaining levels of ghrelin in the fetal testis (Fig. 7a: p < 0.001), GHSR-1a (Fig. 7b: p < 0.05), SCF (Fig. 7c: p < 0.05) and PCNA (Fig. 7d: p < 0.01). At day 30, ghrelin staining levels were about 2 times higher (p < 0.05) than the nadir levels at days 40 and 50. From days 70 to 140, ghrelin staining levels increased reaching peak fetal levels at day 140 that were about 7 times higher than the nadir level at day 40 (p < 0.005). In comparison, GHSR-1a immunostaining intensity levels appeared to have the opposite temporal pattern to ghrelin, with the peak levels at days 40 to 50, which were about 3 to 4 times higher than at any other time during gestation (p < 0.05). The fetal staining intensity pattern for SCF was similar to ghrelin, with immunostaining intensity levels at day 30 being 3 to 4 times higher than the nadir levels at day 40 to 50 of gestation (p < 0.05), whilst the staining intensity pattern for PCNA was similar to GHSR-1a, with immunostaining intensity levels peaking at day 50 to 70, which were about 50 to 80% higher than at days 30, 40 and 100 (p < 0.05), and about 4 times higher than at day 140 (p < 0.01).
Figure 7 Effect of gestational age on the percentage of fetal testicular cells positively immunostained for ghrelin, GHSR-1a, SCF and PCNA. Ghrelin (A), GHSR-1a (B), SCF (C) and PCNA (D) in fetal sheep testes (days 30 – 140 of gestation; n = 7 at each gestational age). Values with different alphabetical superscripts are significantly different to one another: a versus b, b versus c, and c versus d = p < 0.05; a versus c, and b versus d = p < 0.01; a versus d = p < 0.005. Values are means ± S.E.M.
Discussion
The present study provides immunohistochemical evidence for the presence of ghrelin and its functional receptor, growth hormone secretagogue receptor (GHSR-1a), in reproductive tissues of the sheep. Novel data are presented that indicate that ghrelin and its functional receptor, GHSR-1a, are regulated during development of the ovine fetal testis and by seasonal developmental changes in the adult testis. Moreover, the developmental pattern of expression corresponds with the postulate by Barreiro and colleagues [39] that the ghrelin system is linked to the proliferative activity of germ and somatic cells in the testis.
In the sheep brain, ghrelin and GHSR-1a immunoreactivity were demonstrated in the hypothalamic region, including the median eminence (ME), arcuate nucleus (ARC), ventromedial hypothalamus (VMH) and ependymal lining (EL) of the third cerebral ventricle. The presence of ghrelin and GHSR-1a in the hypothalamus is consistent with its role in the regulation of food intake [48]. However, the hypothalamic localisation also corresponds with ghrelin's putative role in the control of LH secretion, as neurons that secrete gonadotrophin-releasing hormone (GnRH) are also located in these hypothalamic regions [49,50]. Central administration of ghrelin has been shown to rapidly suppress pulsatile LH secretion in ovariectomized rats [8]. Ghrelin and GHSR-1a mRNA and protein [2,9-11,51-53] have previously been found in the rat and human hypothalamus. In the sheep hindbrain, ghrelin and GHSR-1a immunostaining was also identified in neuronal cell bodies within the nucleus tract solitarus (NTS), but not in the area postrema. Similarly, an immunohistochemical study in rats by Lin et al. [54] revealed that the GHSR-1a was expressed in the neuronal cells of the NTS and the dorsal motor nucleus of the vagus, but not in the cells of the area postrema. Using c-fos immunohistochemistry, Lawrence et al. [55] demonstrated that central ghrelin administration activated two regions of the brainstem, the NTS and the area postrema. It is tempting to speculate that ghrelin may affect food intake and the neuroendocrine system at the level of the NTS, a central nervous system (CNS) site that receives primary vagal afferent input from the digestive tract and acts as a neuronal relay station to the hypothalamus [56]. However, it could also be possible that ghrelin is simply participating in the central regulation of gastric acid secretion via the vagus system. ICV administration of ghrelin in rats increases gastric acid output in a dose-dependent manner, and vagotomy and the administration of atropine abolishes gastric acid secretion induced by ghrelin [57].
In the sheep anterior pituitary gland, ghrelin and GHSR-1a were ubiquitously expressed. Ghrelin gene expression has been found in the pituitary glands of rodents, pigs and humans [13,21,58]. It is evident from the literature that the main action of ghrelin at the level of the pituitary is the release of GH [59,60]. Indeed, GH secretagogue receptors were first identified in the pituitary by the ability of enkephalin analogues to stimulate GH release [60]. However, ghrelin may also be involved in somatotroph cell differentiation since ghrelin regulates pituitary-specific transcription factor (Pit-1) expression in the rat pituitary [62]. Somatotroph cell-specific expression of the GH gene is dependent on Pit-1 [63,64]. Ghrelin has also been shown to exert a proliferative effect on a rat pituitary somatotroph cell line via the mitogen-activated protein kinase (MAPK) pathway [27]. Interestingly, ghrelin may also play a role in LH secretion at the level of the pituitary as it has been shown that ghrelin stimulates gonadotrophin release from rat pituitary cells in vitro [7]. Particularly evident in the sheep pituitary tissue was the finding that GHSR-1a immunoreactivity in some cells of the pituitary showed intense perinuclear staining. This staining pattern may arise if the polyclonal antibody is detecting the ligand/receptor complex and is consistent with the finding of Camina et al. [65] that the ghrelin/GHSR-1a complex progressively disappears from the plasma membrane after binding of the ligand and accumulates in the perinuclear region.
Ghrelin and GHSR-1a immunostaining were detected in sheep ovarian follicles at all developmental stages, mainly in the granulosa cells. Caminos et al. [13] and Gaytan et al. [14] also found immunoreactivity for ghrelin and its receptor in the rat and human ovary, though only weak ghrelin staining was evident in the follicles. Strong ghrelin and GHSR-1a immunostaining was evident in corpora lutea (CL) of the sheep ovary, similar to the findings in the rat and human CL. Caminos et al. [13] found dynamic changes in the profile of ghrelin expression during the oestrous cycle and throughout pregnancy in rats, suggesting a precise regulation of ovarian expression of ghrelin. Gaytan and colleagues [14] have suggested a potential, yet unproven relationship between GHSR-1a expression and follicle growth, since expression of the receptor in somatic cells derived from ovarian follicles roughly parallels follicular development in the human ovary. There was additional evidence for ghrelin and GHSR-1a peptide expression within the ovarian surface epithelium (OSE) in the present study. Gaytan et al. [66] have also recently shown GHSR-1a peptide expression in human OSE. During ovulatory cycles the OSE is subject to a series of injury and repair processes associated with follicular rupture and CL formation, which involve natural inflammatory events. Pro-inflammatory IL-1α produces an increase in mRNA levels of 11 betahydroxysteriod dehydrogenase type 1 (11βHSD1) in OSE cells, encoding the steroid dehydrogenase that reversibly reduces cortisone to anti-inflammatory cortisol [67]. Tan and colleagues [68] have recently demonstrated the ability of ghrelin to stimulate 11βHSD1 in the human ovary, providing evidence to support the theory that ghrelin plays an immunomodulatory role in OSE cells of the ovary via an anti-inflammatory action.
In the adult sheep testis, strong ghrelin immunostaining was evident in the interstitial area where Leydig cells are localised. Staining was also present in the germ and Sertoli cells, with an indication of increased ghrelin immunoreactivity in the germ cells during the mitotic phases and meiotic pro-phases of the spermatogenic cycle. GHSR-1a protein was detected in the interstitial Leydig cell containing area of the testis, as well as in the Sertoli and germ cells within the tubules, and the pattern of GHSR-1a mRNA expression across the testis indicated that the mRNA was present in the interstitial area and around the periphery of the tubules. Ghrelin immunostaining has been demonstrated in interstitial Leydig cells and, at lower intensity, in Sertoli cells of the rat [15]. Ghrelin and its receptor are present in the human testis, but in contrast to the ovine data, ghrelin protein is not detectable in germ cells at any stage of spermatogenesis [21]. There appears, therefore, to be some species differences in the localisation of ghrelin protein in the testis. Ghrelin has been shown to dose-dependently inhibit testicular testosterone secretion in vitro, and to modulate Leydig cell proliferation in vivo and the expression of relevant testicular genes, such as that encoding stem cell factor (SCF) [69]. In the testicular samples collected from Soay rams maintained in different photoperiods (short day = reproductively active, and long day = reproductively regressed), it was evident that ghrelin and GHSR-1a were up-regulated in the short-day photoperiod. This finding corresponds with the postulated role of ghrelin in the proliferation of somatic and germ cells in the testis [15,30]. It has been suggested that the expression of ghrelin peptide in mature Leydig cells in the rat testis is under the hormonal regulation of pituitary LH [12]. Whether this action is carried out directly, or is mediated by LH-driven locally produced factors, such as testosterone, requires further investigation. It is pertinent that this observation accompanies an increase in testis size which characterises reproductively active animals [40].
In the fetal sheep testis, the present findings indicate that the expression of ghrelin and GHSR-1a protein is linked to gestational age. Differentiation of the fetal gonad begins with the development of the gonadal ridges from thickening of the ventrolateral surface of the embryonic mesonephros. The genital ridge is composed of somatic cells from the mesonephros and migratory primordial germ cells originating from the extraembryonic mesoderm. In the fetal sheep, morphological sexual differentiation of the gonads begins around day 27 of gestation, and germ cell migration is complete just after day 30 [70,71]. In the present study, significant levels of ghrelin and SCF were found in the gonad/mesonephros complex at day 30 of gestation. Ghrelin has been implicated in proliferative activity in a number of tissues, including the gonad [27-29,39]. SCF is also said to be involved in proliferation, germ cell migration and survival [72]. Therefore, ghrelin and SCF may be involved in the early differentiation of the ovine fetal testis. Intratesticular injection of ghrelin has also been shown to decrease the proliferative activity of differentiating immature Leydig cells in the rat, and this response is associated with a decrease in the mRNA levels of SCF, a putative regulator of Leydig cell development [39]. In the fetal sheep, Leydig cell hyperplasia occurs between day 50 and 70 of gestation [73]. In the present study it was noted that at day 50, GHSR-1a and PCNA were significantly up-regulated at the same gestational time-point when SCF protein levels were at their lowest, i.e. at the start of Leydig cell hyperplasia. These data are consistent with the putative involvement of fetal testicular ghrelin and its functional receptor in the paracrine regulation of Leydig and germ cell development, possibly via interactions with SCF and PCNA.
Conclusion
The present data indicate that both components (ligand and receptor) of the ghrelin signalling system are present in tissues of the reproductive axis of the sheep. The ligand and receptor are developmentally regulated in the fetal testis and physiologically regulated by photoperiod in the adult testis. These findings are consistent with both a central endocrine role and a potential peripheral paracrine/endocrine regulatory role for ghrelin in the control of reproductive tissue development and function in sheep. Further studies are needed to identify the precise functional role of the ghrelin system in the reproductive axis.
Authors' contributions
DWM participated in the design of the studies, collection of tissues, all immunological analyses, statistical analyses and drafting the manuscript. JLH participated in the design of the studies, collection of tissues, immunological and in situ analyses of ghrelin and GHSR-1a, statistical analyses and drafting the manuscript. YAB participated in the immunological analysis of GHSR-1a, statistical analysis and drafting the manuscript. UD participated in the immunological analysis of SCF, statistical analysis and drafting the manuscript. AL participated in the immunological analysis of PCNA, statistical analysis and drafting the manuscript. RGL participated in the design of the studies, all immunological analyses, statistical analyses and drafting the manuscript. CLA participated in the design of the studies, collection of tissues, in situ analysis, statistical analyses and drafting the manuscript.
Acknowledgements
The authors of this article would like to thank Lisa Hannah and Patricia Findlay for help with tissue collection, immunohistochemical and in situ analyses, and Dr A. Nigel Brooks at ZENECA Central Toxicology Laboratory, Macclesfield, Cheshire, UK, for kindly providing the fetal testicular tissue samples. The polyclonal antibody for ovine SCF was kindly supplied by Dr Ken McNatty, Wallaceville Animal Research Station, Upper Hutt, New Zealand, and the GHSR-1a cDNA was generously provided by Dr Zoe Archer, Rowett Research Institute.
==== Refs
Hayashida T Murakami K Mogi K Nishihara M Nakazato M Mondal MS Horii Y Kojima M Kangawa K Murakami N Ghrelin in domestic animals: distribution in stomach and its possible role Domest Anim Endocrinol 2001 21 17 24 11524171 10.1016/S0739-7240(01)00104-7
Kojima M Hosoda H Date Y Nakazato M Matsuo H Kangawa K Ghrelin is a growth-hormone-releasing acylated peptide from stomach Nature 1999 402 656 660 10604470 10.1038/45230
Arvat E Broglio F Aimaretti G Benso A Giordano R Deghenghi R Ghigo E Ghrelin and synthetic GH secretagogues Best Pract Res Clin Endocrinol Metab 2002 16 505 517 12464231 10.1053/beem.2002.0213
Cummings DE Shannon MH Roles for ghrelin in the regulation of appetite and body weight Arch Surg 2003 138 389 396 12686525 10.1001/archsurg.138.4.389
Hosoda H Kojima M Kangawa K Ghrelin and the regulation of food intake and energy balance Mol Interv 2002 2 494 503 14993401 10.1124/mi.2.8.494
Sugino T Hasegawa Y Kurose Y Kojima M Kangawa K Terashima Y Effects of ghrelin on food intake and neuroendocrine function in sheep Anim Reprod Sci 2004 82–83 183 194 10.1016/j.anireprosci.2004.05.001
Fernandez-Fernandez R Tena-Sempere M Aguilar E Pinilla L Ghrelin effects on gonadotropin secretion in male and female rats Neurosci Lett 2004 362 103 107 15193764 10.1016/j.neulet.2004.03.003
Furuta M Funabashi T Kimura F Intracerebroventricular administration of ghrelin rapidly suppresses pulsatile luteinizing hormone secretion in ovariectomized rats Biochem Biophys Res Commun 2001 288 780 785 11688975 10.1006/bbrc.2001.5854
Cowley MA Smith RG Diano S Tschop M Pronchuk N Grove KL Strasburger CJ Bidlingmaier M Esterman M Heiman ML Garcia-Segura LM Nillni EA Mendez P Low MJ Sotonyi P Friedman JM Liu H Pinto S Colmers WF Cone RD Horvath TL The distribution and mechanism of action of ghrelin in the CNS demonstrates a novel hypothalamic circuit regulating energy homeostasis Neuron 2003 37 649 661 12597862 10.1016/S0896-6273(03)00063-1
Lu S Guan JL Wang QP Uehara K Yamada S Goto N Date Y Nakazato M Kojima M Kangawa K Shioda S Immunocytochemical observation of ghrelin-containing neurons in the rat arcuate nucleus Neurosci Lett 2002 321 157 160 11880196 10.1016/S0304-3940(01)02544-7
Mozid AM Tringali G Forsling ML Hendricks MS Ajodha S Edwards R Navarra P Grossman AB Korbonits M Ghrelin is released from rat hypothalamic explants and stimulates corticotrophin-releasing hormone and arginine-vasopressin Horm Metab Res 2003 35 455 459 12953161 10.1055/s-2003-41801
Barreiro ML Gaytan F Caminos JE Pinilla L Casanueva FF Aguilar E Dieguez C Tena-Sempere M Cellular location and hormonal regulation of ghrelin expression in rat testis Biol Reprod 2002 67 1768 1776 12444052 10.1095/biolreprod.102.006965
Caminos JE Tena-Sempere M Gaytan F Sanchez-Criado JE Barreiro ML Nogueiras R Casanueva FF Aguilar E Dieguez C Expression of ghrelin in the cyclic and pregnant rat ovary Endocrinol 2003 144 1594 1602 10.1210/en.2002-221058
Gaytan F Barreiro ML Chopin LK Herington AC Morales C Pinilla L Casanueva FF Aguilar E Dieguez C Tena-Sempere M Immunolocalization of ghrelin and its functional receptor, the type 1a growth hormone secretagogue receptor, in the cyclic human ovary J Clin Endocrinol Metab 2003 88 879 887 12574228 10.1210/jc.2002-021196
Gaytan F Barreiro ML Caminos JE Chopin LK Herington AC Morales C Pinilla L Paniagua R Nistal M Casanueva FF Aguilar E Dieguez C Tena-Sempere M Expression of ghrelin and its functional receptor, the type 1a growth hormone secretagogue receptor, in normal human testis and testicular tumors J Clin Endocrinol Metab 2004 89 400 409 14715878 10.1210/jc.2003-031375
Hashizume T Horiuchi M Tate N Nonaka S Kojima M Hosoda H Kangawa K Effects of ghrelin on growth hormone secretion from cultured adenohypophysial cells in cattle Endocr J 2003 50 289 295 12940457 10.1507/endocrj.50.289
Miura H Tsuchiya N Sasaki I Kikuchi M Kojima M Kangawa K Hasegawa Y Ohnami Y Changes in plasma ghrelin and growth hormone concentrations in mature Holstein cows and three-month-old calves J Anim Sci 2004 82 1329 1333 15144072
Sugino T Hasegawa Y Kikkawa Y Yamaura J Yamagishi M Kurose Y Kojima M Kangawa K Terashima Y A transient ghrelin surge occurs just before feeding in a scheduled meal-fed sheep Biochem Biophys Res Commun 2002 295 255 260 12150940 10.1016/S0006-291X(02)00654-X
Sugino T Yamaura J Yamagishi M Ogura A Hayashi R Kurose Y Kojima M Kangawa K Hasegawa Y Terashima Y A transient surge of ghrelin secretion before feeding is modified by different feeding regimens in sheep Biochem Biophys Res Commun 2002 298 785 788 12419323 10.1016/S0006-291X(02)02572-X
Sugino T Yamaura J Yamagishi M Kurose Y Kojima M Kangawa K Hasegawa Y Terashima Y Involvement of cholinergic neurons in the regulation of the ghrelin secretory response to feeding in sheep Biochem Biophys Res Commun 2003 304 308 312 12711315 10.1016/S0006-291X(03)00593-X
Gnanapavan S Kola B Bustin SA Morris DG McGee P Fairclough P Bhattacharya S Carpenter R Grossman AB Korbonits M The tissue distribution of the mRNA of ghrelin and subtypes of its receptor, GHS-R, in humans J Clin Endocrinol Metab 2002 87 2988 2991 12050285 10.1210/jc.87.6.2988
Kojima M Hosoda H Kangawa K Purification and distribution of ghrelin: the natural endogenous ligand for the growth hormone secretagogue receptor Horm Res 2001 56 93 97 11786694 10.1159/000048143
Papotti M Ghe C Cassoni P Catapano F Deghenghi R Ghigo E Muccioli G Growth hormone secretagogue binding sites in peripheral human tissues J Clin Endocrinol Metab 2000 85 3803 3807 11061542 10.1210/jc.85.10.3803
Andreis PG Malendowicz LK Trejter M Neri G Spinazzi R Rossi GP Nussdorfer GG Ghrelin and growth hormone secretagogue receptor are expressed in the rat adrenal cortex: evidence that ghrelin stimulates the growth, but not the secretory activity of adrenal cells FEBS Lett 2003 536 173 179 12586359 10.1016/S0014-5793(03)00051-6
Baldanzi G Filigheddu N Cutrupi S Catapano F Bonissoni S Fubini A Malan D Baj G Granata R Broglio F Papotti M Surico N Bussolino F Isgaard J Deghenghi R Sinigaglia F Prat M Muccioli G Ghigo E Graziani A Ghrelin and des-acyl ghrelin inhibit cell death in cardiomyocytes and endothelial cells through ERK1/2 and PI 3-kinase/AKT J Cell Biol 2002 159 1029 1037 12486113 10.1083/jcb.200207165
Kim MS Yoon CY Jang PG Park YJ Shin CS Park HS Ryu JW Pak YK Park JY Lee KU Kim SY Lee HK Kim YB Park KS The mitogenic and antiapoptotic actions of ghrelin in 3T3-L1 adipocytes Mol Endocrinol 2004 18 2291 301 15178745 10.1210/me.2003-0459
Nanzer AM Khalaf S Mozid AM Fowkes RC Patel MV Burrin JM Grossman AB Korbonits M Ghrelin exerts a proliferative effect on a rat pituitary somatotroph cell line via the mitogen-activated protein kinase pathway Eur J Endocrinol 2004 151 233 240 15296479 10.1530/eje.0.1510233
Xia Q Pang W Pan H Zheng Y Kang JS Zhu SG Effects of ghrelin on the proliferation and secretion of splenic T lymphocytes in mice Regul Pept 2004 122 173 178 15491788 10.1016/j.regpep.2004.06.016
Zhang W Lin TR Hu Y Fan Y Zhao L Stuenkel EL Mulholland MW Ghrelin stimulates neurogenesis in the dorsal motor nucleus of the vagus J Physiol 2004 559 729 737 15272046
Barreiro ML Suominen JS Gaytan F Pinilla L Chopin LK Casanueva FF Dieguez C Aguilar E Toppari J Tena-Sempere M Developmental, stage-specific, and hormonally regulated expression of growth hormone secretagogue receptor messenger RNA in rat testis Biol Reprod 2003 68 1631 1640 12606422 10.1095/biolreprod.102.008862
Mortimer D Lincoln GA Ultrastructural study of regressed and reactivated testes from Soay rams J Reprod Fertil 1982 64 437 42 7069660
Cortelazzi D Cappiello V Morpurgo PS Ronzoni S Nobile De Santis MS Cetin I Beck-Peccoz P Spada A Circulating levels of ghrelin in human fetuses Eur J Endocrinol 2003 149 111 116 12887287 10.1530/eje.0.1490111
Katayama M Nogami H Nishiyama J Kawase T Kawamura K Developmentally and regionally regulated expression of growth hormone secretagogue receptor mRNA in rat brain and pituitary gland Neuroendocrinol 2000 72 333 340 10.1159/000054602
Rindi G Necchi V Savio A Torsello A Zoli M Locatelli V Raimondo F Cocchi D Solcia E Characterisation of gastric ghrelin cells in man and other mammals: studies in adult and fetal tissues Histochem Cell Biol 2002 117 511 519 12107501 10.1007/s00418-002-0415-1
Wierup N Sundler F Circulating levels of ghrelin in human fetuses Eur J Endocrinol 2004 150 405 15012628 10.1530/eje.0.1500405
Chanoine JP Wong AC Ghrelin gene expression is markedly higher in fetal pancreas compared with fetal stomach: effect of maternal fasting Endocrinol 2004 145 3813 3820 10.1210/en.2004-0053
Desai M Gayle D Babu J Ross MG Programmed obesity in intrauterine growth restricted newborns: Modulation by newborn nutrition Am J Physiol Regul Integr Comp Physiol 2004 288 R91 96 15297266
Gualillo O Caminos JE Nogueiras R Seoane LM Arvat E Ghigo E Casanueva FF Dieguez C Effect of food restriction on ghrelin in normal-cycling female rats and in pregnancy Obes Res 2002 10 682 687 12105291
Barreiro ML Gaytan F Castellano JM Suominen JS Roa J Gaytan M Aguilar E Dieguez C Toppari J Tena-Sempere M Ghrelin inhibits the proliferative activity of immature Leydig cells in vivo and regulates stem cell factor messenger RNA expression in rat testis Endocrinol 2004 145 4825 4834 10.1210/en.2004-0732
Archer ZA Findlay PA McMillen SR Rhind SM Adam CL Effects of nutritional status and gonadal steroids on expression of appetite-regulatory genes in the hypothalamic arcuate nucleus of sheep J Endocrinol 2004 182 409 419 15350183 10.1677/joe.0.1820409
Sweeney T Saunders PT Millar MR Brooks AN Ontogeny of anti-mullerian hormone, 3 beta-hydroxysteroid dehydrogenase and androgen receptor expression during ovine total gonadal development J Endocrinol 1997 153 27 32 9135566 10.1677/joe.0.1530027
Tisdall DJ Fidler AE Smith P Quirke LD Stent VC Heath DA McNatty KP Stem cell factor and c-kit gene expression and protein localization in the sheep ovary during fetal development J Reprod Fertil 1999 116 277 291 10615253
Lazarczyk MA Lazarczyk M Grzela T Ghrelin: a recently discovered gut-brain peptide (review) Int J Mol Med 2003 12 279 87 12883642
Smith RG Leonard R Bailey AR Palyha O Feighner S Tan C Mckee KK Pong SS Griffin P Howard A Growth hormone secretagogue receptor family members and ligands Endocrine 2001 14 9 14 11322507 10.1385/ENDO:14:1:009
Mercer JG Moar KM Logie TJ Findlay PA Adam CL Morgan PJ Seasonally inappropriate body weight induced by food restriction: effect on hypothalamic gene expression in male Siberian hamsters Endocrinol 2001 142 4173 4181 10.1210/en.142.10.4173
Tups A Helwig M Khorooshi RM Archer ZA Klingenspor M Mercer JG Circulating ghrelin levels and central ghrelin receptor expression are elevated in response to food deprivation in a seasonal mammal (Phodopus sungorus) J Neuroendocrinol 2004 16 922 928 15584933 10.1111/j.1365-2826.2004.01251.x
Murray TJ Fowler PA Abramovich DR Haites N Lea RG Human fetal testis: second trimester proliferation and steroidogenic capacities J Endocr Metab 2000 85 4812 4817 10.1210/jc.85.12.4812
Kojima M Kangawa K Ghrelin, an orexigenic signaling molecule from the gastrointestinal tract Curr Opin Pharmacol 2002 2 665 668 12482728 10.1016/S1471-4892(02)00220-5
Blache D Fabre-Nys C Venier G Inhibition of sexual behaviour and the luteinizing hormone surge by intracerebral progesterone implants in the female sheep Brain Res 1996 741 117 122 9001713 10.1016/S0006-8993(96)00907-9
Strubbe JH Mein CG Increased feeding in response to bilateral injection of insulin antibodies in the VMH Physiol Behav 1977 19 309 313 607241 10.1016/0031-9384(77)90343-2
Guan XM Yu H Palyha OC McKee KK Feighner SD Sirinathsinghji DJ Smith RG Van der Ploeg LH Howard AD Distribution of mRNA encoding the growth hormone secretagogue receptor in brain and peripheral tissues Brain Res Mol Brain Res 1997 48 23 29 9379845 10.1016/S0169-328X(97)00071-5
Kim MS Yoon CY Park KH Shin CS Park KS Kim SY Cho BY Lee HK Changes in ghrelin and ghrelin receptor expression according to feeding status Neuroreport 2003 14 1317 1320 12876464 10.1097/00001756-200307180-00006
Shuto Y Shibasaki T Wada K Parhar I Kamegai J Sugihara H Oikawa S Wakabayashi I Generation of polyclonal antiserum against the growth hormone secretagogue receptor (GHS-R): evidence that the GHS-R exists in the hypothalamus, pituitary and stomach of rats Life Sci 2001 68 991 996 11212874 10.1016/S0024-3205(00)01001-8
Lin Y Matsumura K Fukuhara M Kagiyama S Fujii K Iida M Ghrelin acts at the nucleus of the solitary tract to decrease arterial pressure in rats Hypertension 2004 43 977 982 14993197 10.1161/01.HYP.0000122803.91559.55
Lawrence CB Snape AC Baudoin FM Luckman SM Acute central ghrelin and GH secretagogues induce feeding and activate brain appetite centers Endocrinol 2002 143 155 162 10.1210/en.143.1.155
Schwartz GJ Moran TH Leptin and neuropeptide Y have opposing modulatory effects on nucleus of the solitary tract neurophysiological responses to gastric loads: implications for the control of food intake Endocrinol 2002 143 3779 3784 10.1210/en.2002-220352
Date Y Nakazato M Murakami N Kojima M Kangawa K Matsukura S Ghrelin acts in the central nervous system to stimulate gastric acid secretion Biochem Biophys Res Commun 2001 280 904 907 11162609 10.1006/bbrc.2000.4212
Salfen BE Carroll JA Keisler DH Endocrine responses to short-term feed deprivation in weanling pigs J Endocrinol 2003 178 541 551 12967345 10.1677/joe.0.1780541
Tolle V Zizzari P Tomasetto C Rio MC Epelbaum J Bluet-Pajot MT In vivo and in vitro effects of ghrelin/motilin-related peptide on growth hormone secretion in the rat Neuroendocrinol 2001 73 54 61 10.1159/000054620
Yamazaki M Nakamura K Kobayashi H Matsubara M Hayashi Y Kangawa K Sakai T Regulational effect of ghrelin on growth hormone secretion from perifused rat anterior pituitary cells J Neuroendocrinol 2002 14 156 162 11849375 10.1046/j.0007-1331.2001.00757.x
Howard AD Feighner SD Cully DF Arena JP Liberator PA Rosenblum CI Hamelin M Hreniuk DL Palyha OC Anderson J Paress PS Diaz C Chou M Liu KK McKee KK Pong SS Chaung LY Elbrecht A Dashkevicz M Heavens R Rigby M Sirinathsinghji DJ Dean DC Melillo DG Van der Ploeg LH A receptor in pituitary and hypothalamus that functions in growth hormone release Science 1996 273 974 977 8688086
Garcia A Alvarez CV Smith RG Dieguez C Regulation of Pit-1 expression by ghrelin and GHRP-6 through the GH secretagogue receptor Mol Endocrinol 2001 15 1484 1495 11518797 10.1210/me.15.9.1484
Bodner M Castrillo JL Theill LE Deerinck T Ellisman M Karin M The pituitary-specific transcription factor GHF-1 is a homeobox-containing protein Cell 1988 55 505 518 2902927 10.1016/0092-8674(88)90037-2
Ingraham HA Chen RP Mangalam HJ Elsholtz HP Flynn SE Lin CR Simmons DM Swanson L Rosenfeld MG A tissue-specific transcription factor containing a homeodomain specifies a pituitary phenotype Cell 1988 55 519 529 2902928 10.1016/0092-8674(88)90038-4
Camina JP Carreira MC El Messari S Llorens-Cortes C Smith RG Casanueva FF Desensitization and endocytosis mechanisms of ghrelin-activated growth hormone secretagogue receptor 1a Endocrinol 2004 145 930 940 10.1210/en.2003-0974
Gaytan F Morales C Barreiro ML Jeffery P Chopin LK Herington AC Casanueva FF Aguilar E Dieguez C Tena-Sempere M Expression of growth hormone secretagogue receptor type 1a, the functional ghrelin receptor, in human ovarian surface epithelium, mullerian duct derivatives, and ovarian tumors J Clin Endocrinol Metab 2005 90 1798 1804 15585554 10.1210/jc.2004-1532
Rae MT Niven D Critchley HO Harlow CR Hillier SG Antiinflammatory steroid action in human ovarian surface epithelial cells J Clin Endocrinol Metab 2004 89 4538 4544 15356059 10.1210/jc.2003-032225
Tan TM-M Jonas KC Thurston LM Vanderpump MP Hardiman PJ Michael AE Ghrelin is a potential paracrine stimulator of 11beta-hydroxysteroid dehydrogenase (11βHSD) in the human ovary Reprod Abst Series 2004 30 P96
Tena-Sempere M Exploring the role of ghrelin as novel regulator of gonadal function Growth Horm IGF Res 2005 15 83 88 15809012 10.1016/j.ghir.2005.02.001
McNatty K Smith P Hudson N Heath D Tisdall DOW BrawTal R Development of the sheep ovary during fetal and early neonatal life and the effect of fecundity genes J Reprod Fert 1995 49 123 135
Peers A Hantzis V Dodic M Koukoulas I Gibson A Baird R Salemi R Wintour EM Functional glucocorticoid receptors in the mesonephros of the ovine fetus Kidney Int 2001 59 425 433 11168924 10.1046/j.1523-1755.2001.059002425.x
De Felici M Pesce M Growth factors in mouse primordial germ cell migration and proliferation Prog Growth Factor Res 1994 5 135 143 7919220 10.1016/0955-2235(94)90001-9
Lea RG Rae MT Murray TJ Brooks AN Rhind SM Fowler PA Miller DW Developmental determinants of steroidogenesis in fetal sheep testis J Endocrin 2000 164 P305
|
16259638
|
PMC1291400
|
CC BY
|
2021-01-04 16:37:13
|
no
|
Reprod Biol Endocrinol. 2005 Oct 31; 3:60
|
utf-8
|
Reprod Biol Endocrinol
| 2,005 |
10.1186/1477-7827-3-60
|
oa_comm
|
==== Front
Reprod HealthReproductive Health1742-4755BioMed Central London 1742-4755-2-91626290110.1186/1742-4755-2-9Research"Near-miss" obstetric events and maternal deaths in Sagamu, Nigeria: a retrospective study Oladapo Olufemi T [email protected] Adewale O [email protected] Adetola O [email protected] Olusoji J [email protected] Maternal and Fetal Health Research Unit, Department of Obstetrics and Gynaecology, Obafemi Awolowo College of Health Sciences/ Olabisi Onabanjo University Teaching Hospital, Sagamu, Ogun State, Nigeria2 Department of Community Medicine and Primary care, Obafemi Awolowo College of Health Sciences/ Olabisi Onabanjo University Teaching Hospital, Sagamu, Ogun State, Nigeria2005 1 11 2005 2 9 9 24 8 2005 1 11 2005 Copyright © 2005 Oladapo et al; licensee BioMed Central Ltd.2005Oladapo et al; licensee BioMed Central Ltd.This is an Open Access article distributed under the terms of the Creative Commons Attribution License (), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
Aim
To determine the frequency of near-miss (severe acute maternal morbidity) and the nature of near-miss events, and comparatively analysed near-miss morbidities and maternal deaths among pregnant women managed over a 3-year period in a Nigerian tertiary centre.
Methods
Retrospective facility-based review of cases of near-miss and maternal death which occurred between 1 January 2002 and 31 December 2004. Near-miss case definition was based on validated disease-specific criteria, comprising of five diagnostic categories: haemorrhage, hypertensive disorders in pregnancy, dystocia, infection and anaemia. The near-miss morbidities were compared with maternal deaths with respect to demographic features and disease profiles. Mortality indices were determined for various disease processes to appreciate the standard of care provided for life-threatening obstetric conditions. The maternal death to near-miss ratios for the three years were compared to assess the trend in the quality of obstetric care.
Results
There were 1501 deliveries, 211 near-miss cases and 44 maternal deaths. The total near-miss events were 242 with a decreasing trend from 2002 to 2004. Demographic features of cases of near-miss and maternal death were comparable. Besides infectious morbidity, the categories of complications responsible for near-misses and maternal deaths followed the same order of decreasing frequency. Hypertensive disorders in pregnancy and haemorrhage were responsible for 61.1% of near-miss cases and 50.0% of maternal deaths. More women died after developing severe morbidity due to uterine rupture and infection, with mortality indices of 37.5% and 28.6%, respectively. Early pregnancy complications and antepartum haemorrhage had the lowest mortality indices. Majority of the cases of near-miss (82.5%) and maternal death (88.6%) were unbooked for antenatal care and delivery in this hospital. Maternal mortality ratio for the period was 2931.4 per 100,000 deliveries. The overall maternal death to near-miss ratio was 1: 4.8 and this remained relatively constant over the 3-year period.
Conclusion
The quality of care received by critically ill obstetric patients in this centre is suboptimal with no evident changes between 2002 and 2004. Reduction of the present maternal mortality ratio may best be achieved by developing evidence-based protocols and improving the resources for managing severe morbidities due to hypertension and haemorrhage especially in critically ill unbooked patients. Tertiary care hospitals in Nigeria could also benefit from evaluation of their standard of obstetric care by including near-miss investigations in their maternal death enquiries.
==== Body
Background
For many years, evaluation of maternal healthcare services aimed at improving the quality of obstetric care has traditionally relied on enquiries into maternal deaths. More recently, review of cases at the very severe end of the maternal morbidity spectrum, described as "near-miss" (those who nearly died), has been found to be a useful complement to investigation of maternal mortality [1,2]. Review of near-miss cases has the potential to highlight the deficiencies as well as the positive elements in the provision of obstetric services in any health system. Unlike in the developed countries, there is limited experience with the use of near-miss reviews as a tool for monitoring the quality of maternity services in developing countries. This is probably as a result of the persistently high levels of maternal mortality that has overshadowed other severe obstetric complications, from which lessons could equally be learned. In spite of the high maternal mortality ratios in many of the centres in resource-poor settings, the actual number of maternal deaths per centre may not allow detailed quantification of associated risk factors and determinants that are locally important. Because near-misses occur much more frequently than maternal deaths, more comprehensive and statistically reliable quantitative analyses that are of value to clinical audit can be rapidly conducted [1-4]].
At Olabisi Onabanjo University Teaching Hospital, Sagamu, a state-owned referral centre in southwest Nigeria, maternal mortality ratio is close to 2000 per 100,000 deliveries [5]. The majority of these maternal deaths are largely unpreventable as they occur in unbooked emergency cases that present too late to the hospital and die shortly after admission. Therefore, isolated enquiry into maternal deaths in this centre is unlikely to yield adequate information when the focus of investigation is on the standard of in-hospital care. As shown in previous studies, inclusion of near-miss review in maternal death enquiry can better inform the quality of obstetric care at different levels of healthcare delivery at more frequent intervals [3,4,6].
In order to provide an insight into the quality of maternal care provided in our institution, we embarked on a retrospective study to determine the frequency of near-miss (severe acute maternal morbidity) and the nature of near-miss events, and comparatively analysed near-miss morbidities and maternal deaths among pregnant women managed in this centre over a 3-year period. The review is expected to serve as a complementary method for auditing the quality of maternal healthcare in this institution.
Methods
Hospital Setting
The study was conducted at the obstetric unit of Olabisi Onabanjo University Teaching Hospital (OOUTH), Sagamu, a publicly funded tertiary care institution, which serves as the major referral centre for other public and private hospitals within Ogun State, in southwest Nigeria. In addition to providing emergency obstetric services to women referred from other centres, the hospital also provides antenatal care and delivery services for both unreferred low and high-risk pregnant women from Sagamu community and neighbouring towns. The centre provides emergency obstetric and gynaecological care 24 hours a day. Although patients are expected to pay for their services, in emergency situations, they are managed within the means of existing resources before funds are made available. Nine consultant obstetricians and an average of 15 registrars and 30 midwives ran the three obstetric units of the hospital during the reviewed period. The hospital provides blood transfusion services from limited stock and relatives of patients are requested to donate or pay for blood when needed. Each unit of blood costs between N1500-N2000 (approximately £6–10) during the reviewed period. The only intensive care unit (ICU) of the hospital is within the main surgical theatre though patients requiring critical care are admitted from other units after paying a certain fee. Proliferation of many private hospitals and traditional birth homes within Ogun State in recent past has limited the total number of deliveries but has relatively increased the frequency of complicated pregnancies and deliveries managed in the obstetric unit. An average of 40–45 deliveries are conducted per month out of which 10–15 are unbooked.
Definition of cases
Near-miss events are defined as acute obstetric complications that immediately threaten a woman's survival but do not result in her death either by chance or because of hospital care she receives during pregnancy, labour or within 6 weeks after termination of pregnancy or delivery [1] while a near-miss case is a woman with at least one near-miss event. For identifying near-miss events, we applied the disease-specific criteria that were employed by Filippi et al [7] in similar hospital settings in West Africa, which are based on five main diagnostic groups. These are haemorrhage (leading to shock, emergency hysterectomy, coagulation defects and/or blood transfusion of ≥ 2 litres); hypertensive disorders in pregnancy (eclampsia and severe pre-eclampsia with clinical/laboratory indications for termination of pregnancy to save the woman's life [8]); dystocia (uterine rupture and impending rupture e.g. prolonged obstructed labour with a previous caesarean section); infection (hyperthermia or hypothermia or a clear source of infection and clinical signs of septic shock) and anaemia (low haemoglobin level: haematocrit < 6 g/dl) or clinical signs of severe anaemia in women without severe haemorrhage.
Life-threatening obstetric conditions refer to maternal complications severe enough to cause near-miss morbidity and maternal death while "critically ill obstetric patients" are women who suffered life-threatening obstetric conditions (i.e. cases of near-miss and maternal death). Maternal death is defined according to the tenth revision of International Classification of Diseases (ICD-10) by the World Health Organization [9].
Study design and identification of cases
Near-miss cases were retrospectively identified among women with pregnancy-related complications admitted into the obstetric units of the hospital between 1 January 2002 and 31 December 2004. Using the provisional and final diagnoses documented in the admission-discharge register of the hospital, case files of women whose diagnoses met the above pre-defined criteria as well as those with the possibility of being associated with severe acute maternal complications were retrieved for scrutiny by the Near-miss Audit Committee comprising three consultant obstetricians and three specialist registrars. Overall, 520 cases were retrieved for scrutiny. For each case of near miss, data were collected on demographic characteristics including gestational age at the time of sustaining the near-miss morbidity, nature of obstetric complication(s) responsible, presence of organ-system dysfunction/failure, ICU admission, timing of near-miss event with respect to admission, fetal outcome in those associated with labour and length of hospital stay. Information on maternal deaths and deliveries conducted during the reviewed period were obtained from the labour/delivery registers and case files from the Medical Records Department. For each case of maternal death, data were collected on the demographic characteristics including gestational age at the time of death and the underlying cause of death.
Data analysis
Data were entered into a computer database using Microsoft Excel spreadsheet and statistical analysis was performed with Epi Info 2002 software (CDC and WHO, 2002)[10]. Results are presented as frequencies, percentages and descriptive statistics. The prevalence of near-miss cases is defined as the number of near-miss cases divided by the number of deliveries in the hospital. The frequencies of near-miss events are reported according to the clinical condition responsible, referral status of the patients and whether the complications were present upon arrival or occurred while on admission at the hospital. The near-miss morbidities were compared with maternal mortality for their demographic features and underlying disease processes. In order to appreciate the standard of care provided for each disease process, we calculated the mortality index for each obstetric condition. This was defined as the number of maternal deaths resulting from a particular obstetric condition divided by the sum of the near-miss morbidities and maternal deaths occurring from such obstetric condition, expressed as a percentage. It reflects the proportion of each life-threatening obstetric condition, which ended in maternal death. Maternal mortality ratio was calculated as the number of maternal deaths per 100,000 deliveries.
Categorical variables were compared with the χ2 or Fisher's exact test when appropriate while continuous variables were compared with the Student's t-test. Differences between data were considered statistically significant when p<0.05.
Results
During the reviewed period, there were 1501 deliveries, 211 near-miss cases and 44 maternal deaths. As shown in Table 1, the demographic characteristics of women who sustained near-miss complications and those who died are comparable. At least half of the women in each group were within the ages of 21 and 30 years. Over one-third of women in each group were nulliparous, in keeping with the parity demographics of our obstetric population.
Table 1 A comparison of the demographic characteristics of women with near-miss morbidity and maternal death.
Near-miss cases Maternal deaths
n = 211 n = 44
Age (years)
≤ 20 24 (11.4) 6 (13.6)
21–25 58 (27.5) 9 (20.5)
26–30 56 (26.5) 13 (29.5)
31–35 47 (22.3) 8 (18.2)
>35 26 (12.3) 8 (18.2)
Range 16–44 18–45
Mean ± SD 28.1 ± 6.1 28.6 ± 5.6
Parity
0 75 (35.5) 15 (34.1)
1–4 113 (53.6) 25 (56.8)
≥ 5 23 (10.9) 4 (9.1)
Range 0–7 0–7
Median 2 1
Booking status
Unbooked at OOUTH 174 (82.5) 39 (88.6)
Gestational age (weeks)
<13 26 (12.3) 2 (4.5)
13–28 13 (6.2) 1 (2.3)
>28 172 (81.5) 41 (93.2)
Percentage in parenthesis
A total of 242 near-miss events were identified among the near-miss cases with frequencies decreasing from 95 in 2002 to 65 in 2004 (Table 2). This implies that 31 women had more than one near-miss morbidity, giving an average of 1.2 near-miss morbidities per case. Table 2 shows that the most common types of near-miss events fall under the diagnostic categories of hypertensive disorders in pregnancy, haemorrhage and dystocia. Hypertensive disorders in pregnancy and haemorrhage were responsible for 61.6 % of all near-miss events. Most events of near-miss due to haemorrhage developed in the later part of pregnancy with 41.1% occurring postpartum. Haemorrhage due to abortion did not cause any near-miss complication over the 3-year period. Near-miss events related to infection and anaemia were the least common.
Table 2 Diagnosis distribution and trend of near-miss events in Sagamu, Nigeria
Criteria 2002 2003 2004 Total
Haemorrhage 26 (28.3) 30 (35.3) 17 (26.2) 73 (30.2)
Early pregnancy 6(6.5) 9 (10.6) 9 (13.8) 24 (9.9)
Ectopic pregnancy 6 9 9 24
Abortion 0 0 0 0
Late pregnancy 20 (21.7) 21 (24.7) 8 (12.3) 49 (20.2)
Placenta praevia 4 2 2 8
Abruptio placentae 5 4 2 11
Postpartum haemorrhage 11 15 4 30
Others 0 0 0
Hypertension 32 (34.8) 21 (24.7) 23 (35.4) 76 (31.4)
Eclampsia 22 8 9 39
Severe preeclampsia 10 13 14 37
Dystocia 17 (18.5) 19 (22.4) 11 (16.9) 47 (19.4)
Uterine rupture 3 4 3 10
Impending rupture 14 15 8 37
Infections 9 (9.8) 5 (5.9) 6 (9.2) 20 (8.3)
Anaemia 8 (8.7) 10 (11.8) 8 (12.3) 26 (10.7)
All near-miss events 92 85 65 242 (100.0)
Percentage in parenthesis
The various causes of maternal deaths between 2002 and 2004 are shown in Table 3. Forty-two (95.5%) of the deaths were direct maternal deaths while 2 (4.5%) were indirect maternal deaths. Maternal deaths were also most commonly due to hypertensive disorders in pregnancy and haemorrhage both responsible for 50.0% of all deaths. Overall, eclampsia was the leading cause of deaths singly accounting for 22.7% of all maternal deaths. Most (8/9) of the deaths due to haemorrhage were cases of postpartum haemorrhage. HIV infection (2.3%) and septic abortion (2.3%) were uncommon causes of maternal deaths during the reviewed period.
Table 3 Causes of maternal deaths in Sagamu, Nigeria (2002–2004)
Causes n = 44 (%)
Haemorrhage 9 20.5
Early pregnancy
Ectopic pregnancy 1 2.3
Abortion - -
Late pregnancy
Placenta praevia
- -
Abruptio placentae
- -
Postpartum haemorrhage
8
18.2
Hypertension 13 29.5
Eclampsia 10 22.7
Severe preeclampsia 3 6.8
Dystocia 6 13.6
Uterine rupture 6 13.6
Infections 8 18.2
Puerperal sepsis 2 4.5
Chorioamnionitis 4 9.1
Septic abortion 1 2.3
*HIV infection 1 2.3
Anaemia (not due to haem- orrhage 5 11.4
Pulmonary embolism 1 2.3
Anaesthetic complication 1 2.3
*Pre-existing medical disease 1 2.3
*Indirect obstetric causes
In Table 4, the disease profile of near-miss morbidities was compared with that of maternal mortality. Besides infectious morbidity, the categories of maternal complications responsible for near-miss and maternal mortality followed the same order of decreasing frequency (viz hypertensive disorders in pregnancy, haemorrhage, dystocia and anaemia). Significantly more women died after developing severe morbidity due to uterine rupture and infection. The leading life-threatening obstetric conditions were hypertensive disorders in pregnancy (31.4%) and haemorrhage (29.0%). Uterine rupture (37.5%) and infection (28.6%) had the highest mortality indices though they were less frequent life-threatening obstetric conditions. The lowest mortality indices were recorded for severe morbidities associated with haemorrhage in early pregnancy and antepartum haemorrhage.
Table 4 A comparison of near-miss events and primary causes of maternal deaths
Complication NME *Maternal deaths P LTOC (%) Mortality index (%)
n (%) n (%) n (%)
Haemorrhage 73 (30.2) 9 (22.0) 0.284 82 (29.0) 11.0
Early pregnancy 24 (9.9) 1 (2.4) 0.145 25 (8.8) 4.0
Ectopic pregnancy 24 1 0.145 25 (8.8) 4.0
Abortion 0 0 0.0 (0.0) 0.0
Late pregnancy 49 (20.2) 8 (19.5) 0.914 57 (20.1) 14.0
Placenta praevia 8 0 0.608 8 (2.8) 0.0
Abruptio placentae 11 0 0.376 11 (3.9) 0.0
Postpartum haemorrhage 30 8 0.217 38 (13.4) 21.1
Hypertension 76 (31.4) 13 (31.7) 0.969 89 (31.4) 14.6
Eclampsia 39 10 0.195 49 (17.3) 20.4
Severe pre-eclampsia 37 3 0.175 40 (14.1) 7.5
Dystocia 47 (19.4) 6 (14.6) 0.467 53 (18.7) 11.3
Uterine rupture 10 6 0.017 16 (5.7) 37.5
Impending rupture 37 0 0.007 37 (13.1) 0.0
Infection 20 (8.3) 8 (19.5) 0.042 28 (9.9) 28.6
Anaemia 26 (10.7) 5 (12.2) 0.787 31 (11.0) 16.1
Total 242 (100.0) 41 (100.0) 283 (100.0) 14.5
*Excluding anaesthetic death (n = 1), pulmonary embolism (n = 1) and pre-existing medical disease (n = 1)
Mortality index = maternal deaths divided by life-threatening obstetric conditions
LTOC: life-threatening obstetric condition
NME: near-miss events
Table 5 shows that there was a significant fall in the prevalence of near-miss cases from 2002 to 2004 (χ2 = 9.01, 2df; p = 0.011) with an overall prevalence of 140.6 per 1000 deliveries. Critically ill obstetric patients constituted 17.0% of all women who delivered during the reviewed period. The frequency of critically ill obstetric patients also decreased significantly between 2002 and 2004 (p = 0.002). Maternal mortality ratio for the 3-year period was 2931.4 per 100,000 deliveries. Though there were fewer deaths in 2004, there was no significant difference between the maternal mortality ratios of 2002 and 2004 (p = 0.236). The overall maternal death to near-miss ratio was 1: 4.8 with no significant difference in this relationship between the years of study. Overall, 17.3% of critically ill obstetric patients died during the 3-year period. Majority (80.6%) of the near-miss cases were referred from other facilities namely traditional birth attendant homes, primary and secondary healthcare units and private hospitals within Ogun State and beyond. Most near-miss cases already had near-miss morbidity upon arrival at OOUTH, Sagamu while only 15.6% of them became near-miss after admission to the hospital. The proportion of near-miss events occurring after admission varied between diagnostic categories; haemorrhage (23.3%), hypertensive disorders (15.8%), dystocia (8.5%), infections (10.0%) and anaemia (11.5%).
Table 5 Frequency and characteristics of near-miss cases and maternal death to near-miss ratios for 2002–2004.
2002 2003 2004 Total
Deliveries (n) 475 545 481 1501
Live births 433 502 443 1378
Near-miss cases (n) 85 71 55 211
Referred from other facility [n (%)] 68 (80.0) 58 (81.7) 44 (80.0) 170 (80.6)
On arrival [n (%)] 71(83.5) 59 (83.1) 49 (89.1) 179 (84.8)
During hospitalisation *15 (17.6) 12 (16.9) 6 (10.9) 33 (15.6)
Near-miss cases per 1000 deliveries 178.9 130.3 114.3 140.6
On arrival 149.5 108.3 101.9 119.3
During hospitalisation 31.6 22.0 12.5 22.0
Maternal deaths (n) 17 16 11 44
MMR/100,000 deliveries 3578.9 2935.8 2286.9 2931.4
Critically ill obstetric patients 102 87 66 255
Maternal death to near-miss ratio 1: 5 1: 4.4 1: 5 1: 4.8
MMR: maternal mortality ratio
*One patient qualified as a near-miss on arrival and during hospitalisation
Only 9 (4.3%) of the near-miss cases were managed in the ICU while organ-system dysfunction/failure were recorded in 19 (9.0%) of them. The nature of organ-system dysfunction/failure and the associated obstetric factors among the near-miss cases are shown in Table 6. The two most commonly affected organ-systems were the renal and vascular systems. Among the 167 near-miss cases that were associated with labour, 37.7% and 6.5% resulted in stillbirths and early neonatal deaths, respectively. Duration of hospital stay for near-miss cases ranged between 2 and 74 days (median 11 days, interquartile range: 8–15 days).
Table 6 Causes of organ-system dysfunction/failure in near-miss cases
Organ-system n = 19* Obstetric causes (n)
Cardiac (pulmonary oedema) 3 Severe pre-eclampsia (2)
Severe anaemia (unrelated to haemorrhage (1)
Coagulation 1 Abruptio placentae
Renal 7 Eclampsia (4)
Severe pre-eclampsia (1)
Septic abortion (1)
Postpartum haemorrhage (1)
Vascular 7 Ruptured ectopic pregnancy (1)
Uterine rupture (1)
Postpartum haemorrhage (5)
Cerebral 4 Eclampsia (4)
Immunologic 1 HIV-related sepsis (1)
*Note that some women had >1 organ-system dysfunction/failure
Discussion
The need to assess the quality of obstetric care in any centre is paramount to understanding the improvement resulting from investment in its maternity services. Up till now, evaluation of obstetric performance in many Nigerian hospitals is limited to investigations of maternal deaths, an indicator that is vulnerable to many flaws in this environment. To the best of our knowledge, ours is the first review in this country that quantitatively examined the quality of obstetric care using alternative indices.
The study shows that severe acute maternal morbidities (near-miss) occur in a considerable percentage of women managed in this obstetric unit. Life-threatening obstetric conditions, including those that resulted in deaths complicated up to 17% of all deliveries during the reviewed period. This implies that obstetricians in this centre were confronted with life-saving emergency situations in almost 1 out of every 6 women who utilised their obstetric services. While the prevalence of our near-miss cases shows some degree of consistency with the reports from other teaching hospitals in West Africa [7], it is several-folds higher than those published from developed countries [11,12]. This disparity is possibly due to differences in definition and identification of cases, which are major limitations in comparison of near-miss data across institutions [13,14]. Studies in industrialised countries commonly use ICU admission or organ-system dysfunction/failure as their criteria for case selection [11,15]. Though organ-system based criteria are regarded as the most specific and least vulnerable to bias [13], we adopted a tested case definition that best fits the circumstances in our environment to allow local improvement in services and comparison of studies in our setting. We conclude that the wide difference in the magnitude of our cases compared to those quoted in high-resource settings is unlikely to be due to overestimation of our near-miss cases since what constitutes near-miss morbidity in any centre is dependent on contextual factors and the figures only depicted the number of women at the verge of dying in their respective prevailing circumstances. This conclusion is further supported by the fact that our near-miss cases were approximately five times as frequent as maternal deaths, similar to the findings in studies that used organ-system based criteria [16].
One of the advantages of the criteria used for our case definition is that it mirrors the major causes of maternal mortality and therefore readily permits comparison that allows assessment of the standard of care with respect to common causes of maternal deaths. As logically expected, there were no major differences in the underlying pathology for near-miss morbidities and maternal deaths indicating that near-miss review can be a useful surrogate of maternal death analysis in this centre. Similar to the findings in many previous studies [7,11,15], hypertensive disorders and haemorrhage were the leading causes of near-miss morbidities accounting for almost two-thirds of all cases. The contribution of these complications to maternal deaths, however, signifies a poor response of our system to modify the major disease profile of its obstetric population. In view of the referral status of most critically ill women, reduction of maternal deaths in this centre therefore requires channelling of resources towards the prevention of haemorrhage and hypertensive disorders at the lower levels of healthcare while strengthening the resources for their treatment in the teaching hospital.
The importance of including near-miss investigations in maternal death audits is demonstrated by this review. Comparison of the disease processes responsible for near-miss and maternal deaths shows that infection, with a mortality index of 28.6%, constituted a significant threat to the survival of affected patients though it was the least frequent cause of life-threatening obstetric conditions. Similarly, it became clear that uterine rupture received the poorest form of care even though it accounted for 5.7% of life-threatening obstetric conditions. The level of care provided for pregnancies complicated by eclampsia also deserves special attention. According to the mortality index, approximately one-fifth of critically ill eclamptic patients died in this centre. Though eclampsia is a known major cause of maternal death worldwide, the poor standard of care for women with eclampsia in our unit may be related to the existing management policy for hypertensive disorders in pregnancy. Up till now, magnesium sulphate, which has been shown to reduce eclampsia-related risk of maternal mortality [17,18], is yet to be adopted for use in this institution. In the light of our findings, efforts to reduce maternal death from hypertensive disorders must include urgent adoption of a clear and up-to-date evidence-based protocol for treating eclampsia.
The management of haemorrhage due to abortion during the reviewed period is commendable as none of such cases caused near-miss morbidity or maternal mortality. Though this may be related to the frequent training of members of staff in manual vacuum aspiration and incorporation of postabortal care into the existing management protocol of abortion in the last five years, it is possible that affected women did not present at the hospital either because they did not survive or because of fear of legal prosecution in cases of criminal abortion. Of the various types of life-threatening obstetric haemorrhage, postpartum haemorrhage constituted the greatest danger to affected women while early pregnancy haemorrhagic complications and antepartum haemorrhage were less risky. This implies that efforts need to be focussed on improving the protocols and resources for combating postpartum haemorrhage, while maintaining and improving the existing preventive measures and treatment strategies for early pregnancy complications and antepartum haemorrhage.
As shown in this study, the lack of proper antenatal care is a major determinant of adverse maternal outcome. Majority of the women with near-miss morbidity arrived at the hospital in critical condition having being referred from both modern and traditional maternity facilities. Though some authors suggest that near-miss upon arrival at the hospital should not be used to assess the quality of care at the admitting facility [7], we believe that the proportion of referred near-miss cases to our obstetric unit reflected our ability to prevent maternal deaths, even in previously unanticipated situations. What is worrisome, however, is the recorded maternal death to near-miss ratio, a useful indicator of the quality of care received by near-miss cases irrespective of their primary source of antenatal or labour care [13]. A maternal death to near-miss ratio of approximately 1: 5 indicates that for every 5 women who survived life-threatening complications in this centre, one maternal death was also recorded. This ratio, which reflects the overall standard of obstetric care, is poorer than 1: 11–22 reported from similar centres in Niger [19], Cote d'Ivoire and Benin [7] respectively and a far cry from the 1: 117–223 reported in Europe [12-14] using the same criteria for case definition.
It is unlikely that the overall substandard level of care in this centre is due to a higher prevalence of life-threatening complications compared to other centres. This is because with the recorded delivery rate, the prevalence of critically ill obstetric patients translates to an average of 7 of such patients being managed per month (255/36 months) and a maternal death to near-miss ratio of 1: 5 is unjustifiable with the existing human resources. Though this level of care could be attributed to other extraneous factors ranging from cost of obstetric services, mismanagement at the sources of referral to lapses in the referral chain, it is the duty of a referral hospital to maintain a good standard of care if the utilisation of obstetric services among the population is to be encouraged. In spite of the decreasing trend in the frequency of near-miss cases over the 3 years, the similarity in the death to near-miss ratios indicates that there was no significant improvement in the level of care over these years. Therefore, the significant fall in the prevalence of near-miss cases is probably a reflection of recent governmental efforts to improve obstetric services at the primary and secondary healthcare units, which are the main sources of near-miss cases managed at the teaching hospital.
Some important issues concerning definitions in near-miss studies were also illustrated in this investigation. It appears that identification of cases based on ICU admission or organ-system failure/dysfunction as used in some studies may underestimate the frequency of severe maternal morbidities in our setting since they occurred in only 4.3% and 9.0% of all near-miss cases, respectively. Though data collection is easy, a major disadvantage of using ICU admission as the criteria for case selection is that it is dependent on factors such as availability, capacity and location of ICU and institutional guidelines for ICU admission [1]. Therefore, such method is unlikely to produce accurate data in a centre like ours where ICU is not available in the labour ward unit and patients can only be admitted to the general ICU after payment of certain fees. Likewise, comparison of the frequencies of women identified to have suffered organ dysfunction with those who died is inconsistent with the usual relationship between near-misses and maternal deaths suggesting that organ dysfunction was probably poorly documented. This problem is most likely related to our retrospective identification of cases which essentially relied on obstetric diagnoses indexed in the admission-discharge register. Cases of organ-system dysfunction are best detected as they occur, and are therefore more reliably identified in prospective studies.
A major limitation of this study is its retrospective nature. Besides the possibility of underestimating the near miss-cases as a result of incomplete documentation in case files, the methodology also discouraged assessment of substandard care with respect to the health workers or health administration and patient-orientated missed opportunities. Evaluation of the circumstances surrounding near-misses and maternal deaths would shed light on avoidable factors and therefore enable more focussed remedial actions. This aspect needs to be considered in subsequent near-miss investigations in this institution.
Conclusion
In summary, our review shows that besides the 44 women who died due to pregnancy-related complications, there were 211 additional women who received critical care during the same period supporting the view that near-miss appraisal provides a larger sample to assess the threat to maternal life. The overall maternal death to near-miss ratio, however, indicates that a significant proportion of critically ill women died, suggesting a suboptimal level of care for life-threatening complications. Since there were little differences in the underlying disease processes causing near-miss and maternal mortality, evaluation of the circumstances surrounding near-miss cases could act as a proxy for maternal death in this centre. Efforts geared towards improvement in the management of near-miss morbidities would definitely go a long way in reducing the present maternal mortality ratio. From the findings of this review, attempts to reduce maternal deaths may best be achieved by developing evidence-based protocols for the management of severe hypertension and haemorrhage especially for critically ill referred patients. In addition, considerable efforts should be made to improve maternal care for infrequent but important life-threatening obstetric conditions such as uterine rupture and infection. Necessary facilities should be made available and training of personnel and emergency drills should be frequently conducted to combat the identified disease processes that received suboptimal care. Although this study did not specifically address avoidable factors, it has nevertheless raised awareness of the deficiencies in the management of serious maternal illnesses. It is apparent from this review that tertiary institutions in Nigeria could also benefit from evaluation of their quality of obstetric care by including near-miss investigations in their maternal death enquiries.
Competing interests
The author(s) declare that they have no competing interests.
Authors' contributions
OOT conceived and designed the study. OOT drafted the manuscript while AOS and AOO critically revised it for intellectual contents. OOT, AOS and AOO were members of the committee that collected the data. OOT and OJD analysed and interpreted the data.
Acknowledgements
We thank Dr. Lale Say of the Department of Reproductive Health and Research, World Health Organization, Geneva, for her review of the manuscript and valuable inputs. We are also grateful to Mr. Abiodun Allison, Mr. Sunday Adeyemi and Mrs. Adebimpe Osunlaja of the Medical Records Department of Olabisi Onabanjo University Teaching Hospital, Sagamu, Nigeria for their assistance in the retrieval of case files.
==== Refs
Ronsmans C Filippi V Reviewing severe maternal morbidity: learning from survivors from life-threatening complications Beyond the Numbers: Reviewing Deaths and Complications to Make Pregnancy Safer 2004 Geneva: World Health Organization 103 124
Pattinson RC Buchmann E Mantel G Schoon M Rees H Can enquiries into severe acute maternal morbidity act as a surrogate for maternal death enquiries? BJOG 2003 110 889 893 14550357
Cochet L Pattinson RC MacDonald AP Severe acute maternal morbidity and maternal death audit- a rapid diagnostic tool for evaluating maternal care S Afr Med J 2003 93 700 702 14635560
Vandecruys HIB Pattinson RC Macdonald AP Mantel GD Severe acute maternal morbidity and mortality in the Pretoria Academic Complex: changing patterns over 4 years Eur J Obstet Gynecol Reprod Biol 2002 102 6 10 12039082 10.1016/S0301-2115(01)00558-9
Sule-Odu AO Maternal deaths in Sagamu, Nigeria Int J Gynecol Obstet 2000 69 47 49 10.1016/S0020-7292(99)00199-X
Gandhi MN Welz T Ronsmans C Severe acute maternal morbidity in rural South Africa Int J Gynecol Obstet 2004 87 180 187 10.1016/j.ijgo.2004.07.012
Filippi V Ronsmans C Gohou V Goufodji S Lardi M Sahel A Saizonou J De Brouwere V Maternity wards or emergency obstetric rooms? Incidence of near-miss events in African hospitals Acta Obstet Gynecol Scand 2005 84 11 16 15603561 10.1111/j.0001-6349.2005.00636.x
Robson SC Edmond DK Hypertension and renal disease in pregnancy Dewhurst's Textbook of Obstetrics and Gynaecology for Postgraduates 1999 6 Oxford: Blackwell Science 166 185
World Health Organization ICD-10: International statistical classification of diseases and health-related problems Tenth Revision 1993 2 Geneva: World Health Organization
Centers for Disease Control and Prevention (CDC) and World Health Organization Epi Info 2002 – Database and statistics software for public health professionals 2002 Atlanta, Georgia, USA: Centers for Disease Control and Prevention
Baskett TF Sternadel J Maternal intensive care and near-miss mortality in obstetrics BJOG 1998 105 981 984
Waterstone M Bewley S Wolfe C Incidence and predictors of severe obstetric morbidity: case-control study BMJ 2001 322 1089 1094 11337436 10.1136/bmj.322.7294.1089
Say L Pattinson RC Gülmezoglu AM WHO systematic review of maternal morbidity and mortality: the prevalence of severe acute maternal morbidity (near miss) Reprod Health 2004 1 3 15357863 10.1186/1742-4755-1-3
Minkauskiene M Nadisauskiene R Padaiga Z Makari S Systematic review on the incidence and prevalence of severe maternal morbidity Medicina 2004 40 299 309 15111741
Murphy DJ Charlett P Cohort study of near-miss maternal mortality and subsequent reproductive outcome Eur J Obstet Gynecol Reprod Biol 2002 102 173 178 11950486 10.1016/S0301-2115(01)00320-7
Mantel GD Buchmann E Rees H Pattinson RC Severe acute maternal morbidity: a pilot definition for a near-miss BJOG 1998 105 985 990
Duley L Henderson-Smart D Magnesium sulphate versus diazepam for eclampsia The Cochrane Database of Systematic Reviews 2003 Art. No.: CD000127. DOI: 10.1002/14651858.CD000127
Duley L Gülmezoglu AM Henderson-Smart DJ Magnesium sulphate and other anticonvulsants for women with pre-eclampsia The Cochrane Database of Systematic Reviews 2003 Art. No.: CD000025. DOI: 10.1002/14651858.CD000025
Prual A Huguet D Gabin O Rabe G Severe obstetric morbidity of the third trimester, delivery and early puerperium in Niamey (Niger) Afr J Reprod Health 1998 2 10 19 10214424
|
16262901
|
PMC1291401
|
CC BY
|
2021-01-04 16:38:15
|
no
|
Reprod Health. 2005 Nov 1; 2:9
|
utf-8
|
Reprod Health
| 2,005 |
10.1186/1742-4755-2-9
|
oa_comm
|
==== Front
RetrovirologyRetrovirology1742-4690BioMed Central London 1742-4690-2-661626290610.1186/1742-4690-2-66Short ReportProcessing sites in the human immunodeficiency virus type 1 (HIV-1) Gag-Pro-Pol precursor are cleaved by the viral protease at different rates Pettit Steve C [email protected] Jeffrey N [email protected] Andrew H [email protected] Ronald [email protected] Department of Medicine, University of North Carolina at Chapel Hill, Chapel Hill, NC, USA2 Department of Biochemistry and Biophysics, University of North Carolina at Chapel Hill, Chapel Hill, NC, USA3 The UNC Center for AIDS Research, University of North Carolina at Chapel Hill, Chapel Hill, NC, USA4 CB7295, Rm 22-006 Lineberger Bldg, UNC Center For AIDS Research, University of North Carolina at Chapel Hill, Chapel Hill, NC 27599-7295, USA5 Department of Pathology, Moores UCSD Cancer Center, 3855 Health Sciences Dr. #0803, La Jolla, CA 92093-0803, USA6 3805-103 Chimney Ridge Pl., Durham, NC, 27713, USA2005 1 11 2005 2 66 66 2 8 2005 1 11 2005 Copyright © 2005 Pettit et al; licensee BioMed Central Ltd.2005Pettit et al; licensee BioMed Central Ltd.This is an Open Access article distributed under the terms of the Creative Commons Attribution License (), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
We have examined the kinetics of processing of the HIV-1 Gag-Pro-Pol precursor in an in vitro assay with mature protease added in trans. The processing sites were cleaved at different rates to produce distinct intermediates. The initial cleavage occurred at the p2/NC site. Intermediate cleavages occurred at similar rates at the MA/CA and RT/IN sites, and to a lesser extent at sites upstream of RT. Late cleavages occurred at the sites flanking the protease (PR) domain, suggesting sequestering of these sites. We observed paired intermediates indicative of half- cleavage of RT/RH site, suggesting that the RT domain in Gag-Pro-Pol was in a dimeric form under these assay conditions. These results clarify our understanding of the processing kinetics of the Gag-Pro-Pol precursor and suggest regulated cleavage. Our results further suggest that early dimerization of the PR and RT domains may serve as a regulatory element to influence the kinetics of processing within the Pol domain.
==== Body
Findings
The retroviral protease (PR) processes the Gag and Gag-Pro-Pol precursors during the assembly of the mature virus particle. The viral structural proteins assume altered conformations after processing, and the viral enzymes become fully active in their processed forms [1-7]. Proper proteolytic processing is necessary for assembly of an infectious particle [3,4,8-10].
Cleavage of Gag is ordered and appears to be regulated, at least in part, by the target site sequence, the presence of spacer domains, and the interaction with RNA [8,9,11,12]. Previous studies showed the five HIV-1 Gag processing sites are cleaved at rates that vary up to 400-fold in vitro [9,13]. Initial cleavage occurs at the p2/NC site followed by an intermediate rate of cleavage at the MA/CA and p1/p6 sites, and final cleavage at the CA/p2 and NC/p1 sites [9,12-16]. A similar pattern of ordered processing appears to occur in infected cells [9,12,17,18].
Processing of the HIV-1 Gag-Pro-Pol precursor by protease in trans is less studied, although the final cleavage products [MA, CA, NC, transframe (TF), PR, RT, IN] are well characterized [19-22]. The HIV-1 Gag-Pro-Pol precursor results from a -1 frameshift event during translation at a site near the 3' end of the gag reading frame to join the gag and pro-pol reading frames [23,24]. For this study, we created by site-directed mutagenesis [25,26] a continuous HIV-1 gag-pro-pol reading frame that would produce a full-length precursor identical in sequence to the viral Gag-Pro-Pol polyprotein precursor [23,27] (Fig. 1A). Intrinsic protease activity was inactivated by a D25A substitution of the catalytic aspartate of the PR domain to produce the final construct GPPfs-PR (Fig. 1A). We expressed the radio-labeled Gag-Pro-Pol using an in vitro transcription/translation strategy [9,28] and monitored cleavage at known processing sites as a function of time after adding 0.25 μg recombinant HIV-1 protease (as described in [13,28,29]) in a reaction volume of 50 μl. Under these conditions the concentration of precursor is approximately 0.1 nM. Products were separated using two different SDS-PAGE systems [30,31] prior to autoradiography.
Figure 1 A. The frameshift mutation in pGPPfs-PR. Above: the sequence of wild type HIV-1 HXB (GenBank:NC001802) molecular clone in the area of translational frameshift in gag-pro-pol is shown. The heptanucleotide slippery sequence required for translational frameshifting is underlined [23, 24]. The adenine that is read twice during frameshifting is shown in bold. The exact site of frameshifting in the wild type virus is variable with 70% of Gag-Pro-Pol product containing Leu as the second residue of the transframe domain (TF) [27]. pGPPfs-PR expressed in vitro in a coupled transcription/translation system [28] gives the predominant Gag-Pro-Pol product. Additional translationally silent substitutions were inserted in the area frameshift to reduce secondary structure and translational pausing during expression. The activity of the intrinsic protease was inactivated by a D25A substitution of the catalytic aspartate. The location of the Gag NC/p1 [53] and pl/p6 [54] sites and the Gag-Pro-Pol NC/TF and TF F440/L441 sites [28, 32, 33, 35] are also shown. Below: an overall schematic pGPPfs-PR. B, C. Processing of the HIV-1 Gag-Pro-Pol precursor in vitro showing the kinetics of processing and the generation of product pairs over time. The full-length Gag-Pro-Pol pr160 precursor containing an inactive protease (by PR D25A mutation of the catalytic aspartate) was generated by transcription and translation of pGPPfs-PR in a rabbit reticulocyte lysate. Purified mature HIV-1 protease was added in trans following the 0' timepoint. Aliquots were removed at the indicated time and the protein products separated by Tris-Glycine SDS-PAGE (B) [30] or by Tris-Tricine SDS-PAGE (C) [31]. Paired products resulting from prior removal of IN followed by partial cleavage at the RT/RH site are denoted with brackets. Molecular mass markers are shown on the left. The molecular masses of the intermediates and final products, as estimated from published sequence or common nomenclature, are also shown. Products are represented in abbreviated form by the N- and C-terminal domains according to the nomenclature of Leis et al. [55]. D. Proposed pathway for the ordered processing of the HIV-1 Gag-Pro-Pol precursor by protease in trans. The Gag-Pro-Pol precursor and the observed predominant processing intermediates are represented as boxes with processing sites denoted as vertical lines. The schematic separates the observed Gag-Pro-Pol cleavages into distinct rates. The initial cleavage at p2/NC is shown with a large arrow and labeled 1. The next cleavages occur with similar rates and are labeled 2 (RH/IN and MA/CA). This cleavage is quickly followed by half-cleavage at the RT/RH site (labeled 3). A series of intermediates between 120 kDa and 88 kDa are accounted for at least in part by early cleavage at the sites upstream of RT (TF F440/L441, TF/PR, PR/RT), and these are indicated with small arrows. The slower cleavages at these sites (labeled 4 and 5) give rise to the later paired products. The molecular masses shown of the intermediates and final products were estimated from published sequence or common nomenclature.
Fig. 1B and 1C show the pattern of cleavage products generated at different time points after the addition of protease in trans. We identified over ten distinct species greater than 50 kDa (Fig 1B). Fig. 1C shows products of lower molecular mass [31]. The combination of two different gel systems allowed for the separation and analysis of the appearance of each product. An initial species of 120 kDa (processing intermediate pi120) was rapidly generated within 2 minutes then disappeared to form distinct intermediates of 88, 81, 76, 75, 67, 62 kDa, and finally the mature RT products p66 and p51 (Fig. 1B, C). We observed a large difference in the rates of appearance of these intermediates. After 6 hours of incubation six processing intermediates remained even though the first cleavage event to generate pi120 occurred within 2 min (Fig 1B), indicating that the sites are cleaved at highly different rates. No observable processing occurred without added protease (data not shown), indicating that processing was due to the added protease. Thus, processing of the Gag-Pro-Pol precursor results in a processing cascade consisting of discrete intermediates.
We have used three strategies to assign the cleavage sites that define the ends of the processing products. The first we assigned the products based on the known processing sites in Gag-Pro-Pol. The size of the pi120 intermediate was consistent with an initial cleavage at the p2/NC site, the same site initially cleaved in the Gag precursor [9,14-16]. Second, we truncated the Gag-Pro-Pol precursor to establish the polarity of the initial cleavage site. We implicated cleavage at the p2/NC site by truncating 116 residues from the C-terminal end of the precursor via linearization of the template by Afl II prior to RNA synthesis in vitro. Protease cleavage of the truncated precursor resulted in a shift of the pi120 intermediate to 110 kDa (data not shown), a size consistent with initial cleavage at the p2/NC site. Third, in order to confirm the site of cleavage and the identification of products we blocked individually blocked cleavage at the p2/NC, TF/PR, PR/RT, RT/RH and RH/IN sites by site-directed mutagenesis as described (data not shown) [9,13]. Each blocking mutation resulted in alternative unprocessed intermediates with a molecular mass consistent with an absence of cleavage at the mutated site. Thus, this approach supported the identification of the cleavage sites and the intermediates presented here. We noted that each site was generally cleaved independently of the other sites by protease in trans. A notable exception was the CA/p2 site which showed enhanced cleavage when the earlier cleaved p2/NC site was blocked (M377I mutation). Previously, we reported similar enhanced cleavage of this site in the Gag precursor with the same blocking mutation at the p2/NC site [9]. There is a series of faint minor products between pi120 and pi88, at 113 kDa, 107 kDa, 100 kDa, and 95 kDa (Fig. 2A) seen at the 2-minute time point. These likely represent a low level of cleavage at all of the known cleavage sites upstream of RT early in the processing cascade. We showed by mutagenesis that 113 kDa intermediate resulted from cleavage at the TF F440/L441 site (Fig. 1A, and 1D) rather than cleavage at the NC/TF (data not shown). The TF F440/L441 site has previously been identified as a processing site by others [32-34] using less than full length Pol precursors, and this site is cleaved by the activated PR within full length Gag-Pro-Pol [17,28,35,36]. Other intermediates in this group are likely accounted for as PR-IN (107 kDa) and RT-IN (97 kDa) products.
We observed four sets of paired intermediates and products (denoted by brackets in Fig. 1B, C). We interpret these pairs to represent intermediates that resulted from full cleavage at the RH/IN site followed by half cleavage at the RT/RH site. Numerous studies have shown that partial cleavage of the RT/RH site in the purified RT-RH homodimer is dependent on the dimerization of the RT domain to induce unfolding of a single RH domain [19,21,22,37-40]. We observed a similar pattern with the full length Gag-Pro-Pol precursor, with IN removed prior to half cleavage of the RT/RH cleavage site, also in agreement with [41] where an E. coli based expression system was used. Thus, by analogy with the results of others, we infer that the RT domain within the expressed Gag-Pro-Pol precursor is dimeric either prior to or immediately after removal of IN. The pi88/pi76 paired products, derived from pi120, appeared initially at the 2 minute time point showing that RH/IN and RT/RH cleavage occur relatively early in the processing cascade. The later and overlapping appearance of the three remaining product pairs showed that subsequent N-terminal processing of the pi88/pi76 pair is ordered, but occurs at more similar rates. The SDS-PAGE system utilized in Fig. 1B allows for separation of the pi76 and pi75 intermediates and shows the disappearance of the pi88/pi76 paired products follows the 20 minute time point. The pi81/pi67 and pi75/pi62 pairs represent later products that likely result from cleavage at the TF F440/L441 and TF/PR sites, respectively. Lastly, the mature p66/p51 products represent final cleavage at the PR/RT site.
Initial cleavage at the p2/NC site also generated a MA-CA-p2 (pi42) product (Fig. 1C). We previously showed that cleavage of p42 in vitro occurs at the MA/CA cleavage site followed by slower cleavage at the CA/p2 site [9,13]. We observe here that the rates of processing of the MA/CA and RH/IN sites are similar as shown by the similar appearance of pi25 CA-p2 and p32 IN (Fig. 1C).
Fig. 1D summarizes a proposed cascade for processing of Gag-Pro-Pol by mature protease in trans. The initial cleavage occurs at the p2/NC site (presumably at the same rate this site is cleaved in Gag), generating the pi120 NC-TF-PR-RT-RH-IN intermediate and the p42 MA-CA-p2 intermediate. The next cleavage removes IN from the C terminus of pi120 by cleavage at RH/IN producing pi88. Removal of IN occurs at a rate similar to cleavage between MA-CA. Cleavage of RH/IN is closely followed by cleavage of the RT/RH site to generate the initial paired pi88 and pi76 NC-TF-PR-RT (RH) products. The presence of these paired products suggests that dimerization of the RT-containing processing intermediate occurred early in the processing cascade, consistent with the results of others who observed a similar cleavage pattern using more fully processed dimeric RT [22,38,40]. Processing at the TF F440/L441 and TF/PR occur next followed by the final cleavage between PR/RT to generate the final mature PR and RT products. Final cleavage of the precursor occurs in the sites flanking the PR domain, suggesting that accessibility to these sites may be restricted via formation of a dimer interface structure similar to that observed in mature protease [42].
The overall pattern and extent of processing differs substantially with protease present in trans compared to the pattern seen with the protease embedded in the precursor, as previously characterized [28,35,36]. Cleavage of the Gag-Pro-Pol precursor by the embedded protease appears to be much more restrictive with cleavages only observed at the p2/NC site and the TF F440/L441 sites. We show here that protease present in trans cleaves all of the Gag-Pro-Pol sites but at varying rates (Figs. 1B, C, D), resulting in a processing cascade. One possibility is that the embedded protease shows restricted site selection due to its location within the precursor.
We infer that the Gag-Pro-Pol precursor was able to dimerize in this expression system. The state of the Gag-Pro-Pol precursor in newly assembled (or assembling) virions could differ. In infected cells, Gag-Pro-Pol may dimerize while moving to the assembly site [43-46] or during assembly, affecting the kinetics of precursor processing. Alternatively, dimerization of Gag-Pro-Pol monomer may be constrained by the excess of Gag during assembly, as suggested by others [47-49]. In that case, the presence of Gag could limit Gag-Pro-Pol dimerization by forming heterodimers, in turn altering the kinetic of processing. These considerations are not mutually exclusive. One of the early cleavage events detailed here (such as cleavage at p2/NC) could also release a truncated precursor from a Gag/Gag-Pro-Pol heterodimer and permit rapid dimerization of the PR and RT domains.
The other feature of the system we have used is the reliance of protease cleavages in trans. Use of trans protease on the full length precursor allows for the clear evaluation of generation of each product, however, this approach is unable to discern the possible cleavage of nascent or truncated products or the effect of an active embedded protease. Expression of Gag-Pro-Pol in vitro with an unmutated protease domain results in rapid autocatalytic cleavage at the p2/NC site and the TF F440/L441 site to produce the 113 KDa intermediate [28,35]. Immediate dimerization in cells of the full length precursor would likely result in premature cleavage [50-52]. Thus, in the context of budding virions there may be an interplay between monomeric versus dimeric Gag-Pro-Pol as substrate, and embedded versus free protease for cleavage. The extent to which these different combinations may alter the order of cleavage and the successful assembly of virus is not known.
We show here that cleavage of the Gag-Pro-Pol processing sites by trans protease occurs at different rates, and we suggest that cleavage is likely regulated, in part, by the dimerization of the protease and RT domains. We and others have shown that timed and ordered cleavage of the HIV-1 Gag precursors is highly regulated and is necessary for the production of an infectious, properly assembled virion. We do not yet know the extent of the requirement for timed cleavage of Gag-Pro-Pol in producing infectious virus. Characterization of the ordered cleavage of Gag-Pro-Pol furthers our understanding of HIV-1 precursor processing and suggests further mechanisms at work in the regulation of HIV-1 assembly.
Competing interests
The author(s) declare that they have no competing interests.
Authors' contributions
JL and SP carried out the experiments. RS and SP drafted the manuscripts and designed the experiments. AK provided helpful discussion and editing of the manuscript.
Acknowledgements
This study was funded by NIH grants R01-AI50485 (to RS) and R01-GM66681 (to AK) in addition to support from the UNC Center For AIDS Research (P30-AI50410).
==== Refs
Craven RC Bennett RP Wills JW Role of the avian retroviral protease in the activation of reverse transcriptase during virion assembly J Virol 1991 65 6205 6217 1717718
Stewart L Vogt VM Reverse transcriptase and protease activities of avian leukosis virus Gag-Pol fusion proteins expressed in insect cells J Virol 1993 67 7582 7596 7693975
Swanstrom R Wills J Varmus HE, Coffin JM and Hughes SH Retroviral gene expression: Synthsis, processing, and assembly of viral proteins. Retroviruses 1997 Cold Spring Harbor, NY, Cold Spring Harbor Laboratory Press 263 334
Xiang Y Ridky TW Krishna NK Leis J Altered Rous sarcoma virus Gag polyprotein processing and its effects on particle formation J Virol 1997 71 2083 2091 9032340
Gross I Hohenberg H Wilk T Wiegers K Grattinger M Muller B Fuller S Krausslich HG A conformational switch controlling HIV-1 morphogenesis Embo J 2000 19 103 113 10619849 10.1093/emboj/19.1.103
Joshi SM Vogt VM Role of the Rous sarcoma virus p10 domain in shape determination of gag virus-like particles assembled in vitro and within Escherichia coli J Virol 2000 74 10260 10268 11024160 10.1128/JVI.74.21.10260-10268.2000
Nandhagopal N Simpson AA Johnson MC Francisco AB Schatz GW Rossmann MG Vogt VM Dimeric rous sarcoma virus capsid protein structure relevant to immature Gag assembly J Mol Biol 2004 335 275 282 14659756 10.1016/j.jmb.2003.10.034
Craven RC Leure PA Erdie CR Wilson CB Wills JW Necessity of the spacer peptide between CA and NC in the Rous sarcoma virus gag protein J Virol 1993 67 6246 6252 8396679
Pettit SC Moody MD Wehbie RS Kaplan AH Nantermet PV Klein CA Swanstrom R The p2 domain of human immunodeficiency virus type 1 Gag regulates sequential proteolytic processing and is required to produce fully infectious virions J Virol 1994 68 8017 8027 7966591
Vogt VM Proteolytic processing and particle maturation Curr Top Microbiol Immunol 1996 214 95 131 8791726
Sheng N Pettit SC Tritch RJ Ozturk DH Rayner MM Swanstrom R Erickson VS Determinants of the human immunodeficiency virus type 1 p15NC-RNA interaction that affect enhanced cleavage by the viral protease J Virol 1997 71 5723 5732 9223458
Wiegers K Rutter G Kottler H Tessmer U Hohenberg H Krausslich HG Sequential steps in human immunodeficiency virus particle maturation revealed by alterations of individual Gag polyprotein cleavage sites J Virol 1998 72 2846 2854 9525604
Pettit SC Henderson GJ Schiffer CA Swanstrom R Replacement of the P1 amino acid of human immunodeficiency virus type 1 Gag processing sites can inhibit or enhance the rate of cleavage by the viral protease J Virol 2002 76 10226 10233 12239298 10.1128/JVI.76.20.10226-10233.2002
Mervis RJ Ahmad N Lillehoj EP Raum MG Salazar FH Chan HW Venkatesan S The gag gene products of human immunodeficiency virus type 1: alignment within the gag open reading frame, identification of posttranslational modifications, and evidence for alternative gag precursors J Virol 1988 62 3993 4002 3262776
Gowda S Stein B Engleman E Identification of protein intermediates in the processing of the p55 HIV-1 Gag precursor in cells infected with recombinant vaccinia virus J Biol Chem 1989 264 8459 8462 2785991
Krausslich HG Ingraham RH Skoog MT Wimmer E Pallai PV Carter CA Activity of purified biosynthetic proteinase of human immunodeficiency virus on natural substrates and synthetic peptides Proc Natl Acad Sci USA 1989 86 807 811 2644644
Lindhofer H von der Helm K Nitschko H In vivo processing of Pr160gag-pol from human immunodeficiency virus type 1 (HIV) in acutely infected, cultured human T-lymphocytes Virology 1995 214 624 627 8553565 10.1006/viro.1995.0074
Almog N Roller R Arad G Passi EL Wainberg MA Kotler M A p6Pol-protease fusion protein is present in mature particles of human immunodeficiency virus type 1 J Virol 1996 70 7228 7232 8794372
Veronese FD Copeland TD DeVico AL Rahman R Oroszlan S Gallo RC Sarngadharan MG Characterization of highly immunogenic p66/51 as the reverse transcriptase of HTLV-III/LAV Science 1986 231 1289 1291 2418504
Copeland TD Oroszlan S Genetic locus, primary structure, and chemical synthesis human immunodeficiency virus protease Gene Anal Tech 1988 5 109 115 3063643 10.1016/0735-0651(88)90010-6
Chattopadhyay D Evans DB Deibel MR Vosters AF Eckenrode FM Einspahr HM Hui JO Tomasselli AG Zurcherneely HA Heinrikson RL Sharma SK Purification and Characterization of Heterodimeric Human Immunodeficiency Virus Type-1 (HIV-1) Reverse Transcriptase Produced by Invitro Processing of P66 with Recombinant HIV-1 Protease J Biol Chem 1992 267 14227 14232 1378437
Tomasselli AG Sarcich JL Barrett LJ Reardon IM Howe WJ Evans DB Sharma SK Heinrikson RL Human immunodeficiency virus type-1 reverse transcriptase and ribonuclease H as substrates of the viral protease Protein Sci 1993 2 2167 2176 7507754
Jacks T Power MD Maslarz FR Luciw PA Barr PJ Varmus HE Characterization of ribosomal frame-shifting in HIV-1 gag/pol expression Nature 1988 331 280 283 2447506 10.1038/331280a0
Reil H Kollmus H Weidle UH Hauser H A heptanucleotide sequence mediates ribosomal frameshifting in mammalian cells J Virol 1993 67 5579 5584 8350413
Bebenek K Kunkel TA The use of native T7 DNA polymerase for site-directed mutagenesis Nucleic Acids Res 1989 17 5408 2668888
Kunkel TA Bebenek K McClary J Efficient site-directed mutagenesis using uracil-containing DNA Methods Enzymol 1991 204 125 139 1943776
Gorelick RJ Henderson LE Myers G, Korber B, Wain-Hobson S, Jeang KT, Henderson L and Pavlakis G Part III: Analyses Human Retroviruses and AIDS 1994 Los Alamos, NM, The Los Alamos National Laboratory 2 5
Pettit SC Gulnik S Everitt L Kaplan AH The dimer interfaces of protease and extra-protease domains influence the activation of protease and the specificity of GagPol cleavage J Virol 2003 77 366 374 12477841 10.1128/JVI.77.1.366-374.2003
Moody MD Pettit SC Shao W Everitt L Loeb DD Hutchison C Swanstrom R A side chain at position 48 of the human immunodeficiency virus type-1 protease flap provides an additional specificity determinant Virology 1995 207 475 485 7886951 10.1006/viro.1995.1107
Laemmli UK Cleavage of structural proteins during the assembly of the head of bacterophage T4 Nature (London) 1970 277 680 685 5432063 10.1038/227680a0
Schagger H von Jagow G Tricine-sodium dodecyl sulfate-polyacrylamide gel electrophoresis for the separation of proteins in the range from 1 to 100 kDa Anal Biochem 1987 166 368 379 2449095 10.1016/0003-2697(87)90587-2
Phylip LH Griffiths JT Mills JS Graves MC Dunn BM Kay J Activities of precursor and tethered dimer forms of HIV proteinase Adv Exp Med Biol 1995 362 467 472 8540359
Louis JM Dyda F Nashed NT Kimmel AR Davies DR Hydrophilic peptides derived from the transframe region of Gag-Pol inhibit the HIV-1 protease Biochemistry 1998 37 2105 2110 9485357 10.1021/bi972059x
Chen N Morag A Almog N Blumenzweig I Dreazin O Kotler M Extended nucleocapsid protein is cleaved from the Gag-Pol precursor of human immunodeficiency virus type 1 J Gen Virol 2001 82 581 590 11172099
Pettit S Everitt L Choudhury S Dunn BM Kaplan AH Initial cleavage of the HIV-1 GagPol precursor by its activated protease occurs by an intramolecular mechanism J Virol 2004 78 8477 8485 15280456 10.1128/JVI.78.16.8477-8485.2004
Pettit SC Clemente JC Jeung JA Dunn BM Kaplan AH Ordered processing of the human immunodeficiency virus type 1 GagPol precursor is influenced by the context of the embedded viral protease J Virol 2005 79 10601 10607 16051852 10.1128/JVI.79.16.10601-10607.2005
Lightfoote MM Colisan JE Folks TM Fauci AS Martin MA Venkatesan S Structural characterization of reverse transcriptase and endonuclease polypeptides of the acquired immunodeficiency syndrome retrovirus J Virol 1986 60 771 775 2430111
Goobar-Larsson L Backbro K Unge T Bhikhabhai R Vrang L Zhang H Orvell C Strandberg B Oberg B Disruption of a salt bridge between Asp 488 and Lys 465 in HIV-1 reverse transcriptase alters its proteolytic processing and polymerase activity Virology 1993 196 731 738 7690504 10.1006/viro.1993.1530
Wang J Smerdon SJ Jager J Kohlstaedt LA Rice PA Friedman JM Steitz TA Structural basis of asymmetry in the human immunodeficiency virus type 1 reverse transcriptase heterodimer Proc Natl Acad Sci USA 1994 91 7242 7246 7518928
Navarro JM Damier L Boretto J Priet S Canard B Querat G Sire J Glutamic residue 438 within the protease-sensitive subdomain of HIV-1 reverse transcriptase is critical for heterodimer processing in viral particles Virology 2001 290 300 308 11883194 10.1006/viro.2001.1188
Sluis-Cremer N Arion D Abram ME Parniak MA Proteolytic processing of an HIV-1 pol polyprotein precursor: insights into the mechanism of reverse transcriptase p66/p51 heterodimer formation Int J Biochem Cell Biol 2004 36 1836 1847 15183348 10.1016/j.biocel.2004.02.020
Weber IT Comparison of the crystal structures and intersubunit interactions of human immunodeficiency and Rous sarcoma virus proteases J Biol Chem 1990 265 10492 10496 2162350
Hill MK Hooker CW Harrich D Crowe SM Mak J Gag-Pol supplied in trans is efficiently packaged and supports viral function in human immunodeficiency virus type 1 J Virol 2001 75 6835 6840 11435562 10.1128/JVI.75.15.6835-6840.2001
Carter CA Tsg101: HIV-1's ticket to ride Trends Microbiol 2002 10 203 205 11973141 10.1016/S0966-842X(02)02350-8
Goff A Ehrlich LS Cohen SN Carter CA Tsg101 control of human immunodeficiency virus type 1 Gag trafficking and release J Virol 2003 77 9173 9182 12915533 10.1128/JVI.77.17.9173-9182.2003
Ono A Freed EO Cell-type-dependent targeting of human immunodeficiency virus type 1 assembly to the plasma membrane and the multivesicular body J Virol 2004 78 1552 1563 14722309 10.1128/JVI.78.3.1552-1563.2004
Park J Morrow CD Overexpression of the gag-pol precursor from human immunodeficiency virus type 1 proviral genomes results in efficient proteolytic processing in the absence of virion production J Virol 1991 65 5111 5117 1870215
Karacostas V Wolffe EJ Nagashima K Gonda MA Moss B Overexpression of the HIV-1 gag-pol polyprotein results in intracellular activation of HIV-1 protease and inhibition of assembly and budding of virus-like particles Virology 1993 193 661 671 7681610 10.1006/viro.1993.1174
Halwani R Khorchid A Cen S Kleiman L Rapid localization of Gag/GagPol complexes to detergent-resistant membrane during the assembly of human immunodeficiency virus type 1 J Virol 2003 77 3973 3984 12634357 10.1128/JVI.77.7.3973-3984.2003
Krausslich HG Human immunodeficiency virus proteinase dimer as component of the viral polyprotein prevents particle assembly and viral infectivity Proc Natl Acad Sci USA 1991 88 3213 3217 2014242
Krausslich HG Specific inhibitor of human immunodeficiency virus proteinase prevents the cytotoxic effects of a single-chain proteinase dimer and restores particle formation J Virol 1992 66 567 572 1727499
Gatlin J Arrigo SJ Schmidt MG Regulation of intracellular human immunodeficiency virus type-1 protease activity Virology 1998 244 87 96 9581782 10.1006/viro.1998.9083
Wondrak EM Louis JM de RH Chermann JC Roques BP The gag precursor contains a specific HIV-1 protease cleavage site between the NC (P7) and P1 proteins Febs Letters 1993 333 21 24 8224164 10.1016/0014-5793(93)80367-4
Henderson LE Sowder RC Copeland TD Oroszlan S Benveniste RE Gag precursors of HIV and SIV are cleaved into six proteins found in the mature virions J Med Primatol 1990 19 411 419 2231692
Leis J Baltimore D Bishop JM Coffin J Fleissner E Goff SP Oroszlan S Robinson H Skalka AM Temin HM Vogt VM Standardized and simplified nomenclature for proteins common to all retroviruses J Virol 1988 62 1808 1809 3357211
|
16262906
|
PMC1291402
|
CC BY
|
2021-01-04 16:36:39
|
no
|
Retrovirology. 2005 Nov 1; 2:66
|
utf-8
|
Retrovirology
| 2,005 |
10.1186/1742-4690-2-66
|
oa_comm
|
==== Front
Respir ResRespiratory Research1465-99211465-993XBioMed Central London 1465-9921-6-1231624625910.1186/1465-9921-6-123ResearchAlcohol reversibly disrupts TNF-α/TACE interactions in the cell membrane Song Kejing [email protected] Xue-Jun [email protected] Luis [email protected] Peter [email protected] Steve [email protected] Jay K [email protected] LSUHSC Gene Therapy Program and the LSUHSC Alcohol Research Center, LSU Health Sciences Center, CSRB Rm. 601, 533 Bolivar St., New Orleans, LA 70112, USA2 Children's Hospital of Pittsburgh/University of Pittsburgh, Rm. 3765, 3705 Fifth Ave., Pittsburgh, PA 15213, USA2005 24 10 2005 6 1 123 123 14 7 2005 24 10 2005 Copyright © 2005 Song et al; licensee BioMed Central Ltd.2005Song et al; licensee BioMed Central Ltd.This is an Open Access article distributed under the terms of the Creative Commons Attribution License (), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
Background
Alcohol abuse has long been known to adversely affect innate and adaptive immune responses and pre-dispose to infections. One cellular mechanism responsible for this effect is alcohol-induced suppression of TNF-α (TNF) by mononuclear phagocytes. We have previously shown that alcohol in part inhibits TNF-α processing by TNF converting enzyme (TACE) in human monocytes. We hypothesized that the chain length of the alcohol is critical for post-transcriptional suppression of TNF secretion.
Methods
Due to the complex transcriptional and post-transcriptional regulation of TNF in macrophages, to specifically study TNF processing at the cell membrane we performed transient transfections of A549 cells with the TNF cDNA driven by the heterologous CMV promoter. TNF/TACE interactions at the cell surface were assessed using fluorescent resonance energy transfer (FRET) microscopy.
Results
The single carbon alcohol, methanol suppressed neither TNF secretion nor FRET efficiency between TNF and TACE. However, 2, 3, and 4 carbon alcohols were potent suppressors of TNF processing and FRET efficiency. The effect of ethanol, a 2-carbon alcohol was reversible.
Conclusion
These data show that inhibition of TNF-α processing by acute ethanol is a direct affect of ethanol on the cell membrane and is reversible upon cessation or metabolism.
Cytokineslipopolysaccharideinflammation
==== Body
Background
Alcohol is a potent immunosuppressive drug and has been widely recognized for many centuries as an important risk factor for the development of infections. Among infection, pneumonia has been highly linked to alcohol abuse. In a recent case-controlled study in Spain, of 50 control subjects and 50 patients with community-acquired pneumonia, high alcohol intake was identified as a risk factor for pneumonia [1].
The cellular mechanisms by which alcohol results in immunosupression have recently been reviewed [2]. One proposed mechanism by which acute alcohol exposure suppresses innate immunity is its dose-dependent suppression of TNF-α (TNF) elaboration by mononuclear phagocytes [3-6]. The suppressive affects of alcohol on TNF elaboration occurs at several levels. Szabo and colleagues have reported decreased nuclear translocation of NF-kB and reduced steady-state levels of TNF mRNA in alcohol exposed human monocytes [3,7]. Acute alcohol exposure also results in a significant post-transcriptional and post-translational suppression of TNF production in both rodent and primate macrophages as measured by steady-state transcript levels of TNF-α mRNA [8-11]. Recently we have shown that this is in part due to the ability of acute alcohol exposure to interfere with TNF interacting with TNF-α converting enzyme (TACE), a member of the disintegrin and metalloproteinase (ADAM) family of proteins [12,13], in the cell membrane [14].
Based on these data we hypothesized that this was a direct affect of ethanol and therefore reversible when ethanol was no longer present in the cell and secondary to the ability of alcohol to intercalate in the cell membrane. If this latter condition were true, than the suppression of TNF/TACE interactions should be related to the hydrophobicity and the chain length of the alcohol. To examine these two hypotheses we used florescence resonance energy transfer (FRET) to assess TNF/TACE interactions in A549 cells as previously described [14] using clinically relevant intoxicating concentrations of ethanol, 0, 50 or 100 mM (or approximately 0, 230 or 460 mg/dl).
Methods
Plasmid and plasmid construction
The full-length coding sequence of human TNF-α was prepared by polymerase chain reaction (PCR) from its cDNA coded in pcDV1 (ATCC#39894) with COOH-terminal deletion using following primers: sense, 5'-AAGCTTGGTACCACCACTATGAGCACTGAAAGCATGATC-3'; antisense, 5'-TGACTAGGATCCCAGGGCAATGATTCCAAAGTAGAC-3'. The PCR product was digested with KpnI and BamHI and inserted in-frame into p3xFLAG-CMV-14 protein expression vector (Sigma, St. Louis, MO) to form a vector: p3xFLAG-TNF-α-CMV-14 which can express a FLAG-tagged human TNF-α, driven by the CMV promoter.
Cell Culture
A549 cells were maintained in D-MEM/F12 supplemented with 10% heat-inactivated fetal bovine serum (FBS), 2 mM L-glutamine.
Transfection of A549 cells
Due to the complex transcriptional and post-transcriptional regulation of TNF in macrophages, to specifically study TNF processing by TACE at the cell membrane we performed transient transfections of A549 cells (which express TACE) with the TNF cDNA. A549 cells were transfected with p3xFLAG- TNF-α-CMV-14 using Lipofectamine™2000 reagent (Invitrogen) as recommended by the manufacturer. Transfected cells were incubated in the 37°C, 5% CO2 humidified incubator for 24 hours to get the greatest protein expression (data not shown).
Twenty-four hours post-transfection, A549 cells were split into 24-well plates, adhered for 3 hours, and randomized to treatments for 1 or 2 hours with fresh medium or medium with 50 mM or 100 mM alcohols of different one to four carbon chain lengths, methanol, ethanol, 1-propanol, and n-butanol. A subgroup of cells randomized to the 2 hour incubation time had media changed to remove the alcohol after 1 hour and cultured for an additional 1 hour washout period. Culture supernatants were then harvested for secreted TNF-α determinations by ELISA and cells were removed from culture plate by incubating with 2 mM EDTA and lysed with 1%NP-40 in phosphate-buffered saline (PBS) containing 1 mM EDTA and proteases inhibitor cocktail (Roche) to measure cell-associated TNF-α as previously described [14]. In certain experiments, RNA was extracted with Tri-zol (Invitrogen) and TNF-α mRNA elves were determined using human TNF primers and the following probe: 5'-FAM-CATCGCCGTCC-TACCAGACCAAG-Black Hole Quencher 1–3' (Biosource, Camarillo, CA). Reactions were run on an i-Cycler (Bio-RAD) and normalized per ng 18 rRNA content.
ELISA
The levels of the cell associated form of TNF-α in cell lysates and the secreted from in culture supernatant were measured by ELISA. Before testing, both cell lysates and culture media were spun by centrifugation for 15 min at 13,000 rpm in a microfuge. ELISA was performed using kit from R&D Systems (Minneapolis, MN) and following the manufacturer's protocol. Cell-associated TNF values were normalized to total protein using a Pierce Protein assay (Pierce, Rockford, IL). Concentrations of EtOH up to 100 mM had no inhibitory effect on the TNF ELISA.
Immunofluorescent Labeling
Transfected A549 cells were adhered to cover slips in culture. After alcohol exposure for 1 or 2 hour or after the washout period, the coverslips were washed in cold PBS to remove excess media and fixed in a 4% solution of 10% methanol-free formaldehyde (Polysciences, Inc., Warrington, PA) in PBS for 15 minutes at room temperature. After thoroughly washing the fixative off with cold PBS, the cells were blocked with 5% rabbit serum in 1% bovine serum albumin (BSA) in PBS for 15 minutes. Subsequently, the blocker was replaced with the first primary antibody, mouse anti-human TACE clone M222 (Amgen, Seattle, WA), diluted in 1% BSA at 10 ug/mL. The coverslips were thoroughly washed in three PBS changes after the 45-minute primary antibody incubation. The antibody was indirectly labeled with a rabbit anti-mouse, Cy-3 conjugated Fab fragment (Jackson Immunoresearch, West Grove, PA) at 6.5 ug/mL in 1% BSA. Next, cells were washed and blocked with 5% mouse serum for 10 minutes to saturate any open antigen binding sites within the first primary antibody. The blocker was replaced with the second, primary antibody mouse anti-human, FITC-conjugated, TNF-α clone Mab11 (BD Pharmingen, San Diego, CA) at a 10 ug/mL dilution in 1% BSA for 30 minutes. All antibody dilutions remained constant throughout experiments and were titrated to be detected at their highest saturation point, hence maximizing to an optimal donor to acceptor ratio. Finally, the coverslips were washed, mounted cell-side down in PBS on slides with spacers, and secured with nail polish.
FRET Imaging and Data Collection
Human A549 cells mounted on coverslips were imaged using a digital fluorescence microscopy system. The system included a 12-bit, chilled charged-coupled device (CCD) Sensicam QE with a 1376 × 1040 resolution and 65% quantum efficiency at 550 nm (The Cooke Corporation, Auburn Hills, MI) on an automated Leica DMRXA microscope (Meyer Instruments, Houston, TX) using a 1.4 NA, 63 × Plan-apochromat objective. FITC and Cy3 were detected using appropriate filters (customized FITC filter cube: excitation 480/40 nm, 505 long-pass dichroic, emission 535/50; Cy3 filter cube: excitation 545/30, 565 long-pass dichroic, emission 620/60) (Chroma Technology Corporation, Brattleboro, VT). Fluorescence was excited with a 75 W Xenon arc lamp. The samples were photobleached with 100 W Mercury source. All components were controlled by Slidebook software (Intelligent Imaging Innovations, Denver, CO), which was used for capture, nearest neighbor deconvolution, and FRET analysis. Image acquisition was adjusted in milliseconds to achieve the maximum CCD camera range. As part of detecting bleed-through, typical exposure times were used to confirm no visualization of FITC-labeled samples with the Cy3 filter cube and vice versa.
All experiments were based on FRET measured by acceptor photo bleaching recovery using FITC as the donor and Cy3 as the acceptor. The Cy3 was determined to be photolabile enough to undergo the pre-measured 3-minute photo-bleaching step for a subsequent <2% initial intensity measurement. The donor had a high enough quantum yield and stability to not fade significantly. All experiments included a donor-only and acceptor-only sample to verify minimal crossover between them, hence no significant energy transfer. Data were collected for 20 different fields from a single cover slip. Each fluorescent channel was collected at six consecutive z-planes measuring 0.3 um each. The software allowed stack auto-alignment between exposures to compensate for thermally induced shifts. The FRET experiment imaging began with an initial image stack of the FITC-labeled protein (in the presence of the Cy3-labeled antibody) obtained with the FITC filter set immediately followed by an image stack captured with the Cy3 filter cube. Next, the sample was exposed to constant illumination for 3 minutes to photo bleach. An image stack of the Cy3 fluorescence after photo bleaching was then obtained followed by another FITC fluorescence image stack using the FITC filter set. The exposure times were unchanged between pre and post-photo bleaching steps. Post-FRET calculation images were deconvolved using a nearest neighbor algorithm for presentation purposes only.
FRET data analysis
Images indicating FRET between the labeled antibodies against TNF and TACE were calculated based in the increase in donor fluorescence after acceptor photobleaching by using the following formula:
E = FITC post photo bleach - FITC pre photo bleach / FITC post photo bleach
Here, E represents FRET efficiency after performing a background subtraction caused by scattered light, auto fluorescence, and dark current of the CCD camera. Fluorescence intensities were reported in arbitrary units and on a pixel-to-pixel basis. This was achieved by digitally performing the pre-photo bleaching channel subtraction from the post-photo bleaching channel (see above formula) and thresholding the difference, labeled as "mean FRET intensity". FRET efficiency data was determined by analyzing at least 50 cells per coverslip.
Statistical methods
All data are presented as mean ± SEM. Significance was estimated using ANOVA followed by Tukey's Multiple Comparison Procedure with p < 0.05 being considered significant.
Results
Reversibility of Ethanol and effect of alcohol chain-length on TNF-α secretion
In pilot experiments, we determined the hourly rate of TNF secretion in transfected A549 cells and found that the rate of secretion of TNF-α was 1400–1550 pg/ml from hour 27–28 and hour 28–29 post transfection (hour 27–28 data in Figure 1A). The addition of 50 or 100 mM EtOH at 27 hours after transfection resulted in a dose-dependent secretion of TNF-α into the medium (Figure 1A). Similar suppression was also observed 2 hours after the addition of ethanol (data not shown). However, cells washed 1 hour after the addition of ethanol, and replaced with fresh media recovered their TNF secretion (Figure 1A). Mean Ct values were 22.3 + .36 per ng rRNA content for TNF message were not significantly altered by the 1 or 2 hour alcohol treatment (data not shown).
Figure 1 Panel A: Acute EtOH reversibly suppresses TNF-α secretion in transfected A549 cells. A549 cells transfected with the hTNF-α cDNA were re-seeded 24 hours after transfection and subjected to the addition of 0, 50 or 100 mM EtOH for 1 hour (n = 6, *. Denoted p < 0.05). For the washout experiments, a subgroup of transfected cells had media changed after one hour to fresh media without alcohol for an additional hour (n = 6, * denotes p < 0.05). Panel B: Effect of one to four carbon alcohols on TNF-α secretion (n = 4–6, * denotes p < 0.05 compared to EtOH). Panel C: Effect of one to four carbon alcohols on cell-associated TNF-α levels (n = 4–6, * denotes p < 0.05 compared to EtOH). Transfected A549 cells were treated as in Materials and Methods and TNF-α was measured in cell lysates as the cell-associated level (all data are per mg of protein). Panel D: Effect of one to four carbon alcohols on TNF processing as assessed by the ratio of shed vs. cell-associated TNF-α (n = 4–6, * denotes p < 0.05 compared to EtOH). Transfected A549 cells were treated as in Materials and Methods and TNF-α was measured in cell lysates and media and graphed as the ratio of the shed over the cell-associated level.
We next examined one to four carbon alcohols at concentrations of 0, 50 and 100 mM. Again we observed a dose-dependent suppression of TNF secretion by ethanol, however, methanol had no affect on TNF secretion (Figure 1B). 1-propanol and n-butanol on the other hand, were more potent than ethanol in suppressing TNF-a secretion (Figure 1B). The decrement in TNF production was not due to loss of cell viability as determined by trypan blue staining (data not shown). Moreover, this suppression was verified to occur at the level of TNF-a secretion as the alcohol with two to four carbons did not suppress cell-associated TNF concentrations (Figure 1C) and longer chain alcohols disproportionately suppressed the ratio of shed versus cell-associated TNF-a (Figure 1D), a measure of TACE efficiency.
Effect of alcohols on TNF/TACE interactions by FRET microscopy
To examine if the chain length of alcohol had a differential activity on TNF/TACE interactions, we performed FRET microscopy as previously described [14]. Using deconvolution microscopy, transfected A549 cells expressed TNF-α (Figure 2B) and TACE (Figure 2A). A549 cells, transfected with the human TNF cDNA, were incubated with 50 mM methanol, ethanol, or 1-propanol for 60 minutes followed by fixation in 4% methanol-free formaldehyde. Experiments were not performed with N-butanol as there was altered cell morphology (vacuolization) with this alcohol. Cells were stained for TNF and TACE and FRET efficiency was measured and calculated from 20 representative fields per group. The TACE (Cy3, acceptor) signal was photobleached (Figure 2 panels c, h, m, r) whereas the TNF (FITC) signal increased (Figure 2, panels d, I, n, s) consistent with FRET. FRET was quantified by using pixel by pixel subtraction of FITC fluorescent intensity post and pre-photobleaching. To depict FRET on the cell surface, the FRET intensity was superimposed over donor intensity as a pseudocolor image (Figure 2 panels u, v, w, and x), where blue is little or no FRET efficiency and red is peak FRET efficiency. Incubation with increasing chain lengths of alcohol resulted in an increase in cell surface sting of TNF-a (Figure 2, panels b, g, l, q) consistent with a suppression of TNF-a cleavage from these cells. This was associated with a decreased in cell surface FRET (Figure 2, panels u, v, w, and x and Figure 3). Consistent with a reversible process, FRET efficiency returned to pre-alcohol control values when the media was changed to non-alcohol containing media for an additional hour (Figure 3).
Figure 2 Immunofluorescent detection, co-localization, and FRET intensity in representative transfected A549 cells. Cells were stained with anti-TACE (Cy3) and anti-TNF-α (FITC) as FRET acceptor and donor, respectively (panels a-t). The staining pattern is suggestive of both cell membrane and membrane vesicle localization. In photo-bleached samples (Panels c, h, m, r) there was less than 2% initial anti-TACE (Cy3) intensity post-photobleaching. Panels u-x: FRET intensity, calculated from the difference between donor pre- and post-photobleaching intensities shown in pseudocolor.
Figure 3 Mean FRET intensity, calculated from the difference between donor (TNF-α) pre- and post-photobleaching intensities in 20 high power fields per group (n = 4, * denotes p < 0.05 compared to 0 mM control).
Discussion
These studies show that acute alcohol exposure, at clinically relevant concentrations (for ethanol), block TNF secretion by posttranslational mechanism by inhibiting protein: protein interactions between TNF and TACE in the cell membrane. There have been many studies to better understand the mechanisms by which alcohol contributes to immunosuppression. Many types of infection, including bacterial pneumonia and, more recently, Hepatitis C have been shown to be adversely affected by concomitant alcohol abuse [15-18]. In addition, alcohol has been shown to be an independent risk factor for bacterial pneumonia and development of the Acute Respiratory Distress Syndrome (ARDS) [19]. In studies of pneumonia and in the setting of trauma, which is a risk factor for ARDS, blood alcohol levels can approach 50–75 mM.
One possible mechanism underlying the immunosuppressive effects of acute alcohol exposure is its effects on innate immunity. In human volunteers and experimental animal models, alcohol dose-dependently suppresses neutrophil recruitment in response to chemotactic stimuli. One mechanism operative in the lung is suppression of the pro-inflammatory cytokine, TNF, which is required for induction of CXC-chemokines critical for neutrophil recruitment [20,21], as well as expression of adhesion molecules on vascular endothelium which is required for neutrophil binding and diapedesis [20].
In addition to the known affects of ethanol on TNF-α transcription [7,22], another mechanism of ethanol-induced TNF secretion also occurs at post-transcriptional levels [11,14]. To avoid issues of transcription, and mRNA stability regulated by the TNF-α 3'-untranslated region [23,24] in these experiments, we expressed TNF under control of the heterologous CMV promoter and a poly-adenylation signal from SV40 in human bronchoalveolar cells (A549) which also express surface TACE [14].
Using protein assays and FRET microscopy, we demonstrated that acute alcohol exposure resulted in dose dependent suppression of TACE-mediated processing of TNF. Moreover the suppression of TNF secretion was recoverable within one hour of removing the ethanol. Methanol which contains a single carbon was ineffective in suppressing TNF in A549 transfectants whereas both 1-propanol (3-carbon) and n-butanol (4-carbon) were more potent in suppressing in TNF secretion. Furthermore, methanol was a weak inhibitor of TNF/TACE interactions as measured by FRET efficiency whereas both ethanol and 1-propanol potently and reversibly inhibited FRET efficiency. Taken together these data show that alcohol disrupts TNF/TACE interactions by an activity which is reversible and related to alcohol-chain length. These data support a model by which ethanol, at clinically relevant concentrations, disrupts the cell membrane in a reversible fashion, resulting in inhibition of interactions between TNF-α and TACE.
The mechanism of the reduction TNF/TACE interactions at present is unclear. The chain-length data suggest the possibility of changes in membrane fluidity [25] induced by EtOH [25] or alterations in raft compartments, or perhaps abnormal extracellular signal-regulated kinase (Erk)-mediated phosphorylation of TACE which is critical for protein kinase C-regulated TrkA cleavage [26]. Although there are no specific studies as of yet examining TNF/TACE interactions in the context of membrane rafts, members of the TNF receptor superfamily, CD40 and CD120a [27,28] as well as alpha secretases [29] have been localized to membrane rafts. TACE has been localized to non-raft compartments but depletion of cholesterol permitted the interactions of TACE with CD30 resulting in enlaced shedding of the CD30 ectodomain [30]. Both acute and chronic alcohol has been shown to increase membrane cholesterol content [31,32] and this may be one mechanism by which ethanol disrupts TNF/TACE interactions. As TACE is involved in the cleavage of multiple cytokine receptors [33,34] as well as cell adhesion molecules such as L-selectin [35,33], this may explain in part the potency alcohol as immunosuppressant of both innate and adaptive immune responses.
Conclusion
Alcohol chain length is clearly a factor in ethanol's ability to inhibit TNF secretion as well as TNF/TACE interactions as assessed by FRET. Moreover the effect of EtOH on this process is reversible. Taken together these data strongly support a model that alcohol disrupts TNF/TACE interaction in the cell membrane.
Competing interests
The author(s) declare that they have no competing interests.
Authors' contributions
KS performed the transient transfections and TNF ELISA, X-J. Z. assisted with the transfections, LM did the FRET analysis, P.O., SN, and JKK assisted in experimental design and drafting the manuscript.
Acknowledgements
This work was supported by R01-AA010384 (J.K.K.) and P60AA009803 (S.N.).
==== Refs
Fernandez-Sola J Junque A Estruch R Monforte R Torres A Urbano-Marquez A High alcohol intake as a risk and prognostic factor for community- acquired pneumonia Arch Intern Med 1995 155 1649 1654 7618989 10.1001/archinte.155.15.1649
Nelson S Kolls JK Alcohol, host defence and society Nature Rev Immunol 2002 2 205 209 11913071 10.1038/nri744
Mandrekar P Catalano D Szabo G Inhibition of lipopolysaccharide-mediated NFkappaB activation by ethanol in human monocytes Int Immunol 1999 11 1781 1790 10545482 10.1093/intimm/11.11.1781
Szabo G Mandrekar P Girouard L Catalano D Regulation of human monocyte functions by acute ethanol treatment: Decreased tumor necrosis factor-α, interleukin-1β and elevated interleukin-10, and transforming growth factor-β production Alcohol Clin Exp Res 1996 20 900 907 8865966
D'Souza NB Bagby GJ Nelson S Lang CH Spitzer JJ Acute alcohol infusion suppressess endotoxin-induced tumor necrosis factor production Alcohol Clin Exp Res 1989 13 295 298 2658671
Nelson S Bagby G Summer WR Alcohol suppresses lipopolysaccharide-induced tumor necrosis factor activity in serum and lung Life Sci 1989 44 673 676 2648095 10.1016/0024-3205(89)90472-4
Verma BK Fogarasi M Szabo G Down-regulation of tumor necrosis factor-alpha activity by acute ethanol treatment in human peripheral blood monocytes J Clin Immunol 1993 13 8 22 8445046 10.1007/BF00920631
Xie J Kolls J Bagby G Greenberg SS Independent suppression of nitric oxide and TNF alpha in the lung of conscious rats by ethanol FASEB J 1995 9 253 261 7540157
Kolls JK Xie J Lei D Greenberg S Summer WR Nelson S Differential effects of in vivo ethanol on LPS-induced TNF and nitric oxide production in the lung Am J Physiol Lung Cell Mol Physiol 1995 268 L991 L998
Stoltz DA Nelson S Kolls JK Zhang P Bohm RP Murphey-Corb M Bagby GJ Effects of in vitro ethanol on tumor necrosis factor-alpha production by blood obtained from simian immunodeficiency virus-infected rhesus macaques Alcohol Clin Exp Res 2002 26 527 534 11981130 10.1097/00000374-200204000-00014
Zhang Z Cork J Ye P Lei D Schwarzenberger PO Summer WR Shellito JE Nelson S Kolls JK Inhibition of TNF-alpha processing and TACE-mediated ectodomain shedding by ethanol J Leukoc Biol 2000 67 856 862 10857859
Black RA Rauch CT Kozlosky CJ Peschon JJ Slack JL Wolfson MF Castner BJ Stocking KL Reddy P Srinivasan S A metalloproteinase disintegrin that releases tumour-necrosis factor- alpha from cells Nature 1997 385 729 733 9034190 10.1038/385729a0
Moss ML Jin SL Milla ME Bickett DM Burkhart W Carter HL Chen WJ Clay WC Didsbury JR Hassler D Cloning of a disintegrin metalloproteinase that processes precursor tumour-necrosis factor-alpha [published erratum appears in Nature 1997 Apr 17;386(6626):738] Nature 1997 385 733 736 9034191 10.1038/385733a0
Zhao XJ Marrero L Song K Oliver P Chin SY Simon H Schurr JR Zhang Z Thoppil D Lee S Acute Alcohol Inhibits TNF-alpha Processing in Human Monocytes by Inhibiting TNF/TNF-alpha-Converting Enzyme Interactions in the Cell Membrane J Immunol 2003 170 2923 2931 12626543
Ruiz M Ewig S Marcos MA Martinez JA Arancibia F Mensa J Torres A Etiology of community-acquired pneumonia: impact of age, comorbidity, and severity Am J Respir Crit Care Med 1999 160 397 405 10430704
Khan KN Yatsuhashi H Effect of alcohol consumption on the progression of hepatitis C virus infection and risk of hepatocellular carcinoma in Japanese patients Alcohol Alcohol 2000 35 286 295 10869250
Harris DR Gonin R Alter HJ Wright EC Buskell ZJ Hollinger FB Seeff LB The relationship of acute transfusion-associated hepatitis to the development of cirrhosis in the presence of alcohol abuse Ann Intern Med 2001 134 120 124 11177315
Lima MP Pedro RJ Rocha MD Prevalence and risk factors for hepatitis C virus (HCV) infection among pregnant Brazilian women Int J Gynaecol Obstet 2000 70 319 326 10967165 10.1016/S0020-7292(00)00209-5
Moss M Bucher B Moore FA Moore EE Parsons PE The role of chronic alcohol abuse in the development of acute respiratory distress syndrome in adults JAMA 1996 275 50 54 8531287 10.1001/jama.275.1.50
Huber AR Kunkel SL Todd RF IIIWeiss SJ Regulation of transendothelial neutrophil migration by endogenous interleukin-8 Sci 1991 254 99 102
Strieter RM Kunkel SL Acute lung injury: the role of cytokines in the elicitation of neutrophils J Investig Med 1994 42 640 651 8521027
Szabo G Mandrekar P Catalano D Inhibition of superantigen-induced T cell proliferation and monocyte IL-1β, TNF-α, and IL-6 production by acute ethanol treatment J Leukocyte Biol 1995 58 342 350 7665990
Baseggio L Charlot C Bienvenu J Felman P Salles G Tumor necrosis factor-alpha mRNA stability in human peripheral blood cells after lipopolysaccharide stimulation Eur Cytokine Netw 2002 13 92 98 11956026
Mijatovic T Houzet L Defrance P Droogmans L Huez G Kruys V Tumor necrosis factor-alpha mRNA remains unstable and hypoadenylated upon stimulation of macrophages by lipopolysaccharides Eur J Biochem 2000 267 6004 6012 10998061 10.1046/j.1432-1327.2000.01676.x
Peters TJ Preedy VR Metabolic consequences of alcohol ingestion Novartis Found Symp 1998 216 19 24 9949785
Diaz-Rodriguez E Montero JC Esparis-Ogando A Yuste L Pandiella A Extracellular Signal-regulated Kinase Phosphorylates Tumor Necrosis Factor alpha-converting Enzyme at Threonine 735: A Potential Role in Regulated Shedding Mol Biol Cell 2002 13 2031 2044 12058067 10.1091/mbc.01-11-0561
Cottin V Doan JE Riches DW Restricted localization of the TNF receptor CD120a to lipid rafts: a novel role for the death domain J Immunol 2002 168 4095 4102 11937569
Vidalain PO Azocar O Servet-Delprat C Rabourdin-Combe C Gerlier D Manie S CD40 signaling in human dendritic cells is initiated within membrane rafts EMBO J 2000 19 3304 3313 10880443 10.1093/emboj/19.13.3304
Ikezu T Trapp BD Song KS Schlegel A Lisanti MP Okamoto T Caveolae, plasma membrane microdomains for alpha-secretase-mediated processing of the amyloid precursor protein J Biol Chem 1998 273 10485 10495 9553108 10.1074/jbc.273.17.10485
von Tresckow B Kallen KJ von Strandmann EP Borchmann P Lange H Engert A Hansen HP Depletion of cellular cholesterol and lipid rafts increases shedding of CD30 J Immunol 2004 172 4324 4331 15034047
Gutierrez-Ruiz MC Gomez JL Souza V Bucio L Chronic and acute ethanol treatment modifies fluidity and composition in plasma membranes of a human hepatic cell line (WRL-68) Cell Biol Toxicol 1995 11 69 78 7583873 10.1007/BF00767492
Paramahamsa M Aparna S Varadacharyulu N Alcohol-induced alterations in blood and erythrocyte membrane in diabetics Alcohol Alcohol 2002 37 49 51 11825857
Peschon JJ Slack JL Reddy P Stocking KL Sunnarborg SW Lee DC Russell WE Castner BJ Johnson RS Fitzner JN An essential role for ectodomain shedding in mammalian development [see comments] Sci 1998 282 1281 1284 10.1126/science.282.5392.1281
Althoff K Reddy P Voltz N Rose-John S Mullberg J Shedding of interleukin-6 receptor and tumor necrosis factor alpha. Contribution of the stalk sequence to the cleavage pattern of transmembrane proteins Eur J Biochem 2000 267 2624 2631 10785383 10.1046/j.1432-1327.2000.01278.x
Feehan C Darlak K Kahn J Walcheck B Spatola AF Kishimoto TK Shedding of the lymphocyte L-selectin adhesion molecule is inhibited by a hydroxamic acid-based protease inhibitor. Identification with an L- selectin-alkaline phosphatase reporter J Biol Chem 1996 271 7019 7024 8636132 10.1074/jbc.271.12.7019
|
16246259
|
PMC1291403
|
CC BY
|
2021-01-04 16:36:25
|
no
|
Respir Res. 2005 Oct 24; 6(1):123
|
utf-8
|
Respir Res
| 2,005 |
10.1186/1465-9921-6-123
|
oa_comm
|
==== Front
Respir ResRespiratory Research1465-99211465-993XBioMed Central London 1465-9921-6-1251625313610.1186/1465-9921-6-125ResearchInfluence of hypoxia on the domiciliation of Mesenchymal Stem Cells after infusion into rats: possibilities of targeting pulmonary artery remodeling via cells therapies? Rochefort Gaël Y [email protected] Pascal [email protected] Nicolas [email protected] Jean-Christophe [email protected] Jorge [email protected] Pierre [email protected] Véronique [email protected] LABPART-EA3852, IFR135, Université François Rabelais, faculté de Médecine, 10 boulevard Tonnellé 370032 TOURS France2 INSERM ESPRI-EA3588, IFR135, Université François Rabelais, faculté de Médecine, 10 boulevard Tonnellé 370032 TOURS France3 Virus, pseudo-virus: morphogenése et antigénicité, EA3856, Université François Rabelais, faculté de Médecine, 10 boulevard Tonnellé 370032 TOURS France4 Architecture du Tissu Osseux – Exercice Physique, EA 3895, Université d'Orléans- BP6749, 45067 Orléans cedex 2 France2005 27 10 2005 6 1 125 125 31 10 2004 27 10 2005 Copyright © 2005 Rochefort et al; licensee BioMed Central Ltd.2005Rochefort et al; licensee BioMed Central Ltd.This is an Open Access article distributed under the terms of the Creative Commons Attribution License (), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
Background
Bone marrow (BM) cells are promising tools for vascular therapies. Here, we focused on the possibility of targeting the hypoxia-induced pulmonary artery hypertension remodeling with systemic delivery of BM-derived mesenchymal stem cells (MSCs) into non-irradiated rats.
Methods
Six-week-old Wistar rats were exposed to 3-week chronic hypoxia leading to pulmonary artery wall remodeling. Domiciliation of adhesive BM-derived CD45- CD73+ CD90+ MSCs was first studied after a single intravenous infusion of Indium-111-labeled MSCs followed by whole body scintigraphies and autoradiographies of different harvested organs. In a second set of experiments, enhanced-GFP labeling allowed to observe distribution at later times using sequential infusions during the 3-week hypoxia exposure.
Results
A 30% pulmonary retention was observed by scintigraphies and no differences were observed in the global repartition between hypoxic and control groups. Intrapulmonary radioactivity repartition was homogenous in both groups, as shown by autoradiographies. BM-derived GFP-labeled MSCs were observed with a global repartition in liver, in spleen, in lung parenchyma and rarely in the adventitial layer of remodeled vessels. Furthermore this global repartition was not modified by hypoxia. Interestingly, these cells displayed in vivo bone marrow homing, proving a preservation of their viability and function. Bone marrow homing of GFP-labeled MSCs was increased in the hypoxic group.
Conclusion
Adhesive BM-derived CD45- CD73+ CD90+ MSCs are not integrated in the pulmonary arteries remodeled media after repeated intravenous infusions in contrast to previously described in systemic vascular remodeling or with endothelial progenitor cells infusions.
arterieshypertension, pulmonaryhypoxialungremodelingmesenchymal stem cells.
==== Body
Background
Recent studies emphasize on the perspective of cellular therapy by intravenous stem cells infusion. The participation of stem cells in several vascular diseases pathogenesis was first proved with haematopoietic stem cells (HSCs). In this regard, following bone marrow engraftment, HSCs were observed in remodeled vascular wall following graft vasculopathy or arteriosclerosis [1]. When integrated to the vascular wall, HSCs differentiate into mature vascular cells with an endothelial or smooth muscle cells phenotype.
Mesenchymal Stem cells (MSCs) are bone marrow non-haematopoietic stem cells that are multipotent and can differentiate into bone, cartilage and connective tissue cells [2-4]. They also differentiate in smooth muscle fibers and could be preferential candidates for vascular cells therapies [5]. Moreover MSCs present many advantages as facility to culture or to transform genetically [6]. Surprisingly few studies focused on the domiciliation of MSCs after in vivo infusion, even though they can be found into different organs after several months in normal animals, proving the in vivo infusion possibility without graft rejection [7]. Barbash et al recently showed a MSCs domiciliation into myocardial infarct area, however only a poor fraction of the cells engrafts the myocardium after systemic infusion [8].
Sustained pulmonary hypertension is a common complication of chronic hypoxic lung diseases. Hypoxic pulmonary hypertension is characterized by sustained pulmonary vasoconstriction and pulmonary vascular wall remodeling, including media and adventitia hypertrophy, without endothelial cells disruption. Furthermore chronic hypoxia has been shown to induce capillary angiogenesis [9]. Recently the participation of stem cells to hypoxia-induced adventitial remodeling has been observed in chronically hypoxic rat lungs [10]. Our hypothesis was that MSCs could domicile into the pulmonary artery remodeled wall and thus participate to hypoxia-induced structural changes.
We studied, for the first time, the bone marrow derived CD45- CD73+ CD90+ MSCs domiciliation after intravenous infusion in a model of chronically hypoxic rats, which induces pulmonary artery hypertension and vascular remodeling. Firstly, MSCs distribution was studied after a unique infusion of MSCs labeled by Indium-111 oxinate. Secondly, distribution was studied after sequential infusions of MSCs, transduced with the enhanced green fluorescent protein (GFP) gene by viral infection, during the three weeks of hypoxia exposure.
Methods
Animals
Six-weeks-old Wistar male rats (n = 26, Harlan) were exposed for 3 weeks to chronic hypoxia in a hypobaric chamber (50 kPa) to lead the development of pulmonary hypertension and were compared to control matched rats (n = 26).
The MSCs engraftment and viability control was performed using 4 hypoxic rats and compared to 4 control rats by a direct in-vivo injection of GFP-labeled MSCs into the right lung parenchyma and checked 3 weeks after normoxic or hypoxic condition housing as described below. The early dynamic distribution of infused radiolabeled MSCs was performed using 6 hypoxic rats and compared to 6 control rats. The long-term distribution of infused GFP-labeled MSCs was performed using 6 other hypoxic rats compared to 6 matched control rats. Finally, 5 hypoxic rats and 5 control rats were also sacrificed for DNA extraction and 5 hypoxic rats and 5 control rats were sacrificed for pulmonary enzymatic digestion and culture (see below).
All animal investigations were carried out in accordance with the Guide for the Care and Use of Laboratory Animals published by the US National Institute of Health (NIH Publications N°85-23, revised 1996) and European Directives (86/609/CEE).
Cell culture
Cell isolation and culture procedures for MSCs have been established and published previously [11,12]. Briefly, femurs were aseptically harvested from 6-weeks-old Wistar rats and the adherent soft tissue was removed. The proximal and distal ends of the femur were excised at a level just into the beginning of the marrow cavity. Whole marrow plugs were obtained by flushing the bone marrow cavity with a 18-gauge needle set with a syringe filled with culture medium composed of Modified Eagle Medium Alpha (α-MEM; Invitrogen) supplemented with 20% fetal calf serum (FCS; Hyclone), with antibiotic solution (penicillin/streptomycin: 1%; Invitrogen) and with antimycotic solution (amphotericin B: 0.01%; Bristol-Myers). The marrow plugs were dispersed to obtain a single cell suspension by sequentially passing the dispersion through 18- and 22-gauge needles. The cells were centrifuged and resuspended with culture medium. After counting in Malassez cells following an acetic acid disruption of red blood cells, nucleated cells were plated at a density of 106/cm2 and incubated at 37°C in a humidified atmosphere of 95% air 5% C02. The first medium change was after 2 days and twice a week thereafter. When these primary MSCs reached 80–90% of confluence, they were trypsinized (trypsin-EDTA, Invitrogen), counted and passaged at a density of 104/cm2. For the first study second-passage MSCs were labeled with 111In-oxine as described below and infused intravenously. For the second study MSCs were GFF-labeled after viral gene transduction after the first passage and were used as the second-passage.
Adherent second-passage MSCs were analyzed by flow cytometry with a FACSCalibur flow cytometer (Becton-Dickinson) using a 488 nm argon laser. Cells were incubated for 60 minutes at 4°C with phycoerythrin- or fluorescein isothiocyanate-conjugated monoclonal antibodies against rat CD45 (clone OX-1), rat CD73 (clone 5F/B9), and rat CD90 (Clone OX-7; all from Becton Dickinson). Isotype-identical antibodies served as controls. Samples were analyzed by collecting 10,000 events on a FACSCalibur instrument using Cell-Quest® software (Becton-Dickinson).
Isotopic labeling and Indium-111 labeled MSCs intravenous infusion
The cells were incubated with 111In-oxine (37 MBq/106 cells) and incubated for 60 minutes as previously described [11]. The radiolabeled MSCs were aliquoted at 107 cells/ml and intravenously infused to hypoxic rats within 1 hour and followed by whole body scintigraphic imaging. Preliminary experiments showed that the viability and growth of these labeled MSCs were not adversely affected by this labeling procedure (data not shown); the level of radioisotope was widely sufficient to produce high quality images taken with a gamma camera and to produce high quality autoradiographic images of organs.
Whole body scintigraphic imaging was performed immediately after infusion and within 15 minutes, 30 minutes, 1 hour, 3 hours, 24 hours and 96 hours thereafter. Planar whole body images were acquired with Helix Elscint scanner (GE Healthcare) using a medium energy collimator. Images were acquired on a 256 × 256 matrix using a window centered at 245 keV. The distance between the chest of animals and the detector was fixed at 65 mm. In analysis of the scintigraphic images, regions of interest (ROIs) were placed over lungs, liver and spleen on anterior incidence, and over kidneys on posterior incidence. The whole body count was determined by the mean counts on both incidences. Total counts in the ROIs were corrected with physical decay of 111In and with body count.
After sacrifice lung, liver, heart, spleen, kidneys and bone marrow were harvested. Organs were weighted and assayed for radioactivity using a Muller counter (Ludlum Measurements), after what they were snap-frozen in liquid nitrogen, whereas cytospins of bone marrow were realized. Sample sections (15 μm) and bone marrow cytospins were exposed to a photographic film within 24–96 hours and autoradiographic films were developed.
GFP labeling, in vivo engraftment and viability controls, and GFP-labeled MSCs intravenous infusions
GFP labeling
MSCs were labeled by green fluorescent protein (GFP) after stable viral gene transduction with LNCX-GFP vector. GFP fluorescence from first-passage transduced MSCs was checked by flow cytometry. Non-specific fluorescence was determined using MSCs that were not transduced. GFP-labeling stability was assayed by flow cytometry using tenth-passage GFP-labeled MSCs.
In-vivo engraftment and viability controls
Animals were lightly anesthetized and GFP-labeled MSCs were injected, at a dose of 2.106 cells, through the rib cage, into the right lung lower lobe. After recovering, animals were housed 3 weeks either in normoxic condition, or hypoxic condition. Animals were sacrificed after the 3 weeks and the lung was harvested, snap-frozen in liquid nitrogen. The frozen sample sections (15 μm) were analyzed by tree-dimensional confocal laser microscopy.
GFP-labeled MSCs intravenous infusions
Second-passage GFP-labeled MSCs were sequentially infused intravenously at the dose of 106 MSCs. The first infusion indicated the first day of the 3 weeks chronic hypoxia. Both hypoxic and control rats were infused twice a week during 3 weeks.
After sacrifice lung, liver, heart, spleen, kidneys and bone marrow were harvested. Organs were weighed and snap-frozen in liquid nitrogen. The frozen sample sections (15 μm) of the different organs were analyzed by tree-dimensional confocal laser microscopy. Data was collected with sequential laser excitation to eliminate bleed through and acquired on a 1024 × 1024 matrix using a 110 μm pinhole and an optical section thickness of 0.31 μm. The system was made up of a FV500 confocal microscope (Olympus) using FluoView500 software and a 488 nm argon laser.
The GFP protein was also researched on frozen sections by immunohistochemistry. Sections of harvested organs were incubated with a rabbit polyclonal antibody against GFP (1/200, Santa Cruz Biotechnology) and were revealed either by a conjugated goat anti-rabbit alexa-594 (1/400, Molecular Probes) or by a conjugated goat anti-rabbit horseradish peroxydase (1/400, Biosource).
Bone marrow homing detection
Cytospins of bone marrow aspirates from control and hypoxic rats were realized 3 days after a unique GFP-labeled MSCs infusion and 3 days after the end of GFP-labeled MSCs infusion during the 3-week hypoxia exposure. The percentage of fluorescent cells was estimated for each rat in five random fields by microscopy using Optimas software (Imasys). Thin slices (12 μm) of frozen bone sections were cut in the metaphysis of tibia from five injected rats. Fluorescence (GFP) was directly observed by confocal microscopy and adipocytes were detected after counterstaining with DAPI (4,6-diamidino-2-phenylindole, AbCys) [13].
Detection of GFP transgene and protein by PCR and western blotting
After sequential infusions, organs were harvested. From each animal, GFP transgene and protein were assayed by PCR and Western blotting.
PCR
Total DNA was extracted using QIAamp DNA Mini Kit (Qiagen, Hilden, Germany) according to the manufacturer's instructions. It was analyzed by PCR for GFP transgene presence using a set of primer generating a 249 bp amplicon: forward, GCGACGTAAACGGCCACAAGTTC and reverse, CGTCCTTGAAGAAGATGGTGCGC. DNA was subjected to PCR for 35 cycles of 94°C for 30 seconds, 58°C for 60 seconds, 72°C for 30 seconds, with a final elongation step of 10 minutes at 72°C.
Western blotting
Organs were crushed by Turrax and homogenized with lysis buffer [1% sodium deoxycholate, 0.1% SDS, 1% triton X-100, 10 mM Tris-HCl (pH 8.0), 150 mM NaCl and an inhibitor protease cocktail (chymotrypsin-, thermolysin-, papain-, pronase-, pancreatic extract- and trypsin-inhibitor; Roche)] and centrifuged at 20,000 g for 1 h. After purifying and concentrating small proteins from each sample (Centriprep Centrigugal Devices YM-30MW, Millipore) with a nominal molecular weight limit of 30 kDa, proteins were separated on a SDS/12% polyacrylamide gel and then transferred to a nitrocellulose membrane (Amersham). Blots were blocks for 2 h at room temperature with 5% (w/v) non-ft dried milk in Tris-buffered saline [10 mM Tris-HCl (pH 8.0) and 150 mM NaCl] containing 0.05% Tween 20. The membrane was incubated overnight at 4°C with rabbit polyclonal antibody against GFP (1/400, Santa Cruz Biotechnology). The blot was then incubated with the conjugated goat anti-rabbit horseradish peroxydase (1/1000, Biosource) 2 h at room temperature. Immunoreactive proteins were detected with the ECL Western blotting detection system (Amersham).
Pulmonary enzymatic digestion
Lung from 5 non-hypoxic and 5 hypoxic MSCs-injected rats were cultured after enzymatic digestion. Briefly, rat lungs were harvested, mechanically dissected and the thin pieces were digested with collagenase (0.5 mg/ml, 1 hour at 37°C, Sigma). After wash, the suspension was passed through a cell strainer to remove undigested block and wash in PBS with FCS (20%, Hyclone). Then, the suspension was incubated in trypsin (30 minutes at 37°C, Invitrogen), wash twice in PBS-FCS, counted, plated and incubated at 37°C in a humidified atmosphere of 95% air 5% C02. The first medium change was after 2 days and twice a week thereafter. The GFP fluorescence was checked after 1 and 2 weeks.
Statistical analysis
Data are presented as mean +/-SEM with statistical significance tested using the two tailed paired t-test or the Mann-Whitney test.
Results
Hypoxia-induced pulmonary arteries remodeling and pulmonary hypertension
The hypoxia-induced pulmonary artery hypertension was checked by echocardiography (data not shown). This is pulmonary artery remodeling model already validated and previously reported by our team [14].
Mesenchymal stem cells
Cultured bone marrow-derived cells had a typical fibroblast-like morphology and were evenly distributed on the plate after 2 days (fig. 1A). Cells attachment was observed at about 3–4 h and 80–90% of confluence was typically reached by day 6–7. The average cell viability, determined by exclusion of trypan blue, was approximately 90%. CD73 and Thy-1/CD90 were expressed in these MSCs whereas the haematopoietic lineage marker CD45 was not (fig. 1B). These growth patterns and surface markers expression were similar to those of normal rat bone marrow-derived MSCs previously described [12]. Retroviral infection of MSCs had not modified their morphology or viability. The GFP-labeling efficiency was about 98% and the labeling stability was assayed until tenth passage (data not shown).
Figure 1 Mesenchymal stem cells used during this study. Typical morphological aspects of mesenchymal stem cells observed through culture flask (A). Mesenchymal stem cells expression of CD73 and CD90 antigens was attested by flow cytometry (B).
Dynamic distribution of radiolabeled-MSCs after a single infusion
The distribution of radioactivity after infusion of the radiolabeled-MSCs was imaged from the end of infusion up to 96 h after. This imaging provides an immediate indication of the initial cells distribution. Since radiolabeled-MSCs intravenous infusion, the radioactivity was first observed to accumulate into the lungs, and gradually, the radioactivity was observed in the liver. At 3 h after cell infusion, the radioactivity was observed in the spleen. Kidneys and bone were widely observed at 24 h (fig. 2).
Figure 2 Early dynamic distribution of mesenchymal stem cells in vivo. Sequential whole body scintigraphies after infusion of indium-111 labeled mesenchymal stem cells were acquired from injection up to 96 h. After pulmonary retention, a liver and spleen repartition was observed. A lung domiciliation was indicated by lungs radioactivity stabilization. Bone radioactivity was linked with bone marrow homing after 24 hours.
In order to quantify the distribution of 111In, the specific radioactivity of each organ was calculated as a percentage of the total body counts related to the organs region of interest (ROIs) counts. The pulmonary radioactivity was about 50–60% (fig. 3) in both hypoxic and control rats from infusion and at 1 h. This pulmonary radioactivity decreased afterwards and stabilized by about 30% in both groups at 3 h after infusion. No significant difference in lungs ROIs counts was observed between hypoxic rats and control rats (tab. 1).
Figure 3 Pulmonary radioactivity. Pulmonary repartition was measured in vivo from lung region of interest counts on scintigraphies at different times after radiolabeled mesenchymal stem cells infusion. Counts were normalized by whole body counts. After 24 hours, radioactivity was stabilized without differences between control and hypoxic groups.
Table 1 Harvested organs radioactivity. The radioactivity repartition in different organs, measured ex vivo after animals sacrifice 96 h after radiolabeled mesenchymal stem cells infusion, was normalized by organ weight and by infused activity. The results were corrected by time decay and are presented as mean +/-SEM.
Control group rats Hypoxic group rats
Lungs 17.22 % ± 6.92 25.26 % ± 2.78
Liver 41.28 % ± 19,62 29.39 % ± 12.42
Spleen 20.23 % ± 13.59 9.07 % ± 2.83
Kidneys 21.16 % ± 13.01 14.75 % ± 6.93
To observe the distribution of the infused-cells in the lungs, autoradiography of lungs sections were performed (fig. 4A). These films showed homogenous distribution of the radioactivity in both groups. Furthermore, radioactivity was not observed in the lumen of large diameter pulmonary arteries, proving that the infused cells were not agglomerated into the pulmonary vessels lumen.
Figure 4 Autoradiographies. Autoradiographies of organs frozen sections were realized after animals sacrifice, by 96 h after radiolabeled mesenchymal stem cells infusion. Lung images showed a homogenous repartition and the absence of radioactivity into main arteries that appeared in negative (A, arrows). Lonely signals on bone marrow cytospins confirmed the mesenchymal stem cells homing and excluded free indium bone uptake (B). In all cases no differences in repartition between control and hypoxic groups were observed (see tab. 1).
Bone marrow from radiolabeled-MSCs infused-rats was also harvested and exposed to autoradiographic film. We therefore showed that infused-MSCs homed in bone marrow at 96 h after infusion in both groups (fig. 4B).
In-vivo engraftment and viability controls
In order to have positive controls of GFP signals for confocal images interpretation, we first directly injected GFP-labeled cells into a freshly harvested lung (fig. 5A) and compared to non-injected freshly harvested lung (fig. 5B).
Figure 5 In-vivo engraftment and viability controls. GFP signals were researched by confocal microscopy on lungs frozen sections. In a first step, GFP-labeled MSCs were directly injected in ex-vivo excised lungs in order to provide positive control (A, arrow) for confocal images interpretation whereas a non-injected freshly harvested lung served as negative control (B). Then, MSCs were directly injected in the right lower lobe of the lung in vivo and rats placed in normoxic or hypoxic conditions for three weeks. Frozen sections of lungs were observed after three weeks in confocal microscopy to provide in vivo positive engraftment and viability controls. Indeed, the injection site was visualized macroscopically (C, arrows) and GFP signals were seen centered on the injection injury (D, arrows). Bar = 50 μm, a indicates artery.
To check the in-vivo engraftment and viability of the MSCs into lungs, we have directly injected GFP-labeled MSCs into the right lower lobe of the lung and housed animals either in normoxic or hypoxic conditions during 3 weeks. The tolerance of these injections was good and no animals died or showed rejection. From confocal microscopy observation centered on the injection injury (fig. 5C), we observed GFP signals proving the lung engraftment capacity and the viability of the MSCs after 3 weeks (fig. 5D). No difference in the appearance of MSCs was observed between hypoxic and non-hypoxic rats.
Distribution of GFP-labeled MSCs after sequential infusions
After sequential infusions during the 3-week hypoxia exposure, we examined the harvested lungs sections from control and hypoxic rats. Only few GFP-labeled MSCs were observed per lung sections in both control and hypoxic rats. Moreover when observed, the GFP-labeled MSCs were localized in the lung parenchyma and rarely close to the vascular lumen in both control (fig. 6A) and hypoxic (fig. 6B, 6C, 6D) rats. To localize these cells, we then performed the GFP detection in lungs using immunohistochemistry and peroxydase revelations (data not shown). No signal linked to MSCs localization was observed into the media of pulmonary arteries. Rarely, GFP-labeled MSCs were observed close to the adventitial layer of remodeled vessels. So we confirm the absence of GFP-labeled cells into the remodeled pulmonary arteries.
Figure 6 Mesenchymal stem cell localization in lungs. GFP-labeled MSCs (arrows) were localized essentially into the pulmonary parenchyma without difference between the non-hypoxic (A) and the hypoxic group (B, C, D). Bar = 50 μm, a indicates artery, b indicates bronchiole.
GFP cells were also and better observed on liver (fig. 7A) and spleen sections (fig. 7B) with the same aspect. No difference in the repartition of GFP-labeled cells was observed in these organs between normoxic and hypoxic groups confirming the absence of pulmonary domiciliation enhanced by hypoxia.
Figure 7 Mesenchymal stem cell localization in liver and spleen. GFP signals were observed in liver (A) and spleen (B) from frozen sections after GFP-labeled mesenchymal stem cells infusions observed in confocal microscopy. Hypoxia did not modify their repartition. Arrows refer to GFP signals. Bar = 50 μm.
The GFP transgene was found in lungs by PCR (fig. 8A) and the GFP protein was recovered in lungs by western blotting (fig. 8B) confirming the presence of GFP-cells into the lungs.
Figure 8 GFP transgene and protein detection. After sequential infusions, lungs were harvested. PCR (A) and western blot (B) confirmed presence of GFP transgene and GFP protein in harvested lungs 96 hours after the last infusion in both groups.
To extract and culture the engrafted GFP-labeled cells from lungs following the same protocol of three-week cell injection and hypoxic exposure, we enzymatically digested lung from 5 control and 5 hypoxic injected rats and cultured. However this experiment failed to obtain cultured GFP-labeled cells suggesting that only few numbers of GFP-labeled cells localized into the lung both in normoxic and hypoxic group.
Bone marrow homing and engraftment
The fluorescent cell ratio was evaluated on bone marrow cytospins by averaging the results of five views fields for each slide (tab. 2). Compared to a single infusion, we observed an increase of fluorescent cell ratios with sequential infusions (tab. 2) while hypoxia appeared to enhance bone marrow homing. Moreover, on slices of rat tibial bone after GFP-labeled MSCs infusion, we observed fluorescent cells localized between adipocytes (fig. 9A) in contrast to non-infused control rats (fig. 9C). Surprisingly, their appearance in some part looked like the surrounding adipocytes counterstained by DAPI (same size and shape) (fig. 9B). We therefore concluded that MSCs are able to home into bone with preserved viability.
Table 2 Bone marrow homing. The evaluation of bone marrow homing, after unique or sequential infusions by evaluation of fluorescent cells ratio on bone marrow cytospins, is presented as mean +/-SEM with statistical significance tested using the Mann-Whitney test.
Control group rats Hypoxic group rats
Unique infusion 2.57 ± 1.48 % 2.76 ± 1.53 % NS
Sequential infusions 13.01 ± 6.41 % 17.27 ± 5.69 % p < 0.05
p < 0.02 p < 0.01
Figure 9 Bone homing. After sacrifice, tibia was harvested and slices were cut out from metaphysis and observed by confocal microscopy (A, B). GFP-labeled cells were observed in the same area than adipocytes counterstained by DAPI (white arrows), whereas no green fluorescent was observed in the area of bone marrow (red arrow). A tibia harvested from a non-injected rat was used as control (C).
Discussion
Mesenchymal stem cells
Bone marrow comprises both haematopoietic and non-haematopoietic cells among these last mesenchymal stem cells can be found. MSCs in culture can be characterized by their adhesivity, fusiform shape and presence of specific membrane surface antigens. In culture after two passages we showed that more than 90% of collected cells were MSCs. Transduction by GFP did not alter these properties. We did not study the effect of gamma irradiation on the MSCs phenotype. One of the results of our study is that no engraftment intolerance was observed. In accordance with previous studies our results demonstrated that infused MSCs could be found several weeks after infusion. In a precedent study, MSCs were isolated from the receptor organs after in vivo infusion and cultured successfully, confirming their viability after domiciliation [15]. Moreover these authors concluded that MSCs could by themselves immuno-privileged. In our study, MSCs morphology and fluorescent labeling were also kept intact and no inflammatory reactions were observed in the surrounding tissue. We then concluded in the absence of graft rejection.
In our study, despite the fact that rats have not been irradiated, we also observed bone marrow homing of MSCs, as previously described for haematopoietic cells after intravenous infusion in immuno-competent animals [16]. This homing could be significantly increased after irradiation [17,18]. In the present study, we observed bone marrow homing of MSCs that was increased by sequential infusions. In bone, some GFP-labeled cells even displayed an adipogenic phenotype, proving in vivo their viability. Chondrogenic differentiation of MSCs has already been observed in vivo in bone after intravenous infusion in neonatal mice [15]. Nevertheless further studies are required to confirm adipogenic differentiation. Finally, we showed in our hypoxic rat model that 3 weeks after the first intravenous infusion, MSCs remain detectable, viable and functional.
Pulmonary domiciliation
In our model, adhesive bone marrow derived CD45- CD73+ CD90+ MSCs were localized into the pulmonary parenchyma. After a first phase of pulmonary arteries retention, some MSCs reached the systemic circulation and were distributed mainly in the spleen, and the liver. These cells are essentially observed into the parenchyma of these organs and their presence was confirmed by the detection of the GFP protein in Western Blotting and by detection of the transgene in PCR analysis from lung samples.
This observed global cells distribution is in agreement with previous study [11]. These results leaded some authors to conclude that the pulmonary retention was not specific and without any precise localization neither in the parenchyma nor in the vasculature and to hypothesize that stem cells infusion induces only passive embolism or endothelium adhesion. In our study, we also failed to culture GFP-labeled cells from injected rat lungs suggesting that only few adhesive bone marrow-derived CD45- CD73+ CD90+ MSCs were localized into the pulmonary parenchyma.
Since our autoradiography results showed clearly the absence of radioactivity in the lumen of large diameter pulmonary artery vasculature, we could exclude the idea of a simple intravascular retention of MSCs after infusion. However, we could not exclude MSC localization into the small pulmonary artery walls. Unfortunately, our isotopic labeling did not allow us to analyze the signal observed in peripheral small vessels. Thus, we performed GFP labeling and immunohistochemistry with peroxydase, which confirmed the absence of GFP-labeled cells into the media of small pulmonary artery, as well as into their lumen.
Effects of hypoxia
Another conclusion of our study is that exposure to chronic hypoxia did not modify the global repartition of MSCs in vivo. We demonstrated that after chronic hypoxia, cells were essentially observed into the lung parenchyma. The repartition of MSCs in spleen, liver and bone was unchanged after hypoxic exposure, which argues for a non-specific pulmonary domiciliation.
In hypoxic model, remodeling occurs in the media but also in the adventitial. In a recent study, Hayashida et al demonstrated after bone marrow transplantation, that donor's stem cells were present in the remodeled adventitia [10]. However in their studies MSCs do not constitute the major component of the stem cells that can engraft in the bone marrow. So, it is unlikely that stem cells observed by Hayashida et al could be mesenchymal stem cells. Moreover, Davie et al showed localizations of infused endothelial progenitor in adventitial vaso-vasorum vessels [19]. In another study using the same chronic hypoxic rat model, it has been shown that endothelial progenitor cells could be observed after intravenous infusion in the pulmonary arterioles wall [20]. In this latter study, GFP signal was observed in the parietal wall. However the authors did not focus on what it was observed in control. In our study, fluorescence was observed on the media layer but without difference between hypoxic and control groups and we concluded that artifacts are linked to auto-fluorescence.
Using monocrotaline model of pulmonary hypertension, Zhao et al [21] founded local endothelial progenitor cell domiciliation after infusion. However, monocrotaline induces disruption of endothelial layer, which could have a positive impact on this domiciliation contrarily to our hypoxic model without endothelial damaged. We can speculate that in our model, endothelium constitutes a barrier stopping the MSCs incorporation. Our study cannot exclude any participation of mesenchymal stem cells to adventitial remodeling but, in our particular experimental conditions, we did not observed any significant and specific recruitment into pulmonary arterial vasculature after hypoxia exposure. This fact is a major limit to MSCs therapies with intravenous injection.
We showed that MSC bone marrow homing is increased by hypoxia. This interesting finding needs to be confirmed by further studies before speculating a specific effect of hypoxia on MSCs homing, mobilization and migration.
Conclusion
The major conclusion of our study is that, in our model of hypoxic pulmonary artery remodeling without endothelial disruption, the adhesive bone marrow-derived CD45- CD73+ CD90+ MSCs are not significantly integrated in the parietal wall after repeated intravenous infusion whereas it has been described in systemic vascular remodeling such graft vasculopathy and arteriosclerosis with endothelial progenitor cells.
List of Abbreviations used
α-MEM modified eagle medium alpha
DAPI 4,6-diamidino-2-phenylindole
DNA deoxyribonucleic acid
ECL enhanced chemiluminescence
FCS fetal calf serum
GFP green fluorescent protein
HSC haematopoietic stem cells
MSC mesenchymal stem cells
PCR polymerase chain reaction
ROI regions of interest
SDF-1 stromal cell derived factor-1
SDS sodium dodecyl sulfate
SEM standard error of mean
Authors' contributions
GYR conducted the majority of the research experiments. PV and JCP helped with the GFP detection. NB carried out the hypoxic model and helped with GFP detection in bone. JD and PC provided the cell culture equipment. VE conceived the experimental study, participated in its design and coordination and conducted the isotopic experiments. GYR and VE participated in writing and preparation of the manuscript. All authors read and approved the final manuscript.
Acknowledgements
Confocal microscopy analysis was done at the PPF Analyse des systèmes biologiques, Université François Rabelais, faculté de Médecine, Tours, France. This work was in part supported by grants from the Conseil Général d'Indre-et-Loire, France, and from IFR135.
==== Refs
Sata M Saiura A Kunisato A Tojo A Okada S Tokuhisa T Hirai H Makuuchi M Hirata Y Nagai R Hematopoietic stem cells differentiate into vascular cells that participate in the pathogenesis of atherosclerosis Nat Med 2002 8 403 409 11927948 10.1038/nm0402-403
Galmiche MC Koteliansky VE Briere J Herve P Charbord P Stromal cells from human long-term marrow cultures are mesenchymal cells that differentiate following a vascular smooth muscle differentiation pathway Blood 1993 82 66 76 8324235
Pereira RF Halford KW O'Hara MD Leeper DB Sokolov BP Pollard MD Bagasra O Prockop DJ Cultured adherent cells from marrow can serve as long-lasting precursor cells for bone, cartilage, and lung in irradiated mice Proc Natl Acad Sci U S A 1995 92 4857 4861 7761413
Prockop DJ Marrow stromal cells as stem cells for nonhematopoietic tissues Science 1997 276 71 74 9082988 10.1126/science.276.5309.71
Ferrari G Cusella-De Angelis G Coletta M Paolucci E Stornaiuolo A Cossu G Mavilio F Muscle regeneration by bone marrow-derived myogenic progenitors Science 1998 279 1528 1530 9488650 10.1126/science.279.5356.1528
Horwitz EM Prockop DJ Fitzpatrick LA Koo WW Gordon PL Neel M Sussman M Orchard P Marx JC Pyeritz RE Brenner MK Transplantability and therapeutic effects of bone marrow-derived mesenchymal cells in children with osteogenesis imperfecta Nat Med 1999 5 309 313 10086387 10.1038/6529
Jorgensen C Djouad F Apparailly F Noel D Engineering mesenchymal stem cells for immunotherapy Gene Ther 2003 10 928 931 12732877 10.1038/sj.gt.3302019
Barbash IM Chouraqui P Baron J Feinberg MS Etzion S Tessone A Miller L Guetta E Zipori D Kedes LH Kloner RA Leor J Systemic delivery of bone marrow-derived mesenchymal stem cells to the infarcted myocardium: feasibility, cell migration, and body distribution Circulation 2003 108 863 868 12900340 10.1161/01.CIR.0000084828.50310.6A
Hyvelin JM Howell K Nichol A Costello CM Preston RJ McLoughlin P Inhibition of Rho-kinase attenuates hypoxia-induced angiogenesis in the pulmonary circulation Circ Res 2005 97 185 191 15961717 10.1161/01.RES.0000174287.17953.83
Hayashida K Fujita J Miyake Y Kawada H Ando K Ogawa S Fukuda K Bone marrow-derived cells contribute to pulmonary vascular remodeling in hypoxia-induced pulmonary hypertension Chest 2005 127 1793 1798 15888860 10.1378/chest.127.5.1793
Gao J Dennis JE Muzic RF Lundberg M Caplan AI The dynamic in vivo distribution of bone marrow-derived mesenchymal stem cells after infusion Cells Tissues Organs 2001 169 12 20 11340257 10.1159/000047856
Lennon DP Haynesworth SE Young RG Dennis JE Caplan AI A chemically defined medium supports in vitro proliferation and maintains the osteochondral potential of rat marrow-derived mesenchymal stem cells Exp Cell Res 1995 219 211 222 7628536 10.1006/excr.1995.1221
Caron M Auclair M Sterlingot H Kornprobst M Capeau J Some HIV protease inhibitors alter lamin A/C maturation and stability, SREBP-1 nuclear localization and adipocyte differentiation Aids 2003 17 2437 2444 14600514 10.1097/00002030-200311210-00005
Bonnet P Bonnet S Boissiere J Le Net JL Gautier M Dumas de la Roque E Eder V Chronic hypoxia induces nonreversible right ventricle dysfunction and dysplasia in rats Am J Physiol Heart Circ Physiol 2004 287 H1023 8 15317673 10.1152/ajpheart.00802.2003
Niyibizi C Wang S Mi Z Robbins PD The fate of mesenchymal stem cells transplanted into immunocompetent neonatal mice: implications for skeletal gene therapy via stem cells Mol Ther 2004 9 955 963 15194062 10.1016/j.ymthe.2004.02.022
Krause DS Theise ND Collector MI Henegariu O Hwang S Gardner R Neutzel S Sharkis SJ Multi-organ, multi-lineage engraftment by a single bone marrow-derived stem cell Cell 2001 105 369 377 11348593 10.1016/S0092-8674(01)00328-2
Jiang Y Jahagirdar BN Reinhardt RL Schwartz RE Keene CD Ortiz-Gonzalez XR Reyes M Lenvik T Lund T Blackstad M Du J Aldrich S Lisberg A Low WC Largaespada DA Verfaillie CM Pluripotency of mesenchymal stem cells derived from adult marrow Nature 2002 418 41 49 12077603 10.1038/nature00870
Rombouts WJ Ploemacher RE Primary murine MSC show highly efficient homing to the bone marrow but lose homing ability following culture Leukemia 2003 17 160 170 12529674 10.1038/sj.leu.2402763
Davie NJ Crossno JTJ Frid MG Hofmeister SE Reeves JT Hyde DM Carpenter TC Brunetti JA McNiece IK Stenmark KR Hypoxia-induced pulmonary artery adventitial remodeling and neovascularization: contribution of progenitor cells Am J Physiol Lung Cell Mol Physiol 2004 286 L668 78 12754186 10.1152/ajplung.00108.2003
Nagaya N Kangawa K Kanda M Uematsu M Horio T Fukuyama N Hino J Harada-Shiba M Okumura H Tabata Y Mochizuki N Chiba Y Nishioka K Miyatake K Asahara T Hara H Mori H Hybrid cell-gene therapy for pulmonary hypertension based on phagocytosing action of endothelial progenitor cells Circulation 2003 108 889 895 12835224 10.1161/01.CIR.0000079161.56080.22
Zhao YD Courtman DW Deng Y Kugathasan L Zhang Q Stewart DJ Rescue of monocrotaline-induced pulmonary arterial hypertension using bone marrow-derived endothelial-like progenitor cells: efficacy of combined cell and eNOS gene therapy in established disease Circ Res 2005 96 442 450 15692087 10.1161/01.RES.0000157672.70560.7b
|
16253136
|
PMC1291404
|
CC BY
|
2021-01-04 16:36:25
|
no
|
Respir Res. 2005 Oct 27; 6(1):125
|
utf-8
|
Respir Res
| 2,005 |
10.1186/1465-9921-6-125
|
oa_comm
|
==== Front
Respir ResRespiratory Research1465-99211465-993XBioMed Central London 1465-9921-6-1271625578410.1186/1465-9921-6-127ResearchPulmonary function and fuel use: A population survey Saha Asim [email protected] Rao N [email protected] PK [email protected] PK [email protected] HN [email protected] Occupational medicine Division, National Institute of Occupational Health, Ahmedabad – 380 016, India2 Respiratory Physiology Division, National Institute of Occupational Health, Ahmedabad – 380 016, India3 Biostatistics Division, National Institute of Occupational Health, Ahmedabad – 380 016, India4 Occupational medicine Division, National Institute of Occupational Health, Ahmedabad – 380 016, India5 Director, National Institute of Occupational Health, Ahmedabad – 380 016, India2005 31 10 2005 6 1 127 127 6 9 2005 31 10 2005 Copyright © 2005 Saha et al; licensee BioMed Central Ltd.2005Saha et al; licensee BioMed Central Ltd.This is an Open Access article distributed under the terms of the Creative Commons Attribution License (), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
Background
In the backdrop of conflicting reports (some studies reported adverse outcomes of biomass fuel use whereas few studies reported absence of any association between adverse health effect and fuel use, may be due to presence of large number of confounding variables) on the respiratory health effects of biomass fuel use, this cross sectional survey was undertaken to understand the role of fuel use on pulmonary function.
Method
This study was conducted in a village of western India involving 369 randomly selected adult subjects (165 male and 204 female). All the subjects were interviewed and were subjected to pulmonary function test. Analysis of covariance was performed to compare the levels of different pulmonary function test parameters in relation to different fuel use taking care of the role of possible confounding factors.
Results
This study showed that biomass fuel use (especially wood) is an important factor for deterioration of pulmonary function (particularly in female). FEV1 (p < .05), FEV1 % (p < .01), PEFR (p < .05) and FEF25–75 (p < .01) values were significantly lower in biomass fuel using females than nonusers. Comparison of only biomass fuel use vs. only LPG (Liquefied Petroleum Gas) use and only wood vs. only LPG use has showed that LPG is a safer fuel so far as deterioration of pulmonary function is concerned. This study observes some deterioration of pulmonary function in the male subjects also, who came from biomass fuel using families.
Conclusion
This study concluded that traditional biomass fuels like wood have adverse effects on pulmonary function.
Biomass fuelsLiquefied Petroleum GasPulmonary function
==== Body
Background
Exposure to indoor air pollution from the combustion of traditional biomass fuels (wood, charcoal, animal dung, and crop wastes) and coal is a significant public health hazard predominantly affecting poor rural and urban communities in developing countries. Large numbers of people are exposed on a daily basis to harmful emissions and other health risks from biomass and coal burning. It is estimated that globally 2.5 to 3 billion people rely on these fuels for everyday household energy needs [1]. The majority of those exposed are women, who are normally responsible for food preparation and cooking, and infants/young children who are usually with their mothers near the cooking area. Although the fraction of global energy from biofuels has fallen from 50 percent in 1900 to around 13 percent currently, this trend has leveled off and there is evidence that biofuel use is increasing among the poor in some parts of the world [1,2]. There is consistent evidence that exposure to biomass smoke increases the risk of a range of common and serious diseases of both children and adults. Chief amongst these are acute lower respiratory infections (ALRI) in childhood, particularly pneumonia [3,4]. Association of exposure with chronic bronchitis and chronic obstructive pulmonary disease is also quite well established, particularly among women [5,6]. In addition there is evidence (mainly from China), that exposure to coal smoke in the home markedly increases the risk of lung cancer, particularly in women [7,8].
So far as deterioration of lung function is concerned there is growing evidence is that biomass fuel use is harmful. Studies on biomass fuel user women [9] (urban [10] and rural [11], both) have shown that they had significantly lower pulmonary function values in comparison to non-biomass fuel users. Adverse effects of biomass fuel use have been observed in female asthmatics [12] as well as in children [13] also. At the same time there are studies where no significant difference of pulmonary function values have been observed when compared between biomass and modern fuel users [14]. Confounding effect of different other factors have been stated to be responsible for this kind of conflicting findings and the need of more such studies including intervention studies have been stressed in order to gather stronger scientific evidence [14,15]. In this backdrop this study was initiated to understand the effect of fuel use on pulmonary function.
Material and methods
This cross sectional survey was conducted in a small village of western India situated 20 kilometers away from the nearest city. This small village had total adult population of 1509. Prevalence of pulmonary function abnormality in India being 10% in adult population, we calculated the sample size for prevalence study keeping acceptable range as 6.5–13.5%. Thus, the minimum sample size for 1% level of significance was calculated as 368. We set our target as 400 persons. Selection of subjects was done from the electoral list (voters' list) of that village by using random numbers generated from Microsoft Excel software. This list being the most frequently updated and most complete list of its kind available in India was thought to be most suitable for this purpose. Among the 400 people, who were approached for study, 369 subjects participated in this study. Necessary ethical clearance was obtained from the institutional ethics committee of National Institute of Occupational Health; India prior to the initiation of this study and informed consent was taken from the concerned study subjects during the study. These subjects were interviewed with a questionnaire to obtain information about their personal characteristics including fuel use. All of them were subjected to pulmonary function test using Spirovit-sp-10 (Schiller Health Care Ltd, Switzerland) to measure the parameters like forced vital capacity (FVC), forced expiratory volume in 1 second (FEV1), FEV1% and forced expiratory flow25–75(FEF25–75). Peak Expiratory Flow Rate (PEFR) was measured using Wright's Peak Flow Meter (Clement and Clarke, UK). FVC and FEV1 were expressed in litres, PEFR in litres/min, FEF25–75 in litres/sec and FEV1% was presented as the ratio of FEV1 and FVC expressed in percentage. Initially, a descriptive analysis was done to observe the personal characteristics of the study subjects as well as to understand the distribution of different fuel use. Afterwards different pulmonary function values were compared in males and females separately with reference to their fuel use taking care of the possible confounders. Analysis was done using SPSS Release 6.1.4 software. Analysis of covariance was performed to compare the levels of different pulmonary function parameters in relation to different fuel users taking care of the role of possible confounding factors. In our study we found that our subjects were users of wood, cattle dung, coal, kerosene and liquefied petroleum gas either alone or in combination. While categorizing we initially named wood and cattle dung users as biomass fuel users. Coal, kerosene and LPG users were treated as separate groups. Afterwards, to evaluate the effects of individual fuels within the biomass fuel group, we have treated wood and cattle dung as separate groups. Four stage analysis was actually done in relation to pulmonary function values while doing ANOCOVA analysis: biomass vs. no biomass analysis, individual fuel group wise analysis, only biomass vs. only LPG (most modern fuel among the lot) analysis and only wood (most important fuel factor according to our study in relation to deterioration of pulmonary function) vs. only LPG analysis. Though individual fuel wise analysis gave us the contribution of all the fuel factors on pulmonary function, we did different sub analyses to show the one to one comparison of different fuel groups (biomass vs. no biomass) or individual fuels (wood vs. LPG). Variables like different fuels (fuel wise yes, no), smoking (ever smoker, never smoker), house type (mud made not so well ventilated, cement and brick made well ventilated) and occupation (dusty/non-dusty) were taken as categorical variables. Other variables like age (yrs), height (centimeters), weight (kilograms) and family size were taken as continuous variables. While analyzing, we accommodated all fuel types along with possible confounding variables simultaneously in the ANOCOVA model in order to examine the effect of fuel variables adjusting for the effects of other variables. Analysis was done separately for male and female subjects.
Results
Mean age of the study subjects was 35.5 (± 14.6) years for females (n = 204) and 38.7 (± 16.9) years for males (n = 165). In case of female subjects almost 14% were <20 years old, 103 (50.5%) subjects were in 20–39 years age group, 51(25%) subjects in 40–59 years age group and 10.8% subjects were more than 59 years old. In case of male subjects almost 13% were <20 years old, 67 (40.6%) subjects were in 20–39 years age group, 47 (28.5%) subjects in 40–59 years age group and 18.2% subjects were more than 59 years old. Mean height was 164.8 cms (± 8.9) for male subjects and 153.2 cms (± 6.4) for female subjects. Similarly mean weight was 51.9 kg (± 11.8) and 46.4 kg (± 9.9) for male and female subjects respectively.
So far as fuel use is concerned, 145 (87.9%) male and 192 (94.1%) female subjects were using biomass (wood, cattle dung) fuel either alone or in combination with other fuels. Liquefied Petroleum Gas, kerosene (a petroleum product widely used in India as a cooking fuel. It is not as purified as LPG), coal, wood and cattle dung was being used by 60 (36.4%), 58 (35.2%), 2 (1.2%), 144 (87.3%) and 16 (9.7%) male subjects respectively either alone or in combination. For female subjects the numbers were 52 (25.5%), 78 (38.2%), 7 (3.4%), 192 (94.1%) and 36 (17.6%) respectively. Only LPG was used by 14 (8.5%) male and 8 (3.9%) female subjects. Similarly wood alone was used by 63 (38.4) male and 72 (35.3%) female subjects and biomass fuel alone was used by 68 (41.4%) male and 90 (44.1%) female subjects (Table 1). All the subjects were using these fuels for more than 10 years. Cooking time was 2–3 hours/day for all the subjects.
Table 1 Distribution of fuel use among the study subjects
Fuel Use Male (N = 165) Female (N = 204)
Either alone or in combination
Biomass (Wood/Animal dung) 145 (87.9) 192 (94.1)
Liquefied Petroleum Gas 60 (36.4) 52 (25.5)
Kerosene 58 (35.2) 78 (38.2)
Coal 2 (1.2) 7 (3.4)
Wood 144 (87.3) 192 (94.1)
Animal dung 16 (9.7) 36 (17.6)
Only Biomass 68 (41.4) 90 (44.1)
Only liquefied petroleum gas 14 (8.5) 8 (3.9)
Only wood 63 (38.4) 72 (35.3)
Table 2 shows the results of ANOCOVA analysis, where comparison of different pulmonary function parameters of female study subjects have been done in relation to different fuel use. Adjustment has been done for age, height, weight, house type, family size and occupation while doing this analysis. In individual fuel wise analysis, it was observed that pulmonary function values were comparatively lower in most of the cases in the users of wood, animal dung, coal and kerosene whereas the values were comparatively higher in case of LPG users in comparison to their respective non-users. When biomass users (either alone or in combination) were compared with non-users similar observations were found. In comparison between only wood and only LPG users, it was found that only wood users were having lower values then only LPG users. Similar findings were observed in comparison between only biomass fuel users and only LPG users. The values of FEV1, PEFR and FEF25–75 were significantly less (p < .01 in FEV1 and p < .05 in others) in wood users than the respective non-users. When biomass users (either alone or in combination) were compared with non-users, values of FEV1 (p < .05), FEV1% (p < .01), PEFR (p < .05) and FEF25–75 (p < .01) were significantly less in biomass fuel users. Similarly, only biomass fuel users were having significantly lower values of FEV1% (p < .01), PEFR (p < .05) and FEF25–75 (p < .01) in comparison to only LPG users and only wood users were having significantly lower values of FEV1% (p < .001), and FEF25–75 (p < .001) in comparison to only LPG users. The difference of PEFR values in the later case was also nearly significant (p = 0.055) statistically.
Table 2 Distribution of pulmonary function values of the female study subjects according to fuel use
Fuel FVC FEV1 FEV1% PEFR FEF25–75
Wood
- Non-user 2.45 ± 0.44 2.26 ± 0.46 91.99 ± 3.98 403.50 ± 59.05 3.06 ± 0.84
- User 2.28 ± 0.62 1.97 ± 0.59** 86.44 ± 8.88 347.91 ± 80.96* 2.40 ± 0.95*
Animal dung
- Non-user 2.31 ± 0.62 2.00 ± 0.55 86.70 ± 8.34 353.24 ± 81.02 2.45 ± 0.94
- User 2.18 ± 0.55 1.89 ± 0.49 87.09 ± 10.67 341.58 ± 80.14 2.42 ± 1.05
Coal
- Non-user 2.29 ± 0.62 1.99 ± 0.55 86.77 ± 8.84 351.61 ± 81.54 2.45 ± 0.96
- User 2.22 ± 0.35 1.93 ± 0.39 86.61 ± 6.72 339.00 ± 59.02 2.30 ± 0.77
Kerosene
- Non-user 2.30 ± 0.65 2.02 ± 0.59 87.58 ± 8.44 353.52 ± 87.19 2.53 ± 1.01
- User 2.27 ± 0.54 1.93 ± 0.46 85.46 ± 9.17 347.40 ± 69.60 2.30 ± 0.85*
LPG
- Non-user 2.26 ± 0.64 1.96 ± 0.56 86.61 ± 8.75 347.53 ± 83.70 2.38 ± 0.96
- User 2.36 ± 0.52 2.06 ± 0.49 87.23 ± 8.87 361.86 ± 71.32 2.62 ± 0.99
Biomass
- Non-user 2.45 ± 0.44 2.26 ± 0.46 91.99 ± 3.98 403.50 ± 59.05 3.06 ± 0.84
- User 2.28 ± 0.62 1.97 ± 0.54* 86.44 ± 8.89** 347.91 ± 80.96* 2.40 ± 0.95**
Only biomass user 2.25 ± 0.68 1.96 ± 0.61 86.90 ± 8.92** 345.50 ± 93.36* 2.41 ± 1.00**
Only LPG user 2.45 ± 0.52 2.29 ± 0.55 93.15 ± 4.09 407.50 ± 71.94 3.22 ± 0.97
Only wood user 2.29 ± 0.70 1.98 ± 0.63 86.13 ± 8.66*** 347.55 ± 97.04 2.37 ± 0.98***
Only LPG user 2.45 ± 0.52 2.29 ± 0.55 93.15 ± 4.09 407.50 ± 71.94 3.22 ± 0.97
* p < 0.05
** p < 0.01
*** p < 0.001
Adjusted for age, height, weight, house type, family size and occupation.
Table 3 shows the comparison of pulmonary function values in case of male subjects. In this case the difference of pulmonary function values were not so prominent as in case of female study subjects. Though the pulmonary function values were not significantly different, while comparing different fuel users with their respective non-users in individual fuel wise analysis, some of the values showed significant difference when only biomass fuel users were compared with only LPG users and when only wood users were compared with only LPG users. FEV1 (p < .05), PEFR (p < .05) and FEF25–75 (p < .05) values were significantly lower in only biomass fuel users as well as in only wood users when compared with only LPG users.
Table 3 Distribution of pulmonary function values of the male study subjects according to fuel use
Fuel FVC FEV1 FEV1% PEFR FEF25–75
Wood
- Non-user 3.33 ± 0.81 2.74 ± 0.76 81.83 ± 9.02 453.40 ± 117.79 2.88 ± 1.04
- User 2.98 ± 0.87 2.48 ± 0.80 82.68 ± 10.20 427.11 ± 122.42 2.71 ± 1.15
Animal dung
- Non-user 3.03 ± 0.88 2.52 ± 0.81 82.55 ± 10.25 431.89 ± 125.79 2.75 ± 1.16
- User 2.94 ± 0.81 2.44 ± 0.70 82.79 ± 8.09 415.75 ± 104.06 2.58 ± 0.94
Coal
- Non-user 3.02 ± 0.87 2.51 ± 0.80 82.42 ± 10.00 429.38 ± 122.28 2.72 ± 1.14
- User 2.64 ± 0.53 2.50 ± 0.42 95.04 ± 3.28 506.00 ± 1.41 3.62 ± 0.35
Kerosene
- Non-user 2.96 ± 0.93 2.47 ± 0.85 82.83 ± 9.68 422.65 ± 130.64 2.72 ± 1.20
- User 3.13 ± 0.75 2.58 ± 0.69 82.12 ± 10.74 444.33 ± 103.40 2.76 ± 1.02
LPG
- Non-user 2.93 ± 0.92 2.45 ± 0.85 82.72 ± 10.41 418.27 ± 128.01 2.71 ± 1.22
- User 3.17 ± 0.74 2.62 ± 0.68 82.32 ± 9.45 451.20 ± 108.14 2.78 ± 0.99
Biomass
- Non-user 3.33 ± 0.81 2.74 ± 0.76 81.83 ± 9.02 453.40 ± 117.79 2.88 ± 1.04
- User 2.98 ± 0.87 2.48 ± 0.79 82.64 ± 10.17 427.50 ± 122.09 2.71 ± 1.15
Only biomass user 2.87 ± 0.98 2.40 ± 0.91* 82.76 ± 10.46 411.28 ± 142.02* 2.70 ± 1.29*
Only LPG user 3.36 ± 0.93 2.74 ± 0.88 80.77 ± 10.25 440.86 ± 128.32 2.88 ± 1.19
Only wood user 2.84 ± 1.01 2.38 ± 0.94* 82.51 ± 10.80 407.14 ± 145.33* 2.67 ± 1.32*
Only LPG user 3.36 ± 0.93 2.74 ± 0.88 80.77 ± 10.25 440.86 ± 128.32 2.88 ± 1.19
* p < 0.05
Adjusted for age, height, weight, smoking, house type, family size and occupation.
Discussion
In a study conducted in Turkey a highly significant (p < 0.00001) reduction of FEV1, FVC, FEV1/FVC and FEF25–75 was observed in case of biomass fuel users [9]. A study conducted in an urban Indian slum showed significantly lower FVC, FEV1, FEV1% and PEFR values in bio-fuel using women in comparison to modern fuel users (kerosene and LPG) [10] whereas a similar study undertaken involving rural Indian women could show the prominent adverse effect of biomass fuel use on FVC only [11]. This deterioration of pulmonary function in biomass fuel users has been attributed to the fact that the amount and concentration of particulate matter and other toxic gases emitted during biomass combustion while cooking are more than those emitted during combustion of LPG [16].
This study has come up with the finding that biomass fuel use (especially wood use) is an important factor for deterioration of pulmonary function. Wood, animal dung, coal and kerosene users (female) were having comparatively lower pulmonary function values than their respective non-users whereas LPG users were having comparatively higher values. Comparison of only biomass fuel vs. only LPG use and only wood use vs. only LPG use in females has showed that LPG is better than others so far as deterioration of pulmonary function is concerned. The pulmonary function parameters affected by biomass fuel use have been FEV1, FEV1%, PEFR and FEF25–75. No effect has been observed on FVC. This study has also observed some effect on the male members of biomass using families (especially wood) while comparing only biomass fuel vs. only LPG use and only wood use vs. only LPG use.
Most of the families being farmers, cooking is usually done in early morning or evening hours to enable both male and female members to go to field for agricultural activities. Otherwise also, cooking in such hours was found to be the custom of this village and thereby both the male and female members of the families were exposed to the effects of cooking fuels because of their presence during cooking hours. However, direct involvement to the act of cooking may have been the reason for more exposure in females, which may have caused more effects in females.
This study has made an effort to derive the effects of different fuel use on pulmonary function adjusting for the effects of the confounders as much as possible. Adjustment of the effect of confounders being extremely important [14,15] in such a study, this effort has been done as meticulously as possible. The effects of age, height, weight, house type (ventilation status), family size (overcrowding) and occupation have been adjusted for. All the subjects reported that they were using the fuel for more than last 10 years and the cooking time/day was 2–3 hours. Lack of specific information on exact cooking time and exact duration of fuel used restricted us from examining the effect of these two variables as confounders. This has been a limitation of this study. Inclusion of large number of variables in the ANOCOVA analysis model may have been another limitation of this study considering its sample size (calculated sample size was for a prevalence study only). With a larger sample size in addition to the effects of individual fuels, effects of all possible fuel combinations could have been studied. Nevertheless, this study has included the combination fuel users also in the analysis (unlike previously reported studies) along with single fuel users and the effect of fuel use on pulmonary function has been estimated not only by comparing individual fuel users vs. non-users but also with the one to one comparison of different fuel users. This study being a cross sectional one also bears the restriction of understanding the temporal relationship of fuel use and pulmonary function deterioration. Selection of subjects from more villages also could have made the findings of this study more generalisable.
Conclusion
This study eventually concludes showing the adverse effects of biomass fuels (especially wood) use on the deterioration of pulmonary function. The findings of this study point towards an important environmental health problem involving mostly the poor women and the children and indicate that the health consequences of exposure from biomass and other solid fuels in developing countries should not be ignored not only because the health burden is high but also because of the fact that such fuels will continue to be used throughout the world by a large number of households in the foreseeable future because of economic reasons.
Competing interests
The author(s) declare that they have no competing interest.
Authors' contributions
AS: Planned, executed the study. Prepared the write up.
NMR: Involved in planning and execution of the study.
PKK: Performed statistical analysis
PKM: Executed the study, helped in literature search.
HNS: Guided in planning, executing the study and in preparing the write up.
Acknowledgements
We hereby acknowledge the help of Dr D. J. Parikh, Mr I. S Makwana, Ms M. Mankad, Ms S. Sheikh and Ms N. Pathak for their help in this study.
==== Refs
World Resources Institute UNEP, UNDP, World Bank: 1998–99 World Resources: a guide to the global environment 1999 Oxford University Press
World Health Organization Health and environment in sustainable development WHO/EHG/978 1997 Geneva, World Health Organization
Smith K Samet J Romieu I Bruce N Indoor air pollution in developing countries and acute respiratory infections in children Thorax 2000 55 518 532 10817802 10.1136/thorax.55.6.518
Ezzati M Kammen D Indoor air pollution from biomass combustion and acute respiratory infections in Kenya: an exposure-response study Lancet 2001 358 9282 619 24 11530148 10.1016/S0140-6736(01)05777-4
Bruce N Perez-Padilla R Alablak R Indoor air pollution in developing countries: a major environmental and public health challenge WHO Bulletin 2000 78 1078 1092
Kiraz K Kart L Demir R Oymak S Gulmez I Unalacak M Ozesmi M Chronic pulmonary disease in rural women exposed to biomass fumes Clin Invest Med 2003 26 5 243 8 14596485
Mumford JL Lung cancer and indoor air pollution in Xuan Wei, China Science 1987 235 217 220 3798109
Smith KR Liu Y Samet J Îndoor air pollution in developing countries Epidemiology of lung cancer Lung biology in health and disease 1993 New York, MarcelDekker
Sumer H Turaclar UT Onarlioglu T Ozdemir L Zwahlen M The association of biomass fuel combustion on pulmonary function tests in the adult population of Mid-Anatolia Soz Praventivmed 2004 49 4 247 53 15357526
Dutt D Srinivasa DK Rotti SB Sahai A Konar D Effect of indoor air pollution on the respiratory system of women using different fuels for cooking in an urban slum of Pondicherry Natl Med J India 1996 9 3 113 7 8664820
Behera D Jindal SK Malhotra HS Ventilatory function in nonsmoking rural Indian women using different cooking fuels Respiration 1994 61 2 89 92 8008994
Behera D Chakrabarti T Khanduja KL Effect of Exposure to Domestic Cooking Fuels on Bronchial Asthma Indian J Chest Dis Allied Sci 2001 43 1 19 25 11370502
Behera D Sood P Singh S Passive smoking, domestic fuels and lung function in north Indian children Indian J Chest Dis Allied Sci 1998 40 2 89 98 9775566
Reddy TS Guleria R Sinha S Sharma SK Pande JN Domestic Cooking Fuel and Lung Functions in Healthy Non-smoking Women Indian J Chest Dis Allied Sci 2004 46 85 90 15072322
Bruce N Neufeld L Boy E West C Indoor biofuel air pollution and respiratory health: the role of confounding factors among women in highland Guatemala Int J Epidemiol 1998 27 3 454 8 9698135 10.1093/ije/27.3.454
WHO, Geneva: Working paper from WHO Consultation – indoor air pollution from biomass fuel 1992
|
16255784
|
PMC1291405
|
CC BY
|
2021-01-04 16:36:25
|
no
|
Respir Res. 2005 Oct 31; 6(1):127
|
utf-8
|
Respir Res
| 2,005 |
10.1186/1465-9921-6-127
|
oa_comm
|
==== Front
Respir ResRespiratory Research1465-99211465-993XBioMed Central London 1465-9921-6-1281626290010.1186/1465-9921-6-128ResearchInhaled tolafentrine reverses pulmonary vascular remodeling via inhibition of smooth muscle cell migration Pullamsetti Soni [email protected] Stefanie [email protected] Hüseyin 1Hü[email protected] Hossein Ardeschir [email protected] Christian [email protected] Norbert [email protected] Beate [email protected] Werner [email protected] Friedrich [email protected] Ralph Theo [email protected] University of Giessen Lung Center (UGLC), Medical Clinic II/V, Giessen, Germany2 Altana Pharma, Constance, Germany2005 1 11 2005 6 1 128 128 17 8 2005 1 11 2005 Copyright © 2005 Pullamsetti et al; licensee BioMed Central Ltd.2005Pullamsetti et al; licensee BioMed Central Ltd.This is an Open Access article distributed under the terms of the Creative Commons Attribution License (), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
Background
The aim of the study was to assess the chronic effects of combined phosphodiesterase 3/4 inhibitor tolafentrine, administered by inhalation, during monocrotaline-induced pulmonary arterial hypertension (PAH) in rats.
Methods
CD rats were given a single subcutaneous injection of monocrotaline to induce PAH. Four weeks after, rats were subjected to inhalation of tolafentrine or sham nebulization in an unrestrained, whole body aerosol exposure system. In these animals (i) the acute pulmonary vasodilatory efficacy of inhaled tolafentrine (ii) the anti-remodeling effect of long-term inhalation of tolafentrine (iii) the effects of tolafentrine on the expression profile of 96 genes encoding cell adhesion and extracellular matrix regulation were examined. In addition, the inhibitory effect of tolafentrine on ex vivo isolated pulmonary artery SMC cell migration was also investigated.
Results
Monocrotaline injection provoked severe PAH (right ventricular systolic pressure increased from 25.9 ± 4.0 to 68.9 ± 3.2 after 4 weeks and 74.9 ± 5.1 mmHg after 6 weeks), cardiac output depression and right heart hypertrophy. The media thickness of the pulmonary arteries and the proportion of muscularization of small precapillary resistance vessels increased dramatically, and the migratory response of ex-vivo isolated pulmonary artery smooth muscle cells (PASMC) was increased. Micro-arrays and subsequent confirmation with real time PCR demonstrated upregulation of several extracellular matrix regulation and adhesion genes, such as matrixmetalloproteases (MMP) 2, 8, 9, 10, 11, 12, 20, Icam, Itgax, Plat and serpinb2. When chronically nebulized from day 28 to 42 (12 daily aerosol maneuvers), after full establishment of severe pulmonary hypertension, tolafentrine reversed about 60% of all hemodynamic abnormalities, right heart hypertrophy and monocrotaline-induced structural lung vascular changes, including the proportion of pulmonary artery muscularization. The upregulation of extracellular matrix regulation and adhesion genes was reduced by nearly 80% by inhalation of the tolafentrine. When assessed in vitro, tolafentrine blocked the enhanced PASMC migratory response.
Conclusion
In conclusion, we demonstrate for the first time that inhalation of combined PDE3/4 inhibitor reverses pulmonary hypertension fully developed in response to monocrotaline in rats. This "reverse-remodeling" effect includes structural changes in the lung vascular wall and key molecular pathways of matrix regulation, concomitant with 60% normalization of hemodynamics.
==== Body
Background
Pulmonary arterial hypertension (PAH) is a severe disabling disease characterized by elevation of pulmonary artery pressure and death attributable to right heart failure [1]. It is a progressive, proliferative vascular disorder resulting from persistent vasoconstriction and structural remodeling of pulmonary vessels. The structural changes include endothelial cell injury, neovascularization of small arteries, smooth muscle cell (SMC) migration and proliferation, and abnormal accumulation of extracellular matrix proteins associated with activation of matrix metalloproteinases (MMPs) [2].
The MMPs are a family of matrix-degrading enzymes that have been implicated in two important processes in vessel wall repair: cellular migration [3], and regulating extracellular matrix composition and content [4]. Based on their substrate specificity, MMPs were subdivided into four groups: i) interstitial collagenases (MMP 1, MMP 8 and MMP 13) that degrade fibrillary collagens; ii) type IV collagenases (MMP 2 and MMP 9) that degrade basement membrane components; iii) stromelysins (MMP 3, MMP 10 and MMP 11) that degrade proteoglycans, fibronectin, laminin, gelatin and the globular proteins of the type IV collagen; and iv) membrane type-MMPs (MT-1, MMP 15, MMP 16 and MMP 17), possessing a broad spectrum of activities.
Of the MMPs, expression of MMP 2 and MMP 9 in particular, which degrade type IV collagen of basement membranes, are increased in the pulmonary vascular bed, during both monocrotaline (MCT) and hypoxia-induced experimental PAH [5]. In addition, MMP 1, an interstitial collagenase, is upregulated in MCT-induced PAH [6]. In recent years, several pharmacological interventions in PAH, including MMP inhibitors, endothelin antagonist, angiotensinogen inhibitors and phosphodiesterase (PDE) inhibitors that target either the MMP cascades or endogenous vascular elastases, proved to be beneficial in experimental PAH models, with most of these agents being applied prior to full establishment of the disease [7-11]. Suppression of vessel wall remodeling was assumed to largely contribute to these beneficial effects.
The phosphodiesterases (PDEs) are a large family of intracellular enzymes that degrade cyclic nucleotides [12,13]. Because of their potential for altering a variety of cellular responses, PDEs are appealing targets for the treatment of PAH [14,15]. Phosphodiesterase 3 and 4 isoenzymes are the essential players co-regulating cAMP catabolism in many organs, including the lung, and were shown to be upregulated in experimental PAH models. Inhibitors of PDE 3 and 4 synergistically promoted the acute pulmonary vasodilation evoked by prostacyclin or its stable analogues in experimental models of PAH [16-18]. Recently, inhalation of the combined -selective PDE3/4 inhibitor tolafentrine has been shown to amplify the vasodilatory effect of inhaled iloprost in patients with PAH [18]. However, no data are currently available regarding effects of long-term inhalation of tolafentrine on hemodynamics and pulmonary vascular remodeling in PAH models.
In the present investigation, we employed monocrotaline (MCT), a toxin derived from plants of the Crotalaria species [19], for pulmonary artery smooth muscle cell hypertrophy and severe pulmonary hypertension in rats [20]. In this model we examined (i) the acute pulmonary vasodilatory efficacy of inhaled tolafentrine (ii) the anti-remodeling effect of long-term inhalation of tolafentrine (iii) the inhibitory effect of tolafentrine on pulmonary artery SMC cell migration, and (iv) the effect of the PDE inhibitor on the expression profile of 96 genes encoding cell adhesion and extracellular matrix regulation. To mimic clinical conditions, inhalation of tolafentrine commenced after pulmonary hypertension was already fully established. Essentially, we found that inhaled tolafentrine reverses hemodynamic abnormalities as well as structural changes, SMC migration and proliferation and matrix remodeling, representing the key features of MCT induced PAH.
Methods
Animal experiments
Experiments were performed on male CD rats 300 – 350 g body weight, Charles River, Sulzfeld, Germany). Pulmonary hypertension was induced by a single subcutaneous injection of monocrotaline (MCT, 60 mg/kg, Sigma, Deishofen, Germany), dissolved in 0.1 M NaOH, adjusted to pH 7.4 with 0.1 M HCl, according to the previous reports [7,10,21,22].
The experiments were performed in accordance with the National Institutes of Health Guidelines on the Use of Laboratory Animals. Both the University Animal Care Committee and the Federal Authorities for Animal Research of the Regierungspräsidium Giessen (Hessen, Germany) approved the study protocol.
Study groups
The animals were classified into the following four groups: 1) rats injected with saline sacrificed after 28 days (Control, n = 8); 2) MCT-injected rats sacrificed after 28 days (MCT[28d], n = 12); 3) MCT-injected sacrificed after 42 days, with nebulized vehicle being administered from day 28 to day 42 (MCT[42d], n = 14); 4) MCT-injected rats sacrificed after 42 days, with nebulized tolafentrine being administered from day 28 to day 42 (MCT[42d]/Tola, n = 10). For acute hemodynamic studies 12 MCT[28d] rats (MCT injected for 28 days) were used (inhalation of saline, n = 6; and inhalation of tolafentrine in two different doses, n = 6 each).
Inhalation of tolafentrine
Four weeks after a single MCT injection, rats were subjected to inhalation of tolafentrine or sham nebulization in an unrestrained, whole body aerosol exposure system as described [23]. For assessment of chronic effects of inhaled saline or tolafentrine (dose deposited in the lungs ~ 120 μg/kg day), 15 min nebulization maneuvers using a jet nebulizer with a constant flow rate of 6 l/min (Pari LC Star, Pari, Starnberg, Germany) were repeated twelve times per day for 2 weeks (day 28 – 42). Particles generated by the jet nebulizer were characterized by a mass median aerodynamic diameter (MMAD) of 2.8 μm and a geometric standard deviation (GSD) of 2.5 (determined by laser difractometric measurements, as described [24]. Preliminary experiments with nebulization of 99Tc-DTPA determined the total lung deposition of nebulized material to range at 0.5 %, in accordance with previous studies in this model [22].
To assess the acute effects of inhaled tolafentrine, hemodynamic testing in response to the drug was performed in animals that had previously received a MCT injection 4 weeks prior to study. Hemodynamics were measured before and after a single inhalation of the tolafentrine (130 and 650 μg/kg min for 10 min) by an ultrasonic nebulizer with MMAD of 4.0 μm and GSD of 2.1 as described previously [22,23,25].
Surgical Preparation, measurement of hemodynamics and tissue preparation
For measurement of hemodynamic parameters, rats were anaesthetized with an i.p. injection of ketamine (9 mg/kg body mass) and medetomidine (100 μg/kg body mass), followed by an i.m. injection of atropine (250 μg/kg body mass). The rats were tracheotomized and ventilated with a frequency of 60 breaths/min. Positive end expiratory pressure was set at 1 cm H2O. A polyethylene catheter was inserted into the left carotid artery to measure arterial pressure. A right heart catheter (PE 50 tubing) was inserted into the right ventricle through the right jugular vein for measurement of right ventricular systolic pressure with fluid-filled force transducers [10,11]. Cardiac output (CO) was measured by thermodilution technique as described [10,11]. Briefly, a thermistor (1.5 F) was placed into the ascending thoracic aorta via the right carotid artery for the measurement of transpulmonary thermodilution cardiac output (Cardiotherm 500-X, Hugo-Sachs Electronic – Harvard Apparatus GmbH, March-Hugstetten, Germany). The CO was averaged from three consecutive determinations and indexed to the weight of the animal to obtain cardiac index (CI). After exsanguination, the left lung was fixed for histology in 10% neutral buffered formalin and the right lung was snap frozen in liquid nitrogen.
Right ventricular hypertrophy
The RV was dissected from the left ventricle (LV) and the septum (S) and weighed to determine the extent of RV hypertrophy as follows: RV/ (LV+S).
Histological examination of the lungs
Paraffin lung sections (3 μm) were double-stained with anti-α-smooth muscle actin antibody (dilution 1:900, clone 1A4, Sigma, Saint Louis, Missouri) and anti-human von Willebrand factor antibody (vWF, dilution 1:900, Dako, Hamburg, Germany). Sections were counterstained with hematoxylin and examined by light microscopy using a computerized morphometric system (Qwin, Leica, and Wetzlar, Germany) for assessing the degree of muscularization of small peripheral pulmonary arteries [10,11,23]. In addition, lung sections were stained for Elastin-Nuclear Fast Red to assess the medial wall thickness.
Categorization of pulmonary arteries based on the degree of muscularization and on the external diameter and the percentage of medial wall thickness were performed as previously described [10,11,23].
Microarray and data analysis
Total RNA extraction was performed using the Trizol reagent (Burlington, Ontario, Canada) according to the instructions given by the manufacturer from control, monocrotaline (MCT[42d]) and monocrotaline plus tolafentrine (MCT[42d]/Tola) treated lungs. Samples were treated with DNase I according to the RNase-free DNase set (QIAGEN, Hilden, Germany). Total RNA (5 μg) from each sample was used to generate Biotin-16-dUTP labeled cDNA and hybridized to the extracellular matrix and adhesion molecules gene array (GEArray Q series, Superarray, MD, USA). After hybridization, the membranes were developed according to the manufacturer's recommendations to yield luminescent signals, which were then captured on X-ray film.
The resulting image data were analyzed for differential gene expression patterns using GE Array Analyzer 1.2 (Super Array Bioscience Corp., Frederick, MD) software. Loading was adjusted on the basis of the intensity of hybridization signals relative to the housekeeping gene GAPDH.
Real-time polymerase chain reaction
Quantitative real-time PCR was performed as described [26]. Briefly, total RNA was isolated from frozen lungs using Trizol Reagent according to the manufacturer's instructions. For the generation of cDNA, equal amounts of RNA from each sample were used as templates for reverse transcription of first-strand cDNA using the qPCR™ Mastermix (Euro-genetec, Seraing, Belgium) according to the manufacturer's protocol. For quantitative real-time RT-PCR analysis, 2 μl cDNA was placed into 50 μl reaction volume containing SYBR Green PCR mix and sequence-specific oligonucleotide primers. The thermal cycle conditions used for all reactions were as follows: activation, 50°C for 2 min; denaturation, 95°C for 10 min; and cycle, 95°C for 10 s and 60°C for 5 s (40 times). Specific primers used for sequence detection were: rGAPDH,5'-GTG ATG GGT GTG AAC CAC GAG-3' (forward) and5'-CCA CGA TGC CAA AGT TGT CA-3'(reverse); forMMP2,5'-ATC TGC AAG CAA GAC ATT GTC TT-3'(forward)and 5'-GCC AAA TAA ACC GAT CCT TGA A-3' (reverse);for MMP8, 5'-CGG GAA GAC ATA CTC TTC GAA-3'(forward)and 5'-CAT GGA TCT TCT TTG ATT GTC G-3' (reverse);for MMP9, 5'-GTA ACC CTG GTC ACC GGA CTT -3'(forward)and 5'-ATA CGT TCC CGG CTG ATC AG-3' (reverse); for MMP10, 5'- GGA GAT GCT CAC TTC GAT GAT-3' (forward)and 5'-CAG CAA CCA GGA ATA AAT TGG-3' (reverse); for MMP11, 5'-TTC TGA GAT TGA TGC TGC TTT C-3'(forward)and 5'-TGT CCA CGA AGG AAG TAG GC-3' (reverse); for MMP12, 5'-GCT GTC ACA ACA GTG GGA GA-3' (forward)and 5'-GTA ATG TTG GTG GCT GGA CTC-3' (reverse); for MMP20, 5'-GAA TAA ACT CTG GGG AAG CAG A-3'(forward)and 5'-TGA TTG GAT TAA GGC CTC GT-3' (reverse); for Icam, 5'-CAG CTG CGC TGT GTT TTG-3'(forward)and 5'-GGA TGG GAG CTG AAA AGT TG-3' (reverse); For Itgax, 5'-GCC TCG AGA CTG GAG ATC AT-3'(forward)and 5'-GGA GAG CTG GGA GCC AGT-3' (reverse);for Ncam1, 5'-ACC ATG AGC TGG ACA AAG GA-3'(forward)and 5'-CAC TGA AGA TGT GCT TCT CGT C-3' (reverse); for Plat, 5'-AAT GAA GGG AGA GCT GTT GTG-3'(forward) and 5'-TCC TCT TCT GAA CCT CCT GTG-3' (reverse); for Serpinb2, 5'-CGA AAG GGA TTT TGT GAT GTC-3' (forward)and 5'-GTG GGA AAT GGG AAT TCG T-3' (reverse). All real-time reactions were carried on an ABI 7700 sequence Detection System (Applied Biosystems, Foster City, CA, USA), and analysis was performed with the accompanying software. At the end of the PCR cycle, a dissociation curve was generated to ensure the amplification of a single product and the threshold cycle time (Ct values) for each gene was determined. Relative mRNA levels were calculated based on the Ct values and normalized to house keeping gene GAPDH.
Migration assay
Pulmonary artery smooth muscle cell (PASMC) migration was examined in Transwell cell culture chambers with gelatin-coated polycarbonate membranes as described previously [27]. For this assay, control and MCT-treated rat PASMCs were isolated and added to the upper well of a Transwell (Corning Costar, Cambridge, MA) at 2 × 105 cells/well. The platelet-derived growth factor – BB (PDGF-BB) was added to the lower chamber at a concentration of 25 ng/ml and cells were allowed to migrate for 8 and 24 hours. In experiments with the combined selective PDE3/4 inhibitor tolafentrine (0.05, 0.1 and 0.3 μM), MCT-treated rat PASMCs were pre-incubated with tolafentrine in the upper well of a Transwell for 30 min before the addition of PDGF-BB into the lower chamber. Migration was quantified by staining the cells with crystal violet (Sigma, Deishofen, Germany) following microscopic cell counts on five random fields in each well. Experiments were performed in triplicate and were repeated at least three times.
Data analysis
All data are given as mean ± SEM. Differences between the groups were assessed by analysis of variance (one-way ANOVA) and Student-Newman-Keuls test for multiple comparisons with a p value < 0.05 regarded to be significant.
Results
Acute vasodilatory effects of aerosolized tolafentrine in MCT treated (28d) rats
Aerosolized tolafentrine reduced right ventricular systolic pressure [RVSP] in MCT[28d] rats in a dose-dependent manner (Figure 1). As depicted, this pulmonary vasodilatation was accompanied by less pronounced decrease in systemic arterial pressure.
Figure 1 Immediate vasodilatory effects of inhaled tolafentrine in monocrotaline-induced pulmonary arterial hypertension. Monocrotaline (MCT[28d]) treated animals received tolafentrine over an inhalation period of 5 min subsequent to catheterization. The decrease in right ventricular systolic pressure (RVSP, in mmHg) and systemic arterial pressure (SAP, in mmHg) in response to the vasodilatory treatment is given. All values are given as mean ± SEM. *, p < 0.05 versus non treated animals
Chronic effects of aerosolized tolafentrine: hemodynamics
After injection of monocrotaline, pulmonary hypertension developed (right ventricular systolic pressure on day 28 = 66.5 ± 3.2 mm Hg (n = 11) and on day 42 = 74.9 ± 5.1 mm Hg (n = 9), as compared to 25.9 ± 4.0 mm Hg in the control animals (n = 10)) (Figure 2). No significant changes in systemic arterial pressure occurred. As compared to control animals (36.5 ± 3.5 ml/min 100 g body weight), cardiac index was slightly decreased on day 28 (31.8 ± 1.3 ml/min/100 g body weight) and significantly decreased on day 42 (28.1 ± 2.5 ml/min/100 g body weight). Aerosolized tolafentrine treatment significantly lowered right ventricular pressure to 48.4 ± 2.1 mmHg (p < 0.05 versus MCT[42d] and MCT[28d]). No significant changes in systemic arterial pressure occurred in animals undergoing long-term tolafentrine inhalation (n = 8), while cardiac output was significantly increased in the MCT[42d]/Tola group.(42.1 ± 5.1 ml/min/100 g body weight; p < 0.05 versus MCT[42d]). The total pulmonary resistance index, calculated from the RVSP and cardiac output values, was even fully normalized in the MCT[42d]/Tola animals.
Figure 2 Influence of inhaled tolafentrine on hemodynamics in monocrotaline – induced pulmonary arterial hypertension. Right ventricular systolic pressure (RVSP, in mmHg), systemic arterial pressure (SAP, in mmHg), cardiac index (CI, in ml min-1 100 g body weight-1) and total pulmonary resistance index (PRI, in mmHg min ml-1 100 g body weight-1) are given. Tolafentrine was applied by repetitive inhalations from day 28 to day 42. All values are given as mean ± SEM. *, p < 0.05 versus control; †, p < 0.05 versus MCT[28d]; ‡, p < 0.05 versus MCT[42d].
Chronic effects of aerosolized tolafentrine: right ventricular hypertrophy
Four weeks after injection of MCT, animals demonstrated significant right heart hypertrophy, as indicated by an increase in the right ventricular to left ventricular plus septum weight ratio (RV/LV+S) from 0.29 ± 0.02 (control animals) to 0.60 ± 0.02 (Figure 3). Rats that received inhaled vehicle for 2 weeks demonstrated further progression of right ventricular hypertrophy (RV/LV+S = 0.71 ± 0.05). Inhaled tolafentrine reversed established right ventricular hypertrophy (MCT[42d]/Tola = 0.47 ± 0.03; p < 0.05 versus MCT[28d] and MCT[42d]) (Figure 3).
Figure 3 Influence of inhaled tolafentrine on right heart hypertrophy. Right to left ventricular plus septum ratio (RV/LV+S) of different treatment groups is given. Tolafentrine was applied by repetitive inhalations from day 28 to day 42. All values are given as mean ± SEM. *, p < 0.05 versus control; †, p < 0.05 versus MCT[28d]; ‡, p < 0.05 versus MCT[42d].
Chronic effects of aerosolized tolafentrine: survival
In this study, rats were randomized to receive vehicle or inhaled tolafentrine beginning four weeks after MCT treatment. In the MCT[28d] group a survival of 78% was noted, while survival decreased to 60% in the MCT[42d] group. In the MCT[42d]/Tola 80 % survived the treatment protocol. However, due to insufficient number of experiments the significant effects on this parameter were not shown in detail.
Chronic effects of aerosolized tolafentrine: histopathology
Elastin staining and subsequent morphometric analysis of pulmonary arteries demonstrated a markedly increased medial wall thickness in both the MCT[28d] and the MCT[42d] groups, when compared with the saline-treated group (Figure 4). In comparison to control animals, the percentage of medial wall thickness of arteries sized between external diameters of 25 to 50 μm was increased significantly from 18.3 ± 0.42 to 28.5 ± 0.33 (MCT[28d]) and 29.1 ± 0.48 (MCT[42d]) (Figure 5A). To a similar extent, the percentage of medial wall thickness of arteries with external diameters of 51–100 μm increased from 17 ± 0.49 (control) to 22.2 ± 0.31 (MCT[28d]) and 25.1 ± 0.77 (MCT[42d]) (Figure 5B). In arteries larger than 100 μm, the corresponding data were 14.8 ± 0.95 (control), 18.1 ± 0.57 (MCT[28d]) and 20.6 ± 1.44 (MCT[42d]) (Figure 5C), respectively. The evaluation of the extent of muscularization of pulmonary arteries with external diameters of 15 to 50 μm demonstrated a significant reduction in non-muscularized pulmonary arteries (62.6 ± 5.5% in controls, 3.7 ± 2.1 % in MCT[28d], 1.6 ± 0.7 % in MCT[42d]). Concomitantly, a significant increase in fully muscularized vessels from 2.4 ± 0.62 (control) to 51.1 ± 6.85 (MCT[28d]) and 71.6 ± 2.63 (MCT[42d]) was noted (Figure 6).
Figure 4 Effect of inhaled tolafentrine on the degree of muscularization and on the medial wall thickness of small pulmonary arteries. Immunohistochemical analysis of lung sections originating from saline (Control), monocrotaline (MCT[42d]) and monocrotaline plus tolafentrine (MCT[42d]/Tola) treated animals. Staining was undertaken for von Willebrand-factor (brown; endothelial cells) and alpha smooth muscle actin (purple; smooth muscel cells) as well as elastin. Scale bar: 20 μm.
Figure 5 Effect of inhaled tolafentrine on medial wall thickness of pulmonary arteries. Measurement of medial wall thickness (given in percentage of total wall thickness) of pulmonary arteries sized from (A) 25 to 50 μm, (B) 51–100 μm and (C) >101 μm. All values are given as mean ± SEM. *, p < 0.05 versus control; †, p < 0.05 versus MCT[28d]; ‡, p < 0.05 versus MCT[42d].
Figure 6 Effect of inhaled tolafentrine on the degree of muscularization of peripheral pulmonary arteries. Percentage of non- (N), partially (P) or fully (M) muscularized pulmonary arteries, related to the total number of pulmonary arteries is given. A total of 60 to 80 intra-acinar vessels were analysed in each single lung from saline (Control), monocrotaline (MCT[28d] and MCT[42d]), and monocrotaline plus tolafentrine (MCT[42d]/Tola) treated animals. Tolafentrine was applied by repetitive inhalations from day 28 to day 42. All values are given as mean ± SEM. *, p < 0.05 versus control; †, p < 0.05 versus MCT[28d]; ‡, p < 0.05 versus MCT[42d].
Most impressively, both, medial wall thickness (22.5 ± 0.65%) and percentage of fully muscularized pulmonary arteries (32.5 ± 5) were significantly reduced by long term treatment of aerosolized tolafentrine, while the percentage of non-muscularized pulmonary arteries significantly increased.
Comparison of migratory responses of control- and MCT-PASMCs
Migration of control and MCT-treated rat PASMCs was examined at 8 and 24 h in the presence of the chemo-attractant PDGF-BB (Figure 7). In the presence of PDGF-BB, the migration rate of PASMCs derived from MCT rats ranged at 207% and 155% of that of the PASMCs derived from control rats (8 hr and 24 hr data, respectively).
Figure 7 Comparison of migratory responses of control and MCT – PASMCs. Migratory responses of control and monocrotaline rat PASMCs were compared at different time points in presence of PDGF-BB (25 ng/ml). The number of migrated cells was counted in 5 random areas per membrane. The results are based on three different independent experiments. All values are given as mean ± SEM. *, p < 0.05 versus control PASMCs at the respective time points.
Effect of tolafentrine: PDGF-Induced MCT- PASMCs migration
Inhibitory effects of tolafentrine on MCT- treated rat PASMCs migration were examined using a migration assay with modified Boyden chamber. The PDGF caused strong enhancement of PASMC migration. Treatment with tolafentrine inhibited PDGF induced PASMC migration in a dose-dependent fashion: at 0.05, 0.1, and 0.3 μM tolafentrine, PASMC migration was reduced by 34 %, 72 %, and 92 % (Figure 8).
Figure 8 Effect of tolafentrine on PDGF-directed PASMCs migration. Tolafentrine at different concentrations (0.05, 0.1, and 0.3 μM) inhibited PDGF-BB (25 ng/ml)-induced MCT-treated rat PASMCs migration. The number of migrated cells was counted in 5 random areas per membrane. The results are based on three independent experiments. All values are given as mean ± SEM. *, p < 0.05 versus PDGF-BB stimulated PASMCs.
Effects of tolafentrine: expression of matrix-degrading proteases
To study the mechanisms underlying tolafentrine-inhibited PASMC migration, lung samples from control, MCT[42d] and MCT[42d]/Tola rats were analyzed on 96 gene arrays encoding key extracellular matrix and adhesion molecules (representative arrays given in Figure 9). The analysis shows expression of 12 of the 96 extracellular matrix and adhesion related genes in the control lungs. Further, a total of 12 genes were differentially expressed and modulated in the MCT-treated lungs. The up-regulated genes include 7 MMPs (MMP 2, MMP 8, MMP 9, MMP 10, MMP 11, MMP 12, MMP 20), 1 serine protease (Plat), 3 cell adhesion molecules (ICAM-1, NCAM-1, Itgax), and 1 protease inhibitor (Serpinb2).
Figure 9 Effect of tolafentrine on MMP and adhesion molecule gene expression – cDNA micro-array analysis. Biotin labeled cDNA probes generated from total RNA from control lungs (Control), monocrotaline (MCT[42d]), and monocrotaline plus tolafentrine (MCT[42d]/Tola) treated lungs were hybridized to the identical extracellular matrix and adhesion molecules gene arrays. Hybridization patterns were assessed by chemiluminescence. Arrays representative of three independent array experiments per group are shown.
To validate these findings, the expression of 12 genes was investigated by quantitative real-time PCR. These results confirmed the array data for most of the proteases investigated (Figure 10). Monocrotaline treatment for four weeks caused an upregulation of the gelatinases MMP 2 and MMP 9 by a factor of 3.39 ± 0.65 and 3.92 ± 0.77, respectively. MMP 8, the interstitial collagenase, was elevated 4.77 ± 1.44 fold with MCT treatment. In addition, the stromelysins MMP 10, MMP 11, MMP 12 and MMP 20 were significantly upregulated by factors of 4.65 ± 1.13, 5.27 ± 1.03, 2.27 ± 0.63 and 6.67 ± 0.75, respectively.
Figure 10 Quantitative real-time RTPCR confirmation of relative changes detected in micro arrays. Relative quantification of mRNAs encoding for MMP 2, MMP 8, MMP 9, MMP 10, MMP 11, MMP 12, MMP 20, Icam, Itgax, Ncam, Plat, and Serpinb2 related to the housekeeping gene GAPDH was undertaken by real-time RTPCR. The lung samples originated from control, monocrotaline (MCT[42d]), and monocrotaline plus tolafentrine (MCT[42d]/Tola) treated animals. All values are given as mean ± SEM. *, p < 0.05 versus control, †, p < 0.05 versus MCT[42d].
Several adhesion molecules were also differentially expressed at the mRNA level. We observed a significantly increased expression of Icam and Itgax (4.34 ± 1.22 and 3.44 ± 0.33 fold). In contrast, Ncam was not changed upon MCT treatment. An upregulation of tissue plasminogen activator (plat) by 3.74 ± 0.83 and plasminogen activator inhibitor-2 (Serpinb2) by 2.68 ± 0.83 foldwith MCT treatment was also observed.
Most impressively, virtually all alterations in extracellular matrix and adhesion molecule gene expression induced by monocrotaline were fully or partially abrogated by tolafentrine (12 genes given in Figure 10).
Discussion
In the present study, we demonstrate that daily repetitive tolafentrine inhalation significantly improved pulmonary hemodynamics and reversed structural and molecular changes underlying MCT induced PAH in rats. Notably, the inhalative therapy was commenced after full establishment of pulmonary hypertension, 4 weeks after application of monocrotaline. Similar to the abnormalities in human PAH, monocrotaline treatment in rats is known to provoke endothelial injury, proliferation, migration and hypercontraction of vascular smooth muscle cells, as well as inflammatory sequelae [28,29]. The animals die due to a progressive increase in precapillary lung vascular resistance with subsequent right heart failure.
A recent study by our group showed that intravenous infusion of the combined selective PDE 3/4 inhibitor (tolafentrine) prevented the development of pulmonary hypertension and right ventricular hypertrophy in response to monocrotaline [11]. However, the complexity and complications associated with the intravenous application of an agent exerting at the same time pulmonary and systemic vasodilation prompted us to evaluate the inhalative route of application in the present study. Moreover, the therapeutic potential of tolafentrine, i.e. its efficacy after full establishment of severe pulmonary hypertension, and its impact on molecular mechanisms closely linked with the structural wall changes were not addressed in the previous study.
For the inhalation therapy, aimed to achieve a selective pulmonary vasodilation, a 15 fold lower dose compared with the intravenous route of application was employed (120 μg/kg day versus 2 mg/kg day). This low inhaled dosage exerted per se no acute effects on systemic hemodynamics but demonstrated selective pulmonary vasodilation as has been demonstrated for several other compounds [16,17]. Further increase in dose results in spill over of the compound in the systemic circulation and systemic side effects as demonstrated by the higher dose of 650 μg tolafentrine/kg min in this study. Since PDE 3 and 4 are expressed in smooth muscle cells throughout the cardiovascular system, tolafentrine was nebulized in a dose which cause some direct lung vasorelaxation but which is still too low to provoke systemic vasodilatory effects. However, the results of an acute test of inhaled tolafentrine under general anesthesia may not reproduce the effects in an awake animal, but long term nebulization demonstrated strong anti-remodeling effects in the setting of MCT-induced PH.
Nevertheless, and notwithstanding the late initiation of the tolafentrine treatment, hemodynamics were dramatically improved after 2 weeks of inhalative tolafentrine treatment: RVSP values were markedly lower than those before onset of treatment, and cardiac index as well as total pulmonary resistance index were also fully normalized. Accordingly, the right heart hypertrophy was found to be largely decreased, as were the structural changes of the lung vasculature evoked by monocrotaline treatment. The increase in the medial thickness of the precapillary lung arteries was reversed, and the high percentage of fully muscularized peripheral pulmonary arteries decreased in response to tolafentrine inhalation. To our knowledge, this is the first time that combined selective PDE 3/4 inhibition (tolafentrine) has been shown to reverse established pulmonary hypertension both with respect to hemodynamics and the structural remodeling of the lung vasculature.
These findings suggest a potent anti-proliferative effect of combined selective PDE 3/4 inhibition in the lung vasculature, as has been suggested by preceding in vitro data. Selective and nonselective cAMP PDE inhibitors have been shown to elicit a concentration-dependent attenuation of mitogen-induced proliferation in rat and human PASMC predominantly via adenylyl cyclase and protein kinase A [30,31]. Furthermore, PASMC receiving the combination of a PDE3 and a PDE4 inhibitor exhibited a significant additive or synergistic anti-mitogenic effect as compared to each PDE inhibitor used on its own.
Interestingly, an enhanced migratory capacity in response to PDGF was observed in PASMCs isolated form lungs of MCT treated rats as compared to matched control lungs. This is reminiscent of the reported higher migratory response of aortic SMCs derived from spontaneously hypertensive rats as compared to normotensive rats [31,32]. Although the mechanisms responsible for the differential migratory responses of SMC are not known, increased collagenase activities might partly explain this phenomenon, since the destruction of extracellular matrix barriers is an important early step for SMC migration [33,34]. Moreover, enhanced MAP kinase activities, such as extracellular signal-regulated kinase (ERK) and p38 kinase [35], might play a role in PDGF-induced vascular SMC migration, which is in line with the observation of increased ERK expression in the precapillary arteries of MCT-treated rats (data not shown).
The present study demonstrates for the first time that selective PDE 3/4 inhibition inhibits the migration of SMCs stimulated with PDGF in a concentration-dependent manner. The PDGF-BB-induced SMC migration was significantly inhibited by 0.05–0.3 μM tolafentrine. This is well in line with the study of T. Horio et al, who reported that adrenomedullin inhibits the migration of aortic SMCs in response to PDGF, probably through a cAMP-dependent process [36]. Downstream effects of elevated cAMP levels, being relevant for the SMC migratory response, may include MAP kinase signaling and cytosolic Ca2+ regulation [37,38].
The anti-proliferative and anti-migratory effects of tolafentrine may be closely linked to matrix regulation. Our group and others have previously demonstrated an increased expression of gelatinases, MMP 2 and MMP 9 in monocrotaline induced pulmonary hypertension [5,11]. In order to approach this field in more depth, gene array analysis of proteinases and adhesion molecules and subsequent confirmation with real time PCR was undertaken in the present study. A strong upregulation of various extracellular matrix and adhesion molecule genes in response to monocrotaline was noted, in particular MMPs (MMP 2, 8, 9, 10, 11, 12, 20), serine proteases (Plat and Serpinb2) and cell adhesion molecules (Icam, Ncam and Itgax). These observations strongly support the notion that increased MMP expression and activity directly correlates with the severity of disease in various experimental forms of pulmonary hypertension. Several mechanisms may be responsible for MMP upregulation in pulmonary arteries during PAH (Figure 11). The MMP expression may be stimulated by growth factors (PDGF, EGF), cytokines, most notably interleukin IL-1α, and physical forces, which may be induced after MCT administration. Several of these factors stimulate a cascade of MAP kinases and PKC and potently activate the transcription factors AP-1 and NF-kB, which are involved in MMP and adhesion molecule expression [39,40]. Most notably, tolafentrine treatment resulted in strong down-regulation or even normalization of the transcription of most of the MMP and adhesion molecule genes that were upregulated in response to MCT. Moreover, the inhibitory effect of tolafentrine on MMPs was functionally confirmed, since it significantly reduced the migratory properties of PASMCs in vitro. This most impressive regression of MMP and adhesion molecule expression under tolafentrine may in part be explained by cAMP-mediated regulation of cytokines and growth factors, being upstream effectors of MMPs (Figure 11). In addition, cAMP has been shown to inhibit the NF-kB transcriptional activity via protein kinase A [41-43]. Such impact might also explain the down-regulation of Icam and Ncam by tolafentrine, as these adhesion molecules are also known to be NF-kB dependent [44].
Figure 11 Potential interactions among various mediator pathways in the regulation of MMP expression and in the subsequent development of PH. Schematic depiction of molecular mechanisms responsible for MMP transcriptional regulation. Growth factors such as tyrosine kinases (TK), endothelin system (ET) and serine elastases (S.Elastases) positively modulate several MMP genes transcription. MMPs that were produced then cleave extracellular matrix (ECM) and thereby promotes SMC migration and proliferation. Augmentation of these processes ultimately leads to pulmonary vascular remodeling and PH. In contrast, mediators that influence cAMP pathway (Prostacyclin analogues (PGI2) and combined PDE 3/4 inhibitors (tolafentrine) and to a less extent nitric oxide (NO) represses MMP gene activation that were positively modulated by various growth factors during development of PH. These effects eventually lead to regression of matrix degradation, SMC migration and proliferation and reverse-remodeling of pulmonary arteries. GF, growth factor; TK, tyrosine kinase; MEK, MAP/ERK kinase; ERK, Extracellular signal-regulated kinase; IP, prostaglandin I2 (prostacyclin) receptor; AC, adenylyl cyclase; PDE3, phosphodiesterase isoenzyme 3; PDE4, phosphodiesterase isoenzyme 4; iPKA, inhibitory protein kinase A; aPKA, activated protein kinase A; S.Elastases, serine elastases; NO, nitric oxide; ATF3, activating transcription factor 3; ET, endothelin; ETR, endothelin receptor; AP1, activator protein 1; CREB, CRE-binding protein; NFkβ, Sp1, Sp1, transcription factor; NF-kB, nuclear factor-kappaB (p65); MMP, matrix metallo proteases; ECM, extracellular matrix.
Conclusion
In conclusion, we demonstrate for the first time that inhalation of a combined selective PDE3/4 inhibitor reverses pulmonary hypertension fully developed in response to monocrotaline in rats. This "reverse-remodeling" effect was true for hemodynamics, structural changes to the lung vascular wall, and key molecular pathways of matrix regulation. We provide evidence that inhibition of pulmonary artery SMC migration and MMP-based matrix regulation play a major role in the beneficial effect of inhaled tolafentrine in severe PAH. Reversal of structural lung vascular remodeling may apparently be achieved in pulmonary hypertension, at least in the experimental model of monocrotaline-induced PAH in rats.
Acknowledgements
This work was supported by the Deutsche Forschungsgemeinschaft (SFB 547), Project C6.
==== Refs
D'Alonzo GE Barst RJ Ayres SM Bergofsky EH Brundage BH Detre KM Fishman AP Goldring RM Groves BM Kernis JT Survival in patients with primary pulmonary hypertension. Results from a national prospective registry Ann Intern Med 1991 115 343 349 1863023
Archer S Rich S Primary pulmonary hypertension: a vascular biology and translational research "Work in progress" Circulation 2000 102 2781 2791 11094047
Bendeck MP Zempo N Clowes AW Galardy RE Reidy MA Smooth muscle cell migration and matrix metalloproteinase expression after arterial injury in the rat Circ Res 1994 75 539 545 8062427
Strauss BH Robinson R Batchelor WB Chisholm RJ Ravi G Natarajan MK Logan RA Mehta SR Levy DE Ezrin AM Keeley FW In vivo collagen turnover following experimental balloon angioplasty injury and the role of matrix metalloproteinases Circ Res 1996 79 541 550 8781487
Frisdal E Gest V Vieillard-Baron A Levame M Lepetit H Eddahibi S Lafuma C Harf A Adnot S Dortho MP Gelatinase expression in pulmonary arteries during experimental pulmonary hypertension Eur Respir J 2001 18 838 845 11757635 10.1183/09031936.01.00084601
Qin Y Zhou A Ben X Shen J Liang Y Li F All-trans retinoic acid in pulmonary vascular structural remodeling in rats with pulmonary hypertension induced by monocrotaline Chin Med J (Engl) 2001 114 462 465 11780404
Cowan KN Heilbut A Humpl T Lam C Ito S Rabinovitch M Complete reversal of fatal pulmonary hypertension in rats by a serine elastase inhibitor Nat Med 2000 6 698 702 10835689 10.1038/76282
Zaidi SH You XM Ciura S Husain M Rabinovitch M Overexpression of the serine elastase inhibitor elafin protects transgenic mice from hypoxic pulmonary hypertension Circulation 2002 105 516 521 11815437 10.1161/hc0402.102866
Giannelli G Iannone F Marinosci F Lapadula G Antonaci S The effect of bosentan on matrix metalloproteinase-9 levels in patients with systemic sclerosis-induced pulmonary hypertension Curr Med Res Opin 2005 21 327 332 15811199 10.1185/030079905X30680
Schermuly RT Kreisselmeier KP Ghofrani HA Yilmaz H Butrous G Ermert L Ermert M Weissmann N Rose F Guenther A Walmrath D Seeger W Grimminger F Chronic sildenafil treatment inhibits monocrotaline-induced pulmonary hypertension in rats Am J Respir Crit Care Med 2004 169 39 45 12958054 10.1164/rccm.200302-282OC
Schermuly RT Kreisselmeier KP Ghofrani HA Samidurai A Pullamsetti S Weissmann N Schudt C Ermert L Seeger W Grimminger F Antiremodeling effects of iloprost and the dual-selective phosphodiesterase 3/4 inhibitor tolafentrine in chronic experimental pulmonary hypertension Circ Res 2004 94 1101 1108 15031263 10.1161/01.RES.0000126050.41296.8E
Beavo JA Cyclic nucleotide phosphodiesterases: functional implications of multiple isoforms Physiol Rev 1995 75 725 748 7480160
Manganiello VC Murata T Taira M Belfrage P Degerman E Diversity in cyclic nucleotide phosphodiesterase isoenzyme families Arch Biochem Biophys 1995 322 1 13 7574662 10.1006/abbi.1995.1429
Ghofrani HA Pepke-Zaba J Barbera JA Channick R Keogh AM Gomez-Sanchez MA Kneussl M Grimminger F Nitric oxide pathway and phosphodiesterase inhibitors in pulmonary arterial hypertension J Am Coll Cardiol 2004 43 68S 72S 15194181 10.1016/j.jacc.2004.02.031
Ghofrani HA Pepke-Zaba J Barbera JA Channick R Keogh AM Gomez-Sanchez MA Kneussl M Grimminger F Nitric oxide pathway and phosphodiesterase inhibitors in pulmonary arterial hypertension J Am Coll Cardiol 2004 43 68S 72S 15194181 10.1016/j.jacc.2004.02.031
Schermuly RT Roehl A Weissmann N Ghofrani HA Schudt C Tenor H Grimminger F Seeger W Walmrath D Subthreshold doses of specific phosphodiesterase type 3 and 4 inhibitors enhance the pulmonary vasodilatory response to nebulized prostacyclin with improvement in gas exchange J Pharmacol Exp Ther 2000 292 512 520 10640287
Schermuly RT Krupnik E Tenor H Schudt C Weissmann N Rose F Grimminger F Seeger W Walmrath D Ghofrani HA Coaerosolization of phosphodiesterase inhibitors markedly enhances the pulmonary vasodilatory response to inhaled iloprost in experimental pulmonary hypertension. Maintenance of lung selectivity Am J Respir Crit Care Med 2001 164 1694 1700 11719312
Ghofrani HA Rose F Schermuly RT Olschewski H Wiedemann R Weissmann N Schudt C Tenor H Seeger W Grimminger F Amplification of the pulmonary vasodilatory response to inhaled iloprost by subthreshold phosphodiesterase types 3 and 4 inhibition in severe pulmonary hypertension Crit Care Med 2002 30 2489 2492 12441759 10.1097/00003246-200211000-00014
Huxtable RJ Activation and pulmonary toxicity of pyrrolizidine alkaloids Pharmacol Ther 1990 47 371 389 2290854 10.1016/0163-7258(90)90063-8
Rosenberg HC Rabinovitch M Endothelial injury and vascular reactivity in monocrotaline pulmonary hypertension Am J Physiol 1988 255 H1484 H1491 3144186
Kato S Sugimura H Kishiro I Machida M Suzuki H Kaneko N Suppressive effect of pulmonary hypertension and leukocyte activation by inhaled prostaglandin E1 in rats with monocrotaline-induced pulmonary hypertension Exp Lung Res 2002 28 265 273 12042029 10.1080/01902140252964357
Nagaya N Okumura H Uematsu M Shimizu W Ono F Shirai M Mori H Miyatake K Kangawa K Repeated inhalation of adrenomedullin ameliorates pulmonary hypertension and survival in monocrotaline rats Am J Physiol Heart Circ Physiol 2003 285 H2125 H2131 14561683
Schermuly RT Yilmaz H Ghofrani HA Woyda K Pullamsetti S Schulz A Gessler T Dumitrascu R Weissmann N Grimminger F Seeger W Inhaled Iloprost Reverses Vascular Remodeling in Chronic Experimental Pulmonary Hypertension Am J Respir Crit Care Med 2005 172 358 363 15879421 10.1164/rccm.200502-296OC
Schermuly R Schmehl T Gunther A Grimminger F Seeger W Walmrath D Ultrasonic nebulization for efficient delivery of surfactant in a model of acute lung injury. Impact on gas exchange Am J Respir Crit Care Med 1997 156 445 453 9279222
Schermuly RT Ghofrani HA Enke B Weissmann N Grimminger F Seeger W Schudt C Walmrath D Low-dose systemic phosphodiesterase inhibitors amplify the pulmonary vasodilatory response to inhaled prostacyclin in experimental pulmonary hypertension Am J Respir Crit Care Med 1999 160 1500 1506 10556112
Fink L Seeger W Ermert L Hanze J Stahl U Grimminger F Kummer W Bohle RM Real-time quantitative RT-PCR after laser-assisted cell picking Nat Med 1998 4 1329 1333 9809560 10.1038/3327
Irani C Goncharova EA Hunter DS Walker CL Panettieri RA Krymskaya VP Phosphatidylinositol 3-kinase but not tuberin is required for PDGF-induced cell migration Am J Physiol Lung Cell Mol Physiol 2002 282 L854 L862 11880313
Olschewski H Olschewski A Rose F Schermuly R Schutte H Weissmann N Seeger W Grimminger F Physiologic basis for the treatment of pulmonary hypertension J Lab Clin Med 2001 138 287 297 11709653 10.1067/mlc.2001.119329
Olschewski H Rose F Grunig E Ghofrani HA Walmrath D Schulz R Schermuly R Grimminger F Seeger W Cellular pathophysiology and therapy of pulmonary hypertension J Lab Clin Med 2001 138 367 377 11753283 10.1067/mlc.2001.119285
Wharton J Davie N Upton PD Yacoub MH Polak JM Morrell NW Prostacyclin analogues differentially inhibit growth of distal and proximal human pulmonary artery smooth muscle cells Circulation 2000 102 3130 3136 11120706
Phillips PG Long L Wilkins MR Morrell NW cAMP phosphodiesterase inhibitors potentiate effects of prostacyclin analogs in hypoxic pulmonary vascular remodeling Am J Physiol Lung Cell Mol Physiol 2005 288 L103 L115 15377497 10.1152/ajplung.00095.2004
Hsieh CC Lau Y Migration of vascular smooth muscle cells is enhanced in cultures derived from spontaneously hypertensive rat Pflugers Arch 1998 435 286 292 9382944 10.1007/s004240050514
Pauly RR Passaniti A Bilato C Monticone R Cheng L Papadopoulos N Gluzband YA Smith L Weinstein C Lakatta EG Migration of cultured vascular smooth muscle cells through a basement membrane barrier requires type IV collagenase activity and is inhibited by cellular differentiation Circ Res 1994 75 41 54 8013081
Johnson JL van Eys GJ Angelini GD George SJ Injury induces dedifferentiation of smooth muscle cells and increased matrix-degrading metalloproteinase activity in human saphenous vein Arterioscler Thromb Vasc Biol 2001 21 1146 1151 11451743
Yang GY Yao JS Huey M Hashimoto T Young WL Participation of PI3K and ERK1/2 pathways are required for human brain vascular smooth muscle cell migration Neurochem Int 2004 44 441 446 14687609 10.1016/j.neuint.2003.07.002
Horio T Kohno M Kano H Ikeda M Yasunari K Yokokawa K Minami M Takeda T Adrenomedullin as a novel antimigration factor of vascular smooth muscle cells Circ Res 1995 77 660 664 7554110
Graves LM Bornfeldt KE Raines EW Potts BC Macdonald SG Ross R Krebs EG Protein kinase A antagonizes platelet-derived growth factor-induced signaling by mitogen-activated protein kinase in human arterial smooth muscle cells Proc Natl Acad Sci U S A 1993 90 10300 10304 7694289
Bornfeldt KE Raines EW Nakano T Graves LM Krebs EG Ross R Insulin-like growth factor-I and platelet-derived growth factor-BB induce directed migration of human arterial smooth muscle cells via signaling pathways that are distinct from those of proliferation J Clin Invest 1994 93 1266 1274 8132765
Rabinovitch M Investigational approaches to pulmonary hypertension Toxicol Pathol 1991 19 458 469 1813989
Tozzi CA Poiani GJ Harangozo AM Boyd CD Riley DJ Pressure-induced connective tissue synthesis in pulmonary artery segments is dependent on intact endothelium J Clin Invest 1989 84 1005 1012 2668338
McCawley LJ Li S Benavidez M Halbleib J Wattenberg EV Hudson LG Elevation of intracellular cAMP inhibits growth factor-mediated matrix metalloproteinase-9 induction and keratinocyte migration Mol Pharmacol 2000 58 145 151 10860936
Peracchia F Tamburro A Prontera C Mariani B Rotilio D cAMP involvement in the expression of MMP-2 and MT-MMP1 metalloproteinases in human endothelial cells Arterioscler Thromb Vasc Biol 1997 17 3185 3190 9409309
Takahashi N Tetsuka T Uranishi H Okamoto T Inhibition of the NF-kappaB transcriptional activity by protein kinase A Eur J Biochem 2002 269 4559 4565 12230568 10.1046/j.1432-1033.2002.03157.x
Yoshimura S Morishita R Hayashi K Yamamoto K Nakagami H Kaneda Y Sakai N Ogihara T Inhibition of intimal hyperplasia after balloon injury in rat carotid artery model using cis-element 'decoy' of nuclear factor-kappaB binding site as a novel molecular strategy Gene Ther 2001 8 1635 1642 11895002 10.1038/sj.gt.3301566
|
16262900
|
PMC1291406
|
CC BY
|
2021-01-04 16:36:25
|
no
|
Respir Res. 2005 Nov 1; 6(1):128
|
utf-8
|
Respir Res
| 2,005 |
10.1186/1465-9921-6-128
|
oa_comm
|
==== Front
Respir ResRespiratory Research1465-99211465-993XBioMed Central London 1465-9921-6-1291626642810.1186/1465-9921-6-129ResearchThe crucial role of particle surface reactivity in respirable quartz-induced reactive oxygen/nitrogen species formation and APE/Ref-1 induction in rat lung Albrecht Catrin [email protected] Ad M [email protected] Andrea [email protected]öhr Doris [email protected] Petra [email protected] Schooten Frederik J [email protected] Paul JA [email protected] Roel PF [email protected] Institut für Umweltmedizinische Forschung (IUF) at the Heinrich-Heine-University Düsseldorf, Germany2 Nutrition and Toxicology Research Institute Maastricht (NUTRIM), Department of Health Risk Analysis and Toxicology, University of Maastricht, The Netherlands2005 2 11 2005 6 1 129 129 21 7 2005 2 11 2005 Copyright © 2005 Albrecht et al; licensee BioMed Central Ltd.2005Albrecht et al; licensee BioMed Central Ltd.This is an Open Access article distributed under the terms of the Creative Commons Attribution License (), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
Persistent inflammation and associated excessive oxidative stress have been crucially implicated in quartz-induced pulmonary diseases, including fibrosis and cancer. We have investigated the significance of the particle surface reactivity of respirable quartz dust in relation to the in vivo generation of reactive oxygen and nitrogen species (ROS/RNS) and the associated induction of oxidative stress responses in the lung. Therefore, rats were intratracheally instilled with 2 mg quartz (DQ12) or quartz whose surface was modified by either polyvinylpyridine-N-oxide (PVNO) or aluminium lactate (AL). Seven days after instillation, the bronchoalveolar lavage fluid (BALF) was analysed for markers of inflammation (total/differential cell counts), levels of pulmonary oxidants (H2O2, nitrite), antioxidant status (trolox equivalent antioxidant capacity), as well as for markers of lung tissue damage, e.g. total protein, lactate dehydrogenase and alkaline phosphatase. Lung homogenates as well as sections were investigated regarding the induction of the oxidative DNA-lesion/oxidative stress marker 8-hydroxy-2'-deoxyguanosine (8-OHdG) using HPLC/ECD analysis and immunohistochemistry, respectively. Homogenates and sections were also investigated for the expression of the bifunctional apurinic/apyrimidinic endonuclease/redox factor-1 (APE/Ref-1) by Western blotting and immunohistochemistry. Significantly increased levels of H2O2 and nitrite were observed in rats treated with non-coated quartz, when compared to rats that were treated with either saline or the surface-modified quartz preparations. In the BALF, there was a strong correlation between the number of macrophages and ROS, as well as total cells and RNS. Although enhanced oxidant generation in non-coated DQ12-treated rats was paralleled with an increased total antioxidant capacity in the BALF, these animals also showed significantly enhanced lung tissue damage. Remarkably however, elevated ROS levels were not associated with an increase in 8-OHdG, whereas the lung tissue expression of APE/Ref-1 protein was clearly up-regulated. The present data provide further in vivo evidence for the crucial role of particle surface properties in quartz dust-induced ROS/RNS generation by recruited inflammatory phagocytes. Our results also demonstrate that quartz dust can fail to show steady-state enhanced oxidative DNA damage in the respiratory tract, in conditions were it elicits a marked and persistent inflammation with associated generation of ROS/RNS, and indicate that this may relate to compensatory induction of APE/Ref-1 mediated base excision repair.
==== Body
Background
Worldwide, millions of people are occupationally exposed to crystalline silica (e.g. quartz) dust. Chronic exposure to quartz can lead to a variety of pulmonary diseases, including silicosis and cancer [1]. Notably however, studies on quartz-induced carcinogenicity have revealed that the quartz hazard is a variable entity [2], as carcinogenic outcomes seem to be inconsistent and show a rather large variation [1]. Indeed, the toxicity of quartz is highly variable and has been demonstrated to largely depend on the reactivity of its particle surface [3]. Currently, there is a large body of experimental evidence showing that modification of the particle surface by either grinding or coating with PVNO or aluminium salts modifies quartz-induced cytotoxicity, genotoxicity, inflammogenicity and fibrogenicity [4-11].
It is now generally accepted that excessive and persistent formation of ROS and RNS plays a major role in quartz-induced silicosis and carcinogenicity [3,12-14]. During quartz exposure, ROS may be generated via two major routes: either from the quartz particles themselves, or indirectly, from the oxidative burst of pulmonary inflammatory cells (i.e. neutrophils and macrophages) that invade the lung upon exposure to quartz. Previously, we and others have demonstrated that the surface characteristics of quartz are involved in both of these pathways, since surface-modification significantly impacts on the generation of ROS by quartz particles in an acellular environment [6,15], as well as on the induction and persistence of pulmonary inflammation [4,7,9,11]. It has been also demonstrated that such quartz-surface modifications directly modify the release of ROS from neutrophils and macrophages upon in vitro treatment with quartz [6,16-18]. Notably however, current evidence for a role of the reactive particle surface on the actual generation of ROS in vivo and the oxidative stress response in lung tissue are merely associative.
In the context of the inflammation-mediated carcinogenic effects of quartz, it should be noted that ROS are on the one hand known to activate redox-sensitive signal transduction cascades, such as nuclear factor kappa B (NFκB) and activator protein (AP-1), which are involved in activation of genes controlling inflammation, proliferation and/or apoptosis [11]. On the other hand, quartz-mediated formation of ROS is considered to drive oxidative DNA damage and associated mutagenesis [13,19]. The importance of inflammation-mediated ROS for quartz mutagenesis was initially provided by Driscoll and co-workers [20]. Using a complementary in vivo and ex vivo/in vitro approach, they showed (a) that quartz particles are mutagenic to rat lung epithelial cells in vivo in association with a persistent pulmonary inflammation, and (b) that BAL cells obtained from such quartz-treated rats are mutagenic to rat lung epithelial cells in vitro. In concordance with these observations, quartz has been shown to induce the premutagenic oxidative DNA adduct 8-OHdG in rat lungs [21,22]. In an in vitro co-incubation model we could then demonstrate that the induction of 8-OHdG in lung epithelial cells by neutrophils can be blocked by antioxidants [23].
To cope with exogenous DNA damage, e.g. as may be induced upon quartz exposure, cells are equipped with multiple DNA repair enzymes. The repair of oxidative DNA lesions such as 8-OHdG, predominantly occurs via the base excision repair (BER) pathway. As such, altered expression of BER enzymes has been proposed as a sensitive marker of the induction of oxidative stress and oxidative DNA damage [24,25]. Among these repair proteins, the expression of the bifunctional protein APE/Ref-1 represents a highly interesting candidate [25]. The protein APE/Ref-1 consists of two functionally independent domains, the highly conserved C-terminus, involved in both the short-patch and long-path pathways of BER, and the completely unconserved N-terminus, which exerts the control of several redox-sensitive transcription factors including NFκB and AP-1. Previously, in vitro studies have shown that asbestos fibres enhance APE/Ref-1 expression in mesothelial cells as well as in alveolar macrophages, which has been linked to its role in oxidative DNA damage repair and ROS-mediated regulation of redox-sensitive transcription pathways, respectively [26,27]. So far, however, the role of quartz-elicited ROS generation on APE/Ref-1 expression in vivo has not been elucidated.
The aim of our present study was to investigate whether inhibition of the surface reactivity of quartz particles modulates inflammation-mediated generation of ROS and RNS in the rat lung in vivo, and whether this impacts on pulmonary toxicity and more specifically, on the expression of the lung tissue sensors of oxidative stress/DNA damage 8-OHdG and APE/Ref-1. Therefore, BAL as well as lung tissue from rat lungs were analysed for pulmonary toxicity, inflammation, ROS/RNS generation and induction of 8-OHdG and APE/Ref-1, seven days after a single instillation with native quartz or quartz from which the surface was coated with either PVNO or AL.
Methods
Chemicals
2-2'-azinobis-(-3 ethylbenzothiazoline-6-sulphonate) (ABTS), dimethyl sulphoxide (DMSO), ethidium bromide, L-glutamine, Ham's F12 medium, Hank's balanced salt solution (HBSS), HEPES buffer, fetal calf serum (FCS), penicillin/streptavidin solution, phosphate buffered saline (PBS), were all obtained from Sigma (Germany). Horseradish peroxidase (HRP), guaiacol, phorbol-12-myristate-13-acetate (PMA), anti-mouse-IgG whole protein HRP-conjugated secondary antibody and the tubulin antibody as well as Diaminobenzidine were also purchased from Sigma (Germany). ABAP (2,2'-azobis-(-2-amidinopropane)HCl was from Polysciences, Warrington, USA. F12-K Nutrient Mixture was obtained from Invitrogen (Germany). Protease inhibitors in form of a Complete™ cocktail were purchased from Roche Diagnostics GmbH (Germany). The Bradford-protein assay was used from BioRad (Germany). ECL-reagent/detection system was obtained from Amersham Bioscience (Germany). The antibody against 8-OHdG was obtained from the Institute of Aging (Japan) and the antibody against APE/Ref-1 (C-4) was purchased from Santa Cruz Biotechnology (USA). For immunohistochemical detection secondary biotinylated horse-anti mouse antibody, the streptavidin-biotin-system (Vectastain Elite Kit) and mouse IgG were used from Vector Laboratories (USA). DePex was used from Serva (Germany). Hoechst 33342 was obtained from Sigma (Germany), and MFP488 goat anti-mouse IgG from MoBiTec (Germany). All other chemicals were from Merck (Germany) and were of highest purity.
Surface treatment of quartz
Surface modification of Quartz (DQ12, batch 6, IUF, Düsseldorf) was performed as described previously [10]. Briefly, DQ12 was coated for 5.5 hours in a 5 mg/ml suspension in a 1% dilution of either PVNO or AL, dissolved in distilled deionised sterile water. Non-coated DQ12 was suspended in water without any further additions. After repetitive washings in sterile water, the particles were finally resuspended in sterile water at a concentration of 5 mg/ml in sterile glass tubes, and allowed to dry under a laminar flow chamber. All quartz processing was performed under sterile conditions, and a single batch of non-coated and coated quartz was prepared and used for the whole study to avoid possible variable coating efficiency. Atomic absorption spectrometry and spectrophotometry revealed coating efficiencies of respectively 11 μg PVNO/mg quartz and 1.6 μg aluminium/mg quartz, and transmission electron microscopy analysis showed that the coating procedures did not cause changes in particle size distribution or aggregation of the DQ12 [11]. For intratracheal (i.t.) instillation, the dried quartz preparations were resuspended in 1 ml of PBS (without Mg++ and Ca++) and sonicated in a water bath (Sonorex TK52; 60 Watt, 35 kHz, 5 min).
Quartz instillation and bronchoalveolar lavage
Specific pathogen free female Wistar rats were used for the study (Janvier, Le Genest St. Isle, France). The animals were housed and maintained in an accredited on-site testing facility, according to the guidelines of the Society for Laboratory Animals Science (GV-SOLAS). Food and water were available ad libitum. When weighing 200–250 g (8 weeks old), animals were anaesthetised with isoflurane and i.t. instillation was performed using a laryngoscope. From the non-coated or coated quartz suspensions (5 mg/ml in PBS) 400 μl were instilled giving a final dose of 2 mg per rat (n = 5 per treatment and endpoint). Control rats were instilled with only PBS. Separate control animals received 22 μg PVNO or 35 μg AL (in PBS), amounts calculated from the coating efficiency of both substances, to assess possible direct effects of coating materials. After 7 days, animals were sacrificed by a single i.p. injection of Na-pentobarbital and subsequent exsanguination via the abdominal aorta. The lung was cannulated via the trachea and BAL was performed in situ by infusing the lungs with 5 ml aliquots of PBS. The BALF was drained passively by gravity and the procedure was repeated four times, giving a total BAL volume of 20 ml. Total cell number in the BAL was analysed using a hemocytometer chamber (Neubauer) and viability was assessed by trypan blue dye exclusion. BAL-cell differential was determined on cytospin preparations stained with May-Grünwald/Giemsa (MGG). The BALF was centrifuged twice (300 g to collect cells, followed by 1000 g to obtain BALF), and the cell-free supernatant was analysed for lung injury parameters, e.g. total protein, LDH and AP, as well as myeloperoxidase (MPO).
Measurement of cytotoxic and inflammogenic bronchoalveolar lavage parameters
Total protein was analysed according to the method described by Lowry. LDH and AP were assayed using diagnostic kits from Merck (Germany). MPO activity in the BALF was assayed according to the method originally described by Klebanoff et al [28]. Briefly, 200 μl of cell-free BALF was mixed with 800 μl MPO assay solution, containing 107.6 ml H2O, 12 ml 0.1 M sodium phosphate buffer, 0.192 ml Guaiacol, 0.4 ml 0.1 M H2O2. The generation of tetra-guaiacol was measured spectrophotometrically (Beckman) at 470 nm and the change of optical density per minute was calculated from the initial rate. The MPO activity was calculated from the formula: U/ml = ΔOD/minute × 0.752 and expressed as mU/ml. One unit of the enzyme is defined as the amount that consumes 1 μmol H2O2 per minute.
Measurement of hydrogen peroxide and nitrite in bronchoalveolar lavage fluid
Freshly obtained BALF was used to measure hydrogen peroxide according to the method of Gallati and Pracht [29]. Therefore, 75 μl of BALF was mixed with 75 μl of a 3,3',5,5'-Tetramethylbenzidine solution (TMB solution EIA, solution A, Bio-Rad), containing 8.5 U/ml horseradish peroxidase. After 10 min incubation at RT, 50 μl H2SO4 (1 M) was added and absorption was measured at 450 nm using a microplate reader (Multiskan, Labsystems). The final concentration of H2O2 was calculated from a standard curve made up in BALF obtained from an untreated rat.
Nitric oxide levels in BALF were determined by analysis of its relative stable metabolite nitrite using the Griess reaction. Briefly, 100 μl of the cell free BALF was incubated with an equal volume of Griess reagent (0.1% naphthylene ethylene-diamide.2HCl, 1% sulfanilamide, 2.5% phosphoric acid) at room temperature for 10 minutes. Absorbance (540 nm) was then determined using a microplate reader and concentrations were calculated from a NaNO2 standard curve.
Measurement of ex vivo hydrogen peroxide by bronchoalveolar lavage cells
Freshly isolated BAL cells obtained from rats exposed to the quartz preparations were used to determine spontaneous and PMA-induced ex vivo H2O2 release. H2O2 generation was measured as described by Pick and Keisari [30]. Therefore, 1.5 × 105 cells were incubated in 24 well plates in 500 μl HBSS containing 8.5 U/ml horseradish peroxidase (HRP) and 0.28 mM phenol red (PRS solution). Cells were incubated with or without PMA (100 ng/ml) for 4h at 37°C, 5% CO2. The reaction was stopped by addition of 10 μl NaOH (1 M), and absorption was read at 610 nm, against a standard curve of H2O2 in PRS solution.
Trolox equivalent antioxidant capacity assay
The TEAC (trolox equivalent antioxidant capacity) assay was performed according to Van den Berg et al. [31], with minor modifications. An ABTS (2-2'-azinobis-(-3 ethylbenzothiazoline-6-sulphonate) radical solution was prepared by mixing 2.5 mM ABAP (2,2'-azobis-(-2-amidinopropane)HCl with 20 mM ABTS solution in 150 mM phosphate buffer (pH 7.4) containing 150 mM NaCl. The solution was heated for 10 min at 70°C and, if necessary, diluted to obtain a solution with an absorbance at 734 nm between 0.68 and 0.72. For measuring antioxidant capacity 100 μl of the cell-free BALF was mixed with 900 μl of the ABTS radical solution. Both native and deproteinized (10% TCA) BALF were tested. The decrease in absorbance at 734 nm 5 minutes after addition of the sample was used for calculating the TEAC. Trolox was used as reference compound. The TEAC of the sample is given as the concentration of a trolox solution that gives a similar reduction of the absorbance at 734 nm.
DNA isolation and analysis of 8-hydroxy-2'-deoxyguanosine by HPLC/ECD
Lung tissue was removed from the animals, chopped into small pieces, aliquots were snap frozen in liquid nitrogen and stored at -80°C until later measurement of 8-OHdG as described previously [23]. Briefly, lung tissue was homogenated and lysed overnight at 37°C in a NEP/SDS solution (75 mM NaCl, 25 mM EDTA, 50 μg/ml proteinase K, 1% SDS). The DNA was dissolved in 5 mM Tris-HCl (pH 7.4) at a final concentration of 0.5 mg/ml. 8-OHdG formation was measured using high performance liquid chromatography with electrochemical detection (HPLC-ECD). Values were expressed as the ratio of 8-OHdG to deoxyguanosine (dG).
Western blotting
Lung tissue was removed from the animals, chopped into small pieces, aliquots were snap frozen in liquid nitrogen and stored at -80°C. For preparation of whole protein, lung tissue was homogenised within lysis buffer (1% NP-40, 0.5% sodium deoxycholate, 0.1% SDS in PBS) containing freshly added protease inhibitors. Homogenate-lysis buffer-mix was incubated for 30 min on ice and spun at 15.000 g for 20 min at 4°C. Protein concentrations were determined by BioRad-Assay (according to the Bradford method). Samples were electrophorezed at equal protein concentrations (10 μg) in 10% SDS-polyacrylamide gels, and transferred onto nitrocellulose membranes. Non-specific protein binding was blocked with 5% dried milk powder and 0.05% Tween-20 in PBS. Detection of the APE/Ref-1 protein was performed using a monoclonal antibody (1:2000) and an anti-mouse-IgG whole protein HRP-conjugated secondary antibody (1:5000). Blots were reprobed with an antibody against tubulin (1:5000) and a secondary anti-mouse-IgG whole protein HRP-conjugated antibody (1:5000) for protein normalisation. Band formation was visualised using an ECL-reagent/detection system. Quantification was performed by computer-assisted densitometry scanning using a documentation system (Bio-Rad, Germany) with appropriate software (Gel-doc, Bio-Rad, Germany). For each time point, samples of 4 animals per treatment group were quantitated.
Lung fixation and immunohistochemistry of 8-OHdG and APE/Ref-1
Lungs of five additional animals per treatment group were instilled in situ with 4% paraformaldehyde/PBS (pH 7.4) under atmospheric pressure, removed, fixed, dehydrated, and paraffin embedded. Serial sections of lungs were mounted on different slides and stained either for 8-OHdG or APE/Ref-1. For the detection of both antibodies basically the same method was applied, except for additional steps of RNA digestion and DNA-denaturation for the detection of 8-OHdG. Since both antibodies are monoclonal mouse antibodies, horse serum was used to block unspecific binding. The sections were then incubated with a primary antibody against 8-OHdG (1:100) or against APE/Ref-1 (1:500). Detection was performed by incubation with a secondary biotinylated horse-anti mouse antibody (1:200) followed by the streptavidin-biotin-complex according to the manufacturer's protocol. Diaminobenzidine (DAB) was used as a substrate, and the slides were counter stained with hematoxylin. After washing with distilled water, slides were dehydrated and covered in DePex. For the negative control, control sections were incubated with mouse IgG instead of the primary antibodies at the same IgG concentrations. Slides were analysed using a light microscope (Olympus BX60).
Quantification of 8-OHdG and Ref-1 staining following immunohistochemistry
For quantification of 8-OHdG or APE/Ref-1 five microscopic areas of the left lung lobe of 4 animals per treatment were randomly selected for analysis at an original magnification of × 1000 (oil immersion). Since the staining for 8-OHdG as well as for APE/Ref-1 predominantly occurred within the cell nucleus, in line with the location of the DNA and the action of its repair, all brown (DAB, positive signal) and blue (hematoxylin, negative signal) stained nuclei were counted and expressed as percentage of total cells. In the lungs of the animals that were treated with the non-coated quartz, we observed specific areas with increased an accumulation of inflammatory cells and early indications of tissue remodelling, likely as a result of the non-uniform lung distribution of the quartz-particles after instillation. Therefore, in this treatment group, quantification of each individual section was performed independently for regions with normal lung architecture and focal pathologically altered regions.
Analysis of expression and subcellular localisation of APE/Ref-1 in representative lung cell lines
In relation to the observed immunohistochemical staining in the lung sections as will be discussed later, APE/Ref-1 protein expression was further evaluated in vitro, using Western blotting in the rat cell lines NR8383 and RLE, representing an alveolar macrophage and type II epithelial cell line, respectively [32,20]. NR8383 cells were cultured in F12-K Nutrient Mixture/15% FCS/penicillin/streptomycin, and RLE cells were cultured in Ham's 12 Mixture/5% FCS/penicillin/ streptomycin. Both cell lines were grown until near confluence and nuclear as well as cytosolic proteins were then prepared by the method of Staal et al. [33]. Briefly, cells were harvested by gentle scraping and then lysed by incubation on ice in Buffer A (10 mM Hepes, 10 mM KCl, 2 mM MgCl2, 1 mM DTT, 0.1 mM EDTA containing freshly added protease inhibitors). After 15 min buffer B was added (Buffer A + 10% NP-40), and lysate was centrifuged at 950 g for 10 min. After collection of the supernatant (cytosolic fraction), the pellet containing cell nuclei was resuspended in buffer C (50 mM Hepes, 50 mM KCl, 300 mM NaCl, 0.1 mM EDTA, 1 mM DTT, 10% glycerol containing freshly added protease inhibitors). This nucleic suspension was incubated on ice by agitation for 20 min, followed by centrifugation at 18,000 g for 10 min. Nucleic proteins from this supernatant were collected and stored like the cytosolic proteins at -80°C until analysis. Before analysis of APE/Ref-1 expression by Western Blotting, protein concentration was determined using the BioRad-Assay (according to the Bradford method) and equal protein amounts of 10 μg were loaded.
For an independent evaluation of the subcellular expression of APE/Ref-1 in the RLE cells, also immunocytochemistry was used, as follows: RLE cells were cultured to near confluence on 4-chamber culture slides (Falcon), and immunocytochemistry was performed using the APE/Ref-1 antibody described before followed by a MFP488 goat anti-mouse IgG antibody. Nuclear counter-staining was performed using Hoechst 33342. Slides were analysed using a fluorescence microscope (Olympus BX60) at an original magnification of × 1000.
Statistical evaluation
Data are expressed as mean ± SD, unless stated otherwise. Statistical analysis was performed using SPSS version 10 for Windows. ANOVA was used to evaluate differences between treatments. Multiple comparisons where evaluated using Tukey's method. A difference was considered to be statistically significant when p < 0.05. Correlation analysis was performed using Pearson's test.
Results
Pulmonary inflammation and toxicity
BAL was used to determine inflammation and toxicity in the rat lungs after the different treatments. Treatment of rats with only 22 μg PVNO or 35 μg AL, amounts calculated from the coating efficiency of both substances, showed no effects on any of the studied BAL parameters (data not shown). However, upon instillation of the three different quartz preparations, the total cell number in the BAL was found to be significantly increased only with the non-coated DQ12 (p < 0.001 vs. control, Table 1). The increased cell number as observed with the native quartz was also reflected by an increase in the total number of alveolar macrophages (p < 0.001 vs. control) as well as by the neutrophil number (p < 0.001 vs. control). Analysis of the percentage of neutrophils, revealed a significant increase following treatment with non-coated DQ12 exposure (p < 0.001 vs. control), as well as following treatment with AL-coated DQ12 exposure (p < 0.001 vs. control), but not with PVNO-coated DQ12. However, compared to the non-coated quartz, both coated preparations showed a significantly lower neutrophil percentage (PVNO: p < 0.001, AL: p < 0.01).
Table 1 Markers of inflammation and toxicity in bronchoalveolar lavage
PBS DQ12 DQ12+PVNO DQ12+AL
Total cells (× 106) 1.2 ± 0.4 14.0 ± 3.8*** 1.6 ± 0.5### 3.7 ± 1.2###
Total AM (× 106) 0.8 ± 0.1 2.8 ± 0.9*** 1.3 ± 0.5## 1.7 ± 0.3#
Total PMN(× 106) 0.008 ± 0.008 9.1 ± 3.1*** 0.1 ± 0.1### 1.4 ± 0.5###
PMN (%) 0.8 ± 0.5 64.9 ± 8.7*** 6.7 ± 6.6### 37.1 ± 4.6*** ##
TP (μg/ml) 21.35 ± 8.61 76.1 ± 56.05 23.7 ± 7.94 31.23 ± 14.06
LDH (U/l) 12.2 ± 4.5 140.3 ± 38.2*** 16.2 ± 5.2### 36.8 ± 12.1###
AP (U/l) 16.53 ± 2.09 22.15 ± 2.50* 12.82 ± 0.68### 15.25 ± 2.04##
Significant differences of the particle instilled animals vs. PBS controls are shown by *** p < 0.001; ** p < 0.01; * p < 0.05. Significant differences of the surface-modified quartzes vs. native quartz are shown by ### p < 0.001; ## p < 0.01; # p < 0.05.
Total protein, LDH and AP were analysed in the BALF to evaluate pulmonary toxicity. None of the treatments showed a significant increase in total protein, although the levels tended to be higher upon treatment with the non-coated quartz. In contrast, quartz-treatment caused a significant increase in LDH (p < 0.001), which could be blocked by both coatings (p < 0.001 compared to DQ12). Similarly, the BALF level of the epithelial toxicity marker AP, which was found to be significantly enhanced by non-coated DQ12 (p < 0.05 compared to control), was found to be reduced by both coatings (PVNO: p < 0.001, AL: p < 0.01).
The activity of MPO was determined in BALF to further evaluate the effect of the different particle preparations on neutrophilic inflammation. MPO activity was found to be significantly increased following exposure to the non-coated DQ12 (p < 0.001 compared to control). Coating with PVNO or AL were both able to inhibit this increase (p < 0.001 compared to DQ12, Fig. 1). On a single animal level, covering all treatments, a significant correlation was found between neutrophil numbers and MPO activity (r = 0.639, p < 0.01) confirming the source-specificity of this enzyme.
Figure 1 Myeloperoxidase activity in BALF of rat lungs 7 days following i.t. instillation of 2 mg DQ12 or DQ12 coated with either AL or PVNO. Data are expressed as mean ± SD (n = 5). *p < 0.01 vs. PBS (ANOVA, Tukey).
Formation of ROS/RNS
As a relative stable marker for ROS production in vivo, H2O2 levels were determined in BALF obtained from the differently treated animals. Exposure to the native quartz was found to result in a significant increase in the steady-state H2O2 concentrations (p < 0.05), whereas both coated preparations failed to do so (Fig. 2A). The concentrations of H2O2 in the BALF were significantly related to the total cell numbers (r = 0.59, p < 0.01). More specifically, BALF H2O2 was also correlated with the total number of neutrophils (r = 0.62, p < 0.01, see Fig. 2B), but not with the total number of macrophages in the BAL.
Figure 2 (A) H2O2 generation in the rat lung. H2O2levels were analysed spectrophotometrically in the BALF obtained from rats exposed to non-coated DQ12 or DQ12 coated with AL or PVNO (7 days after i.t. instillation). Data are expressed as mean ± SD (n = 5). *p < 0.01 vs. PBS (ANOVA, Tukey). (B) Correlation between H2O2 levels and total number of neutrophils in the BAL 7 days after i.t. instillation of 2 mg non-coated DQ12 or DQ12 coated with AL or PVNO.
In addition, we determined ex vivo H2O2 generation by BAL cells from the different treatment groups upon PMA activation. Data are shown in Fig. 3 and are expressed as relative increase (%) of H2O2generation in comparison to the cellular H2O2 generation in the absence of PMA stimulation (spontaneous release). PMA-induced increase in H2O2 release was found to be significantly elevated with the BAL cells obtained from rats that were treated with native DQ12 (p < 0.05), but not with the cells obtained from animals exposed to the coated quartz preparations.
Figure 3 Ex vivo release of H2O2 from PMA-stimulated BAL cells. BAL cells, obtained from rats exposed to non-coated DQ12 or DQ12 coated with PVNO or AL (7 days after i.t. instillation) were cultured in vitro (4 h) with or without PMA (100 ng/ml) to activate their oxidative burst. The graph shows the ratio between spontaneous and PMA-induced H2O2 production, which is expressed as % increase. Data are presented as mean ± SD (n = 3). *p < 0.01 vs. PBS.
In order to determine the generation of nitric oxide in rat lungs following the different particle treatments, levels of its relative stable metabolite nitrite were determined in BALF using the Griess-assay. Animals that were treated with the non-coated DQ12 sample, showed significantly higher BALF nitrite levels indicative of NO production (p < 0.05 compared to the controls), whereas both surface-modified samples did not show any difference from controls (Fig. 4). A significant correlation was found between nitrite levels and the total number of cells in the BAL (r = 0.478, p < 0.05). The correlations between nitrite and total number of macrophages or total number of neutrophils did not reach statistical significance.
Figure 4 Nitrite levels as detected in the BALF obtained from rats exposed to non-coated DQ12, or DQ12 coated with AL or PVNO (7 days after i.t. instillation). Data are expressed as mean ± SD (n = 5). *p < 0.05 vs. PBS.
Total non-enzymatic antioxidant capacity
The TEAC assay was used to determine changes in the total non-enzymatic antioxidant capacity of the BALF. Compared with the lavage fluids from the PBS-treated rats, TEAC levels were significantly increased in the BALF from rats treated with native quartz (p < 0.05, Fig. 5), whereas no significant increase could be observed in the lavage fluids from rats exposed to DQ12 from which the surface was coated with either AL or PVNO. No differences in antioxidant capacity was found in the deproteinized BALF (data not shown), suggesting that all the antioxidant capacity was contained within the protein fraction of the BALF.
Figure 5 Non-enzymatic total antioxidant capacity (TEAC) of BALF obtained from rats 7 days after i.t. instillation of non-coated DQ12 or DQ12 coated with AL or PVNO. Data are presented as mean ± SD (n = 5). *p < 0.01.
Determination of the oxidative stress-induced DNA lesion 8-OHdG in lung tissue
DNA of whole lung homogenate was investigated for the premutagenic DNA adduct and established oxidative stress marker 8-OHdG by HPLC/ECD [21]. Results of this analysis are shown in Fig. 6. As can be seen in the figure, no enhanced 8-OHdG/dG ratios were observed in the lung homogenates from the animals that were treated with native quartz, whereas surprisingly, these ratios tended to be higher in the lung homogenate DNA from the rats that were treated with the coated quartz preparations. In fact, this increase reached a statistical significance for the PVNO-coated quartz (p < 0.05, Fig. 6).
Figure 6 8-OHdG analysis by HPLC/ECD in lung tissue, obtained from rats exposed to 2 mg non-coated DQ12 or DQ12 coated with PVNO or AL (7 days after i.t. instillation). Data are presented as mean ± SD (n = 5). *p < 0.01 vs. PBS.
Using an alternative assay, 8-OHdG was also investigated immunohistochemically in the lung tissue sections. Qualitative visual examination of the staining signal intensity, which appeared to be of a distinct nuclear appearance, did not show any differences between the experimental groups (Fig. 7A–D). Subsequent quantification of the proportion of positive stained nuclei from randomly analysed sections also did not show any difference between the treatments (Fig. 7E). In the animals treated with the non-coated DQ12 multiple focal lesions were observed (Fig. 7B). In order to evaluate whether cellular aggregation might have influenced the results, we performed further analysis in this treatment group, by differentiation between focal and non-focal regions. However, this quantification of 8-OHdG staining did not show any difference between the focal and non-focal regions of this treatment group (Fig. 7E).
Figure 7 (A-D) Representative images of lung sections, obtained from controls (A) or rats exposed to 2 mg non-coated DQ12 (B) or DQ12 coated with PVNO (C) or AL (D), 7 days after i.t. instillation, stained with an antibody against 8-OHdG (original magnification × 400, original magnification of inserts × 1000). E: Positive cells were quantified in five random chosen areas (n = 4) at a magnification of × 1000.
Determination of APE/Ref-1 in lung tissue
Whole lung homogenates of the experimental animals were investigated for the expression of the bifunctional APE/Ref-1 protein by Western blotting. Fig. 8 demonstrates the results of densitometry analysis of the APE/Ref-1 expression of 4 animals per treatment. An increase of APE/Ref-1 protein expression was detected in the group instilled with non-coated quartz compared to the control (p < 0.05). Surface modification by PVNO as well as by AL did not show any difference to the control animals.
Figure 8 Semi-quantitative analysis of APE/Ref-1 Western blots of lung homogenates of 5 animals per group exposed to PBS, 2 mg non-coated DQ12, or DQ12 coated with PVNO or AL 7 days after a single i.t. instillation. Data are presented as mean ± SD (n = 4). *p < 0.01 vs. PBS.
To confirm these results, and to assess its cellular localisation, lung tissue sections were also analysed for APE/Ref-1 expression using immunohistochemistry. In fact, serial lung sectioning approach was used were tissues, analysed before for 8-OHdG, were investigated with the APE/Ref-1 antibody (Fig. 9A–D). Immunohistochemical imaging analysis revealed a distinct nuclear staining which contrasted with a very weak cytosolic staining in various cell types. This pattern of cytosolic versus nuclear staining seemed to be similar for all treatment groups including the control animals (Fig. 9A–D). Specifically, clear positive nuclear staining signals could be observed within alveolar macrophages as well as within alveolar epithelial cells. The overall APE/Ref-1 expression was shown to be increased in lung sections of animals that were treated with non-coated DQ12 (Fig. 9B versus 9A). Subsequently, we analysed the proportion of positive nuclei using the same approach as followed for 8-OHdG quantification. This counting analysis revealed a significant increase in the % of APE/Ref-1 stained nuclei, specifically in the focal lesions with accumulated cells of the native quartz-group (Fig. 9E). In contrast, no enhanced APE/Ref-1 signal was found in the lung sections of animals that received PVNO- or AL-coated DQ12 (Fig. 9C, D, E).
Figure 9 (A-D): Representative images of lung sections, obtained from a control rat (A) or rats exposed to 2 mg non-coated DQ12 (B) or DQ12 coated with PVNO (C) or AL (D), 7 days after i.t. instillation, stained with an antibody against APE/Ref-1 (original magnification × 400, original magnification of inserts × 1000. (E) Positive stained cells were quantified in five random areas of the left lung lobe of four animals per group at a magnification of × 1000.
APE/Ref-1 expression in rat alveolar epithelial and macrophage cell lines in vitro
To further validate our in vivo observations concerning the apparent alveolar macrophage and epithelial APE/Ref-1 expression, we comparatively evaluated its expression in vitro in representative rat cell lines, i.e. NR8383 and RLE. Results of Western blotting analysis of both nuclear and cytosolic protein fractions, revealed a strong nuclear accumulation of APE/Ref-1 in the macrophages as well as in the epithelial cells, whereas in both cell lines only a weak distribution in the cytoplasm was found (Fig. 10A). Reprobing of the blots using an antibody against tubulin, a strong cytoplasmic protein, verified that our nuclear fraction had no cytoplasmic impurities (data not shown). As an independent evaluation of the subcellular distribution pattern of APE/Ref-1 we also performed immunocytochemistry. Results for the RLE cells are shown in Fig. 10B. As can be seen in the figure, this analysis confirmed the predominant nuclear appearance of APE/Ref-1 in these cells. As such, these in vitro findings were in concordance with our in vivo observations, concerning the predominant nuclear staining pattern.
Figure 10 (A) Representative image of Western blot investigation of the expression of APE/Ref-1 protein in nuclear and cytosolic extracts of alveolar macrophages (NR8383) as well as rat lung epithelial cells (RLE) cultured in complete medium. (B-C) Immunocytochemical image of (B) APE/Ref-1 staining and (C) Hoechst nuclear counter-staining in RLE cells. Original magnification of × 1000.
Discussion
The data presented in this paper are part of larger ongoing in vitro and in vivo investigations on the role of surface reactivity in quartz-induced genotoxic, proliferative and fibrogenic effects [10,11,15]. Here we report on the effects of surface modification on quartz-induced generation of ROS and RNS in rat lungs in relation to their involvement in the induction of an oxidative stress response (DNA damage, APE/Ref-1 expression) in the lung tissue. Previously, we and others have shown that coating of quartz-particles with PVNO or AL impairs its ability to elicit pulmonary inflammation (i.e. in vivo), as well as the generation of ROS by neutrophils or macrophages in vitro [4,9-11,16,17]. In the present study we showed for the first time, that modification of the quartz-surface with PVNO or AL also abrogates in vivo formation of ROS and RNS in rat lungs.
Our current observation that exposure to a pure quartz sample (DQ12) causes increased pulmonary levels of ROS and RNS in rat, is in line with earlier studies by others using Min-U-Sil quartz [34,35]. Moreover, we found strong relations between total numbers of phagocytes and RNS-levels as well as between total neutrophil numbers and H2O2-levels in the rat lungs in relation to the different particle treatments. Together, these observations contribute to the general opinion on the crucial impact of inflammatory cell-related processes in particle-induced lung diseases [36].
For a clear discussion on the role of phagocytes in quartz-related oxidant generation, a distinction between ROS and RNS should be made. With respect to ROS, the present study has focused on the detection of H2O2, the relative stable dismutation product of superoxide, which is the initial product of the oxidative burst. It has been established that neutrophils are far more potent superoxide-releasing cells than alveolar macrophages [37]. In agreement with this, both PVNO and AL coatings significantly reduced the quartz-induced neutrophil influx as well as H2O2 levels in the BALF, and both parameters were found to be significantly correlated. In contrast to these observations, previously we showed impaired ROS generation from neutrophils upon in vitro treatment with PVNO-coated quartz, but not AL-coated quartz, when compared to treatment with native quartz [10]. This contradiction possibly illustrates that direct particle-cell interactions, as mainly studied using in vitro experiments, are of a minor relevance in determining neutrophilic ROS release in vivo, i.e. within the lung. It also suggests that surface coatings of quartz primarily affect mechanisms regulating the neutrophil influx into the lung, rather than affecting their subsequent activation. The major contribution of neutrophils as a source of pulmonary H2O2 is, however, further illustrated by our current ex vivo experiments, showing that the only BAL cells from non-coated quartz-treated rats, characterised by the highest proportion of neutrophils, could be significantly activated by PMA.
Apart from ROS, RNS are considered to be a major pool of oxidants that contribute to tissue damage during inflammatory processes. The present study has focused on nitrite, as it is a relative stable metabolite of the initial product NO. In our study the positive correlation between total number of macrophages and nitrite failed to reach statistical significance, suggesting the involvement of an additional cellular source of NO in response to silica, such as alveolar type II epithelial cells (35). In general however, alveolar macrophages are known as the major NO-generating cells in the lower lung, and these cells have been shown to produce much more NO than neutrophils [38]. The major role of alveolar macrophages in particle-related NO-production is even better illustrated by studies from Huffman and colleagues [39], who reported that in response to LPS or silica in rats, up to 100% of NOx produced by BAL phagocytes was derived from alveolar macrophages. Furthermore, it has been demonstrated that exposure to quartz results in a clear increase of iNOS mRNA levels in BAL cells [34,40]. Notably, we (data not shown) and others could not detect any in vitro generation of nitrite by quartz-exposed macrophages [41]. Thus it is likely to suggest that the reduction of nitrite levels in the lungs of coated-quartz treated animals mainly results from an inhibited macrophage influx into the lung, rather than from a direct inhibitory effect of coated-quartz on the NO-generation by the macrophages. This also suggests that other factors, including pulmonary cell-cell interactions play a crucial role in activation of NO-release by macrophages per se [41]. This is illustrated by data from Huffman and colleagues [42], who demonstrated that an interaction between macrophages and recruited neutrophils was a crucial factor in the in vivo NO-production upon quartz exposure.
Oxidative stress is defined as a disturbance in the balance between production of ROS/RNS and antioxidant defence, in favour of the former, which causes potential damage [43]. Thus in order to assess oxidative stress in our system, apart from determining ROS/RNS production in vivo, we also evaluated the in vivo antioxidant protection as well as its possibly resulting damage by BAL analysis of toxicity markers and lung tissue induction of 8-OHdG and APE/Ref-1. Silica exposure has been previously demonstrated to cause increased expression and activity of enzymatic antioxidants [44]. In the current study, we applied the TEAC assay to evaluate the effects of quartz on the total non-enzymatic antioxidant capacity of the lung. It was shown that the increase in antioxidant capacity in the lung was most pronounced upon exposure to non-coated quartz, although this was predominantly associated with the protein fraction of the BALF. Nevertheless, in spite of this increased antioxidant protection, clear pulmonary toxicity (i.e. increased LDH and AP levels in BALF) was demonstrated, suggesting an imbalance between generation of ROS/RNS and protective antioxidant pathways. The present data also provide a possible explanation for our earlier observations on the effects of native versus coated quartz on the induction of DNA strand breakage and NFκB-pathway activation in rat lungs [9-11]. In both processes ROS and RNS are considered as crucial mediators of effect [45-47].
To evaluate the impact of the observed in vivo ROS/RNS generation in relation to the different quartz-preparations that were applied, 8-OHdG induction in lung tissue DNA was analysed. 8-OHdG represents the best investigated oxidative and premutagenic DNA lesion, and accordingly it has been forwarded and used as a marker of oxidative stress [21]. Previously, we showed that coating of DQ12 with PNVO or AL completely abrogates 8-OHdG induction in human lung epithelial cells in vitro [15]. In addition, using an in vitro co-incubation model of activated neutrophils and lung epithelial cells, we also demonstrated 8-OHdG induction in epithelial DNA by neutrophil-generated ROS [23]. As such, hypothetically, oxidative DNA damage by quartz in vivo may occur both or either from the surface-reactivity of the quartz particles themselves and/or from phagocyte-derived ROS during inflammation. Importantly however, using two well-established but independent methods (i.e. HPLC/ECD and immunohistochemistry), we did not observe any significant induction of 8-OHdG in our present study by the uncoated, markedly inflammogenic quartz sample. To further complicate this matter, Seiler and colleagues [48] demonstrated, 21 days after a single i.t. instillation of 1.5 mg DQ12, an increase in 8-OHdG immunoreactivity in the DNA of lung alveolar cells, whereas at 3 days after instillation this increase was not detectable. Yamano and colleagues (1995) in contrast showed relatively rapid-increases in 8-OHdG/dG ratios by HPLC/ECD in lung homogenates of rats following i.t. instillation of 10 mg silica. They observed increased 8-OHdG between 1 and 28 days, although significance was only reached between 1 to 5 days after exposure, but not at 7 or 28 days. Several reasons for the discrepancies among these various studies and our current experiments may be given, including the use of different doses and exposure times, as well as the use of differently "hazardous" species or batches of quartz [2]. In relation to the delayed up-regulation of 8-OHdG as observed by Seiler and co-worker [48], it has been suggested that 8-OHdG induction would predominantly occur in proliferating tissue. However, we did observe neither a clear contrast in visual staining intensities of individuals cells nor any difference in the analysed proportion of positively stained cells for 8-OHdG within the focal versus the non-focal regions of individual lung sections of quartz-treated animals, that would support such possibility.
It has also been suggested that HPLC/ECD analysis may lead to artificial induction of 8-OHdG during the extraction and processing of isolated DNA, which might have lead to increased background levels in control animals and thereby obscuring actual occurring differences [15]. However, this would not explain for the significant positive findings by Yamano and colleagues [21] as well as our unexpected effects with the AL-coated, and more specifically the PVNO-coated quartz samples. A possible explanation for these observations might be that, in contrast to the severe inflammation induced by the non-coated quartz, very mild inflammatory responses as observed with the coated samples, fail to up-regulate compensatory feedback mechanisms such as antioxidants and/or oxidative DNA repair actions. At present we are however very cautious about such interpretation, all the more since in our current study the coating-associated effects on 8-OHdG could not be verified by immunohistochemistry.
The hallmark of our observations with regard to the oxidative DNA damage data in current study, was that despite the occurrence of a marked and persistent inflammation, characterised by a 100-fold increased number of neutrophils and a significant increase in pulmonary ROS/RNS levels, no enhanced steady-state expression of 8-OHdG was found by either method of analysis. We therefore hypothesised that quartz-particles and/or their associated inflammatory effects might have caused a compensatory steady-state induction of BER to prevent exponential increases in oxidative DNA damage during conditions of persistent stress. As such, we decided to evaluate the expression of APE/Ref-1, in view of its established in vivo inducibility, its redox-sensitivity, as well as its rather broad action-spectrum in ROS mediated damage repair, compared to other BER proteins such as the 8-OHdG glycosylase Ogg-1 [25,49]. As an additional advantage, investigation of repair enzymes, including APE/Ref-1, has the advantage to being oxidation-artefact-free indices of in vivo-induced oxidative DNA damage [24].
Our investigations of whole lung homogenate revealed a significant increase in APE/Ref-1 protein expression following instillation of non-coated quartz, but not by the surface-modified quartz samples, in comparison to the control animals. These results were further confirmed by immunohistochemistry, where lung sections from the quartz-treated animals showed increased APE/Ref-1 protein expression in the same lung areas as analysed for 8-OHdG. Subsequent random analysis of the % of positively stained cells showed a significant increase, especially in focal pathologically-altered lung areas. To date, in the field of particle research, the induction of APE/Ref-1 has only been described in vitro for asbestos fibres, namely in mesothelial cells and in alveolar macrophages [26,27]. Here we show for the first time that exposure to respirable quartz-dust leads to enhanced expression of APE/Ref-1 in rat lung in vivo. With regard to its cell-specificity, intense nuclear staining was observed within alveolar macrophages but also epithelial cells, which are known for their involvement in quartz-induced inflammatory, proliferative, as well as genotoxic effects. In support of these in vivo observations, concomitant in vitro analysis of APE/Ref-1 expression in NR8383 alveolar macrophages and RLE lung epithelial cells, showed for both cell types a constitutive, predominantly nuclear expression of this protein. Whether the observed in vivo effects on APE/Ref-1 expression are driven by possible direct action of the reactive quartz surface or indirectly via action of phagocyte-derived products including ROS or RNS remains to be clarified, and will be a major part of our further in vitro investigations. Interestingly in this regard, H2O2 has recently been described to cause nuclear accumulation and de novo synthesis of APE/Ref-1 protein in gastric epithelial cells [40], in a process which could be inhibited by antioxidant pre-treatment. On the other hand, quartz particles may also be directly implicated in line with the in vitro observations with asbestos [26,27]. Whereas in the mesothelial cell study, asbestos was reported to enhance both nuclear and mitochondrial APE/Ref-1 expression in relation to its DNA repair actions [26], in the macrophage studies, the observed asbestos effect were connected to its regulation of redox-sensitive transcription [27]. Similarly, apart from current indications for its role in oxidative DNA damage repair, also current in vivo observations are indicative for a role of APE/Ref-1 in the well-known proliferative and fibrogenic effects of quartz [19]. First of all, the currently observed immunohistochemical staining patterns for APE/Ref-1 are well in line with our previous work, were we showed that quartz, unlike coated-quartz, causes in vivo activation of the NFκB pathway in alveolar macrophages and lung epithelial cells [9,11]. Furthermore, the comparatively stronger nuclear staining of APE/Ref-1 among cells within the focal pathologic lesions compared to non-focal locations as observed 7 days after quartz instillation, also points towards a possible role of this bifunctional protein in quartz-induced proliferation in vivo. Accordingly, in our future studies we will further investigate the complex kinetics of APE/Ref-1 expression as a potential hallmark of quartz pathogenesis in relation to its dual involvement, i.e. oxidative DNA damage repair redox-regulation of inflammatory and proliferative signalling pathways.
Conclusion
The present study showed that coating of the reactive particle surface inhibited quartz-induced production of ROS and RNS in the respiratory tract, a process that was closely associated with a reduced level of inflammatory cells. Obviously, since these endpoints were obtained at the persistent stage of inflammation (i.e. seven days following instillation) one should be cautious to extrapolate our results to the acute inflammatory response by quartz. Pulmonary ROS and RNS release is considered as a crucial and unifying factor in the quartz-induced adverse health effects, including fibrogenicity and carcinogenicity. Despite the fact that non-coated quartz caused a significant ROS/RNS generation and lung tissue damage (i.e. LDH, AP), oxidative DNA damage in the form of 8-OHdG was not increased in the lung. Notably however, the same treatment was found to significantly enhance the expression of APE/Ref-1, a BER pathway protein, also involved in the specific repair of 8-OHdG lesions. On the one hand, our data provide further support that DNA repair enzymes, specifically APE/Ref-1, represent more sensitive and less artefact-prone markers of oxidative stress in models of in vivo oxidant exposure than oxidative DNA damage markers, e.g. 8-OHdG. On the other hand, our observations suggest that during quartz-elicited pulmonary inflammation and associated oxidant generation, pathways of oxidative DNA damage repair may be up-regulated to prevent and/or to compensate for the induction and persistence of oxidative DNA damage and possibly resulting mutagenesis. To confirm this hypothesis, our current investigations are focusing on evaluation of the expression of other BER-pathway related proteins apart from APE/Ref-1, as well as on the specific analysis of the actual activity of BER-repair.
List of abbreviations
2,2'-azinobis-(-3 ethylbenzothiazoline-6-sulphonate) (ABTS), 2,2'-azobis-(-2-amidinopropane)HCl (ABAP), Activator protein 1 (AP-1), alkaline phosphatase (AP), aluminium lactate (AL), apurinic/apyrimidinic endonuclease (APE), base excision repair (BER), bronchoalveolar lavage (BAL), bronchoalveolar lavage fluid (BALF), diaminobenzidine (DAB), dimethyl sulphoxide (DMSO), electron spin resonance (ESR), fetal calf serum (FCS), 8-hydroxy-2'-deoxyguanosine (8-OHdG), Hank's balanced salt solution (HBSS), high performance liquid chromatography with electrochemical detection (HPLC-ECD), horseradish peroxidase (HRPO), intraperitoneal (i.p.), intratracheal (i.t.), lactate dehydrogenase (LDH), May-Grünwald/Giemsa (MGG), myeloperoxidase (MPO), nitric oxide (NO), Nuclear factor kappa B (NFκB), phenol red (PRS solution), phorbol-12-myristate-13-acetate (PMA), phosphate buffered saline (PBS), polyvinylpyridine-N-oxide (PVNO), rat lung epithelial cells (RLE), reactive oxygen and nitrogen species (ROS/RNS), redox effector factor (Ref)-1, trolox equivalent antioxidant capacity (TEAC)
Declaration of competing interests
The author(s) declare that they have no competing interests.
Authors' contributions
CA: General study design and supervision (role of surface coating in quartz pathogenesis), instillation and sectioning, immunohistochemical assays and analysis, preparation of manuscript.
AK: Study design (investigations of oxidative stress), developed/conducted in vivo and ex vivo ROS/RNS assays, preparation of manuscript (equal contributions by CA and AK).
AB: Sectioning, analysis of toxicity markers in BAL
DH: Bronchoalveolar lavage and sectioning, analysis of cellular inflammation in BAL.
PH: Western blotting analyses.
FvS: Contribution to experimental design cf. HPLC/ECD analysis of 8-OHdG.
PB: Initial development and contribution to general study design (role of surface coating in quartz pathogenesis).
RS: Contribution to experimental design, statistics, editing of final version of manuscript.
All authors read and approved the final manuscript.
Acknowledgements
The study was financially supported by the Ministerium of Wirtschaft, Mittelstand, Technologie und Verkehr Nordrhein-Westfalen, the Silikosegesellschaft Nordrhein-Westfalen and the Federal Ministry of Environment. A.M. Knaapen was supported by a postdoctoral fellowship from the Netherlands Organisation for Scientific Research (NWO, grant 916.46.092). Prof. A. Bast is acknowledged for performing the TEAC measurements. The authors wish to thank Dr. K. Unfried for his help with the animal instillation. We want to acknowledge Mrs. A. Winzer, K. Ledermann, C. Weishaupt and V. Suri for their technical support. Dr. K. Driscoll kindly provides us with RLE cells and Dr. S. Diabate with NR8383 cells.
==== Refs
IARC IARC Monograph on the Evaluation of the Carcinogenic Risk of Chemicals to Humans 1997 silica, some silicates, coal dust and para-aramid fibrils. IARC Press, Geneva, Switzerland
Donaldson K Borm PJA The quartz hazard: a variable entity Ann Occup Hyg 1998 42 287 294 9729916 10.1016/S0003-4878(98)00044-1
Fubini B Hubbard A Reactive oxygen species (ROS) and reactive nitrogen species (RNS) generation by silica in inflammation and fibrosis Free Radic Biol Med 2003 34 1507 1516 12788471 10.1016/S0891-5849(03)00149-7
Begin R Masse S Rola-Pleszczynski M Martel M Desmarais Y Geoffroy M LeBouffant L Daniel H Martin J Aluminum lactate treatment alters the lung biological activity of quartz Exp Lung Res 1986 10 385 399 3013607
Brown GM Donaldson K Brown DM Bronchoalveolar leukocyte response in experimental silicosis: modulation by a soluble aluminum compound Toxicol Appl Pharmacol 1989 101 95 105 2552617 10.1016/0041-008X(89)90215-9
Vallyathan V Kang JH Van Dyke K Dalal NS Castranova V Response of alveolar macrophages to in vitro exposure to freshly fractured versus aged silica dust: the ability of prosil 28, an organosilane material, to coat silica and reduce its biological activity J Toxicol Environ Health 1991 33 303 315 1649918
Vallyathan V Castranova V Pack D Leonard S Shumaker J Hubbs AF Shoemaker DA Ramsey DM Pretty JR McLaurin JL Khan A Teass A Freshly fractured quartz inhalation leads to enhanced lung injury and inflammation. Potential role of free radicals Am J Respir Crit Care Med 1995 152 3 1003 1009 7663775
Castranova V Castranova V, Dala NS, Vallyathan V Suppression of the cytotoxicity and fibrogenicity of silica with PVPNO Silica and silica-induced lung diseases 1995 CRC Press, Boca Raton, USA 293 298
Duffin R Gilmour PS Schins RPF Clouter A Guy K Brown DM MacNee W Borm PJA Donaldson K Stone V Aluminium lactate treatment of DQ12 quartz inhibits its ability to cause inflammation, chemokine expression and NF-kappaB activation Toxicol Appl 2001 176 10 17 10.1006/taap.2001.9268
Knaapen AM Albrecht C Becker A Höhr D Winzer A Haenen GR Borm PJA Schins RPF DNA damage in lung epithelial cells isolated from rats exposed to quartz: role of surface reactivity and neutrophilic inflammation Carcinogenisis 2002 23 1111 1120 10.1093/carcin/23.7.1111
Albrecht C Schins RPF Höhr D Becker A Shi T Knaapen AM Borm PJA Inflammatory time course following quartz instillation: role of TNFα and particle surface Am J Respir Cell Mol Biol 2004 31 292 301 15191911 10.1165/rcmb.2003-0300OC
Shi X Ding M Chen F Wang L Rojanasakul Y Vallyathan V Castranova V Reactive oxygen species and molecular mechanism of silica-induced lung injury J Environ Pathol Toxicol Oncol 2001 20 Suppl 1 85 93 11570677
Knaapen AM Borm PJ Albrecht C Schins RP Inhaled particles and lung cancer. Part A: Mechanisms Int J Cancer 2004 109 6 799 809 15027112 10.1002/ijc.11708
Castranova V Signaling pathways controlling the production of inflammatory mediators in response to crystalline silica exposure: role of reactive oxygen/nitrogen species Free Radic Biol Med 2004 37 7 916 925 15336307 10.1016/j.freeradbiomed.2004.05.032
Schins RPF Duffin R Höhr D Knaapen AM Shi T Weishaupt C Stone V Donaldson K Borm PJA Surface modification of quartz inhibits toxicity, particle uptake, and oxidative DNA damage in human lung epithelial cells Chem Res Toxicol 2002 15 1166 1173 12230410 10.1021/tx025558u
Hedenborg M Klockars M Quarz-dust-induced production of reactive oxygen metabolites by human granulocytes Lung 1989 167 23 32 2537915
Nyberg P Polyvinylpyridine-N-oxide and carboxymethyl cellulose inhibit mineral dust-induced production of reactive oxygen species by human macrophages Environ Res 1991 55 157 164 1651224 10.1016/S0013-9351(05)80172-0
Vallyathan V Mega JF Shi X Dalal NR Enhanced generation of free radicals from phagocytes induced by mineral dusts Am J Respir Cell Mol Biol 1992 6 404 413 1312851
Albrecht C Borm PJA Unfried K Signal transduction pathways relevant for neoplastic effects of fibrous and non-fibrous particles Mut Res 2004 553 23 35 15288530
Driscoll KE Deyo LC Carter JM Howard BW Hassenbein DG Bertram TA Effects of particle exposure and particle-elicited inflammatory cells on mutation in rat alveolar epithelial cells Carcinogenesis 1997 18 2 423 430 9054638 10.1093/carcin/18.2.423
Yamano Y Kagawa J Hanaoka T Takahashi T Kasai H Tsugane S Watanabe S Oxidative DNA damage induced by silica in vivo Environ Res 1995 69 102 107 8608769 10.1006/enrs.1995.1031
Nehls P Seiler F Rehn B Greferath R Bruch J Formation and persistence of 8-oxoguanine in rat lung cells as an important determinant for tumor formation following particle exposure Environ Health Perspect 1997 105 Suppl 5 1291 1296 9400740
Knaapen AM Seiler F Schilderman PA Nehls P Bruch J Schins RP Borm PJ Neutrophils cause oxidative DNA damage in alveolar epithelial cells Free Radic Biol Med 1999 27 1–2 234 240 10443941 10.1016/S0891-5849(98)00285-8
Rusyn I Asakura S Pachkowski B Bradford BU Denissenko MF Peters JM Holland SM Reddy JK Cunningham ML Swenberg JA Expression of base excision DNA repair genes is a sensitive biomarker for in vivo detection of chemical-induced chronic oxidative stress: identification of the molecular source of radicals responsible for DNA damage by peroxisome proliferators Cancer Res 2004 64 3 1050 1057 14871837 10.1158/0008-5472.CAN-03-3027
Tell G Damante G Caldwell D Kelley MR The intracellular localization of APE1/Ref-1: more than a passive phenomenon? Antioxid Redox Signal 2005 7 3–4 367 384 15706084 10.1089/ars.2005.7.367
Fung H Kow YW Van Houten B Taatjes DJ Hatahet Z Janssen YM Vacek P Faux SP Mossman BT Asbestos increases mammalian AP-endonuclease gene expression, protein levels, and enzyme activity in mesothelial cells Cancer Res 1998 58 2 189 194 9443389
Flaherty DM Monick MM Carter AB Peterson MW Hunninghake GW Oxidant-mediated increases in redox factor-1 nuclear protein and activator protein-1 DNA binding in asbestos-treated macrophages J Immunol 2002 168 11 5675 5681 12023366
Klebanoff SJ Waltersdorph AM Rosen H Antimicrobial activity of myeloperoxidase Meth Enzymol 1984 105 399 403 6328193
Gallati H Pracht I Horseradish peroxidase: kinetic studies and optimization of peroxidase activity determination using substrates H2O2 and 3,3',5,5'-tetramethylbenzidine J Clin Chem Clin Biochem 1985 23 453 460 3903027
Pick E Keisari YA Simple colorimetric method for the measurement of hydrogen peroxide produced by cells in culture J Immunol Meth 1980 38 161 170 10.1016/0022-1759(80)90340-3
Van den Berg R Haenen GRMM Van den Berg H Bast A Applicability of an improved Trolox equivalent antioxidant capacity (TEAC) assay for evaluation of antioxidant capacity measurements of mixtures Food Chem 1999 66 511 517 10.1016/S0308-8146(99)00089-8
Helmke RJ Boyd RL German VF Mangos JA From growth factor dependence to growth factor responsiveness: the genesis of an alveolar macrophage cell line In Vitro Cell Dev Biol 1987 23 8 567 574 3497918
Staal FJ Roederer M Herzenberg LA Herzenberg LA Intracellular thiols regulate activation of nuclear factor kappa B and transcription of human immunodeficiency virus Proc Natl Acad Sci U S A 1990 87 24 9943 9947 2263644
Carter JD Driscoll KE The role of inflammation, oxidative stress, and proliferation in silica-induced lung disease: a species comparison J Env Pathol Toxicol Oncol 2001 20 suppl 1 33 43 11570672
Porter DW Millecchia L Robinson VA Hubbs A Willard P Pack D Ramsey D McLaurin J Khan A Landsittel D Teass A Castranova V Enhanced nitric oxide and reactive oxygen species production and damage after inhalation of silica Am J Physiol Lung Cell Mol Physiol 2002 283 L485 L493 12114212
Greim H Borm P Schins R Donaldson K Driscoll K Hartwig A Kuempel E Oberdorster G Speit G Toxicity of fibers and particles Inhal Toxicol 2001 13 9 26 27
Kamp DW Dunn MM Sbalchiero JS Knap AM Weitzman SA Contrasting effects of alveolar macrophages and neutrophils on asbestos-induced pulmonary epithelial cell injury Am J Physiol 1994 266 L84 L91 8304471
Padgett EL Pruett SB Rat, mouse and human neutrophils stimulated by a variety of activating agents produce much less nitrite that rodent macrophages Immunology 1995 84 135 141 7534260
Huffman LJ Prugh DJ Millecchia L Schuller KC Cantrell S Porter DW Nitric oxide production by rat bronchoalveolar macrophages or polymorphonuclear leukocytes following intratracheal instillation of lipopolysaccharide or silica J Biosci 2003 28 1 29 37 12682422
Ding SZ O'Hara AM Denning TL Dirden-Kramer B Mifflin RC Reyes VE Ryan KA Elliott SN Izumi T Boldogh I Mitra S Ernst PB Crowe SE Helicobacter pylori and H2O2 increase AP endonuclease-1/redox factor-1 expression in human gastric epithelial cells Gastroenterology 2004 127 3 845 858 15362040 10.1053/j.gastro.2004.06.017
Castranova V Huffman LJ Judy DJ Bylander JE Lapp LN Weber SL Blackford JA Dey RD Enhancement of nitric oxide production by pulmonary cells following silica exposure Environ Health Perspect 1998 106 suppl 5 1165 1169 9788892
Huffman LJ Judy DJ Castranova V Regulation of nitric oxide production by rat alveolar macrophages in response to silica exposure J Toxicol Environ Health A 1998 53 29 46 9447227 10.1080/009841098159457
Sies H Oxidative stress II. Oxidants and Antioxidants 1991 Academic Press, London
Janssen YM Marsh JP Absher MP Hemenway D Vacek PM Leslie KO Borm PJ Mossman BT Expression of antioxidant enzymes in rat lungs after inhalation of asbestos or silica J Biol Chem 1992 267 10625 10630 1316905
Schraufstätter I Hyslop PA Jackson JH Cochrane CG Oxidant-induced DNA damage of target cells J Clin Invest 1988 82 1040 1050 2843565
Spencer JP Jenner A Chimel K Aruoma OI Cross CE Wu R Halliwell B DNA strand breakage and base modification induced by hydrogen peroxide treatment of human respiratory tract epithelial cells FEBS Lett 1995 374 233 236 7589542 10.1016/0014-5793(95)01117-W
Janssen-Heininger YM Macara I Mossman BT Cooperativity between oxidants and tumor necrosis factor in the activation of nuclear factor (NF)-kappaB: requirement of Ras/mitogen-activated protein kinases in the activation of NF-kappaB by oxidants Am J Respir Cell Mol Biol 1999 20 942 952 10226064
Seiler F Rehn B Rehn S Hermann M Bruch J Quartz exposure of the rat lung leads to a linear dose response in inflammation but not in oxidative DNA damage and mutagenicity Am J Respir Cell Mol Biol 2001 24 4 492 498 11306444
Fritz G Grosch S Tomicic M Kaina B APE/Ref-1 and the mammalian response to genotoxic stress Toxicology 2003 193 1–2 67 78 14599768 10.1016/S0300-483X(03)00290-7
|
16266428
|
PMC1291407
|
CC BY
|
2021-01-04 16:36:26
|
no
|
Respir Res. 2005 Nov 2; 6(1):129
|
utf-8
|
Respir Res
| 2,005 |
10.1186/1465-9921-6-129
|
oa_comm
|
==== Front
Theor Biol Med ModelTheoretical Biology & Medical Modelling1742-4682BioMed Central London 1742-4682-2-441626908210.1186/1742-4682-2-44ResearchA mathematical model of the euglycemic hyperinsulinemic clamp Picchini Umberto [email protected] Gaetano Andrea [email protected] Simona [email protected] Susanne [email protected] Geltrude [email protected] CNR-IASI BioMatLab, Rome, Italy2 Department of Biostatistics, University of Copenhagen, Denmark3 Istituto di Medicina Interna e Geriatria, Divisione di Malattie del Ricambio, Università Cattolica del Sacro Cuore, Policlinico Universitario "A. Gemelli", Rome, Italy2005 3 11 2005 2 44 44 5 8 2005 3 11 2005 Copyright © 2005 Picchini et al; licensee BioMed Central Ltd.2005Picchini et al; licensee BioMed Central Ltd.This is an Open Access article distributed under the terms of the Creative Commons Attribution License (), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
Background
The Euglycemic Hyperinsulinemic Clamp (EHC) is the most widely used experimental procedure for the determination of insulin sensitivity, and in its usual form the patient is followed under insulinization for two hours. In the present study, sixteen subjects with BMI between 18.5 and 63.6 kg/m2 were studied by long-duration (five hours) EHC.
Results
From the results of this series and from similar reports in the literature it is clear that, in obese subjects, glucose uptake rates continue to increase if the clamp procedure is prolonged beyond the customary 2 hours. A mathematical model of the EHC, incorporating delays, was fitted to the recorded data, and the insulin resistance behaviour of obese subjects was assessed analytically. Obese subjects had significantly less effective suppression of hepatic glucose output and higher pancreatic insulin secretion than lean subjects. Tissue insulin resistance appeared to be higher in the obese group, but this difference did not reach statistical significance.
Conclusion
The use of a mathematical model allows a greater amount of information to be recovered from clamp data, making it easier to understand the components of insulin resistance in obese vs. normal subjects.
==== Body
Background
With the growing epidemiological importance of insulin resistance states such as obesity and Type 2 Diabetes Mellitus, T2DM, and with increasing clinical recognition of the impact of the so-called metabolic syndrome, the assessment of insulin sensitivity has become highly relevant to metabolic research.
The experimental procedures currently employed to gather information on the degree of insulin resistance of a subject are the Oral Glucose Tolerance Test (OGTT), the Intra-Venous Glucose Tolerance Test (IVGTT), the Euglycemic Hyperinsulinemic Clamp (EHC), the Hyperglycemic Clamp, the insulin-induced hypoglycemia test (KITT), and less commonly used methods based on tracer administration [1-3]. Of these, the EHC is considered the tool of choice in the diabetological community, in spite of its labor-intensive execution, because it is usually considered that the results obtained can be interpreted simply [4,5]. The favor with which the EHC is viewed in this context stems in part from the belief that while mathematical models of the glucose insulin system make untenable assumptions, the EHC approach is relatively assumption-free, or model-independent.
In general, insulin resistance expresses an imbalance between the amount of pancreatic insulin secreted in response to a glucose load and the levels of plasma glucose attained. In other words, in order to obtain the same plasma glucose concentration, higher levels of plasma insulin are necessary in insulin-resistant subjects than in normal controls [6].
The clamp, as usually employed, yields easy-to-compute indices, which are commonly used as measures of insulin resistance. The M value [5] is defined as the average glucose infusion rate over a period of 80–120 minutes from the start of the insulin infusion. The M/I ratio is the ratio of the M value to the average plasma insulin concentration during the same period. If a two-step clamp is performed (though see negative comments [4]) the ΔM/ΔI ratio is defined as the increment of M produced by raising the insulin infusion rate over the corresponding increment of I. The use of these indices, however, makes two fundamental assumptions: first, that at the end of 120' of insulin infusion the experimental subject is at steady state with regard to glucose uptake rate; and second, that the glucose uptake rate increases linearly with increasing insulinemia, either throughout the insulin concentration range (when using the M/I index for characterizing the subject's response) or between successive insulin concentrations reached in the two-step clamp (when using the ΔM/ΔI index). These assumptions are, however, only a first approximation to the real state of things. On the one hand, it has already been shown that if a clamp experiment is continued beyond the customary 2 hours " [...] glucose utilization increases progressively through(out) five hours of moderate hyperinsulinemia." [7]. On the other hand [8], carefully measured average glucose uptake rates at two hours are nonlinearly related to increasing levels of plasma insulin, and from the reported data, glucose uptake may approach a maximal value asymptotically as insulinemia increases. In spite of these observations, the vast majority of experimental diabetologists ([9], [4], [10]) consider the EHC the procedure of choice and many studies have already been conducted using it. It would be interesting to be able to reinterpret this vast mass of observations using a more explicitly quantitative approach. The goal of the present work is to formulate a model of the EHC and fit it to EHC data recorded from human subjects. The structure of the model we have developed allows us to discuss the mechanisms whereby a sufficiently long insulin infusion might be able to increase glucose uptake progressively, and to explore the possible implications of the commonly observed insulin resistance pattern in obese subjects.
Methods
Subjects
Sixteen subjects were enrolled in the study, 8 normal volunteers and 8 patients from the Obesity Outpatient Clinic of the Department of Internal Medicine at the Catholic University School of Medicine. For one normal subject the recorded glycemia values were accidentally lost and this subject was therefore discarded from the following mathematical analysis. The subjects had widely differing BMIs (from 18.5 to 63.6). All subjects were clinically euthyroid, had no evidence of diabetes mellitus, hyperlipidemia, or renal, cardiac or hepatic dysfunction and were undergoing no drug treatments that could have affected carbohydrate or insulin metabolism. The subjects consumed a weight-maintaining diet consisting of at least 250 g of carbohydrate per day for 1 week before the study. Table 1 reports the main anthropometric and metabolic characteristics of the subjects.
Table 1 Anthropometric and metabolic characteristics for lean (BMI ≤ 25) and overweight or obese (BMI > 25) subjects.
Lean subjects (n = 7) Overweight and Obese subjects (n = 8) p
BMI [kg/m2] 20.0 [18.5, 22.7] 37.0 [27.8, 63.6] 0.001
BSA [m2] 1.55 [1.49, 1.73] 2.1 [1.83, 2.38] 0.001
Gfast [mM] 3.67 [3.4, 5.4] 5.2 [4.61, 5.9] 0.024
Ifast [pM] 27.8 [13.9, 49.4] 123.7 [79.2, 152.9] 0.001
Imax [pM] 482.14 [464.5, 526.9] 606.3 [497.3, 683.2] 0.004
Values are median [min, max]. All comparisons were performed by the Mann-Whitney U-test. BSA is the Body Surface Area [m2] calculated via the DuBois formula (BSA = 0.20247 · height 0.725 [m] · weight 0.425 [kg])
The study protocol followed the guidelines of the Medical Ethics Committee of the Catholic University of Rome Medical School; written informed consent was obtained from all subjects.
Experimental protocol
Each subject was studied in the postabsorptive state after a 12–14 h overnight fast. Subjects were admitted to the Department of Metabolic Diseases at the Catholic University School of Medicine in Rome the evening before the study. At 07.00 hours on the following morning, the infusion catheter was inserted into an antecubital vein; the sampling catheter was introduced in the contralateral dorsal hand vein and this hand was kept in a heated box (60°C) in order to obtain arterialized blood. A basal blood sample was obtained in which insulin and glucose levels were measured. At 08.00 hours, after a 12–14 h overnight fast, the Euglycemic Hyperinsulinemic glucose Clamp was performed according to [5]. A priming dose of short-acting human insulin was given during the initial 10 min in a logarithmically decreasing manner so that the plasma insulin was raised acutely to the desired level. During the five-hour clamp procedure, the glucose and insulin levels were monitored every 5 min and every 20 min respectively, and the rate of infusion of a 20% glucose solution was adjusted during the procedure following the published algorithm [5]. Because serum potassium levels tend to fall during this procedure, KCl was given at a rate of 15–20 mEq/h to maintain the serum potassium between 3.5 and 4.5 mEq/l.
Serum glucose was measured by the glucose oxidase method using a Beckman Glucose Analyzer II (Beckman Instruments, Fullerton, Calif., USA). Plasma insulin was measured by microparticle enzyme immunoassay (Abbott Imx, Pasadena, Calif., USA).
Modelling
In order to explain the oscillations of glycemia occurring in response to hyperinsulinization and to continuous glucose infusion at varying speeds, we hypothesized the following system:
where
ω(s) = α2se-αs, Tgx(s) = 0 ∀s∊ [-τg,0] and Tix(0) = Tixb.
Tgx(t) and Tix(t) are (input or forcing) state variables of which the values are known at each time; the state variables and the parameters are defined in tables 2 and 3. The model is diagrammatically represented in Figure 1.
Table 2 Definitions of the state variables.
Variables
G(t) [mM] plasma glucose concentration at time t
I(t) [pM] serum insulin concentration at time t
t [min] time from insulin infusion start
Tgx(t) [mmol/min/kgBW] glucose infusion rate at time t
Tix(t) [pmol/min/kgBW] insulin infusion rate at time t
Tgh(t) [mmol/min/kgBW] net Hepatic Glucose Output (HGO) at time t
Table 3 Definitions of the parameters.
Parameters
Gb [mM] basal glycemia
Ib [pM] basal insulinemia
Txg [mM / min] maximal insulin-independent rate constant for glucose tissue uptake
KxgI [min-1/pM] insulin-dependent apparent first-order rate constant for glucose tissue uptake at insulinemia I
Kxi [min-1] apparent first-order rate constant for insulin removal from plasma
TiG [pM/min/mM] apparent zero-order net insulin synthesis rate at unit glycemia (after liver first-pass effect)
Tixb [pmol/min/kgBW] basal insulin infusion rate, which is given by the measured value of Tix at time zero according to [18]
Tghmax [mmol/min/kgBW] maximal Hepatic Glucose Output at zero glycemia, zero insulinemia
Tghb [mmol/min/kgBW] basal value of Tgh
Vg [L/kgBW] volume of distribution for glucose
Vi [L/kgBW] volume of distribution for insulin
α [#] time constant for the insulin delay kernel ω(·)
τg [min] discrete (distributional) delay of the change in glycemia following glucose infusion
λ [mM-1pM-1] rate constant for Hepatic Glucose Output decrease with increase of glycemia and insulinemia
ρ [#] average delay of insulin effect
Figure 1 Schematic representation of the model (1), (2) and (3).
Equations (1) and (2) express the variations of plasma glucose and plasma insulin concentrations. Equation (3) represents the rate of net Hepatic Glucose Output, starting at maximal HGO at zero glucose and zero insulin and decaying monotonically with increases in both glucose and effective insulin concentrations in the plasma.
The variation of glucose concentration in its distribution space may be attributed to the external glucose infusion rate, liver glucose output and delayed-insulin-dependent as well as insulin-independent glucose tissue uptake. Infused glucose raises glycemia after a delay τg due to the time required to equilibrate the intravenously infused quantity throughout the distribution space. The net HGO is assumed to be equal to Tghb at the beginning of the experiment and to decrease toward zero as glycemia or insulinemia levels increase. Serum insulin, after a delay depending on its transport to the periphery and the subsequent activation of cellular membrane glucose transporters, affects glucose clearance through equation (1) and the glucose synthesis rate through equation (3).
We hypothesize that ω(s) represents the density of the metabolic effect at time t for unit serum insulin concentration at time t - s (s ≤ t). We could choose ω(s) as a single function or as a linear combination of functions (with positive coefficients adding up to unity) from the family of Erlang-functions:
The first two functions of the family are
ω(1) (s) = αe-α s
ω(2) (s) = α2se-α s
We note that while ω(1)(s) is monotonically decreasing, ω(2)(s) increases to a maximum at s = 1/α, then decreases monotonically and asymptotically to zero. We choose the second Erlang-function as our kernel because it is the simplest member of the family with a peak. This embodies the concept that, in order to produce its metabolic effect, insulin has to reach the tissues and activate intracellular enzymatic mechanisms (hence its maximal action on glucose metabolism is delayed) and that natural breakdown of insulin induces a progressive loss of effect of increased concentrations of the hormone as they become more distant in the past. A high α value determines a concentrated kernel corresponding to a fast-rising, fast-decaying effect of insulin on peripheral tissues. We therefore set
and we define as the average time for the metabolic effect of insulin in changing glycemia. The insulin-independent glucose tissue uptake process is modelled as a Hill function rapidly increasing to its (asymptotic) maximum value Txg; thus for glycemia values near 2 mM the insulin-independent glucose tissue uptake is already close to its maximum. This formulation is intended to represent the aggregated apparent zero-order (fixed) glucose utilization mechanism at rest (mainly the brain and heart [11]; W. Sacks in [12] p. 320), with the mathematical and physiological requirement that glucose uptake tends to zero as glucose concentration in plasma approaches zero.
The variation of insulin concentration in its distribution space (equation 2) may be thought of as due to the external insulin infusion, glucose dependent pancreatic insulin secretion and the apparently first-order insulin removal from plasma.
We use steady-state conditions to decrease the number of free parameters to be estimated: at steady state, before the start of the clamp (G = Gb, I = Ib, Tgx = Tix = 0), we have
Therefore the parameters Tghb, Txg, and TiG are completely determined by the values of the other parameters (and ρ is determined from α).
Statistical analysis
The system (1), (2) and (3) has been numerically integrated by means of a fourth order Runge-Kutta scheme; the solutions thus obtained have been fitted by Weighted Least Squares (WLS) separately on each subject's glycemia and insulinemia time-points, estimating only the free parameters Gb, Ib, KxgI, Kxi, Tghmax, Vg, Vi, α, τg, λ. The statistical weight associated with each observed glucose and insulin concentration point has been defined as 1/CV2, where CV is the coefficient of variation, equal to 0.015 for glucose and 0.07 for insulin [13]. The weighted quadratic loss function was minimized by a Nelder-Mead simplex algorithm in order to obtain the WLS parameter estimates for each subject. In order to highlight possible physiological differences among subjects depending on their BMI, two groups were defined: a group consisting of lean subjects (BMI ≤ 25) and a group consisting of overweight or obese subjects (BMI > 25). Comparisons of anthropometric characteristics, metabolic characteristics and model parameter values between these groups were performed by the Mann-Whitney U-test owing to the small number of subjects in each group. Comparisons within groups were performed by the Wilcoxon test for matched pairs.
Results
Table 1 shows anthropometric characteristics (BMI, BSA), measured plasma glucose and insulin concentrations (Gfast, Ifast) in the two groups immediately before the clamp, and the average levels of insulin after 80' of clamp insulinization (Imax). All differences in the characteristics were highly significant, with the median values in the obese/overweight group markedly higher than those in the lean group. Even though there was a significant difference in fasting glycemia between the groups, average levels remained within the norm. However, fasting insulinemia was more than four-fold higher in the obese/overweight group, consistent with what is usually observed in this patient population.
For each parameter fitted and determined, the median, minimum and maximum from the sample of values obtained are reported in Table 4.
Table 4 Estimated and determined parameter values for lean (BMI ≤ 25) and overweight or obese (BMI > 25) subjects.
Lean (n = 7) Overweight or Obese (n = 8) p
Estimated Parameters
Gb [mM] 3.67 [2.80, 4.36] 5.11 [4.52, 5.97] 0.001
Ib [pM] 17.91 [8.59, 63.41] 121.05 [61.55, 256.41] 0.002
KxgI [min-1/pM] 9.94 · [7.1, 21.2] · 10-6 6.34 · [0, 13.3] · 10-6 0.132
Kxi [min-1] 0.039 [0.022, 0.057] 0.029 [0.021, 0.045] 0.203
Tghmax [mmol/min/kgBW] 0.069 [0.05, 0.12] 0.128 [0.026, 0.274] 0.105
Vg [L/kgBW] 0.49 [0.33, 0.90] 0.47 [0.25, 0.67] 0.643
Vi [L/kgBW] 0.4 [0.36, 0.78] 0.39 [0.21, 0.65] 0.487
α [#] 0.017 [0.015, 0.082] 0.024 [0.008, 0.048] 0.908
τg [min] 3.00 [1.00, 11.50] 5.14 [0.50, 9.00] 0.917
λ [mM-1pM-1] 8.9 [1.2, 21.3] · 10-3 3.1 [0.2, 4] · 10-3 0.037
Determined Parameters
Tghb [mmol/min/kgBW] 0.042 [0.028, 0.052] 0.019 [0.009, 0.117] 0.36
Txg [mM / min] 0.085 [0.057, 0.126] 0.046 [0.012, 0.397] 0.203
TiG [pM/min/mM] 0.096 [0.031, 0.29] 0.267 [0.128, 0.668] 0.011
ρ [#] 115.4 [24.3, 136.2] 83.6 [42.1, 267.7] 0.908
Comparisons were performed by the Mann-Whitney U-test. Values are expressed as median [min, max].
The predicted basal glycemia and insulinemia values (Gb, Ib) were close to the observed fasting values and were significantly different between groups (respectively p = 0.001 and p = 0.002). Lean subjects have a greater ability (about 3-fold higher) to reduce hepatic glucose output when glycemia and insulinemia increase (expressed by the parameter λ, p = 0.037). The parameter TiG (glucose-dependent pancreatic secretion of insulin) is also significantly different between groups (p = 0.011) and the insulin synthesis rate in obese/overweight subjects is about three-fold higher than in lean subjects. The delay coefficient τg is of the order of 3 to 5 minutes, which seems a reasonable time for glucose infused through an arm vein to be distributed throughout the body, equilibrate, and be detected by sampling through the arterialized contralateral arm vein.
In Table 5 the measured values of the M/I index over the time periods 80'–120' and 260'–300' are shown for normal and obese/overweight subjects: as expected, the rate of glucose uptake per unit plasma insulin concentration is significantly higher in lean subjects in both the 80'–120' (p = 0.001) and the 260'–300' periods (p = 0.015). However, whereas in lean subjects the M/I value remains stable between the two periods (p = 0.6), in the obese/overweight group it increases significantly (p = 0.02).
Table 5 M/I index values for lean and overweight or obese subjects measured over the 80'–120' and on the 260'–300' time periods.
Lean (n = 7) Overweight or Obese (n = 8) p (M-W U)
M / I (80'–120') 9.75 · 10-5 [6.97, 11.42] · 10-5 2.66 · 10-5 [1.57, 5.2] · 10-5 0.001
M / I (260'–300') 8.9 · 10-5 [4.9, 13.2] · 10-5 3.86 · 10-5 [2.54, 7.44] · 10-5 0.015
p (Wilcoxon) 0.6 0.02
Comparisons between groups were performed by the Mann-Whitney U-test. Comparisons within groups were performed via the Wilcoxon test for matched pairs. Values are expressed as median [min, max].
Figures 2, 3, 4, 5 show the time course of observed and predicted glycemia, observed and predicted insulinemia and glucose infusion rate for four experimental subjects (two lean and two obese).
Figure 2 Composite plot for subject 2 (BMI = 35.9). Observed (◆) and predicted (....) glycemia; observed (o) and predicted (----) insulinemia; glucose infusion rate (solid line). For ease of comparison, the insulin concentrations and the glucose infusion rates are divided by factors of 300 and 0.01 respectively.
Figure 3 Composite plot for subject 6 (BMI = 19.33). Observed (◆) and predicted (....) glycemia; observed (o) and predicted (----) insulinemia; glucose infusion rate (solid line). For ease of comparison, the insulin concentrations and the glucose infusion rates are divided by factors of 300 and 0.01 respectively.
Figure 4 Composite plot for subject 9 (BMI = 63.6). Observed (◆) and predicted (....) glycemia; observed (o) and predicted (----) insulinemia; glucose infusion rate (solid line). For ease of comparison, the insulin concentrations and the glucose infusion rates are divided by factors of 300 and 0.01 respectively.
Figure 5 Composite plot for subject 10 (BMI = 18.6). Observed (◆) and predicted (....) glycemia; observed (o) and predicted (----) insulinemia; glucose infusion rate (solid line). For ease of comparison, the insulin concentrations and the glucose infusion rates are divided by factors of 300 and 0.01 respectively.
Discussion
It was shown in the early '80s [7] that a significant increase of glucose tissue uptake during the euglycemic hyperinsulinemic clamp could be obtained in obese subjects by waiting for up to 4–6 hours. This basic observation, confirmed by the series of obese subjects studied in the present work, challenges the assumption that steady state is attained after 2 hours of the clamp, at least in one patient subpopulation of great metabolic interest. Nolan et al. [14], while performing an isoglycemic hyperinsulinemic clamp, also demonstrated a marked delay in activation of whole-body glucose disposal rate, arterio-venous glucose difference and leg glucose uptake in seven subjects with Type 2 Diabetes Mellitus and in seven obese non-diabetic subjects, as compared to healthy controls.
The concept of insulin resistance as a decreased effect of the hormone on whole body glucose uptake can be made more specific: on the one hand we might wish to measure the speed with which a given level of metabolic response is attained; on the other, we might wish to quantify the maximal response attainable by a suitably raised insulin plasma concentration. It is clear now that when using the classical two-hour clamp, subpopulations of subjects respond within different time frames. Concentrating on the level of response at 2 hours would label subjects with a residual metabolic capacity as insulin-resistant: this may or may not be appropriate depending on the mode of insulin resistance that the physiologist is interested in, whether the speed or the capacity of response. The case of the obese subject represents this ambiguity very well: if by insulin resistance we mean the result of the EHC at 2 hours, that is to say a decreased effect of insulin on whole body glucose uptake under hyperinsulinization with respect to a specific and short time frame, then obese subjects can be adequately diagnosed by the clamp as being generally insulin resistant. If, on the other hand, we abandon the time frame requirement and address the maximal ability to respond to the hormone, then the standard clamp procedure is not adequate since it fails to allow slowly-responding subjects to develop a complete response. A way out of this ambiguity for diagnostic purposes could be to use the parameters of a mathematical model of the metabolic response during the clamp. Hopefully, this model would be able to quantify both the maximal response obtainable by the subject and the rate at which this response is generated. Hence the diabetologist would be offered separate, independent and complementary items of information on which to base the diagnosis.
Given the above considerations, the approach followed in the present work was therefore to construct a deterministic mathematical model of the time course of glucose uptake rate during a clamp experiment.
A series of studies [15-17] demonstrated that insulin-stimulated glucose uptake correlates with the appearance of insulin in lymph fluid, a marker for interstitial insulin, rather than with the appearance of insulin in the circulatory stream. Whether trans-endothelial passage of insulin from the circulation to the interstitial space is the sole or the main mechanism for the delay is debatable, even though it may be rate-limiting in the activation of glucose uptake, since the pancreatic response to glucose should be fast and since, once insulin is in the interstitial space, further endocellular steps are very rapid. In any case, out of the many models we tried in order to explain the observed insulin and glucose concentration time courses, the model that best explains the data includes a delay in the action of plasma insulin in correcting glycemia. Of the many alternative explicit representations of such delay that could have been used, one of the simplest was chosen, a Erlang-function kernel, to simplify the model's mathematical treatment.
It has been shown [14] that Hepatic Glucose Output (HGO) suppression after step insulinization is not immediate, HGO decreasing towards 0 in an approximately exponential manner from its pre-insulinization level. In the present work, HGO was not independently measured by tracer techniques. The model proposed here assumes that the variable representing HGO (identified with the symbol Tgh) falls progressively to a new equilibrium value as delayed insulin increases progressively to its new equilibrium level after a step increase in plasma insulin. In this, our model agrees with Nolan's observation. Further, in the model proposed in the present work, equilibrium Tgh falls exponentially (with parameter λ) as equilibrium insulin increases from baseline to full insulinization levels.
The two parameters Tghmax and KxgI express respectively the maximum Hepatic Glucose Output and the sensitivity of glucose uptake to insulin concentration. Neither was significantly different between lean and obese subjects. However, Tghmax was higher and KxgI was lower in obese subjects, and both these changes would point to a decreased insulin sensitivity in this patient group. While the observed lack of significance may well be a consequence of the limited power of the present study, given the small number of subjects considered, the fact that these two parameters were not much changed in obese subjects while λ was significantly lower again indicates a relative slowness in mounting an appropriate response rather than a relative incapacity to mount a sustained response eventually.
From the modelling point of view, the present study prompts two considerations. The first is that a clamp that is medically very successful (i.e. during which the physician manages to clamp glycemia effectively to within a narrow range) may be less informative about the actual subject's compensation mechanisms than a clamp where imprecise correction of glycemia gives rise to oscillations. The second is that, especially for subjects such as the one reported in Figure 5, where sustained oscillations are produced, random perturbations of the system may give rise to accidental phase shifts. This makes it very hard or impossible to follow the oscillations unless for the model can accommodate random variations of metabolism. Future efforts in modelling the clamp will have to consider this feature.
Conclusion
In conclusion, the present paper describes a possible deterministic modelling of the EHC, which may prove useful for studying obese subjects who show delayed expression of their maximal increase of glucose uptake under insulinization. Considering the amplitude of response independently of the time factor, the whole body capacity of glucose uptake in obese subjects does not appear to be decreased with respect to lean subjects.
Competing interests
The author(s) declare that they have no competing interests.
Authors' contributions
UP: mathematical modeling, statistical analysis, drafting of the manuscript;
ADG: mathematical modeling, drafting of the manuscript;
SP: mathematical modeling;
SD: mathematical modeling;
GM: design of the experiment, collection of data, drafting of the "Experimental protocol" and "Discussion" sections of the manuscript.
All authors read and approved the final manuscript.
==== Refs
Ferrannini E Mari A How to measure insulin sensitivity J Hypertens 1998 16 895 906 9794728 10.1097/00004872-199816070-00001
Wallace TM Matthews DR The assessment of insulin resistance in man Diabet Med 2002 19 527 534 12099954 10.1046/j.1464-5491.2002.00745.x
Starke AA Determination of insulin sensitivity: methodological considerations J Cardiovasc Pharmacol 1992 20 S17 S21 1284139
Zierler K Whole body glucose metabolism Am J Physiol 1999 276 E409 E426 10070005
Defronzo RA Tobin JD Andres R Glucose clamp technique: a method for quantifying insulin secretion and resistance Am J Physiol 1979 237 E214 E223 382871
Felber JP Acheson KJ Tappy L From obesity to diabetes 1993 Chichester: John Wiley and Sons
Doberne L Greenfield MS Schulz B Reaven GM Enhanced glucose utilization during prolonged glucose clamp studies Diabetes 1981 30 829 835 7024021
Thiebaud D Jacot E Defronzo RA Maeder E Jequier E Felber JP The effect of graded doses of insulin on total glucose uptake, glucose oxidaton and glucose storage in man Diabetes 1982 31 957 963 6757014
Ferrannini E Smith JD Cobelli C Toffolo G Pilo A Defronzo RA Effect of insulin on the distribution and disposition of glucose in man J Clin Invest 1985 76 357 364 3894421
Ferrannini E Natali A Bell P Cavallo-Perin P Lalic N Mingrone G Insulin resistance and hypersecretion in obesity J Clin Invest 1997 100 1166 1173 9303923
Olson AL Pessin JE Structure, function, and regulation of the mammalian facilitative glucose transporter gene family Annu Rev Nutr 1996 16 235 256 8839927 10.1146/annurev.nu.16.070196.001315
Lajtha A ed Handbook of Neurochemistry 1969 1 New York: Plenum
Bergman RN Ider YZ Bowden CR Cobelli C Quantitative estimation of insulin sensitivity Am J Physiol 1979 236 E667 E677 443421
Nolan JJ Ludvik B Baloga J Reichart D Olefsky JM Mechanisms of the kinetic defect in insulin action in obesity and IDDM Diabetes 1997 46 994 1000 9166671
Yang YJ Hope ID Ader M Bergman RN Insulin transport across capillaries is rate limiting for insulin action in dogs J Clin Invest 1989 84 1620 1628 2681272
Bergman RN Yang YJ Hope ID Ader M The role of the transcapillary insulin transport in the efficiency of insulin action: studies with glucose clamps and the minimal model Horm Metab Res Suppl 1990 24 49 56 2272626
Ader M Poulin RA Yang YJ Bergman RN Dose-response relationship between lymph insulin and glucose uptake reveals enhanced insulin sensitivity of peripheral tissues Diabetes 1992 41 241 253 1733816
Defronzo RA Ferrannini E Insulin resistance. A multifaceted syndrome responsible for NIDDM, obesity, hypertension, dyslipidemia, and atherosclerotic cardiovascular disease Diabetes Care 1991 14 173 194 2044434
|
16269082
|
PMC1291408
|
CC BY
|
2021-01-04 16:39:25
|
no
|
Theor Biol Med Model. 2005 Nov 3; 2:44
|
utf-8
|
Theor Biol Med Model
| 2,005 |
10.1186/1742-4682-2-44
|
oa_comm
|
==== Front
Thromb JThrombosis Journal1477-9560BioMed Central London 1477-9560-3-161625090810.1186/1477-9560-3-16Original Basic ResearchTowards a standardization of thrombin generation assessment: The influence of tissue factor, platelets and phospholipids concentration on the normal values of Thrombogram-Thrombinoscope assay Gerotziafas Grigoris T [email protected] François [email protected] Joël [email protected] Lena [email protected] Ismail [email protected] Meyer M [email protected] Service d'Hématologie Biologique, Hôpital Hôtel-Dieu de Paris, France2 LCL, Ivry sur Seine, France2005 26 10 2005 3 16 16 12 5 2005 26 10 2005 Copyright © 2005 Gerotziafas et al; licensee BioMed Central Ltd.2005Gerotziafas et al; licensee BioMed Central Ltd.This is an Open Access article distributed under the terms of the Creative Commons Attribution License (), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
Background
Thrombin generation assay was developed several years ago to mimic physiological coagulation mechanisms but it had important limitations. Thrombogram-Thrombinoscope assay using a fluorogenic substrate, allows obtaining thrombin generation curves in non-defibrinated platelet rich plasma (PRP) in a fully automated manner.
Methods
We standardised the methodology of Thrombogram-Thrombinoscope and we evaluated the precision of thrombin generation parameters (lag-time, maximum concentration of thrombin [Cmax], time required to reach Cmax [Tmax] and endogenous thrombin potential ETP) using different concentrations of recombinant human tissue factor, platelets or phospholipids. Normal values of thrombin generation assay were established in optimal experimental conditions.
Results
In the presence of low TF concentrations (final dilution of thromboplastin in plasma: 1/1000–1/2000) the Thrombogram assay showed intra-assay and inter-assay coefficients of variation lower than 9%. Thrombin generation parameters showed an important inter-individual variability and the coefficients of variation ranged from 18% to 50%. In PRP the lag-time, Cmax and Tmax but not the ETP, were influenced by TF concentration. Thrombin generation parameters were not influenced by variations of platelet concentration from 50 × 109/l to 400 × 109/l. The addition of synthetic procoagulant phospholipids in PPP strongly influenced all the parameters of thrombogram. For all the parameters of thrombogram a threshold effect was observed in the presence of phspholipid concentrations equal or higher to 4 μM. In frozen-thawed PRP the lag-time and the Tmax were significantly reduced and the Cmax was increased compared to the fresh PRP, but the ETP, the intra assay and the inter-assay coefficients of variation were similar in both test-systems.
Conclusion
Thrombogram-Thrombinoscope assay performed in fresh or in frozen-thawed PRP has an acceptable precision, with low inter-assay and intra-assay coefficient of variations. The concentration of TF is determinant for the normal values of the studied parameters of thrombin generation. When the assay is performed in PPP, thrombin generation parameters are influenced by the concentration of procoagulant synthetic phospholipids. The optimal experimental conditions were obtained in the presence of 1/1000 final dilution of thromboplastin, a platelet count higher than 50 × 109/l and a synthetic phospholipid concentration higher than 4 μM.
Thrombin generationThrombinoscopeendogenous thrombin potentialthrombinplateletstissue factor
==== Body
Background
The global clotting times (i.e. prothrombin time and partial thromboplastin time) are routinely used for the evaluation of human coagulation but they ignore the procedure of thrombin generation since at the time of clot formation only 3% of prothrombin is activated [1,2]. This is a constant observation either in normal plasma or in the presence of antithrombotic agents (i.e. enoxaparin or fondaparinux) [3]. The global clotting times are mandatory for the detection of clotting factor deficiencies and the biological monitoring of treatment with coumarins and unfractionated heparin but they have a very limited value in the detection of hypercoagulable states [1,2]. Moreover, they are not significantly influenced by the presence of therapeutic concentrations of LMWHs or the indirect FXa inhibitors such as fondaparinux [3]. Measuring of anti-Xa activity in plasma from patients treated with LMWHs has a limited predictive value for the clinical outcome [4] and in some patients' populations (i.e elderly or obese patients) the anti-Xa levels in plasma are poorly correlated with the administered dose of LMWH [5,6].
The study of thrombin generation performed either with clotting based assays or with chromogenic substrates is an old and established tool in blood coagulation research [7-12]. It describes all the phases of thrombin generation process (initiation, amplification and inhibition of thrombin generation as well as the integral amount of generated thrombin). According to the experimental system used, thrombin generation may be influenced by most of the factors playing a role in blood coagulation. However, thrombin generation assessment used to be a laborious and time-consuming method. The most sophisticated version of thrombin generation assay using a chromogenic substrate developed by Hemker's group was less time-demanding than the previous one, since it was fully automated accompanied with a software for the calculation of the area under thrombin generation curve (endogenous thrombin potential) [13]. The most important experimental limitation of this assay was that it could be done only in defibrinated platelet poor plasma (PPP). Thus at least two important components of blood coagulation (i.e. platelets and fibrinogen/fibrin) were absent [14].
The Thrombogram-Thrombinoscope assay is a step forward to the study of the blood coagulation process because, using a fluorogenic substrate, it allows obtaining thrombin generation curves in non-defibrinated platelet rich plasma (PRP) in a fully automated manner [15]. An increasing body of evidence shows that Thrombogram-Thrombinoscope assay may be a useful tool in the diagnosis of acquired or congenital hypercoagulable states and it is sensitive enough, in patients with hemorrhagic diathesis [16, reviewed in 17]. Thrombogram is also sensitive to the presence of antiplatelet and antithrombotic agents [3,18].
Until now, there is an important heterogeneity regarding the pre-analytical conditions and the experimental systems for the assessment of thrombogram. In addition, the endogenous thrombin potential is the sole parameter of thrombogram that has been systematically validated [19-21]. Other parameters of thrombin generation curve given by Thrombogram-Thrombinoscope®, such as the lag-time of thrombin generation onset, the maximum concentration of generated thrombin and the time required to reach the maximum concentration of thrombin have not been systematically validated.
Standardization of thrombin generation using the Thrombogram-Thrombinoscope® method is necessary towards the wide clinical use of the assay. The aim of this study was to determine the influence of preanalytical and experimental conditions on the Thrombogram-Thrombinoscope assay. We examined the intra-assay, the inter-assay and the inter-individual coefficients of variations of thrombin generation lag-time, the maximum concentration of thrombin (Cmax), the time required to achieve maximum thrombin generation (Tmax) and the endogenous thrombin potential (ETP) in fresh PRP and in frozen-thawed PRP after triggering tissue factor (TF) pathway in plasma from healthy volunteers and we established the normal values for each parameter of the thrombogram. We also studied the influence of experimental condition such as the TF, concentration, and the platelet counts or phospholipids concentration on each parameter of thrombin generation curve given by the thrombogram.
Methods
Reagents
Bovine serum albumin (BSA) and Tris-HCl were obtained from Sigma laboratories (St. Louis U.S.A.). Relipidated recombinant tissue factor (Hemoliance® RecombiPlasTin) was from Instrumentation Laboratory (Milan, Italy). The Hemoliance® RecombiPlasTin reagent was reconstituted by addition of 5 ml of NaCl 0.9%, and subsequently diluted in Hepes buffer; pH 7.35 (containing 20 mM Hepes, 140 mM NaCl and 5 mg/ml BSA) for the experiments with the Thrombogram-Thrombinoscope™ assay (Synapse b.v., Maastricht, The Netherlands). Thrombin calibrator was also obtained by Synapse b.v. (Maastricht, The Netherlands). The fluorogenic substrate Z-Gly-Gly-Arg-AMC was obtained from Bachem (Bubendorf, Switzerland). Unfractionated heparin was from Leo Laboratories, commercially available enoxaparin preparation was from Sanofi-Aventis Laboratories (Paris, France) and fondaparinux was from Glaxo Smith Klein (Paris, France). Synthetic phospholipids (DOPS/DOPE/DOPC reconstituted according to manufacturer's instructions in a ratio 1:1:1,25) were obtained from Avanti Polar LipidsInC (Albaster, Alabama, USA).
Normal human plasma
Venous blood was obtained from 22 healthy volunteers non taking any medication, who are working in our laboratory. Blood was collected with atraumatic antecubital veinipuncture, into siliconized vacutainer tubes (Becton Dickinson, Meylan, France) containing buffered trisodium citrate (0.129 mol/l, nine parts of blood and one part of citrate solution). Thrombogram-Thrombinoscope™ assay was performed in platelet rich plasma (PRP) prepared after centrifugation of citrated whole blood for 10 min at 150 g at room temperature. The supernatant PRP was removed and the platelet count was adjusted to 150 × 109/l by dilution with autologous PPP obtained after a further centrifugation of the remaining blood for 15 min at 2000 g. To study the influence of platelet concentration on thrombin generation parameters, the platelet concentrations in PRP, prepared from 4 healthy volunteers, were adjusted to, 10 × 109/l, 50 × 109/l, 100 × 109/l, 150 × 109/l, 400 × 109/l by dilution with autologous PPP. To study the influence of phospholipids concentration on thrombin generation assay, frozen-thawed PPP from three normal volunteers was spiked with increasing concentrations of synthetic procoagulant phospholipids (0, 2, 4 and 8 μM).
Thrombogram-Thrombinoscope assay
In each well of a micro-plate, 80 μl of the studied plasma were mixed with 20 μl of diluted relipidated recombinant tissue factor (Hemoliance® RecombiPlasTin) from Instrumentation Laboratory (Milan, Italy). The RecombiPlasTin reagent was reconstituted by addition of 5 ml of NaCl 0.9%, and subsequently diluted in Hepes buffer pH 7.35; containing 20 mM Hepes, 140 mM NaCl and 5 mg/ml bovine serum albumin. The concentration of TF is not given by the manufacturer and for this reason TF is expressed as final dilution of recombiplastin in the plasma. The following final dilutions of recombiplastn were studied: 1/200, 1/500, 1/1000, 1/2000 which correspond to a range of calculated TF concentration from 3 pM to 30 pM in plasma (TF levels in RecombiPlasTin reagent were measured using an ELISA assay from American Diagnostics Inc; data not shown). No inter-lot variability of recombiplastin preparations was observed (data not shown). Thrombogram was also performed without any addition of TF.
Thrombin generation was triggered by adding a solution containing CaCl2 (16.7 mM final concentration) and the fluorogenic substrate (Z-Gly-Gly-Arg-AMC, 417 μM final concentration). A plate reader fluorometer (Fluoroskan Ascent, ThermoLabsystems, Helsinki, Finland) and the appropriate software (Thrombogram-Thrombinoscope™ assay; Synapse b.v., Maastricht, The Netherlands) were used. Generated thrombin was quantified using a thrombin calibrator kindly offered by Synapse b.v. Plasmas spiked with thrombin calibrator were run in parallel with each cycle of test samples. The following parameters were analyzed using the appropriate software provided by Synapse b.v.: a) the lag time of thrombin generation, b) the time to reach the maximum concentration (Tmax) of thrombin, c) the maximum concentration (Cmax) of thrombin and d) the total amount of thrombin activity assessed as the area under the curve (i.e. the endogenous thrombin potential – ETP).
Precision analysis of Thrombogram-Thrombinoscope assay
The intra-assay variability of each studied parameter of Thrombogram-Thrombinoscope was assessed by repeating the assay 6 times on the same assay plate for each studied dilution of recombiplastin and the inter-assay variability was assessed by repeating the assay 3 times, one after the other on different assay plates. Inter-individual variation was assessed by comparing the results obtained in 22 healthy subjects.
Statistical analysis
Paired and unpaired Student's t-test were used to compare the mean values of TG parameters in frozen and not frozen PRP. SPSS statistical software package was used for statistical analysis. Values are depicted as means ± sd.
Results
Intra-assay variability of thrombogram
When TF was added in PRP, the thrombogram parameters showed acceptable intra-assay variability being equal or lower than 9%. The performance of the assay was independent of TF concentration (Table 1).
Table 1 Precision analysis of each parameter of thrombin generation curve given by Thrombogram-Thrombinoscope assessed in fresh PRP in the presence of increasing TF concentrations or with out any TF addition. CV: coefficient of variation.
Recombiplastin final dilution in plasma and approximate TF concentration Lag-time Tmax Cmax ETP
intra-assay CV inter-assay CV intra-assay CV inter-assay CV intra-assay CV inter-assay CV intra-assay CV inter-assay CV
0 (0 pM) 7% 14% 5% 12% 5% 7% 4% 6%
1/2000 (3 pM) 4% 4% 4% 6% 6% 9% 4% 11%
1/1000 (6 pM) 4% 4% 3% 5% 6% 8% 3% 10%
1/500 (12 pM) 5% 6% 4% 6% 6% 8% 4% 5%
1/200 (30 pM) 6% 9% 5% 9% 3% 5% 4% 8%
Inter-assay variability of thrombogram
All thrombogram parameters showed low inter-assay variability when thrombin generation was assessed in the presence of TF in PRP. Decreasing TF concentration did not significantly influence the inter-assay variability of the thrombogram parameters. In contrast, when no TF was added in the experimental system the inter-assay variability of the lag-time and the Tmax was considerably high (14% and 12% respectively). In contrast the inter-assay variability of the Cmax and the ETP was not significantly influenced by the absence of TF (Table 1).
The intra-assay and inter-assay coefficients of variation of the studied parameters were lower than 5% when thrombin generation was studied in PRP spiked with UFH (0.1 IU/ml to 1 IU/ml), or enoxaparin (0.1 anti-Xa IU/ml to 1 anti-Xa IU/ml) or fondaparinux (0,1 μg/ml to 1 μg/ml) in the presence of low TF concentration (final dilution of recombiplastin 1/2000).
Inter-individual variability of thrombogram
Thrombogram parameters showed an important inter-individual variability and the coefficients of variation ranged from 18% to 45% being even more important when no TF was added. Among the studied parameters, the ETP and the Cmax showed the highest inter-individual variability mainly when no TF was added (Table 2).
Table 2 Inter-individual variability of thrombogram parameters in fresh PRP. Influence of TF concentration (n = 17). CV: coefficient of variations.
Recombiplastin final dilution in plasma and approximate TF concentration Lag-time (% CV) ETP (% CV) Cmax (% CV) Tmax (% CV)
0 (0 pM) 24% 31% 45% 25%
1/2000 (3 pM) 24% 25% 35% 24%
1/1000 (6 pM) 24% 25% 31% 24%
1/500 (12 pM) 22% 25% 27% 20%
1/200 (30 pM) 28% 25% 15% 17%
Influence of TF concentration on thrombogram in fresh PPP and PRP
Thrombin generation was triggered in fresh PRP from healthy individuals without any TF addition as well as in the presence of increasing TF concentrations (Figure 1). When thrombin generation was studied in PRP without any addition of TF a significant amount of thrombin was generated (Cmax = 108 ± 9 nM, ETP = 1094 ± 40 nM × min) with a lag-time of 14 ± 3.3 min and a Tmax of 19.5 ± 4.8 min. At recombiplastin dilution 1/2000, which corresponds to an approximate final concentration of TF of 3 pM, a threshold effect on the duration of the lag-time and the Tmax was observed. Increasing the concentration of TF resulted in an increase of the Cmax of thrombin. In contrast, the ETP was not influenced by the presence of TF, since the difference between the ETP values in the absence of TF and in the presence of recombiplastin dilution 1/2000 (which corresponds to approximately 3 pM final concentration of TF) was not statistically significant (1094 ± 340 nMxmin versus 1316 ± 255 nMxmin respectively; p > 0,05) (Figure 1). Table 3 depicts the normal values of Thrombogram parameters when coagulation is triggered in the presence of 1/1000 final dilution of recombiplastin (which corresponds to approximately 6 pM final concentration of TF).
Figure 1 Effect of TF concentration of thrombin generation parameters in PRP from 17 individuals (values are means ± sd). According to the dosage of TF in RecomBiplasTin reagent performed in our laboratory, thromboplastin 1/1000 final dilution in plasma correspond to 6 pM final TF concentration.
Table 3 Normal values and inter-individual variability of thrombogram parameters in the presence of 1/1000 dilution of TF concentration in fresh PRP and in frozen-thawed PRP, (n = 22; mean ± sd).
Parameters of thrombin generation Fresh PRP Frozen-thawed PRP
Normal value Inter-individual cv Normal value Inter-individual cv
Lag-time (min) 3.6 ± 0.8 24% 2 20%
ETP (nM × min) 1321 ± 330 25% 1550 ± 33 30%
Cmax (nM) 164 ± 50 31% 357 ± 1,66 25%
Tmax (min) 7.4 ± 1.8 24% 4,4 ± 0,1 25%
Effect of platelet concentration on Thrombogram
The effect of platelet count on thrombin generation was studied in the presence of the lowest studied concentration of TF (1/2000 dilution of recombiplastin in plasma), which showed an acceptable precision of the assay.
The lag-time of thrombin generation was similar (8 ± 3 min) in the presence of platelet counts 10 × 109/l and 50 × 109/l. In the presence of platelet count 100 × 109/l the lag-time was significantly shorter (5,8 ± 0,7 min) as compared to that observed in the presence of 50 × 109/l (8 ± 3 min; p < 0,05). No difference of the lag-time was observed in platelet counts ranging from 100 × 109/l to 400 × 109/l (5.5 ± 0.5 min) (Table 4).
Table 4 Influence of platelet concentration on the normal values of thrombin generation parameters assessed by Thrombogram-Thrombinoscope in the presence of 1/2000 final dilution of recombiplastin corresponding to estimated TF concentration of 3 pM. (n = 7; means ± sd, *p < 0,05 versus the respective value in the presence of platelet counts 10 × 109/l.
Platelet concentration (× 109/l) lag-time (min) Tmax (min) Cmax (nM) ETP (nM × min)
400 5 ± 0,5 11 ± 2,7 161 ± 38 1633 ± 81
150 5,5 ± 0,5 11 ± 0,2 98 ± 40 1316 ± 255*
100 5,8 ± 0,7* 13 ± 0,9 72 ± 38 1135 ± 300
50 8 ± 3,3 12 ± 0,2* 63 ± 14* 1095 ± 301
10 8 ± 3 19 ± 2 43 ± 16 894 ± 100
The Tmax in the presence of platelet count 10 × 109/l was 19 ± 2 min being significantly longer as compared to that observed in the presence of platelet count 50 × 109/l (12 ± 0.2 min; p < 0.05). In the presence of platelet counts ranging from 50 × 109/l to 400 × 109/l the Tmax of thrombin generation did not significantly change (Table 4).
The Cmax of thrombin was significantly decreased in the presence of platelet count 10 × 109/l, as compared to that observed in the presence of platelet count 50 × 109/l (43 ± 16 nM and 63 ± 14 nM respectively; p < 0.05). In the presence of platelet concentrations ranging from 50 × 109/l to 400 × 109/l a slight but not statistically significant increase of Cmax was observed (Table 4).
In the presence of platelet count 10 × 109/l the ETP was significantly lower as compared to that observed in the presence of platelet count 50 × 109/l (894 ± 100 and 1095 ± 301 respectively; p < 0,05). In the presence of platelet counts ranging from 150 × 109/l to 400 × 109/l no significant differences of the ETP were observed (Table 4).
Comparison of fresh and frozen-thawed platelet rich plasma for the assessment thrombogram
In thrombogram assessed in frozen-thawed PRP and in fresh PRP without any exogenous addition of TF both the lag-time and the Tmax were shorter by 45% and 51% and the Cmax of thrombin increased by 200% in frozen PPR as compared to the non frozen one (p < 0.05). The ETP was less affected by the freezing, since in frozen-thawed PRP it increased by only 6% as compared to that measured in fresh PRP (p > 0.05) (Figure 2).
Figure 2 Thrombogram parameters in the presence of thromboplastin dilutions in fresh PRP (white bars) and frozen-thawed PRP (shadowed bars) (values are means of 4 experiments).
In thrombogram assessed in the presence of TF, the ETP was the sole parameter, which was not significantly affected by the freezing (Figure 2). In frozen-thawed PRP, in the presence of each studied concentration of TF the lag-time decreased by 33% to 38% as compared to the non frozen one (p < 0.05). The Tmax was more markedly influenced by the freezing procedure mainly when minimal amounts of TF were employed. It was reduced by 47% in 1/1000 final dilution of recombiplastin in plasma, 38% in 1/500 final dilution of recombiplastin and 30% in 1/200 final dilution of recombiplastin as compared to the respective values in fresh PRP (p < 0.05). The freezing also affected the Cmax of thrombin, being significantly higher in frozen-thawed than in fresh PRP. The increase of the Cmax in frozen PRP was 119% in 1/1000 final dilution of recombiplastin, 81% in 1/500 final dilution of recombiplastin and39% in the presence of 1/200 final dilution of recombiplastin; in each case the difference was significant (p < 0.001) as compared to the respective values in fresh PRP.
The freezing procedure did not significantly influence the inter-individual variability of thrombogram assay when TF was added. In contrast freezing resulted in an important diminution of the intra-assay coefficient of variations of the ETP and Cmax when thrombogram was performed without addition of TF (Table 5).
Table 5 Intra-assay coefficients of variation of the studied parameters of thrombogram in fresh and frozen-thawed PRP (n = 4).
Lag-time ETP Cmax Tmax
TF dilution Fresh PRP Frozen-thawed PRP Fresh PRP Frozen-thawed PRP Fresh PRP Frozen-thawed PRP Fresh PRP Frozen-thawed PRP
TF 1/200 0 % 0% 3,5% 2,6% 1% 0,7% 3,5% 0%
TF 1/500 0% 0% 3% 1,6% 1,6% 1,1% 4,1% 4%
TF 1/1000 0% 0% 2,1% 2,1% 5,2% 0,5% 2,9% 2,8%
without TF 3,6% 8,1% 4,2% 2,7% 9,8% 1,7% 4,3% 5,1%
The normal values and the inter-individual coefficient of variations of the studied parameters of thrombin generation after addition of 1/1000 final dilution of recombiplastin (approximately 6 pM TF) in fresh and in frozen-thawed PRP are summarized in Table 3.
The influence of synthetic phospholipid concentration on thrombin generation in PPP
The addition of synthetic procoagulant phospholipids in PPP strongly influenced all the parameters of thrombogram (Figure 3). Both the lag-time and the Tmax of thrombin generation were strongly dependent on the presence of procoagulant phospholipids when no TF was added. In the presence of 3 pM of TF both parameters were influenced to a less significant degree by the concentration of phospholipids. In contrast the impact of the procoagulant phospholipids was important on both the ETP and the Cmax of thrombin either in the absence of TF or in the presence of low (3 pM) or higher concentrations (30 pM). For all the parameters of thrombogram a threshold effect was observed in the presence of phsopholipid concentrations equal or higher to 4 μM. However, since the Cmax appeared a slight, though not significant increase in the presence of synthetic phospholipid concentration of 8 μM, the optimum concentration of phospholipids is situated at 8 μM.
Figure 3 Influence of increasing concentrations of synthetic procoagulant phospholipids added in normal PPP. Thrombin generation was triggered in the presence of thromboplastin diluted in plasma 1/200 (30 pM TF) or 1/2000 (3 pM TF) or without any addition of TF. Values are means of 3 experiments.
Discussion
In the present study thrombin generation was assessed with the automated Thrombogram-Thrombinoscope assay. The precision of thrombogram, was evaluated in the presence of increasing TF concentrations, of synthetic procoagulant phospholipids, platelet count and freezing – thawing of platelet rich plasma from healthy volunteers. Normal values of thrombin generation in defined experimental, conditions which can be applied in routine haemostasis laboratory have also been established.
In the presence of a standardized platelet concentration (150 × 109/l) without any addition of TF all the studied parameters of thrombin generation can be measured. Thus we confirm that the presence of procoagulant phospholipids derived from platelets as in our study, is of major importance for steady thrombin generation [22]. The presence of platelets as is the case in our study, or the in vitro addition of negatively charged phospholipids in PPP [20] creates a more physiologically relevant experimental system.
The amplification of thrombin generation also depends on the concentration of the initial stimulus. According to the old scheme of blood coagulation (the cascade scheme) [23,24] one can trigger either the TF pathway or the contact system (intrinsic pathway) in order to study thrombin generation. However, the classic cascade scheme has a limited physiological relevance. It is now generally accepted that tiny amounts of TF, either blood born (from microparticles or activated monocytes) or exposed at the site of vascular injury, bind to FVII and FVIIa and the complex TF/FVIIa triggers blood coagulation [25]. Therefore we validated the Thrombogram-Thrombinoscope assay in TF-induced coagulation of normal plasma. As it was expected, in PRP the lag-time of thrombin generation onset, the Cmax of thrombin and the Tmax were strongly influenced by the TF concentration. The values of the lag-time, Cmax, Tmax and ETP in PRP did not significantly vary when physiologically relevant concentrations of TF, ranging from 1/2000 to 1/1000 dilution of recombiplastin, were used (which correspond to approximately 3 pM to 6 pM of TF).
Thrombogram assay performed in the presence of TF concentrations (3 pM to6 pM) which are considered to be physiologically relevant, had a low inter-assay and intra-assay coefficients of variation in PRP revealing that the precision of the assay is marginally influenced by the concentration of TF. However, all the studied parameters of thrombogram performed in PRP showed an important inter-individual variability, which was not improved by the addition of TF. Chantarangkul et al [20] showed that the presence of residual platelets into the PPP influences the imprecision of the ETP measurement. Our study shows that the inter-individual variability is important in fresh PRP, as well as in frozen-thawed PRP and is not influenced when the platelet count ranges from 10 × 109/l to 400 × 109/l (data not shown). A significant individual dependent variation in thrombogram has also been reported by Vanschoonbeek et al [26]. Thus it seems that individual functional characteristics of platelets, which are not completely understood today, as well as variations of the concentration of clotting factors and the natural inhibitors of blood coagulation [27] influence thrombin generation profile and they can be detected by the thrombogram assay. Platelet debris or microparticles present in PPP might have a major contribution to the observed significant interindividual variability of thrombin generation assay either in the absence or in the presence of TF since they exert a "non standardized" procoagulant activity. In order to eliminate this unpredictable factor it has been proposed that complete platelet depletion (i.e. by filtration of PPP) or addition of an excess concentration of phospholipids in PPP may improve the inter-individual variability of thrombogram [20]. We have to stress out that in the presence of very low TF concentration (3 pM) or in its absence, the CV values of the interindividual variability of thrombin's Cmax are considerably high. This finding allows to postulate that in conditions where the initial trigger of coagulation process is weak the experimental system is quite unstable. In such conditions the impact of artifacts such as platelet debris or probably the presence of procoagulant material such as microparticles is more important. It has been widely recognized that the major contribution of the Thrombogram-Thrombinoscope methodology is that it allows the study of thrombin generation in a more physiologically relevant system where platelets and fibrinogen/fibrin are present. Whether it is better to omit platelets from the experimental system has to be controlled in more detailed studies in larger groups of normal individuals and of patients with hyper- or hypo-coagulant states. In any case the large inter-individual variation could make more difficult the interpretation of the results.
Another important aspect of Thrombogram-Thrombinoscope assay is that the studied parameters are not significantly influenced by quantitative platelet variations within a quite wide range of platelet concentration (from 50 × 109/l to 400 × 109/l). As we have described in details elsewhere [28] and in accordance with the findings from Chantarangkul et al [20] thrombin generation is significantly reduced when platelet count is lower than 50 × 109/l. Consequently, deviations from the normal platelet concentrations do not significantly influence neither the precision of the assay nor the reference normal values.
Freezing procedure of PRP significantly reduced the lag-time and the Tmax and also increased the Cmax of generated thrombin but it did not significantly affect the ETP. Thus in frozen-thawed PRP thrombin generation is accelerated and the maximum amount of generated thrombin is increased apparently due to cold-induced platelet activation and/or procoagulant phospholipids release. However, the integral amount of thrombin generated in time, expressed by the ETP is not modified. In addition, freezing does not modify the coefficients of variation of all the studied parameters of thrombogram when thrombin generation was triggered in the presence of TF showing once more that the high interindividual variability of the assay is poorly related to technical conditions. Consequently ETP is the single parameter of Thrombogram which can be assessed in frozen-thawed PRP but weather its sensitivity to detect pathological hypercoagulable or hypocoagulable conditions needs to be examined. From a practical point of view, when clotting is triggered in PPP spiked with synthetic procoagulant phospholipids at concentrations higher than 4 μM in the presence of low TF concentration (1/2000 recombiplastin dilution in plasma) results in a similar pattern of thrombogram as that observed when PRP was used. In addition the ensemble of the presented data show that the ETP is influenced more by the concentration of procoagulant phospholipids rather than by the concentration of TF.
In conclusion, the present study shows that Thrombogram-Thrombinoscope assay performed in fresh platelet rich plasma has an acceptable precision, with low inter-assay and intra-assay coefficient of variations. The concentration of TF is determinant for the normal values of all the studied parameters except the endogenous thrombin potential. The addition of physiologically relevant concentrations of TF in platelet rich plasma (1/1000 to 1/2000 final dilution of recombiplastin) simulates a more natural experimental system from which however, the white and red blood cell as well as the rheological conditions are absent. Assessing thrombin generation in frozen-thawed PRP induces an important bias on the normal values of all the parameters of thrombogram except the ETP. In contrast, the thrombogram parameters obtained in frozen-thawed PPP supplemented with synthetic procoagulant phospholipids at concentrations equal or superior than 4 μM and physiologically relevant concentrations of TF is quite similar to that obtained in PRP in the presence of the same TF concentration. Regarding all the parameters of thrombin generation, the optimal concentration of the synthetic phospholipids employed in the present study is situated at 8 μM. The use of a standardized thrombin calibrator, which runs in parallel with the studied samples, allows the quantitative expression of thrombin in nM and ETP in nM × min. In addition, thrombogram software can reliably describe the distinct phases of thrombin generation (i.e the initiation phase, the Tmax and the maximum concentration of the generated thrombin). Among the studied parameters of thrombin generation given by the Thrombogram-Thrombinoscope assay, the ETP is less influenced by the concentration of TF as well as by the pre-analytical conditions (fresh PRP or frozen-thawed PRP) but it is strongly influenced by the concentration of phospholipids when it is performed in PPP. The influence of commercially available citrate preparations on thrombogram has to be evaluated in future studies. Thrombogram-Thrombinoscope assay is sensitive to the presence of UFH, enoxaparin and fondaparinux with low intra-assay and inter-assay variability. However, we have shown that the lag-time and the velocity of the propagation phase are more relevant than the ETP when an indirect FXa inhibitor (fondaparinux) or rFVIIa are present in therapeutic concentrations [3,28]. The method can be useful for the study of different antithrombotic agents whatever their mechanism of action and their target (LMWHs, direct and indirect inhibitors of FXa and direct inhibitors of thrombin [3,17,29,30]. Thus the ensemble of the parameters of thrombin generation can be potentially used for the laboratory diagnosis and biological of a large spectrum of hyper- and hypo-coagulant states. The high inter-individual variability of thrombogram is a topic that has to be studied in details in a large collaborative study.
List of abbreviations
Cmax: maximum concentration of thrombin
CV: coefficient of variation
ETP: endogenous thrombin potential
LMWH: Low Molecular Weight Heparin
PPP: Platelet Poor Plasma
PRP: Platelet Rich Plasma
TF: Tissue Factor
Tmax: time required to reach the maximum concentration of generated thrombin
Authors' contributions
GTG conceived of the study, organized its design and coordination, interpreted the data and drafted the manuscript. FD contributed to the design of the study and interpretation of the data. JB participated in the design of the study, the acquisition of the data and the statistical analysis. LL participated in the experiments with the synthetic phospholipids. IE participated to the critical reading of the manuscript. MMS participated to the design and coordination of the study, the interpretation of the data and contributed to the drafting of the manuscript.
Acknowledgements
The authors acknowledge Jonathan Mnich and Daphné Pierreeugene for their skilful technical assistance.
==== Refs
Hemker HC Beguin S Phenotyping the clotting system Thromb Haemost 2000 84 741 51
Mann KG Butenas S Brummel K The dynamics of thrombin formation Arterioscler Thromb Vasc Biol 2003 23 17 25 12524220 10.1161/01.ATV.0000046238.23903.FC
Gerotziafas GT Depasse F Chakroun T Van Dreden P Samama MM Elalamy I Comparison of the effect of fondaparinux and enoxaparin on thrombin generation during in vitro clotting of whole blood and platelet rich plasma Blood Coag Fibrinol 2004 15 149 56 10.1097/00001721-200403000-00006
Bara L Planes A Samama MM Occurrence of thrombosis and haemorrhage, relationship with anti-Xa, anti-IIa activities, and D-dimer plasma levels in patients receiving a low molecular weight heparin, enoxaparin or tinzaparin, to prevent deep vein thrombosis after hip surgery Br J Haematol 1999 104 230 40 10050702 10.1046/j.1365-2141.1999.01153.x
Mahe I Drouet L Chassany O Grenard AS Caulin C Bergmann JF Low molecular weight heparin for the prevention of deep venous thrombosis: a suitable monitoring in elderly patients? Pathophysiol Haemost Thromb 2002 32 134 6 12372928 10.1159/000065216
Heizmann M Baerlocher GM Steinmann F Horber FF Wuillemin W Anti-Xa activity in obese patients after double standard dose of nadroparin for prophylaxis Thromb Res 2002 106 179 81 12297122 10.1016/S0049-3848(02)00188-3
Provenzale L Casolo P Bifani I Value of an original thrombogram in the early diagnosis of thrombosis and the recognition of potential thrombosis Policlinico Chir 1952 59 257 79 13013843
Macfarlane RG Biggs R A thrombin generation test J Clin Pathol 1953 6 3 7 13034911
Pitney WR Dacie J A simple method of studying the generation of thrombin in recalcified plasma J Clin Pathol 1953 6 9 13 13034912
Seegers WH A personal perspective on hemostasis and thrombosis (1937–1981) Semin Thromb Hemost 1981 7 177 207 7034202
Houbouyan L Padilla A Gray E Longstaff C Barrowcliffe TW Inhibition of thrombin generation by heparin and LMW heparins: a comparison of chromogenic and clotting methods Blood Coagul Fibrinolysis 1996 7 24 30 8845459
Hemker HC Willems GM Beguinb S A computer assisted method to obtain the propthrombin activation velocity in whole plasma independent of thrombin decay processes Thromb Haemost 1986 56 9 17 2430342
Hemker HC Wielders S Kessels H Beguin S Continuous registration of thrombin generation in plasma, its use for the determination of the thrombin potential Thromb Haemost 1993 70 617 24 7509511
Kumar R Beguin S Hemker HC The effect of fibrin clots and clot-bound thrombin on the development of platelet procoagulant activity Thromb Haemost 1995 74 962 8 8571330
Hemker HC Giesen PL Ramjee M Wagenvoord R Beguin S The thrombogram: monitoring thrombin generation in platelet-rich plasma Thromb Haemost 2000 83 589 91 10780322
Wielders S Mukherjee M Michiels J Rijkers DT Cambus JP Knebel RW Kakkar V Hemker HC Beguin S The routine determination of the endogenous thrombin potential, first results in different forms of hyper- and hypocoagulability Thromb Haemost 1997 77 629 36 9134633
Hemker HC Giesen P Al Dieri R Regnault V de Smedt E Wagenvoord R Lecompte T Béguin S Calibrated Automated Thrombin Generation Measurement in Clotting Plasma Pathophysiol Haemost Thromb 2003 33 4 15 12853707 10.1159/000071636
Reverter JC Beguin S Kessels H Kumar R Hemker HC Coller BS Inhibition of platelet-mediated, tissue factor-induced thrombin generation by the mouse/human chimeric 7E3 antibody. Potential implications for the effect of c7E3 Fab treatment on acute thrombosis and "clinical restenosis" J Clin Invest 98 863 74 1996 Aug 1 8698879
Regnault V Beguin S Lecompte T Calibrated automated thrombin generation in frozen-thawed platelet-rich plasma to detect hypercoagulability Pathophysiol Haemost Thromb 2003 33 23 9 12853709 10.1159/000071638
Chantarangkul V Clerici M Bressi C Giesen PLA Tripodi A Thrombin generation assessed as endogenous thrombin potential in patients with hyper- or hypocoagulability Haematologica 2003 88 547 54 12745274
Chantarangkul V Clerici M Bressi C Tripodi A Standardisation of the endogenous thrombin potential measurement : how to minimise the effect of residual platelets in stored plasma Br J Haematol 2004 124 355 7 14717784 10.1046/j.1365-2141.2003.04771.x
Allen GA Wolberg AS Oliver JA Hoffman M Roberts HR Monroe DM Impact of procoagulant concentration on rate, peak and total thrombin generation in a cell based model syste J Thromb Haemost 2004 402 13 15009455 10.1111/j.1538-7933.2003.00617.x
Davie EW Ratnoff OD Waterfall sequence of intrinsic blood coagulation Science 1964 145 1310 2 14173416
MacFarland RG An enzyme cascade in the blood clotting mechanism, and its function as a biochemical amplifier Nature 1964 202 498 9 14167839
Davie EW A brief historical review of the waterfall/cascade of blood coagulation J Biol Chem 2003 278 50819 32 14570883 10.1074/jbc.X300009200
Vanschoonbeek K Feijge MA Van Kampen RJ Kenis H Hemker HC Giesen PL Heemskerk JW Initiating and potentiating role of platelets in tissue factor-induced thrombin generation in the presence of plasma: subject-dependent variation in thrombogram characteristics J Thromb Haemost 2004 2 476 84 15009466 10.1111/j.1538-7933.2004.00618.x
Butenas S van't Veer C Mann KG "Normal" Thrombin Generation Blood 1999 94 2169 78 10498586
Gerotziafas GT Chakroun T Elalamy I Depasse F Arzoglou P Samama MM The role of platelets and recombinant factor VIIa on thrombin generation, platelet activation and clot formation after TF pathway activation in human plasma Thromb Haemost 2004 91 977 85 15116259
Nagashima H Studies on the different modes of action of the anticoagulant protease inhibitors DX-9065a and Argatroban. I. Effects on thrombin generation J Biol Chem 2002 277 50439 44 12496240 10.1074/jbc.M208222200
Bostrom SL Hansson GF Kjaer M Sarich TC Effects of melagatran, the active form of the oral direct thrombin inhibitor ximelagatran, and dalteparin on the endogenous thrombin potential in venous blood from healthy male subjects Blood Coagul Fibrinolysis 2003 14 457 62 12851531 10.1097/00001721-200307000-00005
|
16250908
|
PMC1291409
|
CC BY
|
2021-01-04 16:36:24
|
no
|
Thromb J. 2005 Oct 26; 3:16
|
utf-8
|
Thromb J
| 2,005 |
10.1186/1477-9560-3-16
|
oa_comm
|
==== Front
BMC Infect DisBMC Infectious Diseases1471-2334BioMed Central London 1471-2334-5-931625314310.1186/1471-2334-5-93Research ArticleClinical presentation and prognostic factors of Streptococcus pneumoniae meningitis according to the focus of infection Østergaard Christian [email protected] Helle Bossen [email protected] Susanne [email protected] National Center for Antimicrobials and Infection Control, Statens Serum Institut, Artillerivej 5, Copenhagen S, Denmark2 Department of Bacteriology, Mycology and Parasitology, Statens Serum Institut, Artillerivej 5, Copenhagen S, Denmark3 Department of Epidemiology, Statens Serum Institut, Artillerivej 5, Copenhagen S, Denmark2005 27 10 2005 5 93 93 27 7 2005 27 10 2005 Copyright © 2005 Østergaard et al; licensee BioMed Central Ltd.2005Østergaard et al; licensee BioMed Central Ltd.This is an Open Access article distributed under the terms of the Creative Commons Attribution License (), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
Background
We conducted a nationwide study in Denmark to identify clinical features and prognostic factors in patients with Streptococcus pneumoniae according to the focus of infection.
Methods
Based on a nationwide registration, clinical information's was prospectively collected from all reported cases of pneumococcal meningitis during a 2-year period (1999–2000). Clinical and laboratory findings at admission, clinical course and outcome of the disease including follow-up audiological examinations were collected retrospectively. The focus of infection was determined according to the clinical diagnosis made by the physicians and after review of the medical records.
Results
187 consecutive cases with S. pneumoniae meningitis were included in the study. The most common focus was ear (30%), followed by lung (18%), sinus (8%), and other (2%). In 42% of cases a primary infection focus could not be determined. On admission, fever and an altered mental status were the most frequent findings (in 93% and 94% of cases, respectively), whereas back rigidity, headache and convulsion were found in 57%, 41% and 11% of cases, respectively. 21% of patients died during hospitalisation (adults: 27% vs. children: 2%, Fisher Exact Test, P < 0.001), and the causes of death were due to neurological – and systemic complications or the combination of both in 8%, 5% and 6% of cases, respectively. Other causes (e.g. gastrointestinal bleeding, incurable cancer) accounted for 2% of cases. 41% of survivors had neurological sequelae (hearing loss: 24%, focal neurological deficits: 16%, and the combination of both: 1%). The mortality varied with the focus of the infection (otogenic: 7%, sinusitic: 33%, pneumonic: 26%, other kind of focus: 50%, no primary infection focus: 21%, Log rank test: P = 0.0005). Prognostic factors associated with fatal outcome in univariate logistic regression analysis were advanced age, presence of an underlying disease, history of headache, presence of a lung focus, absence of an otogenic focus, having a CT-scan prior to lumbar puncture, convulsions, requirement of assisted ventilation, and alterations in various CSF parameters (WBC <500 cells/μL, high protein levels, glucose levels<1 mmol/L, low CSF/blood glucose levels), P < 0.05. Independent prognostic factor associated with fatal outcome in multivariate logistic regression analysis was convulsions (OR: 4.53, 95%CI: (1.74–11.8), p = 0.002), whereas presence of an otogenic focus was independently associated with a better survival (OR: 6.09, 95%CI: (1.75–21.2), P = 0.005).
Conclusion
These results emphasize the prognostic importance of an early recognition of a predisposing focus to pneumococcal meningitis.
==== Body
Introduction
Although treatment regimens for Streptococcus pneumoniae meningitis are still improving (e.g. adjunctive therapy with corticosteroids [1]), the mortality rate has not changed over half a century and remains as high as ~25% with neurological sequelae in up to half of survivors [2]. Pneumococcal meningitis is secondary to a primary infection focus (e.g. ear focus in ~30%, lung focus in ~25%, sinusitic focus in ~10%), and bacteraemia is present in up to 3/4 of all cases [3-6]. Consequently, the causes of death from pneumococcal meningitis may be multifactoral and due to both neurological complications (e.g. brain herniation, seizures) and systemic complications (e.g. septic shock, multiorgan dysfunction) [5,7,8]. Therefore, predisposing condition such as the focus of the infection may be of significant importance for the outcome of the disease [5,6,9]. However, no previous studies have to our knowledge addressed clinical features of pneumococcal meningitis according to the focus of infection.
Several risk factors associated with a poor clinical outcome of pneumococcal meningitis such as advanced age, presence of an underlying disease, pneumonia or bacteraemia, decreased mental status, delay in initiation of antibiotic therapy, and alterations in various CSF parameters (e.g. low CSF glucose levels, low CSF WBC) have been identified in previous studies, but have been inconsistent findings [3-5,9-17]. However, previous studies have been relatively small in size or did not apply multivariate statistics in their risk factor analysis, and may not be comparable due to differences in study population (e.g. children vs. adults, patients admitted to intensive care units vs. all cases), which may well explain conflicting results.
The aim of the present nationwide study of 187 consecutive Danish cases with pneumococcal meningitis over a two-year period (1999–2000) was to investigate clinical features according to the focus of infection. Moreover, we wanted to clarify whether the focus is an independent prognostic factor for the outcome of pneumococcal meningitis.
Methods
Identification of patients
Pneumococcal meningitis is a notifiable disease in Denmark, and all cases are reported to the Department of Epidemiology, Statens Serum Institut (SSI). In addition, all pneumococcal isolates obtained from CSF and blood are sent to the National Reference Laboratory at the Streptococcus Unit, SSI for serotyping and confirmatory antibiotic susceptibility testing. If a case with a positive CSF isolate were not reported initially, a request was sent out to ensure the reporting of the case. Therefore, the study represents a nationwide collection of consecutive cases with pneumococcal meningitis, and all notified patients with onset of pneumococcal meningitis during the period 1. January 1999 to 31. December 2000 were included in the study. Pneumococcal meningitis was defined as a CSF culture with S. pneumoniae or CSF pleocytosis (≥ 10 leukocytes/mL) in conjunction with a blood culture of S. pneumoniae [18].
Data collection
Clinical and laboratory information is provided with the notification form (i.e. diagnosis, time of onset of symptoms, clinical features, vaccination status, time of lumbar puncture and the results of CSF microscopy and CSF/blood culture), and hospital discharge records have prospectively been collected from all reported cases. Additional medical records including laboratory findings at admission, clinical course and outcome of the disease, as well as results from follow-up audiological examinations were collected retrospectively. Admission and discharge records including time of death were retrospectively controlled in the Danish civil registration database and in the national hospital registration database (Grønne system).
A total of 178 out of 188 isolates (95%) obtained from CSF and blood of the reported cases were referred to the National Reference Laboratory at the Streptococcus Unit, SSI for serotyping and confirmatory antibiotic susceptibility testing. Serotyping was performed by the Quelling reaction using type-specific pneumococcal rabbit antisera (Pneumosera®, SSI). Testing for penicillin susceptibility was performed with oxacillin (1 μg disk, AB Biodisk Solna, Sweden) with subsequent determination of MIC values for all isolates with reduced susceptibility using the E-test (AB Biodisk). The results of susceptibility testing performed at the local hospital laboratory were also collected from 7 isolates that were not sent to SSI.
Definition of the focus
The focus of infection was determined according to the clinical diagnosis made by the physicians, and was confirmed respectively by review of the medical records. 1) The presence of an otogenic focus was determined by otoscopic examination, often performed by an otologist, and did not require the isolation of pneumococci from middle ear fluid. 2) The presence of a sinusitic focus was determined after examination by an otologist and confirmed by cranial radiography of the sinuses including CT-scan/MR-scan, and did not require the isolation of pneumococci from sinus fluid. 3) The presence of a pneumonic focus was determined by clinical examination together with confirmatory chest radiography, and did not require isolation of pneumococci from sputum. 4) The presence of other foci was determined by clinical signs of infection together with isolation of pneumococci in samples from these foci. 5) No primary infection focus was considered, when no primary infection foci could be detected, and when pneumococci were not isolated from other body fluids than CSF and/or blood.
Ethics
All protocols were approved by the local scientific ethic committee and the Danish Data Protection Agency (#2002-41-2278).
Statistical analysis
All results are given as medians and interquartile range. Fisher's exact test was used for analysis of categorical data. For analysis of continuous data, comparison between two groups was performed by use of the non-parametric Mann-Whitney test and between more than two groups by use of the non-parametric Kruskal-Wallis test. For analysis between groups according to the focus, a Bonferroni correction of 10 was used to compensate for multiple comparisons. Survival was estimated by the methods of Kaplan-Meier and compared by the Log rank test. Relative risk for progression to death was calculated using univariate and multivariate logistic regression analysis. Only variables with more than 80% of values available and with a P-value <0.2 were tested in the multivariate analysis. P < 0.05 was considered significant.
Results
Identification of patients
A total of 187 cases with pneumococcal meningitis were identified and included in the study; one case patient was excluded, because the patient had no evidence of meningitis, and the isolation of an unencapsulated pneumococcal strain from the CSF was considered a laboratory contamination (growth in 1 out of 3 cultures). A lumbal puncture was performed on all 187 cases, except one case patient, where the diagnosis was established postmortem (autopsy showed purulent meninges and the growth of pneumococci was obtained from a swap taken from the meninges). A total of 176 out of 186 cases had a positive CSF culture with pneumococci, whereas 10 cases had a positive blood culture with pneumococci and biochemical evidence of meningitis including a CSF WBC count >10 leukocytes/mL. No significant difference in routine CSF parameters was observed between cases with or without a positive CSF culture (WBC: 1799 (290–4343) vs. 4288 (85–4713); protein: 2.9 (1.5–5.8) vs. 2.6 (1.2–8.0); glucose: 0.9 (0.3–2.6) vs. 3.5 (0.7–4.2), respectively, P > 0.05). The annual incidence rate of pneumococcal meningitis in Denmark was 1.73 cases per 100.000 persons.
Clinical characteristics according to the focus (see Table 1)
Table 1 Characteristics of 187 patients with S. pneumoniae meningitis according to the focus of the infection.
% or median (25/75 percentiles) and (No/total) All cases
(N = 187) Otogenic focus
(N = 57) Sinusitic focus
(N = 15) Pneumonic focus
(N = 33) Other foci
(N = 4) No primary infection focus
(N = 78)
Sex (female/male) 91/96 34/23 9/6 15/18 3/1 30/48
Age (in years) 55 (22–69) 56 (37–71) 66 (39–73) 58 (44–70) 75 (55–81) 49 (1–64)
<16 years of age 24% (45/187) 21% (12/57) 7% (1/15) 9% (3/33) 0% (0/4) 37% (29/78)
Predisposing condition* 12 % (23/185) 11% (6/57) 27% (4/15) 6% (2/33) 0% (0/4) 15% (11/76)
Underlying disease§ 18% (33/185) 14% (8/57) 20% (3/15) 36% (12/33) 25% (1/4) 12% (9/76)
Clinical features on admission
Fever 93% (155/166) 98% (55/56) 92% (12/13) 100% (27/27) 75% (3/4) 95% (63/66)
Headache 41% (48/116) 44% (17/39) 63% (5/8) 40% (6/15) 0% (0/3) 39% (20/51)
Back rigidity 57% (86/151) 65% (34/52) 75% (9/12) 41% (9/22) 25% (1/4) 54% (33/61)
Decreased consciousness 94% (165/176) 98% (53/54) 86% (12/14) 93% (28/30) 100% (4/4) 92% (68/74)
Convulsion
(Debut before/
after admission) 31% (54/175)
11% (19/175)
17% (29/175) 23% (13/56)
5% (3/56)
18% (10/56) 33% (5/15)
20% (3/15)
13% (2/15) 41% (12/29)
10% (3/29)
31% (9/29) 50% (2/4)
0% (0/0)
50% (2/4) 31% (22/71)
14% (10/71)
17% (12/71)
Duration of symptoms 2 days (2–4) (111/187) 2 days (1–4) (53/57) 3 days (2–7) (11/15) 3 days (2–8) (12/33) 3 days (1–5) (3/4) 2 days (2–3) (37/76)
Mechanical ventilation 58% (97/168) 49% (25/51) 50% (6/12) 77% (24/31) 100% (4/4) 54% (38/70)
CT-scan before lumbar puncture 11% (21/187) 7% (4/57) 13% (2/15) 12% (4/33) 0% (0/4) 14% (11/78)
Steroid therapy 16% (26/163) 14% (7/51) 8% (2/13) 19% (5/26) 0% (0/4) 17% (12/69)
Paraclinical findings CSF WBC (cells/μL) 1842 (291–4419) (149/187) 2844 (914–4553)## (50/57) 119 (41–5489) (12/15) 497 (66–2708) (24/33) 38 (1–235) (4/4) 2475 (850–4650)¤¤ (59/78)
CSF protein (g/L) 2.7 (1.4–5.8) (123/187) 3.1 (1.7–5.9) (44/57) 4.0 (1.0–9.0) (10/15) 2.8 (1.0–6.0) (18/33) 8.2 (0.5–10) (3/4) 2.3 (1.3–4.4) (48/78)
CSF glucose (mmol/L) 0.9 (0.4–2.8) (129/187) 0.9 (0.4–2.6) (45/57) 0.7 (0.1–3.0) (11/15) 0.9 (0.3–3.2) (21/33) 2.4 (0.6–3.7) (3/4) 0.9 (0.4–2.9) (49/78)
CSF/blood glucose ratio 0.1 (0.04–0.4) (93/187) 0.1 (0.05–0.3) (29/57) 0.05 (0.01–0.4) (9/15) 0.07 (0.01–0.4) (18/33) 0.5 (0.09–0.9) (2/4) 0.2 (0.04–0.5) (35/78)
Positive CSF culture 95% (176/186) 93% (53/57) 100% (15/15) 91% (29/32) 100% (4/4) 96% (75/78)
Blood WBC (109 cells/L) 17.3 (10.5–25.7) (134/187) 17.6 (10.5–24.1) (29/57) 20.0 (8.9–26.2) (12/15) 12.6 (7.9–22.9) (23/33) 10.2 (6.7–19) (3/4) 17.9 (14.1–27.7) (50/78)
Positive blood culture 67% (124/186) 74% (42/57) 60% (9/15) 72% (23/32) 100% (4/4) 60% (46/78)
Decreased penicillin susceptibility 6% (10/183) 7% (4/57) 7% (1/14) 0% (0/32) 0% (0/4) 7% (5/75)
Death during hospitalisation 21% (39/187) 7% (4/57)** 33% (5/15) 26% (12/33) 50% (2/4) 21% (16/78)
1. Neurological causes# 8% (16/187)$ 3.5% (2/57) 13% (2/15) 12% (4/33) 0% (0/4) 10% (8/78)
2. Systemic causes¤ 5% (9/187)$ 0% (0/57) 13% (2/15) 9% (3/33) 25% (1/4) 4% (3/78)
3. Other causes& 2% (3/187) 0% (0/57) 0% (0/15) 9% (3/33) 0% (0/4) 0% (0/78)
Combination of 1 and 2 6%(11/187)$ 3.5% (2/57) 7% (1/15) 6% (2/33) 25% (1/4) 7% (5/78)
Sequelae 41% (57/138) 54% (26/48) 22% (2/9) 45% (9/20) 100% (1/1) 32% (19/60)
1. Hearing loss 24% (34/138) 33% (16/48) 11% (1/9) 20% (4/20) 0% (0/1) 22% (13/60)
2. Neurologic abnormalityˆ 16% (22/138) 21% (10/48) 11% (1/9) 25% (5/20) 100% (1/1) 8% (5/60)
Combination of 1 and 2 1% (1/138) 0% (0/48) 0% (0/9) 0% (0/20) 0% (0/1) 2% (1/60)
Number of days hospitalised among survivors 13 (10–20) 13 (10–19) 11 (10–36) 22 (13–32) 44 13 (11–16)
* Predisposing condition was defined a previous head trauma, liquorrhoea, dura disruption etc.
§Underlying disease was defined as previous splenectomy, presence of immunodeficit, cancer, diabetes mellitus, alcoholism, or the use of immunosuppressive drugs.
# Includes brain herniation, cerebrovascular complications.
¤ Includes septic shock, multiple-organ dysfunction.
& Includes withdrawal of care due to incurable cancer (1 patient), gastrointestinal bleeding (+/- Bilroth II operation, 2 patients).
ˆNeurologic abnormality was defined as presence of aphasia, ataxia and paresis at discharge.
$ 1 patient also had gastrointestinal bleeding/Billroth II operation.
##Significant difference vs. sinusitic cases, pneumonic cases, and cases with other foci (Mann Whitney test with Bonferonis correction P < 0.05).
¤¤ Significant difference vs. sinusitic cases, and cases with other foci (P < 0.05).
**Significant difference vs. pneumonic cases (Fisher Exact test with Bonferonis correction, P = 0.01).
A total of 57 out of 187 patients with pneumococcal meningitis (30%) were identified as having an otogenic focus (10 patients had mastoiditis), 15 out of 187 patients had sinusitis (8%), 33 out of 187 patients had pneumonia (18%), 4 out of 187 patients had an alternative foci (2%, e.g. septic arthritis, subphrenic abscess), whereas no primary infection focus could be found in 78 out of 187 patients (42%). A total of 13 out of the 78 patients, where no primary infection focus could be determined, had no reports of an otoscopic examination. However, none of these patients had a history of earache or suppuration from the ear, and a CT-scan showed no ear focus in 4 of these patients. Six patients with an otogenic focus had an additional focus (1 patient had sinusitis and 5 patients had pneumonia; one of these patients died), whereas 3 patients with sinusitis also had pneumonia (one of these patients died). Other predisposing conditions than the focus of the infection (e.g. previous head trauma, liquorrhoae, dura disruption) were present in 12% of cases with the highest frequency among cases with a sinusitic focus (27%), among cases, where no primary infection focus was found (15%) and among cases with an otogenic focus (11%). The retrospective review of medical records showed that 8 out of 187 cases were initially misclassified in the prospective registration.
There was no significant difference in clinical and demographic characteristic between the 5 focus groups (see Table 1), except with differences in the case fatality rate between groups (see below) and in CSF WBC (Kruskal Wallis test, P < 0.001). CSF WBC counts were significantly higher in otogenic cases than in sinusitic cases, than in pneumonic cases, and than in cases with other foci, and were significantly higher in cases, where no primary infection focus was determined as compared to sinusitic cases and cases with other foci (see Table 1, Mann Whitney test with Bonferonies correction, P < 0.05). The median age of the patients was not significantly different between groups, and the focus of infection was only in part dependent on age groups (<16 vs. ≥ 16 years): otogenic focus (27% (12/45) vs. 32% (45/142), respectively, P = 0.58), sinusitic focus (2% (1/45) vs. 10% (14/142), respectively, P = 0.12), pneumonic focus (8% (3/45) vs. 20% (30/142), respectively, P = 0.03), other foci (0% (0/45) vs. 3% (4/142), respectively, P = 0.57), and no primary infection focus (64% (29/45) vs. 35% (49/142), respectively, P = 0.0005). Not surprisingly was headache more frequently reported in adults than in children, and adults had more frequently an underlying disease, required more frequently assisted ventilation and had a poorer outcome than children (see Table 2, P < 0.05).
Table 2 Characteristics of 187 patients with S. pneumoniae meningitis according to age groups
% (No/total) or Median (25/75 percentiles) Adults (≥ 16 years) (N = 142) Children (<16 years) (N = 45)
Sex (female/male) 71/71 20/25
Age 61 years (50–71) 12 month (7–18)
Predisposing condition* 13% (19/140) 9% (4/45)
Underlying disease§ 23% (32/140) 2% (1/45)**
Clinical features on admission
Fever 97% (119/123) 95% (41/43)
Headache 57% (47/82) 3% (1/34)**
Back rigidity 61% (67/110) 46% (19/41)
Decreased consciousness 96% (131/137) 87% (34/39)
Convulsion
(Debut before/
after admission) 28% (37/130)
6% (8/130)
22% (29/130) 37% (17/45)
24% (11/45)**
13% (6/45)
Duration of symptoms 2 days (2–4) (83/142) 2 days (2–6) (28/45)
Mechanical ventilation 67% (86/128) 28% (11/40)**
CT-scan preceding lumbar puncture 13% (19/142) 4% (2/45)
Steroid therapy 7% (8/121) 43% (18/42)**
Paraclinical findings CSF WBC (cells/μL) 2475 (122–4659) (109/142) 1690 (775–2622) (40/45)
CSF protein (g/L) 3.7 (2.1–6.9) (91/142) 1.6 (0.8–2.2)** (32/45)
CSF glucose (mmol/L) 0.9 (0.3–2.5) (95/142) 1.4 (0.4–3.2) (34/45)
CSF/blood glucose ratio 0.09 (0.02–0.3) (72/142) 0.2 (0.07–0.6) (21/45)
Positive CSF culture 93% (131/141) 100% (45/45)
Blood WBC (109 cells/L) 15.9 (10.1–24.8) (105/142) 20.3 (12.8–29.2) (29/45)
Positive blood culture 67% (94/141) 67% (30/45)
Decreased penicillin susceptibility 4% (5/139) 11% (5/45)
Death during hospitalization 27% (38/142) 2% (1/45)**
1. Neurological causes# 11% (15/142)$ 2% (1/45)
2. Systemic causes¤ 6% (9/142)$ 0% (0/45)
3. Other causes& 2% (3/142) 0% (0/45)
Combination of 1 and 2 8% (11/142) 0% (0/45)
Sequelae 52% (50/96) 17% (7/42)**
1. Hearing loss 30% (29/96) 12% (5/42)
2. Neurologic abnormalityˆ 22% (21/96) 2.5% (1/42)
Combination of 1 and 2 0% (0/96) 2.5% (1/42)
Number of days hospitalised among survivors 15 (11–22) 11 (10–14)**
* Predisposing condition was defined a previous head trauma, liquorrhoea, dura disruption etc.
§Underlying disease was defined as previous splenectomy, presence of immunodeficit, cancer, diabetes mellitus, alcoholism, or the use of immunosuppressive drugs.
# Includes brain herniation, cerebrovascular complications.
¤ Includes septic shock, multiple-organ dysfunction.
& Includes withdrawal of care due to incurable cancer (1 patient), gastrointestinal bleeding (+/- Bilroth II operation, 2 patients). ˆNeurologic abnormality was defined as presence of aphasia, ataxia and paresis at discharge.
$ 1 patient also had gastrointestinal bleeding/Billroth II operation.
**Significant difference vs. adult cases (Mann Whitney test or Fisher Exact Test, P < 0.05.
On admission, fever and an altered mental status were present in almost all cases with pneumococcal meningitis (93% and 94%, respectively), whereas back rigidity, headache, and convulsion were significantly less frequent findings (57%, 41%, and 11%, respectively, P < 0.0001). Twenty-one case patients (11%) had a diagnostic CT-scan before the lumbar puncture. These patients, except one patient, who received antibiotic therapy with ceftriaxone, were initially suspected of having another intracranial disease than meningitis (e.g. apoplexy), resulting in a delay in the establishment of a correct diagnosis and initiation of antibiotic therapy. Initial or empiric antibiotic therapy included penicillin and/or a 3. generation cephalosporin and was changed according to the susceptibility of the pathogen to penicillin in most cases. Adjunctive therapy with corticosteroids was given to 16% of cases with a significantly higher treatment rate among children than among adults (see Table 2, P < 0.001).
Case fatality rate
The case fatality rate during the first 100 days after admission varied according to the focus of infection (Log rank test, P = 0.0005, see Figure 1). Patients with an otogenic focus had a significant lower case fatality rate than non-otogenic cases (within 14 days: 7% vs. 21%, Log-rank test: P = 0.03, within 31 days: 7% vs. 22%, P = 0.02, and within 100 days: 7% vs. 29%, P = 0.002). Adult patients had a significantly higher case fatality rate than children (see Figure 2, and Table 2, P = 0.0001). The case fatality rate according to serotype distribution is shown in Figure 3, and a trend was seen for different case fatality rates among serotypes. The most frequent serotypes according to the focus were as follows: Otogenic focus: serotype 8(6), 3(4), 6B(4), 19F(4), 12F(3), and 7F(3). Sinusitic focus: serotype 7F(3), 6B(2), 10A(2), and 35F(2). Pneumonic focus: serotype 4(5), 14(4), 7F(3), 8(3), and 9V(3). Other kind of focus: serotype 6A(1), 12F(1), 14(1), and 22F(1). No primary infection focus: serotype 7F(8), 6B(7), 12F(6), 23F(6), 6A(5), 9V(5), and 18C(5). Ten out of 184 isolates (5%) had reduced susceptibility for penicillin (serotype 5 (2), 6B (1), 9N (1), 9V (1), 14 (1), 15B (2), and 19F (2); no data available for 3 isolates). No significant differences between case fatality rate penicillin susceptibility were observed (P > 0.05).
Figure 1 Kaplan Meier Survival curve of 187 patients with S. pneumoniae meningitis in Denmark 1999–2000 according to the focus of the infection. Otogenic focus vs. pneumonic focus, sinusitic focus, other foci, and no primary infection focus: Log rank test: P = 0.0002, 0.008, <0.0001, and 0.03, respectively. Other foci vs. no primary infection focus: P = 0.01.
Figure 2 Age distribution and mortality of 187 patients with S. pneumoniae meningitis in Denmark 1999–2000.
Figure 3 Serotype distribution and mortality of 187 patients with S. pneumoniae meningitis in Denmark 1999–2000. *10 isolates were not capsular serotyped.
Two out of every 3 deaths due to pneumococcal meningitis occurred within the first week after admission, and all patients died during hospitalisation, except 2 patients; 1 with severe brain damage (vegetative state) and 1 one with severe cardiovascular complications, who died 57 and 61 days after admission, respectively. These two patients were transferred to a nursing home for a short period before time of death/readmission and death at the hospital. The causes of death during hospitalisation was as follows: 41% died due to neurological causes (e.g. brain herniation, cerebrovascular complications), 23% died due to systemic causes (e.g. septic shock, multiple-organ dysfunction), 8% died due to other causes (e.g. gastrointestinal bleeding, incurable cancer) and 28% died due to a combination of systemic and neurological complications. Gastrointestinal bleeding was observed in a total of 5 patients dying from pneumococcal meningitis (2 patients had a Billroth II operation); none of these were treated with steroids. The median time to death due to neurological -, systemic -, other complications or the combination of neurological and systemic complications was 2 days (3–30), 7 days (4–9), 47 days (11–61), and 6 days (3–10), respectively (Kruskal Wallis test P = 0.07).
Prognostic factors
Patients with advanced age, with an underlying disease, without an otogenic focus, with a lung focus and with convulsions, as well as patients, who had a CT-scan before lumbar puncture, and who required assisted ventilation were at increased risk for fatal outcome (within 100 days) in univariate logistic regression analysis (P < 0.05, see Table 3). Also, an association was found between case fatality rate and various alterations in CSF parameters (WBC< 500 cells/μL, high protein levels, glucose levels <1 mmol/L, low CSF/blood glucose ratios), P < 0.05. In the multivariate analysis (not including assisted ventilation therapy and variables with less than 80% available values), prognostic factors for fatal outcome due to pneumococcal meningitis were convulsions, whereas presence of an otogenic focus was independently associated with a better survival (see Table 3, P < 0.05). Stepwise inclusion of other possible prognostic factors did not weaken the association between a better survival and presence of an otogenic focus (data not shown) Independent prognostic factors for fatal outcome during hospitalisation were similar to the 100 days calculation (data not shown), and independent risk factors for dying within 31 days after admission were having a CT-scan preceding lumbar puncture (OR: 4.13 (1.36–12.6), P = 0.01), and absence of an otogenic focus (OR: 3.61 (1.09–12.0), P = 0.036). Not surprisingly, assisted ventilation was independently associated with fatal outcome (OR: 9.45 (3.18–28.1), P < 0.001) and when included in the multivariate analysis, convulsion and presence of an otogenic focus were still significant associated with a higher and lower case fatality rate at 100 days, respectively (OR: 3.15 (1.11–8.94), P = 0.03 and OR: 6.68 (1.71–26.1), P = 0.006, respectively).
Table 3 Prognostic clinical parameters for fatal outcome# due to S. pneumoniae meningitis
Univariat analysis Multivariate
% or median (25/75 percentiles) and (No/total) Non-survivors N = 41 Survivors N = 146 OR¤ P-value OR¤ P-value
Sex (female) 51% (21/41) 48% (70/146) 1.14 (0.57–2.28) 0.71
Age (in years) 67 (56–74) 50 (1–67) 1.036 (1.02–1.06) <0.001 1.03 (1.00–1.07) 0.05
Age ≤ 16 years 2% (1/41) 30% (44/146) 0.06 (0.008–0.44) 0.006 0.36 (0.02–7.13) 0.50
Predisposing condition* 10% (4/40) 13% (19/145) 0.74 (0.24–2.30) 0.60
Underlying disease§ 30% (12/40) 15% (21/145) 2.53 (1.12–5.74) 0.03 1.72 (0.58–5.08) 0.33
Admission during the first half of the year 83% (34/41) 68% (99/146) 2.30 (0.95–5.58) 0.06 1.27 (0.41–3.97) 0.68
Fever 95% (35/37) 97% (125/129) 0.56 (0.10–3.19) 0.61
History of headache 62% (13/21) 38% (35/95) 2.79 (1.05–7.38) 0.04
Back rigidity 55% (16/29) 57% (70/122) 0.91 (0.40–2.07) 0.84
Decreased consciousness 98% (40/41) 93% (125/135) 3.20 (0.40–25.8) 0.46
Convulsion 53% (19/36) 25% (35/139) 2.32 (1.07–5.04) 0.03 4.53 (1.74–11.8) 0.002
Mechanical ventilation& 90% (35/39) 48% (62/129) 9.45 (3.18–28.1) <0.001
CT-scan preceding lumbar puncture 22% (9/41) 8% (12/146) 3.14 (1.22–8.01) 0.02 2.62 (0.82–8.41) 0.11
Steroid therapy 9% (3/32) 18% (23/131) 0.49 (0.14–1.73) 0.27
Bacteraemia 78% (31/40) 64% (93/146) 1.96 (0.87–4.43) 0.11 2.02 (0.70–5.84) 0.19
Otogenic focus 10% (4/41) 36% (53/146) 0.19 (0.06–0.56) 0.003 0.16 (0.05–0.57) 0.005
Lung focus 39% (16/41) 17% (25/146) 3.10 (1.45–6.63) 0.004 1.46 (0.53–3.97) 0.46
CSF WBC (109 cells/L) 0.32 (0.07–3.1) (33/41) 11 (0.7–4.7) (116/146) 0.88 (0.77–1.01) 0.06
500 cells/μL 55% (18/33) 22% (26/116) 4.15 (1.84–9.36) 0.001
CSF protein (g/L) 4.7 (2.7–8.9) (27/41) 2.4 (1.3–4.8) (96/146) 1.19 (1.07–1.33) 0.001
CSF glucose (mmol/L) 0.4 (0.1–1.6) (29/41) 1.0 (0.4–3.1) (100/146) 0.83 (0.63–1.08) 0.17
<1 mmol/L 76% (22/29) 52% (52/100) 3.41 (1.33–8.69) 0.01
CSF/blood glucose ratio 0.05 (0.01–0.11) (21/41) 0.18 (0.06–0.5) (72/146) 0.01 (0.00–0.36) 0.01
Blood WBC (109 cells/L) 14.8 (9.5–24.1) (28/41) 17.7 (10.8–26.8) (106/146) 0.98 (0.94–1.02) 0.34
# Within 100 days after admission.
¤ OR was calculated per additional units for continuous data.
* Predisposing condition was defined a previous head trauma, liquorrhoea, dura disruption etc.
§Underlying disease was defined as previous splenectomy, presence of immunodeficit, cancer, diabetes mellitus, alcoholism, or the use of immunosuppressive drugs.
& If need for assisted ventilation was included in the multivariate analysis convulsions, absence of otogenic focus, and need for assisted ventilation were of significant importance for fatal outcome (p < 0.05, data not shown).
Sequelae among survivors of pneumococcal meningitis
Sequelae were present in 41% (57/138) of survivors with various degree of hearing loss in ~25% of survivors, and motor sequelae such as ataxia and paresis were present in ~16% of survivors. Among survivors, adult patients had significantly more frequent sequelae than children (see Table 2, P < 0.001), and the median age was significantly higher among patients with – than without sequelae (57 years (50–71) vs. 29 years (1–61), respectively, P < 0.001). Sequelae among survivors were significantly more frequent in patients, who required assisted ventilation (65% (37/57) vs. 23% (15/65), P < 0.001), in patients, who were not treated with corticosteroids 47% (47/100) vs. 13% (3/23), P = 0.004), and patients with sequelae had significantly higher CSF protein levels (3,0 (2.1–4.6) vs. 1.6 (1.0–5.5), P = 0.016 and lower CSF glucose levels 0.7 (0.3–1.6) vs. 1.9 (0.6–3.4), P = 0.002, respectively) than in patients without sequelae.
Discussion
The present nationwide study of 187 consecutive cases with pneumococcal meningitis in Denmark over a 2-year period (1999–2000) shows that the overall case fatality rate was 21% with a 10-fold higher mortality rate among adults than among children, which is comparable with recent reports from other industrialised countries [5,6,9,19-21]. The case fatality rate according to age groups seems to resemble previous findings obtained from patients with pneumococcal meningitis admitted to a tertiary Danish hospital during the period 1966–1976 [4] indicating that the epidemiology as well as the mortality of pneumococcal meningitis has not changed significantly over a 30–40 years period. Also, the present study confirms that predisposing conditions (e.g. dura leakage and associated foci) are found in 2 out of every 3 patients with pneumococcal meningitis [3-5]. An otogenic focus was the most frequent focus (30%) followed by a pneumonic focus (18%), a sinusitic focus (8%), whereas no primary infection focus was found in 42% of cases with pneumococcal meningitis. A high proportion of cases were children, when no primary infection focus was detected, despite of thoroughly examination for other infection foci, which could indicate that nasopharyngeal colonisation with pneumococci may be a risk factor for developing pneumococcal meningitis particular in children.
The case fatality rate varied according to the focus of infection with a significantly lower case fatality rate among otogenic cases (7%) as compared to non-otogenic cases of pneumococcal meningitis (27%). Others have reported similar trends for a better outcome of otogenic pneumococcal meningitis [4]. A lung focus was associated with a higher case fatality rate (26%) as reported by other studies [4,5,14]. Underlying diseases could be a likely explanation, why patients with a lung focus have a poorer outcome than patient with an otogenic focus (36% vs. 14%, respectively, with Bonferonis correction: P = 0.19). However, an otogenic focus remained an independent prognostic factor after inclusion of other prognostic factors in the multivariate analysis.
Whilst several studies have investigated prognostic factors in pneumococcal meningitis, most of these studies have been smaller in size than the present study or have not applied multivariate statistics in their risk factor analysis [3-5,9-17]. The identification of other independent risk factors for fatal outcome than the focus of the infection in the multivariate analysis (convulsions) and in the univariate analysis (advanced age, having an underlying disease, presence of a lung focus, having a CT-scan prior to lumbar puncture, need for assisted ventilation, alterations in various CSF cytochemical parameters (e.g. low WBC counts, high protein and glucose levels)) confirms in part findings inconsistently obtained in previous studies [3-6,9-17].
Almost all patients with pneumococcal meningitis had fever and an altered mental status on admission (~94%), whereas nucal rigidity, the "classic sign" of meningitis, was present in 57% of patients. Consequently, some patients were admitted to the hospital suspected of having other kind of intracranial disease (e.g. apoplexy) than meningitis, and a diagnostic CT-scan was performed before the lumbar puncture with a subsequent delay in establishing the meningitis diagnosis and initiation of antibiotic therapy. Others have previously shown that a delay in the initiation of antibiotic therapy worsen the clinical outcome of community-acquired bacterial meningitis [22]. Thus, the results support international recommendation that patients presenting with fever and altered mental status should promptly have a diagnostic lumbar puncture or be given empirical antibiotic therapy and corticosteroids, and be sampled for blood culture before the CT-scan [23].
Less than 10% of patients were infected with pneumococcal strains with reduced susceptibility to penicillin, which was not related to a poorer outcome. Similar results have previously been found in countries with low or high levels of penicillin resistance [6,12,24]. Previously, we have shown that serotype-related differences (between serotype 1, 3, and 9 V) in mortality of pneumococcal meningitis exists [25]. The results from the present study may indicate that a relationship to mortality may exist for other serotypes, however, this has to be addressed in a larger patient population. The serotype coverage rate by the 7-, 9-, and 11-valent pneumococcal conjugate vaccine were 67%, 72%, and 86%, respectively, among children less than 2 years of age, and by the 23-valent vaccine it was 92% among adults, which is in accordance with previous published results obtained in Denmark during the period 1995–1999 [26]. No reliable data could be extracted from the present study on the vaccination status of the individual patients against pneumococcal infection due to incomplete data registration, but a more frequent use of pneumococcal vaccination may likely reduce the incidence of pneumococcal meningitis in Denmark, since vaccination of children and adults has reduced the risk for developing invasive pneumococcal disease in other countries [27,28].
The causes of death in patients with pneumococcal meningitis are multifactoral and involve both neurological and systemic complications [5]. Three patients died due to other causes (two patients due to gastrointestinal bleeding and one patient, because active therapy was stopped due of incurable cancer). Interestingly, no significant differences were observed in the causes of death between the five different groups of infection foci. Two out of every three patients died within the first week after admission, however, death directly related to pneumococcal meningitis still occurred up to 96 days after admission. Others have previously found that a mortality at 14 days as the endpoint is optimal to use when studying community-acquired bacterial meningitis [18]. Our results here and in a previous study suggest a longer study period, when studying case fatality in pneumococcal meningitis [25].
The frequency of sequelae due to pneumococcal meningitis was 41% with hearing loss in 26% of survivors and neurological sequelae in 16% of survivors, confirming previous results [1,4,6,10]. Sequelae occurred more frequently in adults than in children, most likely because pneumococcal meningitis is a less severe disease in children than in adults, as shown in the present study, or due to a higher susceptibility for the host to develop sequelae with increasing age. On the other hand, sequelae were detected more frequently in otogenic cases than in non-otogenic cases, despite a lower mortality rate in otogenic cases. However, there is one major bias in the evaluation of sequelae among meningitis patients, in particular with the evaluation of hearing loss, since it is not possible to compare the outcome with baseline parameters obtained before the onset of meningitis, which may have a significant impact on the interpretation of the results.
Whereas a beneficial effect of adjunctive therapy with corticosteroids on mortality did not reach statistical significance, surviving patients, who were treated with corticosteroids, had a significantly lower risk of developing sequelae than patients, who were not treated with corticosteroids. International guidelines and results from well-designed randomised trials have recommended the use of corticosteroids as adjunctive therapy of pneumococcal meningitis, when administered before or together with the first dose of antibiotics [1,23,29], but information's on the exact timing of corticosteroid administration could not be extracted from the medical records in the present study. Moreover, the majority of patients, who received corticosteroids in the present study, were children, because the study period was before the publication of the results of the European Dexamethasone in Adulthood Bacterial Meningitis Study in 2002 [1] and before the recommendation in Denmark to use corticosteroids as adjunctive therapy in adult bacterial meningitis [30]. Indeed, this affected the results significantly, and no beneficial effect of corticosteroid therapy was observed, when sub-analysis among children or adult cases were performed (P > 0.05, data not shown). However, the case fatality rate of pneumococcal meningitis will be expected to be lower in future studies performed after the implementation of corticosteroids as adjunctive therapy to all cases of bacterial meningitis.
Although new knowledge in the pathogenesis of pneumococcal meningitis has been obtained experimentally [31], the exact infection route into the CNS is not fully determined. Pneumococci may spread directly from an adjacent focus or via the haematogenously route into the CNS. In the present study, bacteraemia was present in 2 out of every 3 cases, and no significant difference in frequency of bacteraemia was observed between the 5 different groups of infection foci, indicating that pneumococci predominantly may spread haematougenously. However, bacteraemia also occurs secondary to meningitis, as shown in animal models after intracisternal inoculation [25]. Therefore, clinical studies with quantitative CSF and blood cultures from cases with different foci of the infection are required to provide further insight in the pathogenesis of pneumococcal meningitis.
The present study has limitations because not all data was collected prospectively, and several variables were unavailable from a number of cases leading to a decreased statistical power of the multivariate analysis. Also, we could not detect an association between degree of consciousness and case fatality rate, most likely because Glasgow Coma Scale tests were not performed on the patients, and thereby missing one of the most important predictors of a poor clinical outcome [5,6]. The results of the present study should be confirmed in future studies (i.e. in subanalysis of pneumococcal cases from prospective studies with less missing values [6]).
Conclusion
Our results emphasize the prognostic importance of early recognition of a primary infection focus in pneumococcal meningitis. Meningitis should also be considered in patients presenting without nucal rigidity but with fever and altered mental status. Finally, a diagnostic lumbar puncture or start of antibiotic therapy and blood culture sampling should not be delayed by a CT-scan.
Competing interests
The author(s) declare that they have no competing interests.
Contribution of authors
CØ provided the scientific idea of the present study, designed the study, collected patient data, made the statistical analysis, and drafted the manuscript. HBK provided serotype data and participated in the design of the study. SS provided patient data from the notification reports and participated in the design of the study. All authors approved the final manuscript.
Pre-publication history
The pre-publication history for this paper can be accessed here:
Acknowledgements
The authors would like to thank Thomas Benfield and Niels Høiby for critical review of the manuscript.
==== Refs
de Gans J van de Beek D Dexamethasone in adults with bacterial meningitis N Engl J Med 2002 347 1549 1556 12432041 10.1056/NEJMoa021334
Swartz MN Bacterial meningitis--a view of the past 90 years N Engl J Med 2004 351 1826 1828 15509815 10.1056/NEJMp048246
Auburtin M Porcher R Bruneel F Scanvic A Trouillet JL Bedos JP Regnier B Wolff M Pneumococcal meningitis in the intensive care unit: prognostic factors of clinical outcome in a series of 80 cases Am J Respir Crit Care Med 2002 165 713 717 11874820
Bohr V Rasmussen N Hansen B Gade A Kjersem H Johnsen N Paulson O Pneumococcal meningitis: an evaluation of prognostic factors in 164 cases based on mortality and on a study of lasting sequelae J Infect 1985 10 143 157 4008963 10.1016/S0163-4453(85)91585-3
Kastenbauer S Pfister HW Pneumococcal meningitis in adults: spectrum of complications and prognostic factors in a series of 87 cases Brain 2003 126 1015 1025 12690042 10.1093/brain/awg113
van de Beek D de Gans J Spanjaard L Weisfelt M Reitsma JB Vermeulen M Clinical features and prognostic factors in adults with bacterial meningitis N Engl J Med 2004 351 1849 1859 15509818 10.1056/NEJMoa040845
Koedel U Scheld WM Pfister HW Pathogenesis and pathophysiology of pneumococcal meningitis Lancet Infect Dis 2002 2 721 736 12467688 10.1016/S1473-3099(02)00450-4
van de Beek D de Gans J Dexamethasone and pneumococcal meningitis Ann Intern Med 2004 141 327 15313761
McIntyre PB Macintyre CR Gilmour R Wang H A population based study of the impact of corticosteroid therapy and delayed diagnosis on the outcome of childhood pneumococcal meningitis Arch Dis Child 2005 90 391 396 15781931 10.1136/adc.2003.037523
Kornelisse RF Westerbeek CM Spoor AB van der HB Spanjaard L Neijens HJ de Groot R Pneumococcal meningitis in children: prognostic indicators and outcome Clin Infect Dis 1995 21 1390 1397 8749621
Richter RW Brust JC Pneumococcal meningitis at Harlem hospital N Y State J Med 1971 71 2747 2754 5287068
Fiore AE Moroney JF Farley MM Harrison LH Patterson JE Jorgensen JH Cetron M Kolczak MS Breiman RF Schuchat A Clinical outcomes of meningitis caused by Streptococcus pneumoniae in the era of antibiotic resistance Clin Infect Dis 2000 30 71 77 10619736 10.1086/313606
Bruyn GA Kremer HP de Marie S Padberg GW Hermans J van Furth R Clinical evaluation of pneumococcal meningitis in adults over a twelve-year period Eur J Clin Microbiol Infect Dis 1989 8 695 700 2506035 10.1007/BF01963754
Hoen B Viel JF Gerard A Dureux JB Canton P Mortality in pneumococcal meningitis: a multivariate analysis of prognostic factors Eur J Med 1993 2 28 32 8258002
Kirkpatrick B Reeves DS MacGowan AP A review of the clinical presentation, laboratory features, antimicrobial therapy and outcome of 77 episodes of pneumococcal meningitis occurring in children and adults J Infect 1994 29 171 182 7806880 10.1016/S0163-4453(94)90698-X
Kragsbjerg P Kallman J Olcen P Pneumococcal meningitis in adults Scand J Infect Dis 1994 26 659 666 7747088
Weiss W Figueroa W Shapino WH Flippin HF Prognostic factors in pneumococcal meningitis. Arch Intern Med 1967 120 517 524 6054584 10.1001/archinte.120.5.517
McMillan DA Lin CY Aronin SI Quagliarello VJ Community-acquired bacterial meningitis in adults: categorization of causes and timing of death Clin Infect Dis 2001 33 969 975 11528567 10.1086/322612
Robinson KA Baughman W Rothrock G Barrett NL Pass M Lexau C Damaske B Stefonek K Barnes B Patterson J Zell ER Schuchat A Whitney CG Epidemiology of invasive Streptococcus pneumoniae infections in the United States, 1995-1998 - Opportunities for prevention in the conjugate vaccine era JAMA 2001 285 1729 1735 11277827 10.1001/jama.285.13.1729
Rendi-Wagner P Georgopoulos A Kundi M Mutz I Mattauch M Nowak J Mikolasek A Vecsei A Kollaritsch H Prospective surveillance of incidence, serotypes and antimicrobial susceptibility of invasive Streptococcus pneumoniae among hospitalized children in Austria J Antimicrob Chemother 2004 53 826 831 15073163 10.1093/jac/dkh211
Ispahani P Slack RC Donald FE Weston VC Rutter N Twenty year surveillance of invasive pneumococcal disease in Nottingham: serogroups responsible and implications for immunisation Arch Dis Child 2004 89 757 762 15269078 10.1136/adc.2003.036921
Aronin SI Peduzzi P Quagliarello VJ Community-acquired bacterial meningitis: Risk stratification for adverse clinical outcome and effect of antibiotic timing Ann Intern Med 1998 129 862 869 9867727
Tunkel AR Hartman BJ Kaplan SL Kaufman BA Roos KL Scheld WM Whitley RJ Practice guidelines for the management of bacterial meningitis Clin Infect Dis 2004 39 1267 1284 15494903 10.1086/425368
Kellner JD Scheifele DW Halperin SA Lebel MH Moore D Le Saux N Ford-Jones EL Law B Vaudry W Outcome of penicillin-nonsusceptible Streptococcus pneumoniae meningitis: a nested case-control study Pediatr Infect Dis J 2002 21 903 910 12394810 10.1097/00006454-200210000-00004
Østergaard C Brandt C Konradsen HB Samuelsson S Differences in survival, brain damage and CSF cytokine kineticts due to meningitis caused by three different Streptococcus pneumoniae serotypes. Evaluation in humans and in two experimental meningitis models. J Infect Dis 2004 190 1212 1220 15346330 10.1086/423852
Konradsen HB Kaltoft MS Invasive pneumococcal infections in Denmark from 1995 to 1999: epidemiology, serotypes, and resistance Clin Diagn Lab Immunol 2002 9 358 365 11874878 10.1128/CDLI.9.2.358-365.2002
Cutts FT Zaman SM Enwere G Jaffar S Levine OS Okoko JB Oluwalana C Vaughan A Obaro SK Leach A McAdam KP Biney E Saaka M Onwuchekwa U Yallop F Pierce NF Greenwood BM Adegbola RA Efficacy of nine-valent pneumococcal conjugate vaccine against pneumonia and invasive pneumococcal disease in The Gambia: randomised, double-blind, placebo-controlled trial Lancet 2005 365 1139 1146 15794968 10.1016/S0140-6736(05)71876-6
Jackson LA Neuzil KM Yu OC Benson P Barlow WE Adams AL Hanson CA Mahoney LD Shay DK Thompson WW Effectiveness of pneumococcal polysaccharide vaccine in older adults N Engl J Med 2003 348 1747 1755 12724480 10.1056/NEJMoa022678
van de Beek D de Gans J McIntyre P Prasad K Corticosteroids in acute bacterial meningitis Cochrane Database Syst Rev 2003 CD004305 12918010
Møller K Pedersen SS Skinhøj P Dexamethasone to adults with acute bacterial meningitis Ugeskr Laeger 2003 165 243 244 12555711
Orihuela CJ Gao G Francis KP Yu J Tuomanen EI Tissue-specific contributions of pneumococcal virulence factors to pathogenesis J Infect Dis 2004 190 1661 1669 15478073 10.1086/424596
|
16253143
|
PMC1295586
|
CC BY
|
2021-01-04 16:28:17
|
no
|
BMC Infect Dis. 2005 Oct 27; 5:93
|
utf-8
|
BMC Infect Dis
| 2,005 |
10.1186/1471-2334-5-93
|
oa_comm
|
==== Front
BMC Infect DisBMC Infectious Diseases1471-2334BioMed Central London 1471-2334-5-971626290910.1186/1471-2334-5-97Research ArticleThe changing epidemiology of pediatric aseptic meningitis in Daejeon, Korea from 1987 to 2003 Lee Kyung-Yil [email protected] David [email protected] Hyung-Shin [email protected] Ja-Hyun [email protected] Mi-Hee [email protected] Jin-Han [email protected] Byung-Churl [email protected] Department of Pediatric, College of Medicine, The Catholic University of Korea, Seoul, Korea2 School of Paediatrics and Child Health, University of Western Australia, Perth, Australia2005 2 11 2005 5 97 97 15 3 2005 2 11 2005 Copyright © 2005 Lee et al; licensee BioMed Central Ltd.2005Lee et al; licensee BioMed Central Ltd.This is an Open Access article distributed under the terms of the Creative Commons Attribution License (), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
Background
Aseptic meningitis is a relatively frequent childhood disease and virologic data suggest that enteroviruses are the commonest etiologic agents. We evaluated the epidemiologic characteristics of aseptic meningitis in Daejeon, South Korea from 1987 to 2003.
Methods
2201 medical records of children with aseptic meningitis admitted to The Catholic University of Korea, Daejeon St Mary's Hospital were retrospectively analyzed.
Results
Outbreaks of aseptic meningitis were observed in 1990, 1993, 1996, 1997, 2001 and 2002. The age distribution of cases was relatively uniform, with a higher incidence in those aged < 1 year and 4–7 years. The male-to-female ratio was 2:1. There was a higher incidence of disease in the summer (May to August, 74.1% of total). Comparison of the largest epidemics in 1997 and 2002 showed significant differences in the incidence in those < 1 year (11.8% vs. 4.4%, respectively; P = 0.001). Neurologic sequelae were observed in 0.7% of the patients.
Conclusion
Aseptic meningitis, rare before the 1980s in Korea, has since become a common clinical entity. Since 1990, outbreaks of aseptic meningitis have occurred every 1 to 3 years in Daejeon in keeping with Korea-wide epidemics. The frequency of disease affecting children less than one year of age may reflect herd immunity to the epidemic strain.
==== Body
Background
Aseptic meningitis is a relatively common disease of childhood and recent data suggest that it is most frequently caused by enteroviruses [1,2]. Although globally aseptic meningitis occurs year-round, there is a marked seasonality, with an incidence peak in the summer months in temperate climates. In Korea, epidemics of aseptic meningitis have occurred approximately every 3 years since the early 1990s. Whilst there have been a number of studies addressing the clinical characteristics and the causative agents of these epidemics [3-15], long term epidemiologic studies of aseptic meningitis are rare. A variety of enteroviruses have been suggested as the predominant causative agents for each Korean epidemic, with enterovirus 71 or echovirus 30 in 1990 [6], echovirus 9 in 1993 [7], echovirus 30 in 1997 [11], echovirus 6 in 1998 [11], coxsackievirus B5 in 2001 and echovirus 13 in 2002 [12-14].
In this study, we describe the epidemiologic characteristics of pediatric aseptic meningitis during a 17 year period, from 1987 to 2003, in Daejeon, South Korea, with particular emphasis on the two largest epidemics, in 1997 and 2002.
Methods
We retrospectively analyzed a total of 2201 medical records of pediatric patients with aseptic meningitis patients admitted to The Catholic University of Korea, Daejeon St. Mary's Hospital from January 1987 to December 2003. Daejeon, located in central Korea, is one of the largest cities with a population of over 1.4 million. The majority of the patients with aseptic meningitis were admitted to one of four general hospitals in Daejeon. There were no significant changes of medical facilities or the social environment during the study period. Epidemic years were defined as those with more than 100 patients per year and at least a two-fold increase in cases compared to the previous year. The diagnosis of aseptic meningitis was made on the basis of: (i) clinical symptoms and signs of meningitis, such as fever, vomiting, headache, and meningeal irritation, (ii) cerebral spinal fluid (CSF) pleocytosis (≥ 5 leukocytes/mm3) with normal CSF protein and sugar levels and (iii) negative results on bacterial culture and latex particle agglutination test. Studies to identify viral pathogens were not routinely undertaken in these hospitals, especially during epidemics, as the available data suggests that causative agents other than enteroviruses are relatively rare [6,7,11-15]. Other non-enteroviral causes of meningitis occurring during the study period were excluded from the current analysis. These included cases of mumps meningitis (n = 66; 1989–1998) [16], bacterial meningitis (n = 40; 1992–2002) [17] and herpes meningo-encephalitis (n = 3). We analyzed the age and gender of patients, together with monthly and annual frequency. In addition, we evaluated the frequency of neurological complications among the patients hospitalized for more than 10 days. The epidemiologic features of the patients from the two years with the highest incidence, 1997 and 2002, were also compared.
Results
Age and sex distribution
2201 children fulfilled the study criteria. The mean age was 6.0 ± 3.9 years (range 2 weeks to 15 years). The highest incidence was in those aged 4 to 7 years (44.1% of total cases) and in those less than 1 year old (10% of total), with the remaining cases evenly distributed across the remaining ages (Fig. 1). There were 1470 males and 731 females, giving a male-to-female ratio of approximately 2:1. Although enteroviral studies were not performed routinely during the study period, enteroviruses were identified in a proportion of patients in the 1997 and 2002 epidemics (Table 1). The mean duration of hospitalization was 5.5 ± 1.7 days. There were no fatalities.
Figure 1 (A) Age distribution of aseptic meningitis, 1987–2003. (B) Annual cases of aseptic meningitis, 1987–2003.
Table 1 Enteroviruses isolated from each nationwide epidemic of aseptic meningitis in Korea.
Year Cases Specimen(s) Enteroviruses isolated (n) References
1990 118 Serum Suspected Enterovirus 71 & EV 30 6
1993 93 CSF EV 9 (60), Others (17) 7
1996 210 Stool, CSF EV 9 (11), CV B1 (6) 11
1997 493 Stool, CSF EV 30 (54), EV 6 (15), CV B5 (9), Others (3) 11
1997 33 CSF EV 30 (6) *
1998 341 Stool, CSF EV 6 (36), CV B2 (11), Others (29) 11
2002 371 CSF EV 13 (18), EV 9 (15), EV 6 (10), Others (24) 13
2002 13 Stool, CSF Phylogenetic analysis of EV 13 14
2002 29 CSF EV 13 (5), EV 6 (1) *
CSF, Cerebrospinal fluid; EV, Echovirus; CV, Coxsackivirus;
*, Unpublished data of our study
Annual incidence
The number of aseptic meningitis cases per year ranged from 17 (0.8% of total cases) in 1987 to 489 (21.7%) in 1997; the average was 129 cases per year. The greatest number of cases were observed in 1997, 2002 and 1993, with 489 (21.7%) of total, 366 (16.6%) and 257 (11.7%) patients, respectively. Outbreaks occurred approximately every three years in 1990, 1993, 1996, 1997, 2001 and 2002 (Fig. 1).
Monthly and seasonal frequencies
There was a striking seasonal pattern, with almost three quarters of cases occurring during the summer months (May to August) (Fig. 2). Of the total of 2201 cases, 540 (24.5%) presented in June, 378 (17.2%) in July, 371 (16.9%) in May, and 341 (15.5%) in August (Fig. 2).
Figure 2 (A) Monthly cases of aseptic meningitis, 1987–2003. (B) Comparison of age distribution between in 1997 and in 2002.
Comparison of epidemiologic features of the epidemics in 1997 and 2002
1997 and 2002 had the largest epidemics of aseptic meningitis in the study period. There was no significant difference between in these two large epidemics in the age of children and the monthly distribution. However, in 1997, there was a higher frequency in those < 1 year of age than in 2002 (11.8% vs. 4.4%, P = 0.001, X2 test) (Fig. 2).
Neurologic sequelae
Medical records of the 127 children hospitalized for ≥ 10 days were analyzed for possible complications. Although other complications of the gastrointestinal or respiratory system were noted, only neurological complications were evaluated in this study. Neurological complications were noted in 16 patients (0.7% of total). All children with neurological complications had abnormal mental status or abnormal neurological symptoms and signs following their aseptic meningitis and had been neurologically normal previously. Seizures were observed in 5 children, of whom two developed status epilepticus. Transient amnesia was observed in 2 children; one had antegrade amnesia, and the other had retrograde amnesia. One child had an inappropriate secretion of anti-diuretic hormone and mild hydrocephalus. One child showed transient signs of quadriplegia. Thirteen of the 16 children with neurological complications were followed up for at least 2 months. No permanent neurological sequelae were observed.
Discussion
Although there have been a number of studies of aseptic meningitis in Korea [3-15], there are few epidemiologic studies of this scope and duration. Historically, aseptic meningitis appears to have been relatively unusual in Korea prior to the 1980s, but this may partly reflect ascertainment bias and diagnostic methods. Chung et al. described 104 cases of childhood meningitis from 1960–1967, with 41 cases of aseptic meningitis cases and 63 bacterial meningitis cases [3]. Rheu et al. reported an increasing incidence of aseptic meningitis in the Wonju area of Korea from 1966 to 1983; among the total of 200 cases of aseptic meningitis, there were 1–3 cases per year from 1966–1971, 6–23 cases per year in 1972–1982, and 40 cases in 1983 [4]. Similar findings are reported from Busan during the 1980s, with 5–19 cases annually between 1980 and 1988, 30 cases in 1989 and 56 cases in 1990 [5]. These results suggest that the nationwide epidemics of aseptic meningitis in Korea started in the early 1990s. The striking change in epidemiology at this time, with the occurrence of epidemics, may partly reflect improved diagnosis and increased awareness. In addition, changes in herd immunity may have contributed to an increased population susceptibility to epidemics of virus transmitted by the fecal-oral route. Public health improvements that accompanied this period of marked economic growth in Korea may have reduced sporadic exposure to enteroviruses and led to a large non-immune population. Similar patterns have been observed in the rapidly declining seroprevalence of hepatitis A in Korea. Hepatitis A, which is a kind of enteroviruses and also spread by the fecal-oral route, currently has a seroprevalence in Korean children < 15 years old of almost zero [18,19].
In this study, epidemics of aseptic meningitis in Daejeon mirrored the nationwide epidemics in 1990 [5,6], 1993 [7,8], 1996 [9], 1997 [10], 2001 and 2002 [12-14]. A nationwide study of 5090 patients in 1993 showed that the epidemic of aseptic meningitis commenced in the southern regions in spring, and then gradually moved to the northern regions (including Seoul), reaching a peak incidence in summer and then waned by late fall [8]. Similar patterns have been observed in other epidemics in Korea [5,6] and may reflect the geographical and social environment with a large and mobile population contributing to the nationwide spread of the disease. Aseptic meningitis is generally commoner in males with a male-to-female ratio of 1.2–2.3 to 1 [4-13], as observed in the current study.
Globally the age distribution during epidemics of aseptic meningitis varies, possibly due to different causative agents and specific herd immunity that results from varying socio-economic environments and other factors. For example, in the United States, the peak age for children with aseptic meningitis is reported to be < 1 year old [1,2,20], whereas in five South African epidemics in the 1980s, the mean age of children with echovirus meningitis was 4–5 years of age, whereas those with coxsackie B meningitis were most commonly <1 year old [21]. In Japan, Yamashita et al. analyzed 8595 cases of aseptic meningitis from 1981–1991. There were two peak ages of < 1 year and 4–7 years, with varying age distributions according to the causative virus in each epidemic [22]. Although Korea and Japan have similar geographical, racial and socio-economic environments, the period of marked Korean economic growth occurred later than in Japan. The Korean epidemiologic data from the 1990s is similar to those from the 1980s in Japan [22,23], possibly reflecting similar changes in important determinants of epidemiology.
The pattern of age distribution in an aseptic meningitis epidemic may reflect herd immunity from past infection with the same causative virus. Aseptic meningitis is rare in younger adults in Korea, whereas adults aged 20 and 40 years have an incidence similar to children in the United States [20]. In addition, there were differences in incidence in infants (< 1 year old) with each epidemic, suggesting that herd immunity from previous infections with enteroviruses is reflected in protective transplacental antibodies. Maternal (transplacental) antibodies are detected in over half of the infants aged 6 months, and may persist for up to 12 months after birth. Thus a high incidence in both infants and adults suggests a new epidemic strain with low herd immunity.
Previous studies in Korea have reported that enteroviruses predominate in each nationwide epidemic (Table 1). Enteroviruses include more than 70 serotypes, but only a few serotypes typically cause aseptic meningitis in any given community and during any given year. This may be due to differences in the background rate of infection in a community (herd immunity), in host immunity and possibly in the viral strain's neurotropism. Recently, echovirus 13 meningitis epidemics have been reported in Europe in 2000 [24,25], in the United States in 2001 [26], and in Korea in 2002 [12-14]. The genotype of echovirus 13 isolated in Korea in 2002 is almost identical to that isolated in Japan and Germany [14]. Thus, aseptic meningitis epidemics from an enteroviral strain may spread globally.
Children with enteroviral meningitis generally recover without complications, but rarely the disease can cause neurologic sequelae [1,2,8]. The clinical manifestations and outcome may differ with the enterovirus serotype. For example, a more severe clinical course and worse outcome with enterovirus 71 infections have been reported both in Korea and in other countries [27,28]. In the current study, 16 children (0.7%) with aseptic meningitis had neurological complications; 2 children in 1990, 4 in 1993, 4 in 1997, and 1 child each in six other years. No single epidemic resulted in a significantly larger number of neurologic sequelae, although the denominator was small as the outcome was generally excellent.
Conclusion
Aseptic meningitis appears to have been a rare disease in Korea during the 1960s and 1970s. Since 1990, outbreaks of aseptic meningitis have occurred every 1–3 years in Daejeon in keeping with nation-wide epidemics. The changing epidemiology may reflect improvements in public health, characteristics of the predominant etiologic agents and changes in herd immunity. In particular, the incidence of aseptic meningitis in children less than one year of age in each epidemic may be indicative of the agent-specific immunity of the parental generation.
Competing interests
The author(s) declare that they have no competing interests.
Authors' contributions
KYL designed the study and drafted the manuscript. JHH, MHL and JHK participated in the data collection and analysis, HSL analyzed the final data. BCL participated in the supervising the execution of the study. DB assisted with data interpretation and in drafting the manuscript. All authors read and approved the final version of the manuscript.
Pre-publication history
The pre-publication history for this paper can be accessed here:
==== Refs
Connolly KJ Hammer SM The acute aseptic meningitis syndrome Infect Dis Clin North Am 1990 4 599 622 2277191
Sawyer MH Enterovirus infections: diagnosis and treatment Pediatr Infect Dis J 1999 18 1033 91 10608620 10.1097/00006454-199912000-00002
Chung YW Bang DH Lee KS Clinical studies on acute meningitis in infancy and childhood J Korean Pediatr Soc 1968 11 295 302
Rheu SH Park SB Lim BK Kim JS Clinical studies of aseptic meningitis J Korean Pediatr Soc 1984 27 1176 83
Park KW Choi DY Kim SW A clinical study of aseptic meningitis J Korean Pediatr Soc 1991 34 1400 8
Cho EY Kang MK Hong SJ Kim KS Park YS Park IS Moon HN Hong CY Epidemics of aseptic meningitis in Seoul area during 1989–90 J Korean Pediatr Soc 1991 34 1565 72
Cho JY Kim HJ Jung GY Pang JK Lee DB Epidemic aseptic meningitis in 1993: focused on virus culture J Korean Pediatr Soc 1995 38 901 906
Oh SH Lee MS Kang JH Kim CH Park JY Shon YM Lee WJ Chun CS Shin SM Report of nationwide epidemiology of aseptic meningitis outbreaks in 1993 in Korea J Korean Pediatr Soc 1996 39 42 52
Chung JA Kim YJ Choi HJ Chung WK An epidemic of aseptic meningitis in summer 1996 and global analysis and comparison of it with 1993 J Korean Pediatr Soc 1997 40 1081 9
Park YH Kim WJ Son BH Kim SW Global analysis of aseptic meningitis in Pusan area in 1997 J Korean Pediatr Infect Dis 1998 5 115 20
Kim KS Kim JE Cheon DS Chung YS Park JK Kang YH Lee YS Jee YM Yoon JD Lee YJ Kim DS Kim MB Na BK Song CY Lee KH An epidemiologic study of enterovirus as causative agents of aseptic meningitis between 1993 and 1998 in Korea Korean J Infect Dis 1999 31 382 9
Kim HJ Cheong HK Jung C Lee KM Jee YM Kim DW Lee DS Kim DK Choi SM Clinical and viologic study of aseptic meningitis Korean J Pediatr 2004 47 392 8
Kim CK Ha TY Lee JH Yoon JD Kim YD Jee YM Park SK Jung JY A clinical study of aseptic meningitis in Ulsan from May to July, 2002 J Korean Child Neurol Soc 2003 11 328 34
Cheon DS Lee J Lee K Lee S Park K Ahn J Jee Y Yoon J Cho H Isolation and molecular identification of echovirus 13 isolated from patients of aseptic meningitis in Korea, 2002 J Med Virol 2004 73 439 42 15170640 10.1002/jmv.10543
Joo CH Ahn J Seo I Kim YK Kim D Hong H Lee H Characterization of nonpolio enteroviruses recovered from patients with aseptic meningitis in Korea Intervirology 2005 48 97 103 15812181 10.1159/000081735
Kang HD Lee KY Cha SW Yoon KN Lee DJ Kang JH Whang KT An outbreak of mumps in Daejeon, Korea, 1998 J Korean Pediatr Infect Dis 1999 6 239 44
Kim HJ Lee JW Lee KY Lee HS Hong JH Han SH Whang KT Causative organisms in children with bacterial meningitis (1992–2002) J Korean Pediatr Soc 2003 46 1085 8
Lee KY Song KH Kang JH Seroepidemiology of Hepatitis A in Daejeon, Korea, 1996 J Korean Pediatr Soc 1998 41 53 61
Kang JH Lee KY Kim CH Sim D Changing hepatitis A epidemiology and the need for vaccination in Korea Asian Pac J Allergy Immunol 2004 22 237 42 15783137
Khetsuriani N Quiroz ES Holman RC Anderson LJ Viral meningitis-associated hospitalizations in the United States, 1988–1999 Neuroepidemiolgy 2003 22 345 52 10.1159/000072924
McIntyre JP Keen GA Laboratory surveillance of viral meningitis by examination of cerebral fluid in Cape town, 1981–9 Epidemiol Infect 1993 111 357 71 8405162
Yamashita K Miyamura K Yamadera S Kato N Akatsuka M Hashido M Inouye S Yamazaki S Enteroviral aseptic meningitis in Japan, 1981–1991 Jpn J Med Sci Biol 1992 45 151 61 1337925
Yamashita K Miyamura K Yamadera S Kato N Akatsuka M Hashido M Inouye S Yamazaki S Epidemics of aseptic meningitis due to echovirus 30 in japan. A report of the national epidemiological surveillance of infectious agents in Japan Jpn J Med Sci Biol 1994 47 221 39 7715095
Bottner A Daneschnejad S Handrick W Schuster V Liebert UG Kiess W A season of aseptic meningitis in Germany: epidemiologic, clinical and diagnostic aspects Pediatr Infect Dis J 2002 21 1126 32 12488662
Diedrich S Schreier E Aseptic meningitis in Germany associated with echovirus type 13 BMC Infect Dis 2001 1 14 11591222 10.1186/1471-2334-1-14
Kirschke DL Jones TF Buckingham SC Craig AS Schaffner W Outbreak of aseptic meningitis associated echovirus 13 Pediatr Infect Dis J 2002 21 1034 8 12442025 10.1097/00006454-200211000-00011
Ho M Chen ER Hsu KH Twu SJ Chen KT Tsai SF Wang JR Shih SR An epidemic of enterovirus 71 infections in Taiwan New Engl J Med 1999 341 929 35 10498487 10.1056/NEJM199909233411301
Moon WY Kim KS Park YS Moon HN Hong CY Suh DC Yu SJ Seong IY A clinical study on aseptic meningitis combined with polio-like paralysis J Korean Pediatr Soc 1993 36 485 93
|
16262909
|
PMC1295587
|
CC BY
|
2021-01-04 16:28:14
|
no
|
BMC Infect Dis. 2005 Nov 2; 5:97
|
utf-8
|
BMC Infect Dis
| 2,005 |
10.1186/1471-2334-5-97
|
oa_comm
|
==== Front
BMC NeurolBMC Neurology1471-2377BioMed Central London 1471-2377-5-201626289010.1186/1471-2377-5-20Research ArticleFamily history and stroke outcome in a bi-ethnic, population-based stroke surveillance study Lisabeth Lynda D [email protected] Melinda A [email protected] Devin L [email protected] Ken [email protected] Lewis B [email protected] Stroke Program, University of Michigan Medical School, Ann Arbor, Michigan, USA2 Stroke Institute, University of Pittsburgh Medical Center, Pittsburgh, Pennsylvania, USA3 Department of Epidemiology, University of Michigan School of Public Health, Ann Arbor, Michigan, USA2005 31 10 2005 5 20 20 26 5 2005 31 10 2005 Copyright © 2005 Lisabeth et al; licensee BioMed Central Ltd.2005Lisabeth et al; licensee BioMed Central Ltd.This is an Open Access article distributed under the terms of the Creative Commons Attribution License (), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
Background
The genetic epidemiology of ischemic stroke remains relatively unstudied, and information about the genetic epidemiology of ischemic stroke in populations with significant minority representation is currently unavailable.
Methods
The Brain Attack Surveillance in Corpus Christi project (BASIC) is a population-based stroke surveillance study conducted in the bi-ethnic community of Nueces County, Texas, USA. Completed ischemic strokes were identified among patients 45 years or older seen at hospitals in the county between January 1, 2000 – December 31, 2002. A random sample of ischemic stroke patients underwent an in-person interview and detailed medical record abstraction (n = 400). Outcomes, including initial stroke severity (NIH stroke scale), age at stroke onset, 90-day mortality and functional outcome (modified Rankin scale ≥2), were studied for their association with family history of stroke among a first degree relative using multivariable logistic and linear regression. A chi-square test was used to test the association between family history of stroke and ischemic stroke subtype.
Results
The study population was 53.0% Mexican American and 58.4% female. Median age was 73.2 years. Forty percent reported a family history of stroke among a first degree relative. Family history of stroke was borderline significantly associated with stroke subtype (p = 0.0563). Family history was associated with poor functional outcome in the multivariable model (OR = 1.87; 95% CI: 1.14–3.09). Family history was not significantly related to initial stroke severity, age at stroke onset, or 90-day mortality.
Conclusion
Family history of stroke was related to ischemic stroke subtype and to functional status at discharge. More research is needed to understand whether stroke subtype would be a useful selection criterion for genetic association studies and to hypothesize about a possible genetic link to recovery following ischemic stroke.
==== Body
Background
Data from different sources, including animal and family history studies, suggest that genetic factors play a role in ischemic stroke [1-4]. Despite the challenges associated with studying the genetics of a heterogeneous disorder, several polymorphisms have been identified for ischemic stroke with small to modest effects [5]. Most recently, researchers using data from an Icelandic population have identified a locus on chromosome 5 thought to be associated with ischemic stroke in humans [6]. While the focus has been on identifying genetic determinants of stroke risk, many aspects of the genetic epidemiology of ischemic stroke remain relatively unstudied. Aside from the relationship between a positive family history of stroke and stroke risk, aspects of ischemic stroke for which genetic factors are plausible, such as stroke subtype, stroke severity and stroke outcomes, have not been thoroughly investigated.
Further, information about the genetic epidemiology of ischemic stroke in populations with significant representation from minority groups, such as Hispanic Americans, is currently unavailable. Differences in stroke risk, age at onset of stroke and family history of stroke between Hispanics and non-Hispanic whites raise the possibility of genetic differences by ethnicity [7,8]. The objective of this study was to provide an overview of the relationship between family history of stroke and the epidemiology of ischemic stroke including associations with stroke subtype, age at stroke onset, stroke severity, mortality, and functional status following stroke using data from a bi-ethnic study population of ischemic stroke cases aged greater than 44 years.
Methods
BASIC is a population-based stroke surveillance study conducted in Nueces County, Texas, USA. Methods of the Brain Attack Surveillance in Corpus Christi (BASIC) Project have been reported [9]. Due to its distance from Houston and San Antonio, investigation in Nueces County allows for complete case capture for first medical contact in acute stroke. The population size of the county is 313,645, and 95% of the population resides within the city of Corpus Christi. Non Hispanic whites (NHW) comprise 38% of the population, and Mexican Americans (MA) comprise 56%.
Acute cerebrovascular events (completed ischemic strokes, transient ischemic attacks, intracerebral hemorrhages (ICH), and subarachnoid hemorrhages (SAH)) were identified among patients 45 years and over who were seen (including those seen in the emergency department and not admitted) at one of the seven area hospitals between January 1, 2000 through December 31, 2002 using active and passive surveillance methods. Cases that did not present to a hospital were identified through a simple random sample from 45 primary care physicians from the community and four nursing homes, and from all 11 neurologists practicing in the county. Cerebrovascular events were validated by fellowship-trained stroke neurologists based on published criteria and blinded to subjects' race-ethnicity and age [10]. Only ischemic strokes were included in this study. For those individuals with multiple ischemic strokes (n = 5) within the time period, only the first was considered for this analysis. Therefore, the final population for analysis incorporated incident strokes as well as recurrent strokes that occurred during the study time period, but was limited to one stroke per individual.
Outcomes
For this analysis, five outcomes were studied for their possible association with family history of stroke as outlined in Table 1. How each outcome was measured and defined for this analysis is described below.
Table 1 Outcomes and end points investigated for possible associations with family history of stroke.
Outcome End Point
Ischemic stroke subtype TOAST criteria (large artery atherosclerosis, cardioembolism, small vessel occlusion, stroke of other determined etiology, and stroke of undetermined etiology) plus additional category of non-lacunar stroke of unknown etiology[13]
Age at stroke onset Continuous age
Initial stroke severity (NIHSS) Continuous NIHSS
Functional outcome (mRS) mRS ≤1 vs mRS >1
Mortality Death ≤ 90 days vs Alive at 90 days
NIHSS = NIH stroke scale, mRS = modified Rankin scale
Interview methodology and family history data
A random sample comprising two-thirds of patients with validated cerebrovascular events was asked to participate in an in-person interview with the goal of achieving a 50% sample of stroke cases. The response rate for interview was 84% [11]. The interview contained questions regarding family history of stroke among first degree relatives (parents, siblings, and children). Patients unable to answer appropriately to a series of orientation questions asked before the interview had a proxy interview, in the presence of the patient whenever possible. Proxy interviews were conducted with the person who knew the patient's daily activities and medical history the most. A previous analysis revealed the agreement between patient/proxy interviews for six critical elements (insurance status, use of a routine physician, history of hypertension, history of diabetes, current smoking status, educational level, and trust in doctors and nurses) ranged from 84% to 100%. Interviews were performed in English or Spanish depending on the patients' language preferences.
A random sample of the patients interviewed was selected to undergo a more detailed medical record abstraction, including reports of medical testing performed and information necessary to classify ischemic strokes according to subtype. The random sample with both interview data and detailed medical record abstraction was used for this analysis.
Ascertainment of outcomes
Ischemic stroke cases were classified by two fellowship-trained stroke neurologists into five stroke subtype categories according to the criteria of the Trial of ORG 10172 in Acute Stroke Treatment (TOAST) Study: large artery atherosclerosis, cardioembolism, small vessel occlusion, stroke of other determined etiology, and stroke of undetermined etiology [12]. An additional category of non-lacunar stroke of unknown etiology was developed that comprised large strokes that had insufficient evidence for categorization into large artery atherosclerosis or cardioembolism [13]. The inter-rater agreement between the two neurologists for determination of subtype was high (kappa 0.80, p < 0.001). Neurologists were blinded to the subject's ethnicity and age.
Initial stroke severity was measured using the NIH stroke scale (NIHSS). The NIHSS was retrospectively abstracted from the chart in accordance with the validated method of Williams et al [14]. Functional outcome was assessed at discharge using a modified Rankin scale (mRS) calculated by the abstractors using data in the medical record. For the purposes of this analysis, the mRS, which ranges from 0–6, was dichotomized to represent good (0–1) and poor (2–6) functional outcome. This categorization of the mRS has been shown to be more powerful than other choices for cut points for this scale [15].
Deaths among the stroke cases were identified for the time period January 1, 2000 through December 31, 2003 by four methods: 1) surveillance of hospitalized or emergency department stroke cases, 2) Texas Department of Health (TDH), 3) Social Security Death Index (SSDI), and 4) Nueces County coroner. If a stroke case died in-hospital, the death was recorded by the abstractors during surveillance. Otherwise, deaths were captured using methods 2–4 above. TDH provided death certificates for Texas residents electronically, with a one-year lag period to ensure complete capture. Demographic data from the medical record, including first name, last name, social security number, date of birth, and permanent address, was crossed-referenced with the TDH death certificate database. At least three of the five items must have been identical for the BASIC stroke case to be considered a match with the mortality case from TDH. In an attempt to capture deaths that occurred outside of Texas and not captured by TDH, we linked cases not identified as having died using the first two data sources to the SSDI. We also routinely screened records from the coroner. All deaths reported to the coroner have a death certificate filed with TDH.
Statistical analysis
Frequencies and percents were calculated for categorical variables. Means and medians were calculated for continuous variables. A chi-square test was used to test the association between family history of stroke among a first degree relative and ischemic stroke subtype. Multivariable logistic regression was used to test the association between a positive family history of stroke among a first degree relative and functional outcome (dichotomized mRS) and 90-day mortality. Covariates were selected for inclusion in the models in a pre-specified fashion based on their plausible relationship with the given outcome or their ability to confound family history. In each logistic model, the number of covariates was limited based on the rule of 10 which requires at least 10 least frequent outcomes for each degree of freedom [16]. For the model predicting functional outcome the following covariates were included: age, gender, ethnicity, hypertension, diabetes, atrial fibrillation, coronary heart disease, NIHSS, and stroke subtype. For the model predicting 90-day mortality, the following covariates were included: age, gender, ethnicity, and NIHSS.
Multivariable linear regression was used to test the association between a positive family history of stroke among a first degree relative and initial stroke severity (NIHSS) and age at stroke onset. Again, covariates were selected for inclusion in the models in a pre-specified fashion. For the models predicting NIHSS and age at stroke onset the following covariates were included: gender, ethnicity, hypertension, diabetes, atrial fibrillation, coronary heart disease, high cholesterol, smoking, and stroke subtype. Age was also included in the model for NIHSS.
Family history was modeled dichotomously in all models (yes/no). Ethnicity (MA vs. NHW), gender (female vs. male) and the history variables (smoking, high cholesterol, coronary artery disease, diabetes, atrial fibrillation, hypertension) were treated dichotomously. Age was treated continuously. Stroke subtype was modeled as a series of indicator variables with cardioembolic strokes as the referent.
Chi-square tests were used to test the association between use of proxy subjects and family history and to compare the demographic characteristics of cases with and without proxies. To assess whether the use of proxy subjects confounded the relationship between family history and the five outcomes, a covariate representing proxy subject (yes/no) was also added to each of the multivariable models and the degree to which the point or parameter estimate changed was evaluated. A greater than 10% change was considered to represent confounding. A chi-square test was used to test the association between use of proxy subjects and stroke subtype.
The project was approved by the Institutional Review Boards at the University of Texas, Houston and University of Michigan, and each Nueces County hospital.
Results
There were 2,550 validated cerebrovascular events between January 1, 2000 and December 31, 2002. Ischemic strokes constituted 1,477 of the 2,550 events. One-hundred six cases who were not MA or NHW (7 Asian, 89 African American, 10 unknown race/ethnicity) were excluded due to small sample size. Eight-hundred seventeen of these cases were interviewed and 405 of the cases were subtyped based on documentation gathered during the extended medical record abstraction. Five recurrent events were eliminated resulting in 400 ischemic stroke cases for the final analysis.
Of the 400 cases interviewed, 39.5% reported a positive family history of stroke among a first degree relative. Eleven percent (n = 44) were unsure of a family history of stroke and for three cases (0.8%), this question was not answered. Table 2 displays baseline characteristics for stroke subjects with complete family history data (n = 353). The study population was 53.0% (n = 187) MA and 58.4% (n = 206) female. Median age was 73.2 years (IQR: 65.2–79.7). There were no differences in the prevalence of stroke risk factors among those with and without a positive family history of stroke. Women were more likely to report a family history of stroke compared with men as we have previously reported [8].
Table 2 Demographics and risk factors for ischemic stroke cases with complete family history data (n = 353).
Family History
Yes (n = 158) No (n = 195)
N % n % p-value
MA 90 57.0 97 49.7 0.1773
NHW 68 43.0 98 50.3
Females 108 68.4 98 50.3 0.0006
Males 50 31.7 97 49.7
Risk Factors
Hypertension 120 76.0 138 70.8 0.2759
Diabetes 76 48.1 80 41.0 0.1838
Atrial Fibrillation 18 11.4 26 13.3 0.5836
Coronary Heart Disease 51 32.3 71 36.4 0.4423
Hyperlipidemia 25 15.8 44 22.6 0.1128
Smoking 40 25.3 60 30.8 0.2589
History of Stroke/TIA 62 39.2 66 33.9 0.2952
MA = Mexican American, NHW = non-Hispanic white
The distribution of ischemic stroke subtype among the 400 cases was 21.0% cardioembolic, 14.3% large vessel, 19.3% small vessel, 22.8% large strokes with insufficient evidence for categorization into large artery atherosclerosis or cardioembolism, 21.5% undetermined etiology, and 1.3% other determined etiology. Family history of stroke was borderline significantly associated with ischemic stroke subtype (p = 0.0563), with family history less frequent among those with large vessel strokes and more frequent among those with small vessel disease (Table 3).
Table 3 Distribution of ischemic stroke subtypes by family history (n = 353).
Family History
Ischemic Stroke Subtype Yes No
n % n %
Cardioembolic 37 23.4 32 16.4
Large vessel 16 10.1 35 18.0
Non-lacunar/large stroke* 34 21.5 44 22.6
Small vessel 39 24.7 33 16.9
Undetermined 29 18.4 49 25.1
Other 3 1.9 2 1.0
Total 158 100.0 195 100.0
χ2 (df 5) = 10.76, p = 0.0563
* Non-lacunar strokes of unknown etiology comprised of large strokes with insufficient evidence for categorization into large artery atherosclerosis or cardioembolism
Results from the multivariable logistic regression models are displayed in Table 4. Among the 400 cases, 16.0% died within 90 days. Among those with a positive family history, 16.5% died within 90 days, and among those with no history, 12.3% died. Family history showed a positive association with 90-day mortality (OR = 1.55, 95% CI: 0.77–3.13), although this association did not reach significance in the multivariable model.
Table 4 Association of family history of stroke with functional outcome and mortality (n = 353).
End Point OR 95% CI
Rankin ≥ 21 1.87 (1.14, 3.09)
90-day Mortality2 1.55 (0.77, 3.13)
OR = odds ratio, CI = confidence interval
1 Adjusted for age, gender, ethnicity, hypertension, diabetes, atrial fibrillation, coronary artery disease, initial stroke severity, ischemic stroke subtype
2 Adjusted for age, gender, ethnicity, initial stroke severity
Among the 353 cases with family history data, 61.5% had a mRS ≥ 2. Sixty-nine percent of those with a positive family history had poor functional outcome compared to 55.4% in those with no family history. Family history was positively associated with poor functional outcome in the multivariable model (OR = 1.87; 95% CI: 1.14–3.09) adjusted for the other covariates.
Results from the multivariable linear regression models are displayed in Table 5. Median age at stroke onset was 71.4 years among those with a family history and 73.9 years among those with no family history. Family history was associated with a younger age at stroke onset in the multivariable model (β = -1.38, standard error = 1.18), but this relationship was not statistically significant. Median NIHSS was 4 among those with a family history and 3 among those with no family history. Results from the multivariable model (Table 5) suggest a mean NIHSS difference of 0.41 between those with and without a family history, but this association did not reach significance.
Table 5 Association of family history of stroke with initial stroke severity and age at stroke onset (n = 353).
End Point β Std Error
NIH stroke scale1 0.41 0.72
Age at stroke onset2 -1.38 1.18
Std Error = standard error
1 Adjusted for age, gender, ethnicity, hypertension, diabetes, atrial fibrillation, coronary artery disease, high cholesterol, smoking, ischemic stroke subtype
2 Adjusted for gender, ethnicity, hypertension, diabetes, atrial fibrillation, coronary artery disease, high cholesterol, smoking, ischemic stroke subtype
Use of proxies
Thirty-seven percent (n = 147) of the interviews were conducted with proxies. In 80.7%, the proxy subject was a spouse (31.7%) or child (49.0%) of the patient. The use of proxies varied with age of the stroke case (p < 0.0001). Median age of cases with a proxy interview was 78.8 years and median age of cases without a proxy interview was 69.6 years. Use of proxies did not differ by gender or ethnicity of the case.
Family history did not vary by use of proxy subjects (p = 0.7560). However, the use of a proxy did appear to confound the relationship between family history and age at stroke onset as evidenced by the change in the parameter estimate with inclusion of this covariate (Table 6). Family history remained insignificant (p = 0.2817) in this model regardless of inclusion of the variable for proxy subject. Use of proxy subjects was significantly related to stroke subtype (p < 0.0001).
Table 6 Comparison of point/parameter estimates for family history before and after adjustment for use of proxy subjects (n = 353).
Original Model Model with Proxy
End Point OR/Parameter Estimate for Family History OR/Parameter Estimate for Family History
Rankin ≥ 2 1.87 1.90
90-day Mortality 1.55 1.49
NIH stroke scale 0.41 0.42
Age at stroke onset -1.38 -1.19
Discussion
In our study population of ischemic stroke cases, family history of stroke among a first degree relative was associated with poor functional outcome, defined as a modified Rankin scale score ≥ 2 at discharge, adjusted for initial stroke severity and other potential confounders. The associations of family history with initial stroke severity and 90-day mortality, although not significant, together with the association with poor functional outcome at hospital discharge, suggested family history was related to more severe strokes among the cases. Initial stroke severity and functional outcome are known to be highly correlated and were significantly associated in our data [17,18]. Possible explanations for the lack of a significant association of family history with initial stroke severity include reduced power to detect a difference due to our sample size and possible clustering or downward bias of NIHSS score assignments due to the retrospective designation of these scores. Familial factors that influence poor short-term recovery after stroke could explain the finding of a positive association of family history with poor functional outcome at discharge and its lack of association with initial NIHSS. Additional large-scale studies are needed to confirm our findings and to suggest a potential genetic link with recovery after stroke.
There may be differences in genetic susceptibility to stroke based on ischemic stroke subtype. We found a borderline significant association with the distribution of subtype and family history, with the data suggesting that cases with small vessel disease were more likely to have a family history of stroke. This result is consistent with results from hospital-based studies of stroke cases and controls [19-21]. Our data also suggested that cases with large vessel disease were less likely to have a family history. Previous studies have found a greater prevalence of family history among those with large vessel disease [19-21]. In our study population, large vessel disease was present in 14.3% of cases. This percent is less than other study populations and perhaps contributes to our different findings. We also had 21.5% of cases with undetermined etiology possibly due to less extensive evaluation documentation in community hospitals. There may also be a genetic susceptibility to the same stroke type in the case as that of the family member [19], but we were unable to explore this question as the stroke type of the family member is unknown. First degree relatives with stroke who are parents of the case are often deceased by the time of the offspring's stroke, adding to the challenge of classifying the family members' stroke type.
Previous studies have found associations between family history of stroke and earlier age at stroke onset [20,22] suggesting age may be a useful selection criteria for future ischemic stroke genetics studies. Our findings with regard to age at stroke onset were in this direction but not significant after multivariable adjustment for confounders and independent predictors.
Strengths of the current study include: 1) the population-based design and ascertainment of all cases presenting for medical attention to a hospital, including those seen in the emergency department but not subsequently hospitalized, 2) the inclusion of ischemic strokes only, as not all types of stroke are equally likely to be heritable, 3) the detailed subtyping of ischemic strokes to allow for comparison of family history across subtypes, and 4) the inclusion of a significant proportion of Mexican Americans, a minority population with increased risk of stroke. Further, we adjusted for potential confounding factors including stroke risk factors, stroke severity, demographics and stroke subtype when sample size and the frequency of the outcome permitted.
Some limitations of this study warrant discussion. Family history data was based on self-report by stroke cases and was not validated through source documentation or interviews with family members. We interviewed a proxy subject in lieu of a patient interview when the patient could not answer a series of orientation questions. In the majority of these cases (84%), the interview was with a spouse, sibling, or child who would likely have knowledge of the family history. Further, the use of proxy subjects did not appear to confound the relationship between family history and the outcomes with the exception of age at stroke onset; however, family history remained an insignificant predictor of age at stroke regardless of proxy use. Some stroke cases were unaware of their family history (11%), and this could have biased our findings regarding family history; however, the effect is likely to be small. We did not have data on family size, which may impact family history of stroke, and were unable to account for this factor in our analysis. We also did not have information on age of the first-degree relative at the time of the stroke. Stroke subtypes were classified using data from abstracted medical charts including radiology reports so misclassification of subtype may have occurred because of lack of full information. Furthermore, the TOAST criteria may not be an accurate reflection of stroke mechanism; although a superior classification system is not available [23]. The mRS was retrospectively assessed at discharge by data abstractors, a method that has not been previously validated, but all abstractors were trained by the same individual with training reinforced by a study neurologist. The BASIC study population is over 50% Mexican American and thus differs from the general US population. Findings from this study population may not be generalizable. Several end points were assessed in our analysis and thus readers are cautioned regarding the multiple comparisons, although the number of comparisons was not large at five.
Conclusion
In our study population, we found that a positive family history of stroke among a first degree relative was related to ischemic stroke subtype and also to functional status at discharge. More research is needed to understand whether stroke subtype would be a useful selection criterion for genetic association studies and to hypothesize about a possible genetic link to recovery following ischemic stroke.
Competing interests
The author(s) declare that they have no competing interests.
Authors' contributions
LDL conceived of this analysis from the BASIC study, performed statistical analysis, and drafted the manuscript. MAS prepared the data for analysis and critically reviewed the manuscript. DLB helped to interpret the data analysis and draft the manuscript. KU carried out the ischemic stroke subtyping and critically reviewed the manuscript. LBM is the principal investigator of BASIC. He helped to conceive of this analysis, performed ischemic stroke validation and subtyping, helped interpret the data analysis, and critically reviewed the manuscript. All authors read and approved the final manuscript.
Pre-publication history
The pre-publication history for this paper can be accessed here:
Acknowledgements
Grant Support: This study was funded by NIH RO1 NS38916.
==== Refs
Rubattu S Ridker P Stampfer MJ Volpe M Hennekens CH Lindpaintner K The gene encoding atrial natriuretic peptide and the risk of human stroke Circulation 1999 100 1722 1726 10525492
Brass LM Isaacsohn JL Merikangas KR Robinette CD A study of twins and stroke Stroke 1992 23 221 223 1561651
Liao D Myers R Hunt S Shahar E Paton C Burke G Province M Heiss G Familial history of stroke and stroke risk. The Family Heart Study Stroke 1997 28 1908 1912 9341694
Kiely DK Wolf PA Cupples LA Beiser AS Myers RH Familial aggregation of stroke. The Framingham Study Stroke 1993 24 1366 1371 8362432
Casas JP Hingorani AD Bautista LE Sharma P Meta-analysis of genetic studies in ischemic stroke: thirty-two genes involving approximately 18,000 cases and 58,000 controls Arch Neurol 2004 61 1652 1661 15534175 10.1001/archneur.61.11.1652
Gretarsdottir S Sveinbjornsdottir S Jonsson HH Jakobsson F Einarsdottir E Agnarsson U Shkolny D Einarsson G Gudjonsdottir HM Valdimarsson EM Einarsson OB Thorgeirsson G Hadzic R Jonsdottir S Reynisdottir ST Bjarnadottir SM Gudmundsdottir T Gudlaugsdottir GJ Gill R Lindpaintner K Sainz J Hannesson HH Sigurdsson GT Frigge ML Kong A Gudnason V Stefansson K Gulcher JR Localization of a susceptibility gene for common forms of stroke to 5q12 Am J Hum Genet 2002 70 593 603 11833004 10.1086/339252
Morgenstern LB Smith MA Lisabeth LD Risser JM Uchino K Garcia N Longwell PJ McFarling DA Akuwumi O Al-Wabil A Al-Senani F Brown DL Moye LA Excess stroke in Mexican Americans compared with non-Hispanic Whites: the Brain Attack Surveillance in Corpus Christi Project Am J Epidemiol 2004 160 376 383 15286023 10.1093/aje/kwh225
Lisabeth LD Kardia SL Smith MA Fornage M Morgenstern LB Family history of stroke among Mexican-American and non-Hispanic white patients with stroke and TIA: Implications for the feasibility and design of stroke genetics research Neuroepidemiology 2004 24 96 102 15459516 10.1159/000081056
Piriyawat P Smajsova M Smith MA Pallegar S Al-Wabil A Garcia NM Risser JM Moye LA Morgenstern LB Comparison of active and passive surveillance for cerebrovascular disease: The Brain Attack Surveillance in Corpus Christi (BASIC) Project Am J Epidemiol 2002 156 1062 1069 12446264 10.1093/aje/kwf152
Asplund K Tuomilehto J Stegmayr B Wester PO Tunstall-Pedoe H Diagnostic criteria and quality control of the registration of stroke events in the MONICA project Acta Med Scand Suppl 1988 728 26 39 3202029
Smith MA Risser JM Lisabeth LD Moye LA Morgenstern LB Access to care, acculturation, and risk factors for stroke in Mexican Americans: the Brain Attack Surveillance in Corpus Christi (BASIC) project Stroke 2003 34 2671 2675 14576374 10.1161/01.STR.0000096459.62826.1F
Adams HPJ Bendixen BH Kappelle LJ Biller J Love BB Gordon DL Marsh EE Classification of subtype of acute ischemic stroke. Definitions for use in a multicenter clinical trial. TOAST. Trial of Org 10172 in Acute Stroke Treatment Stroke 1993 24 35 41 7678184
Uchino K Risser JM Smith MA Moye LA Morgenstern LB Ischemic stroke subtypes among Mexican Americans and non-Hispanic whites: the BASIC Project Neurology 2004 63 574 576 15304600
Williams LS Yilmaz EY Lopez-Yunez AM Retrospective assessment of initial stroke severity with the NIH Stroke Scale Stroke 2000 31 858 862 10753988
Young FB Lees KR Weir CJ Strengthening acute stroke trials through optimal use of disability end points Stroke 2003 34 2676 2680 14563964 10.1161/01.STR.0000096210.36741.E7
Harrell FEJ Lee KL Mark DB Multivariable prognostic models: issues in developing models, evaluating assumptions and adequacy, and measuring and reducing errors Stat Med 1996 15 361 387 8668867 10.1002/(SICI)1097-0258(19960229)15:4<361::AID-SIM168>3.0.CO;2-4
Demchuk AM Tanne D Hill MD Kasner SE Hanson S Grond M Levine SR Predictors of good outcome after intravenous tPA for acute ischemic stroke Neurology 2001 57 474 480 11502916
Johnston KC Connors AFJ Wagner DP Knaus WA Wang X Haley ECJ A predictive risk model for outcomes of ischemic stroke Stroke 2000 31 448 455 10657421
Polychronopoulos P Gioldasis G Ellul J Metallinos IC Lekka NP Paschalis C Papapetropoulos T Family history of stroke in stroke types and subtypes J Neurol Sci 2002 195 117 122 11897241 10.1016/S0022-510X(01)00691-8
Jerrard-Dunne P Cloud G Hassan A Markus HS Evaluating the genetic component of ischemic stroke subtypes: a family history study Stroke 2003 34 1364 1369 12714707 10.1161/01.STR.0000069723.17984.FD
Jood K Ladenvall C Rosengren A Blomstrand C Jern C Family history in ischemic stroke before 70 years of age: the Sahlgrenska Academy Study on Ischemic Stroke Stroke 2005 36 1383 1387 15933254 10.1161/01.STR.0000169944.46025.09
Jousilahti P Rastenyte D Tuomilehto J Sarti C Vartiainen E Parental history of cardiovascular disease and risk of stroke. A prospective follow-up of 14371 middle-aged men and women in Finland Stroke 1997 28 1361 1366 9227684
Landau WM Nassief A Editorial Comment - Time to Burn the TOAST Stroke 2005 36 902 904 15800251
|
16262890
|
PMC1295588
|
CC BY
|
2021-01-04 16:28:53
|
no
|
BMC Neurol. 2005 Oct 31; 5:20
|
utf-8
|
BMC Neurol
| 2,005 |
10.1186/1471-2377-5-20
|
oa_comm
|
==== Front
Biomed Eng OnlineBioMedical Engineering OnLine1475-925XBioMed Central London 1475-925X-4-611626288810.1186/1475-925X-4-61ResearchSurface pretreatment for prolonged survival of cemented tibial prosthesis components: full- vs. surface-cementation technique Marx Rudolf [email protected] Mutaz [email protected] Dieter Christian [email protected] Fritz Uwe [email protected] Thorsten [email protected] Department of Prosthetic Dentistry, Section of Dental Materials, University Hospital of the University of Technology, Aachen, Germany2 Department of Orthopedic Surgery, University Hospital of the University of Technology, Aachen, Germany2005 31 10 2005 4 61 61 9 5 2005 31 10 2005 Copyright © 2005 Marx et al; licensee BioMed Central Ltd.2005Marx et al; licensee BioMed Central Ltd.This is an Open Access article distributed under the terms of the Creative Commons Attribution License (), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
Background
One of few persisting problems of cemented total knee arthroplasty (TKA) is aseptic loosening of tibial component due to degradation of the interface between bone cement and metallic tibial shaft component, particularly for surface cemented tibial components. Surface cementation technique has important clinical meaning in case of revision and for avoidance of stress shielding. Degradation of the interface between bone cement and bone may be a secondary effect due to excessive crack formation in bone cement starting at the opposite metallic surface.
Methods
This study was done to prove crack formation in the bone cement near the metallic surface when this is not coated. We propose a newly developed coating process by PVD layering with SiOx to avoid that crack formation in the bone cement. A biomechanical model for vibration fatigue test was done to simulate the physiological and biomechanical conditions of the human knee joint and to prove excessive crack formation.
Results
It was found that coated tibial components showed a highly significant reduction of cement cracking near the interface metal/bone cement (p < 0.01) and a significant reduction of gap formation in the interface metal-to-bone cement (p < 0.05).
Conclusion
Coating dramatically reduces hydrolytic- and stress-related crack formation at the prosthesis interface metal/bone cement. This leads to a more homogenous load transfer into the cement mantle which should reduce the frequency of loosening in the interfaces metal/bone cement/bone. With surface coating of the tibial component it should become possible that surface cemented TKAs reveal similar loosening rates as TKAs both surface and stem cemented. This would be an important clinical advantage since it is believed that surface cementing reduces metaphyseal bone loss in case of revision and stress shielding for better bone health.
==== Body
Background
One of the few still persistent problems of cemented TKA is aseptic loosening of the cemented prosthesis components due to hydrolytic degradation of the interface between bone cement and metal and/or bone, respectively. Thanks to improved surgical techniques, continued development of arthroplasty components and innovative fixation methods for the cemented components, long-time survival rates of at least 90% are reported for 10 to 15 years of follow-up [1,2]. Therefore, total knee arthroplasty is a highly successful procedure, with the percentage of patients requiring revision relatively small [4-7]. However, when considering the large number of these procedures performed annually, a small percentage of failures making revision necessary constitutes a significant number of patients. Annually, 22,000 TKAs are revised in the USA [1] and 35,000 TKAs are revised worldwide [2]. Failure of total knee arthroplasty is devastating to the patient and frustrating for the surgeon. The cost associated with revision surgery is substantial.
The leading causes for failure of TKAs are (only rates above 15 % are quoted) polyethylene wear (25 %), aseptic loosening (24.1 %), instability (21.2 %) and infection (17.5%) [3]. That rate for aseptic loosening (24.1 %) comprises loosening of cemented and uncemented TKAs, the figure for loosening of uncemented TKAs is about 21 % and the figure for surface cementing technique (the undersurface of the tibial base plate is cemented but the stem (keel) is not cemented) is about 10.5 % after an average interval of four years [1].
Under long-term clinical aspects surface cementing technique with press-fit seems to be particularly adequate and the restricted cemented area should be an important advantage. Although the cement application area is limited to the surface sufficient stability is believed to be achievable. This question, however, is controversially disputed in literature, since the loosening rate associated with surface cemented components is with 10.5 % [1] significantly larger than the overall loosening rate for cemented tibial components (3 %; [3]).
On revision fully cemented stems are difficult to remove and revision is accompanied by increased metaphyseal bone loss. Moreover, fully cemented stems promote stress shielding because the tight contact of bone cement and bone in the proximal region of the tibia means additional stiffness which is not physiological [8]. Stress shielding can cause an osteopenia and osteolysis type of bone loss and disuse osteoporosis [9,10]. Therefore, cement application restricted to the surface area may become the technique of fixation of choice in total knee arthroplasty [8] provided the increased rates of aseptic loosening corresponding to this technique can be reduced. This is our goal.
The question of cemented vs. cement-free components is here not addressed. Particularly regarding the tibia plateau cement-free tibial components have poor primary stability (microscopic movements in the range of 30 – 100 μm may occur) and this hinders bone from growing in. Thus the cemented fixation is considered to be the golden standard for tibial components provided a technology is available which improves the stability of only surface cemented tibia components without partially or completely filling up the space between stem and bone with bone cement. Instead press-fit is applied.
As with all cyclically loaded systems improving retention stability and reducing debonding tendency means reducing crack, void and flaw frequencies in the interfaces of TKA/bone cement/bone [11]. Cracks originating at the interface metal/bone cement may also influence stability at the interface bone cement/bone, since due to subcritical crack growth promoted by static and in particular due to cyclic loading cracks starting at the interface metal/bone cement can extent to the interface bone cement/bone (distance to be bridged shorter than 2 mm [10]: Fig. 3 reveals one crack which has already extended to 1 mm). Bone cement has brittle properties and crack branching can play an important part. This is similar to other brittle materials like ceramics and glasses [12,13].
Figure 3 Cracks at the metal (left hand)-cement (right hand) interface made visible using the fluorescence penetration technique.
"Craquelées" in PMMA full dentures are a well known phenomenon in dentistry. Therefore, both interfaces behave not isolated and independent, they interact instead. Reducing crack frequency at the interface metal/bone cement must have positive influence on the interface bone cement/bone.
Since in vivo conditions of the human body (37°C, 100% humidity) strongly favor hydrolysis, and due to the strong interaction of the metal oxide layer of the metal surface of the TKA with polar water molecules, the bone cement/metal interface is prone to hydrolysis [16]. The water causes hydrolytic degradation of the metal/bone cement interface, when it is not conditioned to this environment [15-20]. The water molecules diffuse through the permeable PMMA and easily reach the interface to form a moisture film at the metal-cement interface even when the PMMA layer and the interface is free of cracks [14]. Any cracks or fissures will accelerate the process via capillary action. Thus, the initially solid bond between the oxide layer of the metal surface and the bone cement is eroded and becomes mechanically unstable as the metal-cement interface is progressively hydrolyzed. The well known radio lucent lines discussed frequently in literature are signs of that phenomenon [21,22].
This hydrolytic debonding, which is largely independent of design of the prosthesis, greatly favors microscopic movements in the metal-cement interface even during normal walking due to repeated pressure and corresponding shear stresses. Step by step interface debonding accompanied by increasing appearance of radio lucent lines results in a vicious circle of cement degradation with cement wear and foreign body-induced osteolysis with eventually mechanical (aseptic) loosening of the TKA.
Hence loosening rates of surface cemented tibial components can be reduced if attention is focused on the central and most common failure mechanism: crack formation and its avoidance [11-13].
This study was done to evaluate a new coating process of cemented tibial components which appears to be suitable to reduce the frequency of crack formation drastically. A biomechanical model for vibration fatigue trials was designed to simulate the physiological and biomechanical conditions of the human knee joint to make the cracks visible.
Methods
The new coating method [14,24] consists in silicoating the metal surface by sublimating a stream of silicon monoxide particles onto the metal surface under a high vacuum (p = 10-5 mbar and better; PVD method). The particle source was heated to 1130°C. Nevertheless "pre"heating of the tibial component is not to be feared because a distance of about 100 mm between source and substrate was kept and the shutter window between source and substrate was constricted to about 10 mm. The SiOx layer is held in place by adhesion (Van der Waals bonds, hydrogen bonds, ionic bonds) and by mechanical microbonds. A silane layer (3-methacryloxypropyltrimethoxysilan) is then applied to the silicoated metal surface [15], and both layers are finally coated with a polymer protective varnish consisting of MMA (methylmethacrylate), UDMA (urethanedimethacrylate) and TEGDMA (triethylene glycol dimethylacrylate) to ensure covalent bonding between the coating and the bone cement. The MMA/UDMA/TEGDMA varnish is then hardened by UV light in a desiccator under vacuum. The coating is stable at normal environmental conditions even it is finally gamma-sterilized at up to 25 kGy. That layer has protective function because it protects the highly reactive silicate/silane coating from all chemical and physical contamination [14]. When applying bone cement to the varnish it is integrated into the bone cement by chemical reaction.
In order to access the clinical suitability of this patented (European and US patents [24]) coating process for TKA [25], size T3 Columbus tibial components (CoCrMo alloy, for geometry of stem refer to Fig. 2; Aesculap, Tuttlingen, Germany) were tentatively cemented in composite sawbones for biomechanic studies (Sawbones Europe, Malmö, Sweden; note that we did not use "foam bones"; surface cementing technique applied). Cementation to cadaver bone was considered not to be feasible for hygienic reasons since we did not succeed in finding a partner which was ready to carry out the saw cuts to be described below. The prosthesis was subjected to a cyclical stress test under near-physiological conditions (37°C, 0.9% NaCl solution) in accordance with ISO 14243. Cyclic loading is considered to resemble the physiological load during walking and it promotes the inception of cracks at voids and flaws and their growth with time in brittle materials [12,13].
Figure 2 Tibial components ready for surface cementation onto the Technobones (below). Note that the central hole was machined such that there was no "press fit" between the TKA and the Technobones. This ensured maximum stress in the interface.
This pilot study showed that the sawbones did not withstand a stress of 6 kN. When designing the experiment this load was considered to be adequate since its level did not only reflect the physiological conditions but also supplied an overload to test whether the coating could withstand even larger loads. Retrospective, a load of 3 kN would have been a better choice (corresponds to a person of 75 kg of weight [26]). Results for sawbones are closer to the physiological situation.
We switched over to very stable "technobones" which could withstand the stress of 6 kN. The "technobones" consisted of Technovit® 4002 (Heraeus-Kulzer, Werheim, Germany). Since its main component besides styrol and dimethylanilin is MMA good bonding to bone cement is assured due to near chemical neighborhood. Technovit "powder" and Technovit "liquid" were mixed at a weight ratio of 2:1 and poured into a mold to achieve similar contours as known from the sawbones. In order to improve their wettability by the bone cement when mounting the tibial components the proximal surface of the "technobones" was coated with the varnish described above.
Young's moduli of the sawbones were E = 15 GPa and E = 0.5 GPa for the cortical surface and the spongeous inner part, respectively. In order to check whether young's modulus of the "technobones" was similar to that of the sawbones we prepared five rods (length: 55 mm, width: 6 mm, thickness: 3 mm) which were subjected to a tensile stress test (universal testing machine Zwick 030, Zwick, Ulm, Germany). Using the standard formula Young's modulus was found to be E = 1.4 ± 0.6 GPa, similar to the spongeous inner parts of the sawbones.
The vibration fatigue test was considered to be a suitable biomechanical model (geometry of load: refer to Fig. 1). It was done using a servo pneumatically controlled 6kN two-column tabletop tester (Dyna-Mess, Aachen, Germany). The samples (tibial components cemented to the "technobones": Fig. 2; surface cementing technique applied, no press-fit, stem neither conditioned nor cemented) were kept in a plexiglass chamber under physiological conditions, i.e., at 37°C and at 100% relative humidity while being immersed in a 0.9% NaCl solution. This environmental conditions ensured hydrolytic conditions similar to the human body.
Figure 1 Load geometry for servopneumatically controlled 6kN two-column tabletop tester.
The test arrangement was mounted on ball bearings for free movement in the anterior-posterior direction and to eliminate all transverse forces while allowing rotation. Inclination of the tibial plane was 10°.
Each tibial implant was exposed to 106 alternating (sine) stress cycles, with a minimum and maximum load of 150 N and 6,000 N, respectively and a cycle frequency of 3 Hz to simulate the physiological conditions in the knee over an average walking period of one year. Total test duration for one component was 3.86 days (about 93 hours). A total number of 8 coated (before PVD-layering abraded by Al2O3-blasting (grain size 120 μm) and 12 uncoated (F22 (Ra = 3.41 μm) = 8, Silbermatt (Ra = 0.49 μm) = 2, B60 (Ra = 1,96 μm) = 2; P1 and P2) tibial components were tested. After the vibration fatigue trials, the components were removed and kept in a wet atmosphere at room temperature, as drying was shown reducing crack frequencies and restoring bonding by virtue of water leaving the interface.
For evaluation, the prostheses were sliced vertically using a high-pressure water jet (IPT, Frauenhofer-Insitut, Aachen, Germany), cleaned from debris and air-dried. Afterwards the surface was microscopically examined for cracks at the metal-cement interface using the fluorescent penetration technique (Fig 3). A penetration fluid (MET-L-CHEK Penetrant FBP 913, HELLING, Heidgraben, Germany) with a high capillarity and a low surface tension enters cracks in the surface. Scanning the area with a UV light microscope (Leica DMR X, Leica, Wetzlar, Germany; approx. wavelength 360 nm) revealed cracks even as narrow as 0.25 μm. To detect gaps the surface was investigated using standard microscopic technique (reflected-light microscopy).
Results
The length and maximum width of all cracks and gaps were recorded. Fig. 3 shows an example of a crack area and Fig. 4 an example of a gap. However, the overall number of gaps detected was much smaller than the number of cracks. Hence we focused to the number and quality of cracks utilizing their numbers and quality as criterions to differentiate between coated and uncoated samples.
Figure 4 Gap at the metal (above)-cement (below) interface.
The coated components showed almost no cracks. The cumulative crack length of them was 937 μm while the uncoated components showed cumulative crack lengths of 13,999 – 27,351 μm (Table 1).
Table 1 Cumulative crack lengths for various prostheses
Prosthesis Length [μm]
Silbermatt (m = 2) 13,999
F22 (m = 8) 24,693
B60 (m = 2) 27,351
Coated (m = 8) 937
The differences in cumulative crack lengths between the firstly tested "P1-B60" components (13,180 μm) and the secondly tested "P2-B60" (41,521 μm) components are due to the fact that both the P1-B60 and P2-B60 interfaces detached from the metal surface during the trial (the P1-B60 interface in an early stage and the P2-B60 interface towards the end). There were clearly more lateral than medial cracks.
The coated components showed the best overall performance (Table 2, 0.8%). There were no significant differences between uncoated, Silbermatt (Ra = 0.49 μm, 1.7%) and uncoated F22 (Ra = 3.41 μm, 2.5%) components, although the latter both as a group differed greatly from the B60 (Ra = 1.96 μm, 69%) components. The B60 (m = 2) components also showed a significant difference between medial (49.3%) and lateral (87.7%) cracks, which the other types did not (Silbermatt (m = 2): 1.7% vs. 1.7%; F22 (m = 8): 2.2% vs. 2.7%; coated (m = 8): 0.9% vs. 0.8%).
Table 2 Relative proportion of gap formation by prosthesis type
Coated Prostheses all 0.8 %
Uncoated Prostheses Silbermatt (m = 2) 1.7 %
F22 (m = 8) 2.5 %
B60 (m = 2) 69.0 %
Since there were more short than long cracks, distribution was asymmetrical. This suggested a Weibull statistical analysis (examples and graphical method: Fig. 5 and Fig. 6; summary of results: Table 3) which yielded for Silbermatt components a median of 260 μm, a modal value of 175 μm and the lowest number of cracks (52). The F22 had a median of 239 μm, a modal value of 205 μm and 112 cracks. The B60 components, the interfaces of which were partially debonded on a macroscopic scale during the vibration fatigue trial, had a median of 170 μm, a modal value of 133 μm and 177 cracks (refer also to Fig. 5 and Fig. 6). All the components types have almost the same Weibull shape parameter m = 2. "Median" means that half of the crack lengths will be less than this value and half will be larger. The modal-value gives the crack length with the highest probability and the shape parameter, m, tells how blurred the distribution is, i.e. if all the crack lengths tend to be not very close to a certain value, the distribution will have a low shape parameter m, and the distribution appears to be very flat and blurred.
Figure 5 Prosthesis with surface specification B60: Weibull distribution of crack lengths. Modal value π = 120 μm (crack with maximum propability P). Median value of crack length δ = 163 μm (P = 63%; from Fig. 6). Note that the values given in Tab. 3 for B60 are slightly different (170 μm and 133 μm for Median and Modal value, respectively) since those values represent the arithmetic means of characteristic values determined for all prosthesis of this type.
Figure 6 Prosthesis with surface specification B60: Weibull plot of crack length. Median value of crack length μ = 163 μm. Modal value of crack length π = 120 μm (from Fig. 5). Modul m = 2,1.
Table 3 Characteristic Weibull parameters of uncoated prostheses
Prostheses Area Median δ [μm] Modal π [μm] Number n
PSilbermatt all 260 175 52
PF22 all 239 205 112
PB60 all 170 133 177
Discussion
Our findings show that mechanical degradation in combination with hydrolytic load of the interface metallic tibial surface/bone cement is an important failure mechanism for aseptic loosening of surface cemented tibial prosthesis components. From previous investigations we have learnt [14] that the bond strength between bone cement and unconditioned samples of CoCrMo is unstable against hydrolysis [15,16] and stability is limited by retention due to surface roughness. PVD conditioned samples (silicoating, silane, coating varnish), however, reveal stable bond strengths even in a humid surrounding [23]. These results collected from investigations on a sample geometry which is most suitable to measure bond strengths under tension load (adhesion means resistance against normal loads; these are aligned perpendicular to the adhesion surface) appear to become resembled by our investigations on real tibial components. Without PVD layering and varnish implant stability is limited by the retentive micro- and macro mechanical anchoring of the PMMA polymer to the corundum blasted metal surface. Since no chemical bonding has taken place, the metal surface is prone to hydrolytic degradation [14]. In addition cracks and gaps arise in the bone cement, starting from irregularities in a surface near PMMA layer and from voids in the balk of the PMMA [11]. Irregularities in a surface near the PMMA (e.g.) layer are mainly due to the retention mechanism which is typical for a material combination of a rough metal and a plastic layer. The corundum grains at high speed blasted onto the surface of the metal score deeply into the surface, creating furrows which are bordered by rough, upturned edges. The burrs cut into the plastic and if loaded they provide weak links (predetermined breaking points). The weak links result in cracks and extend due to cyclic load in analogy to subcritical crack propagation in ceramics leading to further deterioration of the cement layer, PMMA abrasion and (in vivo) to foreign body-induced resorption at the bone matrix, increasing the micro movements in a vicious circle, eventually making replacement necessary.
In the contrary, when the rough surface is PVD layered by a SiOx film the surface becomes much smoother and the burrs and edges become masked by this layer. Moreover, there is no longer the need of a rough surface with deep cuts. A textured surface for an increase of the effective surface is sufficient. The weak links disappear then or its number is reduced. The surface is no longer prone to emit cracks into the PMMA cement in abundance.
Hence, in order to reduce or even prevent the described mechanisms of degradation and loosening, the prosthesis surface of knee implants must be modified to provide a more durable, hydrolysis-proof interface between bone cement and metal. Strong covalent bonds between bone cement and the prosthesis prevent micro movements at the interface, and hence formation of cracks and aseptic loosening of the implant.
Similar problems are encountered in the field of prosthodontics with veneering crowns and bridges by plastics, where the solution was a firmly adhering layer of silicate [23]. After applying this principle to cemented hip endoprosthesis, the present investigation showed that it became possible to create a durable, hydrolysis-resistant interface between the metallic prosthesis and the bone cement polymer, thus probably improving the life of endoprosthesis at a rate not yet known [25].
The vibration fatigue trials at 106 stress cycles and after 7 days in physiological (0,9 %) NaCl solution, done to simulate downhill walking, showed that under standardized and reproducible in-vitro conditions, the coating system withstands the typical mechanical stress present under physiological conditions. The results were confirmed by interface analysis. No cracks were found in three of the eight coated components, the other showed only minimal cumulative crack lengths (937 μm), whereas the uncoated components F22 showed significantly more and longer cracks (Table 1). Comparing the surface roughness of the uncoated components, it was seen that the Silbermatt components exhibited lower cumulative crack lengths than the F22 components while the B60 components showed a fatal crack spectrum and partial macroscopically debonding of the metal-bone interface.
Analyzing the crack locations along the medial and lateral lines of symmetry clearly revealed more lateral than medial cracks. Parting from the mediolateral distribution of load (67.5% medially and 32.5% laterally), it becomes apparent that the cracks are caused more by tensile than by compressive stress. These results agree with reports in the literature. Bone cement has a higher compressive than tensile strength (80 MPa vs. 30 MPa), the latter thus being critical in causing material failure [25].
The coated components showed the best overall performance, with a gap formation ratio of 0.8%, whereas the uncoated components showed more variable rates (Table 2). Due to the partial debonding and interface deterioration of the B60 components, the crack rate was 69%. The results clearly show that the new coating system is well suited for preventing hydrolytic debonding at the metal-bone interface.
Since there were more smaller than longer cracks, an asymmetrical crack spectrum originated. This suggested to endeavor a Weibull analysis of crack distribution. This kind of analysis revealed to be very suitable for this purpose because it enabled us to list median and modal values characterizing the crack spectra of the uncoated components investigated.
By the coated components we could not employ the Weibull distribution because they revealed almost no cracks. By the other component types one sees that the Silbermatt ones have done best. They have the smaller modal value and the lowest number of cracks in comparison to the F22 one (Table 3).
Earlier efforts to decrease the rate of aseptic loosening, however for femoral components, resulted in a considerably higher rate of failure with failure occurring earlier. The components were designed to have a more roughened or textured surface and a PMMA precoating [27,28]. Note that under chemical aspects there is no reason why a PMMA precoating should have a higher resistance against hydrolysis than a surface near PMMA layer which is part of the cementing. To make a surface more rough is of no use because the above described effect of burrs will be more pronounced and the failure due to crack development and crack propagation will occur earlier.
Conclusion
There is a dilemma to be solved: Cementing the stem means improved retention and additional orientation for the tibial component. However stress shielding occurs which is an unfavorable condition for a healthy bone. In case of revision additional bone is lost.
Only cementing the surface of the tibial component means largely reduced overall retention area and loss of orientation for the tibial component. This is reflected by the loosening rate of 10.5 % after an average interval of four years [1]. The rate may be diminished by "press fitting" the stem.
In the present contribution we have demonstrated how this dilemma can be solved: gaining additional retention for the tibial surface by application of a layer system which stabilizes the retention under hydrolytic and cyclic load and which helps avoiding the initiation of cracks. Then cementing of the stem can be abandoned and a more physiological load of the tibia head is reached. Stabilizing the interface metal/bone cement also means indirectly stabilizing the interface bone cement/bone since the frequency of cracks which emanate from the metal surface and propagate through the bone cement to the bone surface is dramatically reduced by the proposed and patented layer system.
Is should be kept in mind that for the presently discussed results instead of natural bones we utilized artificial "techno"bones. Therefore, a multicentre study with several university clinics is going ahead to validate these results in vivo.
A particular interesting aspect will be to notice whether also the frequency of radio lucency is be reduced for coated tibial components.
Authors' contributions
RM conceived in the study, participated in the design, carried out the statistical analyses and drafted the manuscript. MQ carried out experimental work. DCW conceived in the study and participated in the design. FUN coordinated the work. TM carried out experimental and analytical work and participated in drafting the manuscript. All authors read and approved the final manuscript.
Acknowledgements
The project using the Vacuum Vapor Deposition Method to improve the long-term bonding strength between an orthopedic metal endoprosthesis and PMMA bone cement has been supported by the German Research Community (DFG, project Ma 1093/4-1,2). It was realized by close cooperation between the Department of Dental Prosthetics, Section of Dental Materials, Department of Orthopedics, both University Hospital of the University of Technology, Aachen and B Braun Aesculap, Tuttlingen.
==== Refs
Sharkey PF Hozack WJ Rothman RH Shastri S Jacoby SM Why Are Total Knee Arthroplasties Failing Today? Clin Orthop 2002 404 7 13 12439231
Sharkey PF Hozack WJ Rothman RH Shastri S Jacoby SM Why Are Total Knee Arthroplasties Failing Today? John Insall Award. Program and abstracts of the Combined Speciality Day meeting of the AAHKS Dallas, Texas February 16, 2002
Austin MS Sharkey PF Hozack WJ Rothman RH Knee Failure Mechanisms After Total Knee Arthroplasty Techniques in Knee Surgery 2004 3 55 59 10.1097/00132588-200403000-00008
Gonzalez MH Mekhail AO The Failed Total Knee Arthroplasty: Evaluation and Etiology J Am Acad Orthop Surg 2004 12 436 446 15615509
Rand JA Trousdale RT Ilstrup DM Harmsen WS Factors Affecting the Durability of Primary Total Knee Prostheses J Bone Joint Surg Am 2003 85 259 265 12571303
Scuderi GR Pagnano MW The rationale for posterior cruciate substituting total knee arthroplasty J Orthop Surg 2001 9 81 88
Gill GS Joshi AB Long-term results of Kinematic Condylar knee replacement J Bone Joint Surg [Br] 2001 83-B 355 358 10.1302/0301-620X.83B3.11288
Peters CL Craig MA Mohr RA Bachus KN Tibial Component Fixation With Cement. Full-Versus Surface-Cementation Techniques Clin Orthop 2003 409 158 168 12671498
Van Loon CJ de Waal Malefijt MC Buma P Verdonschot N Veth RP Femoral bone loss in total knee arthroplasty. A review Acta Orthop Belg 1999 65 154 163 10427796
Nyman JS Hazelwood SJ Rodrigo JJ Martin RB Yeh OC Long stemmed total knee arthroplasty with interlocking screws: a computational bone adaption study J Orthop Res 2004 22 51 57 14656659 10.1016/S0736-0266(03)00159-1
Spiegelberg StH Ruberti JW Braithwaite GGC System and methods for reducing interfacial porosity in cements US Patent 6,884,264 issued on 04/26/2005
Green DJ An introduction to the mechanical properties of ceramics 1998 Cambridge, Cambridge University Press
Munz D Fett T Ceramics – Mechanical Properties, Failure Behaviour, Materials Selection 1999 Berlin Heidelberg, Springer
Erli HJ Marx R Paar O Niethard FU Weber M Wirtz DC Surface pretreatments for medical application of adhesion BioMedical Engineering OnLine 2003 2 15 33 14561228 10.1186/1475-925X-2-15
Plueddemann EP Silane coupling agents 1982 New York, Plenum Press
Kinloch AJ Adhesion and Adhesives 1987 London, Chapman and Hall
Habenicht G Kleben. Grundlagen, Technologien, Anwendungen 1990 Berlin, Springer
Brewis DM Comyn J Tegg JL The durability of some epoxide adhesive-bonded joints on exposure to moist warm air Int J Adhesion and Adhesives 1980 1 35 39 10.1016/0143-7496(80)90032-9
Marx R Ursachen für das Versagen von Klebeverbindungen bei Adhäsivbrücken Schweiz Monatssch Zahnmed 1987 97 1081 1086
Marx R Kerschbaum Th Auswahl der Werkstoffe Adhäsivprothetik München, Urban und Schwarzenberg 1995 107 120
Ritter MA Herbst SA Keating EM Faris PM Radiolucency at the bone-cement interface in total knee replacement J Bone Joint Surg 1994 76A 60 65 8288666
Ecker ML Lotke PA Windsor RE Cella JP Long-term results after total condylar knee arthroplasty. Significance of radiolucent lines Clin Orthop 1987 216 151 158 3815942
Musil R Tiller HJ Der Kunststoff-Metall-Verbund in der zahnärztlichen Prothetik Leipzig, VEB JA Barth 1988
Marx R Fischer H Work piece and method for producing and utilizing said work piece US Patent 6,777,028 issued on 08/17/2004
Wirtz DC Eine neue Beschichtungsmethode für zementierte Femurschaftimplantate zur hydrolysestabilen Optimierung des Metall-Knochenzement-Verbundes Aachen, Aachener Beiträge zur Medizin, Band 25, Wissenschaftsverlag Mainz, Aachen 2002 ISBN 3-86073-945-x
Bergmann G In vivo Messung der Belastung von Hüftimplantaten Habilitationsschrift, Freie Universität Berlin 1994
Ong A Wong KL Lai M Garino JP Steinberg ME Early failure of precoated femoral components in primary total hip arthroplasty J Bone Joint Surg Am 2002 84 786 792 12004022
Strickland AB Chen YH Andriacchi TP Miller J The initial fixation of porous coated tibial components evaluated by the study of rigid body motion under static load Trans Orthop Res Soc 1988 13 476 481
|
16262888
|
PMC1295589
|
CC BY
|
2021-01-04 16:37:36
|
no
|
Biomed Eng Online. 2005 Oct 31; 4:61
|
utf-8
|
Biomed Eng Online
| 2,005 |
10.1186/1475-925X-4-61
|
oa_comm
|
==== Front
Plant MethodsPlant Methods1746-4811BioMed Central London 1746-4811-1-91627091010.1186/1746-4811-1-9MethodologyCombining metal oxide affinity chromatography (MOAC) and selective mass spectrometry for robust identification of in vivo protein phosphorylation sites Wolschin Florian [email protected] Wolfram [email protected] Max Planck Institute of Molecular Plant Physiology, 14424 Potsdam, Germany2005 1 11 2005 1 9 9 27 8 2005 1 11 2005 Copyright © 2005 Wolschin and PWeckwerth; licensee BioMed Central Ltd.2005Wolschin and PWeckwerth; licensee BioMed Central Ltd.This is an Open Access article distributed under the terms of the Creative Commons Attribution License (), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
Background
Protein phosphorylation is accepted as a major regulatory pathway in plants. More than 1000 protein kinases are predicted in the Arabidopsis proteome, however, only a few studies look systematically for in vivo protein phosphorylation sites. Owing to the low stoichiometry and low abundance of phosphorylated proteins, phosphorylation site identification using mass spectrometry imposes difficulties. Moreover, the often observed poor quality of mass spectra derived from phosphopeptides results frequently in uncertain database hits. Thus, several lines of evidence have to be combined for a precise phosphorylation site identification strategy.
Results
Here, a strategy is presented that combines enrichment of phosphoproteins using a technique termed metaloxide affinity chromatography (MOAC) and selective ion trap mass spectrometry. The complete approach involves (i) enrichment of proteins with low phosphorylation stoichiometry out of complex mixtures using MOAC, (ii) gel separation and detection of phosphorylation using specific fluorescence staining (confirmation of enrichment), (iii) identification of phosphoprotein candidates out of the SDS-PAGE using liquid chromatography coupled to mass spectrometry, and (iv) identification of phosphorylation sites of these enriched proteins using automatic detection of H3PO4 neutral loss peaks and data-dependent MS3-fragmentation of the corresponding MS2-fragment. The utility of this approach is demonstrated by the identification of phosphorylation sites in Arabidopsis thaliana seed proteins. Regulatory importance of the identified sites is indicated by conservation of the detected sites in gene families such as ribosomal proteins and sterol dehydrogenases. To demonstrate further the wide applicability of MOAC, phosphoproteins were enriched from Chlamydomonas reinhardtii cell cultures.
Conclusion
A novel phosphoprotein enrichment procedure MOAC was applied to seed proteins of A. thaliana and to proteins extracted from C. reinhardtii. Thus, the method can easily be adapted to suit the sample of interest since it is inexpensive and the components needed are widely available. Reproducibility of the approach was tested by monitoring phosphorylation sites on specific proteins from seeds and C. reinhardtii in duplicate experiments. The whole process is proposed as a strategy adaptable to other plant tissues providing high confidence in the identification of phosphoproteins and their corresponding phosphorylation sites.
Protein phosphorylationNano-ESISeedsIMACMOACNeutral lossMS3
==== Body
Background
The proteome of different developmental stages of any kind of organism reflects more directly than the genome or the transcriptome the metabolic specialisation for the actual developmental state. In plants several studies on the proteome of different developmental stages have been conducted [1]. Seed dormancy plays a crucial role in the life cycle of plants and its proteome reflects the metabolic processes during this important developmental period. However, investigation on posttranslational modifications of the proteins gives an even more detailed view on the complex nature of seed metabolism. Protein phosphorylation has been widely described as a major regulatory protein posttranslational modification influencing many important processes in living cells [2-4]. Thus, measuring protein phosphorylation is essential to reveal regulatory and signal pathways. However, the study of protein phosphorylation confronts the researcher with several hurdles.
A complicating fact is that many proteins are not only phosphorylated at one site, but on multiple sites and that each modification seems to have different regulatory functions [5,6]. Therefore, detection of protein phosphorylation and identification of phosphorylation sites are needed for the understanding of protein regulation.
Traditionally, phosphorylation is detected by specific antibodies and/or by incorporating radioactive [32P]orthophosphate into proteins. However, while immunolabelling is often unspecific, incorporation experiments using radioactivity might result in artificial phosphorylation events and impose waste disposal problems. Only recently has it become possible to reliably detect protein phosphorylation by resolving the proteins of interest on a gel and submitting the gel to fluorescent phosphate specific dyes followed by a staining of total protein [7-11].
The low abundance of phosphoproteins and the accurate identification of the specific phosphorylation sites, however, still impose problems.
Because of their sensitivity and resolving power, selective enrichment of phosphorylated peptides/proteins and liquid chromatography coupled to mass spectrometry (LC-MS) based methods are now widely used for phosphorylation site identification (for review see [4,12,13]).
The low abundance of phosphorylated peptides can be circumvented in part by enrichment of the phosphopeptides with IMAC (Immobilised metal affinity chromatography) after tryptic digestion of phosphoproteins. While numerous publications exist describing the enrichment of phosphopeptides from animal sources considerably less studies have been published for plant tissue (for review see [14]). These broad range studies focus for example on thylakoid proteins and plasma membrane proteins of A. thaliana [15,16] or on the moss Physcomitrella patens [17].
However, relying on just one peptide for protein identification (as is commonly done after phosphopeptide enrichment) is prone to the identification of false positives. What is more, peptides phosphorylated on serines or threonines tend to loose their phosphate group during the fragmentation process in the mass spectrometer thus further complicating correct assignment. This drawback can be circumvented by the enrichment of complete phosphoproteins since this approach leads to the identification of several peptides per protein and therefore enhances the reliability of protein identification.
Recently, we reported a novel method for the enrichment of phosphorylated proteins out of complex mixtures termed MOAC (metal oxide affinity chromatography) [18]. This method adds another tool to the repertoire of methods for the identification of phosphorylated proteins and might help to overcome some of the specificity problems associated with IMAC. In this initial study we showed that MOAC can be used to enrich phosphorylated proteins from plant leaf tissue [18].
In the present study we use MOAC for the enrichment of phosphorylated proteins from A. thaliana seeds and for proteins from C. reinhardtii cell cultures and show that the method is suitable and reproducible for different kinds of samples.
However, for the identification of the exact sites of phosphorylation further steps are necessary. As mentioned above the ionization efficiency and the quality of a phosphopeptide spectra is sometimes not good enough for reliable identification and even more difficult is the determination of the exact phosphorylation site. To circumvent these problems phosphate groups can be replaced by beta-elimination and Michael addition with more stable residues thereby increasing the ionization efficiency and improving the fragmentation behaviour [5,19-23]. Yet, these approaches sometimes result in unwanted side reactions and are difficult to perform on complex mixtures. Another promising method is the analysis of peptide sequences by electron transfer dissociation [24] but this requires sophisticated modification of the mass spectrometer not yet widely available.
The approach we employed to identify phosphorylation sites makes use of the neutral loss of H3PO4 during MS2 fragmentation. The dominant neutral loss peak observed in many MS2 spectra derived from serine/threonine phosphorylated peptides [25] is routinely broken down in an additional MS3 step and putative phosphopeptides are automatically detected in a single LC/MS run [26]. The resulting MS3 spectra are often informative enough to identify the correct peptide and the phosphorylation site in a database search. However, considering the information of both MS2 and MS3 data is sometimes necessary to obtain unambiguous identification [27,28].
In the present work we combine the MOAC enrichment of phosphoproteins and selective mass spectrometry for a detailed study of protein phosphorylation. Phosphorylation site identification is demonstrated for enriched seed proteins and is achieved using data-dependent MS2 as well as neutral-loss triggered MS3 fragmentation on a linear ion trap mass spectrometer. The proposed strategy is suitable to identify putative phosphoprotein candidates with high sequence coverage and at the same time it allows identification of corresponding protein phosphorylation sites. To avoid false positive identification of phosphorylation sites only hits with congruent MS2 and MS3 spectra were considered in this study. Careful validation of the data led to the identification of 16 phosphorylation sites in nine seed proteins, some of them known to be phosphorylated also in the mammalian system such as ribosomal proteins. A comparison with microarraydata showed that these proteins are mainly seed specific.
Results
General strategy for the identification of serine/threonine phosphorylation in plants
An overview on the general strategy for the identification of serine /threonine phosphorylation in plants is shown in Figure 1. The strategy includes enrichment of phosphoproteins using MOAC, phosphorylation detection using fluorescent dye technology, and determination of the phosphorylation site with neutral loss driven MS3. Most important, each step of the procedure gives further confidence for a robust identification of phosphoproteins and phosphorylation sites. For a detailed description see the following sections.
Figure 1 Proposed strategy for robust identification of serine/threonine phosphorylation in plants. In case of low stoichiometry as for the in vivo situation enrichment of phosphoproteins is necessary. To cope with this a novel enrichment procedure called metaloxide affinity chromatography (MOAC) is used in the strategy [18]. The whole approach is applicable to identify phosphorylation sites out of complex samples. The library of in vivo sites may be used for screening of protein phosphorylation dynamics using triple-quadruple mass spectrometry [6].
MOAC enrichment of phosphoproteins
Proteins were extracted with phenol and enriched for phosphoproteins using metaloxide affinity chromatography (MOAC) [18]. Proteins were separated using SDS-PAGE and visualised with a phosphate-specific stain followed by staining with coomassie R-250 (Figure 1). Similar amounts of total protein from samples taken before and after MOAC (Figure 2B) are accompanied by strong differences in the phosphostain signal indicating clear enrichment of phosphoproteins out of the complex sample (Figure 2A). Bands corresponding to enriched seed proteins with a strong signal in the phosphostain were excised, digested with trypsin, and analysed by mass spectrometry (see below).
Figure 2 A. thaliana seed proteins and C. reinhardtii proteins before and after MOAC. M: marker S: sample (before MOAC) E: eluate (after MOAC); A: A. thaliana seed proteins, Coomassiestaining; B: A. thaliana seed proteins, Phosphostaining. C: C. reinhardtii proteins, Coomassiestaining; D: C. reinhardtii proteins, Phosphostaining. The labelled marker protein in the phosphostain is the phosphorylated protein Ovalbumin.
Identification of in vivo phosphorylation sites in A. thaliana seed proteins
Following the proposed strategy (see Figure 1) we identified after careful validation 16 phosphorylation sites in 9 proteins using the combined information of MS2 and MS3 data. The advantage of using the combined data is exemplified by the MS2 and MS3 spectra in Figure 3. In the first MS2 fragmentation step a very intense fragment ion stemming from a dominant neutral loss of phosphoric acid is seen. This spectrum alone is often not suitable for database search. On the other hand it is a distinct indication for a phosphopeptide. The dominant neutral loss fragment triggers in a second scan event an MS3 fragmentation step. The combination of MS2 and MS3 leads to the clear phosphorylation site identification with increased confidence based on the observed neutral loss of phosphoric acid [27,28]. A further level of confidence is provided by the combination of protein identification (see Figure 1 and protein sequence coverage in table 1) and the corresponding protein phosphorylation site in one LC/MS analysis. This information is missing in strategies where only phosphopeptides are enriched and protein identification is based solely on the detection of one phosphopeptide. The phosphorylation sites are shown in Table 1. The sites of the two ribosomal proteins are known to be phosphorylated in mouse [29]. For some of the proteins identified after MOAC, MS2 data suggested phosphorylation but MS3 data were lacking. All these ambiguous results were not included in the table. The analysis of gene expression data revealed that five of the identified proteins are apparently seed specific (table 1). Notably, phosphorylation is most probably not confunded with o-sulfonation in these experiments since sulfated peptides exhibit the characteristic loss of 80 Da instead of 98 Da during collision induced dissociation [30], which does not lead to the triggering of an MS3 spectrum.
Figure 3 MS2 spectrum and the cognate neutral loss MS3 spectrum of the phosphopeptide LGYTGENGGGQSEpSPVKDETPR exemplifying the additional sequence information gained of an MS2 spectrum compared to a MS3 spectrum. A: MS2 spectrum (Xcorr: 2.424); B: MS3 spectrum (Xcorr: 3.248). Exact localisation of the phosphorylation site is only possible with the MS3 spectrum in this case. Only ions above the noise signal were annotated.
Table 1 Identified phosphorylation sites in seed proteins. Gene expression data were derived from a previous study [38].
Accession nr. Description Protein sequence coverage Site of phosphorylation (designated as "p") Seed specific expression
At5g52300.1 low-temperature-responsive 65 kD protein (LTI65)/desiccation-responsive protein 29B (RD29B) 65.8% (Mr 65971 Da) MKVTDEpSPDQKSR
ESDINKNpSPARFGGESK
LPLSGGGpSGVKETQQGEEK
LGYTGENGGGQSEpSPVKDETPR
GAVTSWLGGKPKpSPR
EAHQEPLNpTPVSLLSATEDVTR +
At3g12960.1 expressed protein similar to seed maturation protein from [Glycine max] 93% (Mr 9464 Da) DIKDIKGTRTDDpSPR.- +
At1g01100.1 60S acidic ribosomal protein P1 (RPP1A) 81.2% (Mr 11162 Da) KKDEPAEEpSDGDLGFGLFD.- -
At2g27710.1 60S acidic ribosomal protein P2 (RPP2B) 87% (Mr 11444 Da) KEEKEEpSDDDMGFSLFE.- -
At5g40420.1 glycine-rich protein/oleosin 42.2% (Mr 21279 Da) HFQFQpSPYEGGR +
At1g07985.1 Expressed protein 46.5% (Mr 16461 Da) KLVDKVVGSSSPTNIHpSK
At5g50600.1 short-chain dehydrogenase/reductase (SDR) family protein similar to sterol-binding dehydrogenase steroleosin from [Sesamum indicum] 61% (Mr 39087 Da) STLYPESIRTPEIKpSD.-
ELGpSPNVVTVHADVSKPDDCRR +
At1g29350.1 expressed protein 14.6% (Mr 90879 Da) SGpSTHFSSTDSGNFQGK No data
At4g25580.1 stress-responsive protein-related contains weak similarity to Low-temperature-induced 65 kDa protein 57.8 % (Mr 66520 Da) RGAPTLTPHNTPVSLLpSATEDVTR
GAPTLpTPHNTPVSLLSATEDVTR +
Reproducibility of the method
The enrichment process was repeated using A. thaliana leaf and C. reinhardtii proteins. Similar patterns were observed in SDS PAGE analysis combined with phospho- and coomassiestaining (data not shown). Two prominent bands, one of the enriched seed sample at about 65 kDa (corresponding to the mw of the protein with the highest number of phosphorylation sites identified in the first experiment), and one of the enriched C. reinhardtii sample (in duplicate) at about 56 kDa, were selected for testing the reproducibility of phosphorylation site identification. With the seed protein we tested analytical reproducibility after storage of the sample for 1 month (see table 2). Reproducibility of the whole procedure including MOAC enrichment is demonstrated by the repeated identification of phosphorylation sites in the 56 kDa C. reinhardtii phosphoprotein (table 2).
Table 2 Reproducibility test of phosphorylation site identification. +: detected; -: not detected.
Protein At5g52300.1 low-temperature-responsive 65 kD protein jgi code 153417 putative protein
Mw [Da] 65971 56515
Experiment 1 2 1 2
Sequence coverage [%] 65.8 67.5 29. 1 28.0
Peptide 1 GAVTSWLGGKPKpSPR + KLESAApTVAER +
Peptide 2 LGYTGENGGGQSEpSPVKDETPR + VAVAPPSRPGpSGK +
Peptide 3 LPLSGGGpSGVKETQQGEEK + SGpSAKVAVAPSR -
Peptide 4 EAHQEPLNpTPVSLLSATEDVTR +
Peptide 5 ESDINKNpSPARFGGESK -
Peptide 6 MKVTDEpSPDQKSR -
Phosphorylation of sterol dehydrogenase
One phosphorylation site was identified as serine 95 in an isoform of the A. thaliana short chain sterol dehydrogenases gene family. We aligned protein sequences belonging to the same family from A. thaliana (At) and Sesamum indicum (Si) (Figure 4). Shown is the region surrounding the phosphorylation site. The region of interest is part of a proposed NADP+ binding domain [31] and displays high homology. The serine phosphorylation site is conserved in 6 out of 8 sequences. The two remaining sequences (At4 and Si1) show no serine at the site of interest and might belong to a separate group.
Figure 4 Sequence alignment of different short chain sterol dehydrogenases. Si: Sesamum indicum; At: Arabidopsis thaliana. The phosphopeptide is shown in red and the site of phosphorylation in bold.
Discussion
The approach described in this work is suitable to reliably identify and routinely screen for serine/threonine phosphorylation in plant proteins. Several lines of evidence are integrated into the strategy thus making unambiguous identification of protein phosphorylation possible: (i) Enrichment for phosphoproteins based on the affinity of phosphate to MOAC, (ii) a specific phosphostain reveals phosphorylation of the proteins and confirms enrichment, (iii) gel separation of the proteins helps to guarantee high confidence in protein identification, and (iv) a highly selective method based on mass spectrometry specific for phosphorylation is used for site identification.
Phosphorylated proteins are enriched by MOAC, a method that can be easily adapted to suit the sample of interest since it is inexpensive and the components needed are widely available. For demonstration MOAC phosphoprotein enrichment was applied to A. thaliana leaf proteins [18], A. thaliana seed material and Chlamydomonas reinhardtii cell cultures (this study, Figure 2).
The following staining with a phosphospecific fluorescent dye is a quick and easy to use method to detect protein phosphorylation and its changes on gels. However, mass spectrometry-based site identification leads to more detailed information about site specific regulation. This holds especially true for proteins phosphorylated at multiple sites [5,6]. While it is not possible with the fluorescent stain to define if a protein is singly or multiply phosphorylated and on which amino acid the phosphate moiety is located, this can be done by mass spectrometry [32]. Consequently, enriched proteins are digested and the peptides are analysed in experiments in which an additional fragmentation event (MS3) is triggered when a peptide looses phosphoric acid during the first fragmentation step (MS2). The combined information of MS2 and MS3 data is then used to obtain high quality data about the peptide sequence and its phosphorylation site (see also Figure 3A and 3B).
Since this is a proof of concept study we did not aim at identifying all phosphorylation sites present in the enriched fraction but at setting up a robust and convenient workflow for the analysis of in vivo protein phosphorylation in plants. The major drawback of the method in its current state is that preferably phosphorylation sites of high abundant phosphoproteins are detected and that often the protein can be assigned reliably but data on the phosphopeptides are lacking. However, this might be circumvented by separating complex mixtures using established chromatography prior to MOAC and/or by coupling a phosphopeptide enrichment step to the protein enrichment. Another problem could be dephosphorylation occurring during or prior to gel separation or during sample storage. This might also explain differences in the reproducibility test. Nevertheless, most of the tested sites could be reproducibly found in a second experiment thus reconfirming the robustness of the method. However, if the sample amount is not limiting separation and digestion after enrichment might also be performed in solution without SDS-PAGE. This, of course would lead to a missing confirmation step in the strategy.
Albeit the phosphorylation of A. thaliana seed proteins probably plays a crucial role in seed development and dormancy [33,34], to our knowledge this is the first time a broad approach has been used to identify phosphorylation sites in seed proteins. In more than half of the identified seed derived phosphopeptides (9 out of 16) the identified phosphorylation sites are directly neighboured by a proline. Additionally, in three peptides the sites are located adjacent to aspartic acid residues. This could be due to enhanced cleavage at proline and aspartic acid residues during the fragmentation process in the mass spectrometer which has been described before [35]. Interestingly, we did find dominant neutral loss in the respective MS2 spectra, even though it was reported recently that proline and aspartic acid containing phosphopeptides exhibit a less pronounced neutral loss of H3PO4 during fragmentation in a Q-Tof mass spectrometer [36]. This apparent difference might be explained by the different instrument types used in the studies or by the special nature of the investigated phosphopeptides. Both serine-proline (SP) and serine-aspartate (SD,E) containing phosphorylation sites are postulated as putative kinase substrates [15]; SP is a MAP-kinase motif which coincides with the importance of MAP-kinases in cellular processes [37] (see also ). A majority of the identified phosphoproteins appear to be seed specific since their expression is reported to be highly dominant if not exclusively expressed during this developmental stage (see table 1 and [38]).
Multisite phosphorylation seems to be quite common as indicated by the fact that more than one phosphorylation site was found in 3 of the 9 proteins. The sites identified on the two ribosomal proteins are the same c-terminal sites identified in mouse [29] thus adding another evidence for the conservation of phosphorylation sites throughout different species [39]. Phosphorylation of these ribosomal P-proteins at their c-terminal end has also been proposed for Saccharomyces cerevisiaea, Rattus norvegicus, Trypanosoma cruzei, and Zea mays [40]. It has been shown that phosphorylation of these proteins leads to accelerated degradation in yeast [41] and this might also hold true for A. thaliana. Interestingly, another two of the identified phosphopeptides, belonging to proteins otherwise unrelated to the ribosomal P-proteins were also found to be phosphorylated at their c-terminal end indicating either the accessibility of c-terminal phosphopeptides for enrichment, digestion and detection or a general pattern, putatively for protein degradation.
The identification of a phosphorylated protein with homology to a short chain sterol dehydrogenase (Sop2) seems to be especially interesting for seed development. These proteins are thought to play a vital role in plant signal transduction elicited by sterols [31,42] and therefore their phosphorylation/dephosphorylation might have large implications on seed development. Serine 95 in the peptide ELGpSPNVVTVHADVSKPDDCRR derived from Sop2 (which we showed to be phosphorylated in A. thaliana) is highly conserved in five out of six homologues in A. thaliana as well as in one out of two homologues in Sesamum indicum (see Figure 4). It is located in the NADP+ binding domain of the protein [31] and might be important for enzyme specificity and selectivity.
Methods
Chemicals
All chemicals were from Sigma (München, Germany). The aluminium hydroxide was purchased as aluminium hydroxide hydrate (ordering nr. A-1577; Sigma).
Seeds
A. thaliana (ecotype Columbia) seeds were taken from an in-house seed stock.
Chlamydomonas reinhardtii culture
C. reinhardtii (wildtype CC-125) was grown for 7 days in 16/8 hour light/dark regime in liquid culture containing TAP Medium [43]. Cells were harvested in the light period and centrifuged for 10 min at 4000 rpm. The supernatant was discarded and the Pellet was used for the extraction of proteins.
Denatured protein extraction from A. thaliana seeds and C. reinhardtii
A. thaliana seed proteins and C. reinhardtii proteins were extracted by adding a mixture of three volumes buffer-saturated phenol (15 ml) and one volume 50 mM Hepes-KOH, pH 7.2 containing 1% β-mercaptoethanol, 40 % sucrose, and 40 mM NaF (5 ml) to 2 g of seed material ground in liquid nitrogen or to a pellet derived from 100 ml of C. reinhardtii culture, respectively. After mixing for 20 minutes at 4°C, protein was precipitated out of the phenol phase with five volumes ice-cold acetone over night at -20°C. The pellet was washed twice with 100% ice-cold methanol and air dried for 15 min.
Enrichment of phosphoproteins using Metal Oxide Affinity Chromatography (MOAC)
The pellet obtained by denatured protein extraction was dissolved in 1.5 ml incubation buffer containing 30 mM MES, 0.2 M potassium glutamate, 0.2 M sodium aspartate, 0.25% Chaps, and 8 M urea with a pH of 6.1. 1.5 ml of a 0.5 mg/ml protein solution was used for 30 min incubation with 80 mg of AlOH3 at 4°C (AlOH3 was washed before once with the incubation buffer described above). The incubation was followed by five washing steps of 1.6 ml and one step of 0.8 ml. Finally, proteins were eluted from the matrix by incubation with 800 μl of 100 mM potassium pyrophosphate buffer pH 9.0 containing 8 M urea for 30 min at RT. Proteins were precipitated with methanol/chloroform prior to gel loading as described by Wessel & Fluegge [44] and about one half of the eluted protein fraction was loaded onto the gel.
Determination of protein content
Protein concentrations were determined via the dye-binding method of Bradford as described previously [45], using ovalbumin as a standard. Each measurement was made in triplicate and the mean values were used.
SDS-PAGE, Phosphostaining, and Coomassie staining
Pellets derived from the methanol/chloroform precipitation were dissolved in 2 × SDS sample buffer (90 mM Tris, pH 6.9, 20% Glycerin, 2% SDS, 0.02% bromophenol-blue, and 100 mM DTT) and approximately 30 μg were subjected to SDS-PAGE (800 μl MOAC eluate fraction was precipitated and dissolved in 40 μl of 1 × SDS sample buffer). After the separation proteins were stained with Pro-Q diamond stain (Invitrogen, Karlsruhe, Germany) essentially following the instructions in the manual. For specificity, we found exchanging fixation solution once (after 30 min) and leaving the gels in the fixation solution over night to be crucial steps. Phosphorylation was visualised using a chemdoc station and a 550 nm filter. After visualisation the phosphostain gels were washed three times with ddH2O and stained with coomassie.
In-gel tryptic digest
Seed protein spots exhibiting a strong signal after phosphostaining were excised and digested over night with trypsin as described before [46]. Peptides were extracted from the gel in three consecutive steps using increasing percentages of acetonitrile (5 %, 50 %, 90% each containing 1% formic acid).
Protein and phosphorylation site identification using nano LC-linear-iontrap-MS
Peptides were loaded directly onto a ProteoSpher® Micro column (0.5 mm × 15 mm) at a flow rate of 3 μl/min and separated in a 85 min gradient from 80% A (0.1% formic acid, and 2.5 % acetonitrile in water) to 100 % B (0.1% formic acid in methanol). Separation and measurements were performed with a nano-LC-pump (Agilent 1100) coupled to an LTQ ion trap (Thermoelectron) with a nano-ESI-source. The voltage was applied directly to the analyte solution using a T-piece. To identify tryptic peptides, phosphopeptides and phosphorylation sites, automatic data-dependent acquisition was performed consisting of a full scan (m/z 400–2000), a subsequent MS2, and a neutral loss scan of 98, 49, or 32.7 (H3PO4 for the +1, +2, and +3 charged ions, respectively) in the five most abundant MS2 fragments. An MS3 scan was automatically collected on the corresponding neutral loss fragments of the MS2 scan events. Peptides were identified by searching the spectra against an A. thaliana database or against the C. Chlamydomonas database version 2.0 containing trypsin and keratin sequences using the Sequest algorithm (ThermoElectron, Dreieich, Germany) and filtering the results with an Xcorr of 2.0, 2.5, and 3.5 for singly, doubly, and triply charged ions, respectively. Protein hits were accepted when at least three peptides with the corresponding Xcorr criteria described above were identified. The spectra derived from phosphopeptides were verified manually (charge state and identification of MS2 and MS3 spectra were checked for their concordance).
Sequence alignment
Sequences were derived from NCBI and alignment was performed using ClustalW with the default settings.
Authors' contributions
FW carried out the optimization of MOAC with seed material, phosphorylation site identification using MS2 and MS3 and participated in writing and drafting the manuscript. WW drafted the conception of this study, advised throughout the project and participated in writing and drafting the manuscript.
Acknowledgements
We thank Megan McKenzie for revising the manuscript. We thank the Max Planck Society for financial support and Mirko Glinski, Stefanie Wienkoop, and Mark Stitt for valuable discussion. Finally, we thank Daniel Karcher, Dietrich Köster, and Christian Bölling for providing C. reinhardtii, Regina Stark and Karsten Oelkers for technical assistance.
==== Refs
Agrawal GK Yonekura M Iwahashi Y Iwahashi H Rakwal R System, trends and perspectives of proteomics in dicot plants Part II: Proteomes of the complex developmental stages J Chromatogr B Analyt Technol Biomed Life Sci 2005 815 125 136 15652803
Rubin CS Rosen OM Protein phosphorylation Annu Rev Biochem 1975 44 831 887 166607 10.1146/annurev.bi.44.070175.004151
Ma H Protein phosphorylation in plants: enzymes, substrates and regulators Trends Genet 1993 9 228 230 8397456 10.1016/0168-9525(93)90075-S
Glinski M Weckwerth W The Role of Mass Spectrometry in Plant Biology Mass Spectrometry Reviews 2005 in press
Glinski M Romeis T Witte CP Wienkoop S Weckwerth W Stable isotope labeling of phosphopeptides for multiparallel kinase target analysis and identification of phosphorylation sites Rapid Commun Mass Spectrom 2003 17 1579 1584 12845583 10.1002/rcm.1093
Glinski M Weckwerth W Differential Multisite Phosphorylation of the Trehalose-6-phosphate Synthase Gene Family in Arabidopsis thaliana: A Mass Spectrometry-based Process for Multiparallel Peptide Library Phosphorylation Analysis Mol Cell Proteomics 2005 4 1614 1625 16030010 10.1074/mcp.M500134-MCP200
Martin K Steinberg TH Goodman T Schulenberg B Kilgore JA Gee KR Beechem JM Patton WF Strategies and solid-phase formats for the analysis of protein and peptide phosphorylation employing a novel fluorescent phosphorylation sensor dye Comb Chem High Throughput Screen 2003 6 331 339 12769676
Schulenberg B Goodman TN Aggeler R Capaldi RA Patton WF Characterization of dynamic and steady-state protein phosphorylation using a fluorescent phosphoprotein gel stain and mass spectrometry Electrophoresis 2004 25 2526 2532 15300772 10.1002/elps.200406007
Schulenberg B Arnold B Patton WF An improved mechanically durable electrophoresis gel matrix that is fully compatible with fluorescence-based protein detection technologies Proteomics 2003 3 1196 1205 12872220 10.1002/pmic.200300440
Steinberg TH Agnew BJ Gee KR Leung WY Goodman T Schulenberg B Hendrickson J Beechem JM Haugland RP Patton WF Global quantitative phosphoprotein analysis using Multiplexed Proteomics technology Proteomics 2003 3 1128 1144 12872214 10.1002/pmic.200300434
Goodman T Schulenberg B Steinberg TH Patton WF Detection of phosphoproteins on electroblot membranes using a small-molecule organic fluorophore Electrophoresis 2004 25 2533 2538 15300773 10.1002/elps.200406008
Salih E Phosphoproteomics by mass spectrometry and classical protein chemistry approaches Mass Spectrom Rev 2004 DOI 10.1002/mas.20042
Zeller M Konig S The impact of chromatography and mass spectrometry on the analysis of protein phosphorylation sites Anal Bioanal Chem 2004 378 898 909 14673564 10.1007/s00216-003-2391-2
Laugesen S Bergoin A Rossignol M Deciphering the plant phosphoproteome: tools and strategies for a challenging task Plant Physiol Biochem 2004 42 929 936 15707831 10.1016/j.plaphy.2004.09.007
Nuhse TS Stensballe A Jensen ON Peck SC Phosphoproteomics of the Arabidopsis plasma membrane and a new phosphorylation site database Plant Cell 2004 16 2394 2405 15308754 10.1105/tpc.104.023150
Aro EM Rokka A Vener AV Determination of phosphoproteins in higher plant thylakoids Methods Mol Biol 2004 274 271 285 15187286
Heintz D Wurtz V High AA Van Dorsselaer A Reski R Sarnighausen E An efficient protocol for the identification of protein phosphorylation in a seedless plant, sensitive enough to detect members of signalling cascades Electrophoresis 2004 25 1149 1159 15095459 10.1002/elps.200305795
Wolschin F Wienkoop S Weckwerth W Enrichment of phosphorylated proteins and peptides from complex mixtures using metal oxide/hydroxide affinity chromatography (MOAC) Proteomics 2005 [Epub ahead of print]
Weckwerth W Willmitzer L Fiehn O Comparative quantification and identification of phosphoproteins using stable isotope labeling and liquid chromatography/mass spectrometry Rapid Communications in Mass Spectrometry 2000 14 1677 1681 10962490 10.1002/1097-0231(20000930)14:18<1677::AID-RCM84>3.0.CO;2-N
Jaffe H Veeranna Pant HC Characterization of serine and threonine phosphorylation sites in beta-elimination/ethanethiol addition-modified proteins by electrospray tandem mass spectrometry and database searching Biochemistry 1998 37 16211 16224 9819213 10.1021/bi981264p
McLachlin DT Chait BT Improved beta-elimination-based affinity purification strategy for enrichment of phosphopeptides Anal Chem 2003 75 6826 6836 14670042 10.1021/ac034989u
Goshe MB Conrads TP Panisko EA Angell NH Veenstra TD Smith RD Phosphoprotein isotope-coded affinity tag approach for isolating and quantitating phosphopeptides in proteome-wide analyses Anal Chem 2001 73 2578 2586 11403303 10.1021/ac010081x
Klemm C Schroder S Gluckmann M Beyermann M Krause E Derivatization of phosphorylated peptides with S- and N-nucleophiles for enhanced ionization efficiency in matrix-assisted laser desorption/ionization mass spectrometry Rapid Communications in Mass Spectrometry 2004 18 2697 2705 15487013 10.1002/rcm.1676
Syka JE Coon JJ Schroeder MJ Shabanowitz J Hunt DF Peptide and protein sequence analysis by electron transfer dissociation mass spectrometry Proc Natl Acad Sci U S A 2004 101 9528 9533 15210983 10.1073/pnas.0402700101
DeGnore JP Qin J Fragmentation of phosphopeptides in an ion trap mass spectrometer J Am Soc Mass Spectrom 1998 9 1175 1188 9794085 10.1016/S1044-0305(98)00088-9
Olsen JV Mann M Improved peptide identification in proteomics by two consecutive stages of mass spectrometric fragmentation Proc Natl Acad Sci U S A 2004 101 13417 13422 15347803 10.1073/pnas.0405549101
Giorgianni F Beranova-Giorgianni S Desiderio DM Identification and characterization of phosphorylated proteins in the human pituitary Proteomics 2004 4 587 598 14997482 10.1002/pmic.200300584
Jin WH Dai J Zhou H Xia QC Zou HF Zeng R Phosphoproteome analysis of mouse liver using immobilized metal affinity purification and linear ion trap mass spectrometry Rapid Commun Mass Spectrom 2004 18 2169 2176 15378723 10.1002/rcm.1604
Shu H Chen S Bi Q Mumby M Brekken DL Identification of phosphoproteins and their phosphorylation sites in the WEHI-231 B lymphoma cell line Mol Cell Proteomics 2004 3 279 286 14729942 10.1074/mcp.D300003-MCP200
Medzihradszky KF Darula Z Perlson E Fainzilber M Chalkley RJ Ball H Greenbaum D Bogyo M Tyson DR Bradshaw RA Burlingame AL O-sulfonation of serine and threonine: mass spectrometric detection and characterization of a new posttranslational modification in diverse proteins throughout the eukaryotes Mol Cell Proteomics 2004 3 429 440 14752058 10.1074/mcp.M300140-MCP200
Lin LJ Tzen JT Two distinct steroleosins are present in seed oil bodies Plant Physiol Biochem 2004 42 601 608 15331088 10.1016/j.plaphy.2004.06.006
Wolschin F Lehmann U Glinski M Weckwerth. W An integrated strategy for identification and relative quantification of site-specific protein phosphorylation using liquid chromatography coupled to MS2/MS3 Rapid Commun Mass Spectrom 2005 in press
Borisjuk L Rolletschek H Radchuk R Weschke W Wobus U Weber H Seed development and differentiation: A role for metabolic regulation Plant Biology 2004 6 375 386 15248120 10.1055/s-2004-817908
Walker-Simmons MK Protein kinases in seeds Seed Science Research 1998 8 193 200
Wysocki VH Tsaprailis G Smith LL Breci LA Mobile and localized protons: a framework for understanding peptide dissociation J Mass Spectrom 2000 35 1399 1406 11180630 10.1002/1096-9888(200012)35:12<1399::AID-JMS86>3.0.CO;2-R
Salek M Di Bartolo V Cittaro D Borsotti D Wei J Acuto O Rappsilber J Lehmann WD Sequence tag scanning: a new explorative strategy for recognition of unexpected protein alterations by nanoelectrospray ionization-tandem mass spectrometry Proteomics 2005 5 667 674 15714472 10.1002/pmic.200401152
Samaj J Baluska F Hirt H From signal to cell polarity: mitogen-activated protein kinases as sensors and effectors of cytoskeleton dynamicity J Exp Bot 2004 55 189 198 14673033 10.1093/jxb/erh012
Schmid M Davison TS Henz SR Pape UJ Demar M Vingron M Scholkopf B Weigel D Lohmann JU A gene expression map of Arabidopsis thaliana development Nat Genet 2005 37 501 506 15806101 10.1038/ng1543
Weckwerth W Selbig J Scoring and identifying organism-specific functional patterns and putative phosphorylation sites in protein sequences using mutual information Biochem Bioph Res Co Biochem Bioph Res Co 2003 307 516 521 10.1016/S0006-291X(03)01182-3
Rodriguez-Gabriel MA Bou G Briones E Zambrano R Remacha M Ballesta JP Structure and function of the stalk, a putative regulatory element of the yeast ribosome. Role of stalk protein phosphorylation Folia Microbiol (Praha) 1999 44 153 163 10588050
Nusspaumer G Remacha M Ballesta JP Phosphorylation and N-terminal region of yeast ribosomal protein P1 mediate its degradation, which is prevented by protein P2 Embo J 2000 19 6075 6084 11080154 10.1093/emboj/19.22.6075
Lin LJ Tai SS Peng CC Tzen JT Steroleosin, a sterol-binding dehydrogenase in seed oil bodies Plant Physiol 2002 128 1200 1211 11950969
Harris EH The Chlamydomonas Sourcebook Academic Press, Inc, San Diego, CA, USA 1989
Wessel D Flugge UI A method for the quantitative recovery of protein in dilute solution in the presence of detergents and lipids Anal Biochem 1984 138 141 143 6731838 10.1016/0003-2697(84)90782-6
Bradford MM Rapid and Sensitive Method for Quantitation of Microgram Quantities of Protein Utilizing Principle of Protein-Dye Binding Analytical Biochemistry 1976 72 248 254 942051 10.1016/0003-2697(76)90527-3
Otto A Thiede B Muller EC Scheler C WittmannLiebold B Jungblut P Identification of human myocardial proteins separated by two-dimensional electrophoresis using an effective sample preparation for mass spectrometry Electrophoresis 1996 17 1643 1650 8957197 10.1002/elps.1150171027
|
16270910
|
PMC1295590
|
CC BY
|
2021-01-04 16:37:29
|
no
|
Plant Methods. 2005 Nov 1; 1:9
|
utf-8
|
Plant Methods
| 2,005 |
10.1186/1746-4811-1-9
|
oa_comm
|
==== Front
PLoS MedPLoS MedpmedplosmedPLoS Medicine1549-12771549-1676Public Library of Science San Francisco, USA 1628855310.1371/journal.pmed.0020322Student ForumOtherMedical EducationMedical EducationMedical CareersAcademic MedicineMedicine in Developing CountriesWhy Medical Students Are Crucial to the Future of Research in South Asia Student ForumAslam Fawad *Shakir Murtaza Qayyum Muhammad Ahad Fawad Aslam and Murtaza Shakir are final-year medical students at the Aga Khan University in Karachi, Pakistan. Fawad Aslam is also the vice president of Aga Khan University's Student Research Forum. Muhammad Ahad Qayyum is a final-year medical student at the Punjab Medical College in Faisalabad, Pakistan.
Competing Interests: FA is the vice president of the Student Research Forum at Aga Khan University. MS and MAQ have declared that they have no competing interests.
*To whom correspondence should be addressed. E-mail: [email protected] 2005 29 11 2005 2 11 e322Copyright: © 2005 Aslam et al.2005This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are properly credited.One long-term strategy for promoting health research in developing countries is to target medical students early in their careers.
==== Body
Health research is essential to improving health care [1]. Unfortunately, health research has a low priority in the developing world. In all disciplines of science and technology, India and Pakistan combined have 208 researchers per million citizens [2,3], as compared to 4,526 researchers per million citizens in the United States [4]. The published research output from South Asia is small—South Asian health researchers accounted for only 1.2% of all papers within the Institute for Scientific Information database from 1992–2001 [5]. Developing countries must therefore enhance their research capacity to efficiently address the growing burden of both communicable and non-communicable diseases [6].
Engaging Medical Students in Health Research
One long-term strategy for promoting health research is to target medical students early in their careers. Most of the research to date on the effectiveness of such a strategy has been done in Western settings. This research has shown that research experience as a medical student is strongly associated with postgraduate research involvement [7,8].
Student research can also contribute to the published output of an institution. In Germany, for example, medical students authored 28% of the publications of one institution, including first authorship in 7.8% of papers [9]. Nothing can be more motivating for a student than to get published.
Even if the experience of doing research as a student does not lead to a later career in academic medicine, research experience can help improve students' skills in searching and critically appraising the medical literature, independent learning, and writing research papers [10,11]. Such exposure to research as a student can also help to identify future careers, establish important contacts, and secure better residency positions. Given the many benefits of doing a research project as a student, not surprisingly, 97% of students considered research as a useful alternative to electives [11].
Student research is not without its problems. Good mentorship, for example, is a vital component of effective student research, and inadequate mentoring can lead to discontentment with research. Other problems include lack of time, neglect of routine studies and deterioration of clinical skills due to more time being spent on research activities, and inadequate project management [12]. The perceived competitiveness and greater demands of a career in academic medicine and lower salary compared to private practice may deter students from research. Another concern is that students may work simply as junior laborers with no role in designing the research or in critical thinking during the research process. For example, they may be told to review, say, 200 charts, and hand over the data to the principal investigator. Those who criticize medical student research would also argue that student papers are rarely cited, and thus are of limited utility.
Student Research in South Asia
Student research is dependent on national research activity. Since there is still only a limited research infrastructure in many developing countries, this means that opportunities for medical student research are limited. Research is not considered a part of the medical curriculum in many of these countries.
In one Indian study, for example, 91% of interns reported no research experience in medical school [13]. Thus, students in India rarely get exposed to research at this crucial stage in their academic development when such exposure could encourage further research after qualification. Faculty across South Asia themselves seldom engage in research owing to inadequate training in research, lack of incentives, work overload, poor pay, and minimal research demand in clinical practice from patients, fellow physicians, and policymakers. Consequently, students are deprived of mentors and role models. Medical training in general in South Asia does not emphasize the importance of research to medical practice. More than two-thirds of the postgraduate trainees at one Pakistani institution, for example, reported reading scientific journals only once in six months or more [14].
Taking Action
What can be done to increase medical students' involvement in research in South Asia (Figure 1)? First and foremost, the research infrastructure needs extensive improvement, and the meager funding for research must be boosted, so that there will be a healthier research culture in which students can participate. There also need to be effective international agreements to halt the “brain drain” of academic clinicians from low-income to high-income countries [15], since this migration robs medical students of role models.
Figure 1 Factors That Can Help Encourage South Asian Medical Students to Choose a Career in Academic Medicine
In addition, students need to be “sensitized” to research—that is, they should be made aware of why research is so crucial to health care. The medical curriculum must begin to incorporate and emphasize evidence-based medicine. To stimulate students' interest in research, we believe that they should undertake a mandatory course on research skills, along with a compulsory community project. Such projects undertaken in the primary-care setting will allow students to execute all the steps of a research project, from conception to final report writing, thereby narrowing the gap between theory and practice. Elective slots should be available to those students who are interested in pursuing further research. Young researchers need to be encouraged, for example, by being awarded scholarships. Prior research experience should be seen as valuable when recruiting postgraduate trainees. Only when students are sensitized in this way—and provided that they also have sufficient flexibility and security to pursue a career in academic medicine—will they choose career paths involving research.
Conclusion
At our own medical schools, we believe that medical students are becoming more enthusiastic about getting involved in research, which is encouraging. Some efforts to promote student research are already underway in South Asia. For example, student conferences and research workshops are being held in major cities of Pakistan, and some medical journals, including the Journal of Postgraduate Medicine (India) and the Journal of the College of Physicians and Surgeons of Pakistan, have introduced student sections. The students at Aga Khan University in Pakistan, which has a well-established research infrastructure, have won awards for their projects at international student conferences and have published widely in indexed journals. Given the right amount of support, medical students' interest in research can be successfully nurtured.
We are grateful to Fahim Jafary for critically reviewing the manuscript and being a worthy mentor. We are also thankful to our colleagues Abdul Waheed and Abdul Moiz Khan for providing valuable input.
Citation: Aslam F, Shakir M, Qayyum MA (2005) Why medical students are crucial to the future of research in South Asia. PLoS Med 2(11): e322.
==== Refs
References
Global Forum for Health Research 10/90 report on health research 2003–2004 2004 Geneva Global Forum for Health Research Available: http://www.globalforumhealth.org/site/002__What%20we%20do/005__Publications/001__10%2090%20reports.php . Accessed 26 September 2005
United Nations Educational, Scientific and Cultural Organization Institute for Statistics Country profile: India 2005 Quebec United Nations Educational, Scientific and Cultural Organization Institute for Statistics Available: http://www.uis.unesco.org/countryprofiles/html/EN/countryProfile_en.aspx?code=3560.htm . Accessed 26 September 2005
United Nations Educational, Scientific and Cultural Organization Institute for Statistics Country profile: Pakistan 2005 Quebec United Nations Educational, Scientific and Cultural Organization Institute for Statistics Available: http://www.uis.unesco.org/countryprofiles/html/EN/countryProfile_en.aspx?code=5860.htm . Accessed 26 September 2005
United Nations Educational, Scientific and Cultural Organization Institute for Statistics Country profile: United States 2005 Quebec United Nations Educational, Scientific and Cultural Organization Institute for Statistics Available: http://www.uis.unesco.org/countryprofiles/html/EN/countryProfile_en.aspx?code=8400.htm . Accessed 26 September 2005
Sadana R D'Souza C Hyder AA Chowdhury AM Importance of health research in South Asia BMJ 2004 328 826 830 15070643
Buddha Basnyat Rajapaksa LC Cardiovascular and infectious diseases in South Asia: The double whammy BMJ 2004 328 781 15070612
Segal S Lloyd T Houts PS Stillman PL Jungas RL The association between students' research involvement in medical school and their postgraduate medical activities Acad Med 1990 65 530 533 2383337
Reinders JJ Kropmans TJ Cohen-Schotanus J Extracurricular research experience of medical students and their scientific output after graduation Med Educ 2005 39 237 15679693
Cursiefen C Altunbas A Contribution of medical student research to the Medline-indexed publications of a German medical faculty Med Educ 1998 32 439 440 9743810
Houlden RL Raja JB Collier CP Clark AF Waugh JM Medical students' perceptions of an undergraduate research elective Med Teach 2004 26 659 661 15763861
Frishman WH Student research projects and theses: Should they be a requirement for medical school graduation? Heart Dis 2001 3 140 144 11975783
Diez C Arkenau C Meyer-Wentrup F The German medical dissertation—Time to change? Acad Med 2000 75 861 863 10965870
Chaturvedi S Aggarwal OP Training interns in population-based research: Learners' feedback from 13 consecutive batches from a medical school in India Med Educ 2001 35 585 589 11380862
Aslam F Qayyum MA Mahmud H Qasim R Haque IU Attitudes and practices of postgraduate medical trainees towards research—A snapshot from Faisalabad J Pak Med Assoc 2004 54 534 536 15552293
Dovlo D Taking more than a fair share? The migration of health professionals from poor to rich countries PLoS Med 2005 2 e109 10.1371/journal.pmed.0020109 15916458
|
16288553
|
PMC1297528
|
CC BY
|
2021-01-05 10:39:18
|
no
|
PLoS Med. 2005 Nov 29; 2(11):e322
|
utf-8
|
PLoS Med
| 2,005 |
10.1371/journal.pmed.0020322
|
oa_comm
|
==== Front
PLoS MedPLoS MedpmedplosmedPLoS Medicine1549-12771549-1676Public Library of Science San Francisco, USA 1628855410.1371/journal.pmed.0020361Correspondence and Other CommunicationsGenetics/Genomics/Gene TherapyOtherScience PolicyEpidemiology/Public HealthStatisticsGeneral MedicineCommunication in Health CareEditorial policies (including conflicts of interest)Medical journalsTruth, Probability, and Frameworks CorrespondenceWren Jonathan D
1
1University of OklahomaNorman, OklahomaUnited States of AmericaE-mail: [email protected]
Competing Interests: The author has declared that no competing interests exist.
11 2005 29 11 2005 2 11 e361Copyright: © 2005 Jonathan D. Wren.2005This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are properly credited.
Why Most Published Research Findings Are False
==== Body
James T. Kirk: Harry lied to you, Norman. Everything Harry says is a lie. Remember that, Norman: Everything he says is a lie.
Harry Mudd: Now I want you to listen to me very carefully, Norman: I… am… lying.
—Star Trek, the episode “I, Mudd”
Although John P. A. Ioannidis [1] brings up several good points about over-reliance on formal—yet arbitrary—statistical cutoffs and bias against the reporting of negative results, his claim that most published research findings are false is somewhat paradoxical. Ironically, the truer his premise is, the less likely his conclusions are. He, after all, relies heavily on other studies to support his premise, so if most (i.e., greater than 50%) of his cited studies are themselves false (including the eight of 37 that pertain to his own work), then his argument is automatically on shaky ground. As mentioned in the PLoS Medicine Editorial [2], scientific studies don't offer truth, per se. Even when studies appear in the best journals, they offer probabilistic assertions. Ioannidis's statement that “the probability that a research finding is indeed true depends on the prior probability of it being true” [1] is really begging the question; this, after all, is the problem. We cannot know such probabilities a priori, and guessing at such probabilities and/or parameters (as he does in his single nucleotide polymorphism [SNP] association example) surely could not be less biased than any statistical test of significance. The key problem in Ioannidis's positive predictive value (PPV) formula to calculate the post-study probability that a relationship is true (PPV = [1 − β]R/[R − βR + α], where R is the ratio of true relationships to no relationships) is that one can postulate a near-infinite number of non-relationships. Just extending his SNP example, why assume each SNP acts independently? This is not unreasonable, given that schizophrenia is clearly not inherited in a Mendelian pattern. So rather than 99,990 SNPs not being associated with schizophrenia, we have potentially 99,990n not associated, where n is the number of potentially interacting SNPs. As n grows, R becomes very small very quickly, and PPV becomes effectively zero. Taken to the extreme, this would imply that all empirical studies are fruitless. One of the most important factors in moving toward the truth, which was not discussed, is fitting discoveries into a framework. Optimally, if a relationship is true, it should have more than one implication, permitting validation from multiple angles. For example, an SNP causally associated with schizophrenia must affect something on the molecular level, whether genomic, transcriptional, post-transcriptional, translational, or post-translational. In turn, these molecules should interact differently with each other, with other molecules within the cell, within a tissue, and/or with the system as a whole. If Norman, the android from Star Trek mentioned in the beginning quote, had been equipped with the capacity to evaluate statements within a framework, he never would have short-circuited as a result of Kirk's paradox. He could have entertained the possibility that either Kirk was lying about Harry or Harry's statement was incomplete (i.e., lying about what?) Similarly, repeatedly re-examining any particular finding to resolve the true/not true paradox via statistical arguments alone can short-circuit our patience. We should instead seek to identify the framework by which implications of the finding can be tested, and I would argue that the more important the finding, the more testable implications it has.
Citation: Wren J (2005) Truth, probability, and frameworks. PLoS Med 2(11): e361.
==== Refs
References
Ioannidis JP Why most published research findings are false PLoS Med 2005 2 e124 10.1371/journal.pmed.0020124 16060722
PLoS Medicine Editors Minimizing mistakes and embracing uncertainty PLoS Med 2005 2 e272 10.1371/journal.pmed.0020272 16120013
|
16288554
|
PMC1297534
|
CC BY
|
2021-01-05 11:13:37
|
no
|
PLoS Med. 2005 Nov 29; 2(11):e361
|
utf-8
|
PLoS Med
| 2,005 |
10.1371/journal.pmed.0020361
|
oa_comm
|
==== Front
PLoS MedPLoS MedpmedplosmedPLoS Medicine1549-12771549-1676Public Library of Science San Francisco, USA 1628855510.1371/journal.pmed.0020366PerspectivesOtherEpidemiology/Public HealthObstetrics/GynecologyPrimary CareSexual HealthUrologyWomen's HealthSexual HealthWomen's HealthMen's HealthMedicine in Developing CountriesGeneral Practice/Family Practice/Primary CareWhere Do People in Nigeria Get Their Contraception? PerspectiveLadipo Oladapo A Oladapo A. Ladipo is the president and chief executive officer of the Association for Reproductive and Family Health (http://www.arfh.org), Ibadan, Nigeria. E-mail: [email protected]
Competing Interests: The author declares that no competing interests exist.
11 2005 29 11 2005 2 11 e366Copyright: © 2005 Oladapo A. Ladipo.2005This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are properly credited.
Sources of Contraceptive Commodities for Users in Nigeria
Knowing where users obtain different contraceptive methods is useful for planning service delivery, argues Ladipo, who examines a new study on contraception in Nigeria.
==== Body
In this issue of PLoS Medicine, Oye-Adeniran and colleagues report a new community study in Nigeria in which contraceptive users were surveyed about their sources of family planning information and contraception [1]. Enquiry about the source of contraceptive is a standard part of family planning surveys that assess knowledge, attitudes, and practices, and these surveys are commonly used in Nigeria and elsewhere [2,3]. The underlying proposition is that knowing where users obtain different contraceptive methods is useful for planning service delivery. The associated question—why do people choose a particular source for obtaining contraceptives (for example, a pharmacy rather than a clinic)—generally receives only passing attention.
Over the years, there has been little variation in the pattern of contraceptive sourcing. Nonclinic facilities are the main source for obtaining condoms and the oral contraceptive pill, while the intrauterine contraceptive device and injectable contraceptives—both of which require health provider intervention—are predominately obtained from the clinic setting. Over time, neither the clinic nor the nonclinic sources have developed to their full potential as family planning providers, despite years of family planning programming in both public and private sectors.
The New Study
Oye-Adeniran and colleagues surveyed 2,001 persons aged 14–49 from four states of Nigeria—Anambra from the southeast, Oyo from the southwest, Kaduna from the northwest, and Bauchi from the northeast. These states were randomly selected for the study, one each from Nigeria's four health zones. A multistage cluster sampling design was used to select the respondents. Of the 2,001 people surveyed, 1,647 (82.3%) were sexually active, out of whom 244 were found to be using contraceptive methods at the time of the study, giving a contraceptive prevalence rate of 14.8%.
Knowing where users obtain different contraceptive methods is useful for planning service delivery.
The study had three major findings: (1) friends are the predominant source of information on contraception; (2) young people tend to prefer chemists (pharmacists), while older people prefer government and private hospitals as sources of contraception; and (3) Catholics prefer to avoid public health institutions. These findings are similar to those obtained by Nigeria's 2003 Demographic and Health Survey [4].
What Issues Does the Study Raise?
There are several issues arising from this study. First, sources of contraception in the public sector are not set up to work with nonclinic sources, although a referral system may well suit the needs of clients. Second, the roles of the limited number of modern contraceptive methods available and of service providers in a sexually active person's choice of contraceptive method are not yet fully resolved [5,6]. Third, the cultural disposition against providing contraceptives to young people in Nigeria, in spite of their sexual activity, makes the young wary of public health institutions. At such institutions, they are likely to confront adult health providers who will often frown on their sexual activity. The added advantage of the chemist for both young people and Catholics may well be connected to their need for anonymity, which the chemist provides, but which public health institutions, with their formal procedures for documentation, counseling, and service provision, deny. Above all, the findings reinforce the point that clinics at all levels of the health delivery system are characterized by unfriendliness to youths.
A major limitation of Oye-Adeniran and colleagues' study is the relatively small number of respondents who were using contraceptive methods. This limits extrapolation of the study to the general population, which may have different demographic and socioeconomic characteristics than the study sample.
How Can We Reach Young People?
It is obvious that the provision of youth-friendly clinics still eludes both policymakers and service providers. The cultural predisposition against family planning services being made available to youths is a major barrier to reaching young people. Until the sexual activity and associated risks among young people are acknowledged by the adult population, and the provision of contraceptives is seen as a solution rather than the problem, the implementation of youth-friendly clinics and their utilization will be compromised.
Although Oye-Adeniran and colleagues mentioned the preference of the adult married population for public sector services, there is as yet no clear identification of the level of satisfaction that this group derives from the services provided. The very presence of older clients at public clinics may well be a deterrent to youth patronage. The service characteristics of the chemists that make chemist shops such attractive locations for youths could reward in-depth study [7,8]. A likely direction of such investigation will be the extent to which the anonymity and lack of curiosity or the absence of intrusive counseling in chemist shops are preferred by young people to the formal procedures in public health institutions.
Citation: Ladipo OA (2005) Where do people in Nigeria get their contraception? PLoS Med 2(11): e366.
==== Refs
References
Oye-Adeniran A Adewole IF Umoh AV Oladokun A Gbadegesin A Sources of contraceptive commodity for users in Nigeria PLoS Med 2005 2 e306 10.1371/journal.pmed.0020306 16218768
Chaya N Helsing K Conly SR Contraceptive choice: Worldwide access to family planning. 1997 report on progress towards world population stabilization [wallchart] 1997 Washington (D.C.) Population Action International
Khalifa M Mahran M El-Zanaty FH Way AA Choice of family planning service provider in Egypt Perspectives on fertility and family planning in Egypt. Results of further analysis of the 1992 Egypt Demographic and Health Survey 1995 Calverton (Maryland) Macro International 143 167
National Population Commission Federal Republic of Nigeria Nigeria Demographic and Health Survey 2003 2003 Calverton (Maryland) Macro International Available: http://www.measuredhs.com/pubs/pdf/FR148/00FrontMatter.pdf . Accessed 29 September 2005
Adinma JI Agbai AO Nwosu BO Contraceptive choices among Nigerian women attending an antenatal clinic Adv Contracept 1998 14 131 145 9820931
Konje JC Oladini F Otolorin EO Ladipo OA Factors determining the choice of contraceptive methods at the Family Planning Clinic, University College Hospital, Ibadan, Nigeria Br J Fam Plann 1998 24 107 110 9855717
Jones EF Bulatao RA Palmore JA Ward SE Accessibility as a determinant of contraceptive method choice Choosing a contraceptive: Method choice in Asia and the United States 1989 Boulder (Colorado) Westview Press 57 77
Harden A Ogden J Sixteen to nineteen year olds' use of, and beliefs about, contraceptive services Br J Fam Plann 1999 24 141 144 10023099
|
16288555
|
PMC1297536
|
CC BY
|
2021-01-05 10:39:17
|
no
|
PLoS Med. 2005 Nov 29; 2(11):e366
|
utf-8
|
PLoS Med
| 2,005 |
10.1371/journal.pmed.0020366
|
oa_comm
|
==== Front
PLoS MedPLoS MedpmedplosmedPLoS Medicine1549-12771549-1676Public Library of Science San Francisco, USA 1628855710.1371/journal.pmed.0020379Correspondence and Other CommunicationsInfectious DiseasesOtherEpidemiology/Public HealthHealth PolicyObstetrics/GynecologyPediatricsWomen's HealthInfectious DiseasesInternational healthMedicine in Developing CountriesPublic HealthResponse to Amir Attaran CorrespondenceMcArthur John W Sachs Jeffrey D Schmidt-Traub Guido
1
1United Nations Millennium ProjectNew York, New YorkUnited States of AmericaE-mail: [email protected]
Competing Interests: The authors have declared that no competing interests exist.
11 2005 29 11 2005 2 11 e379Copyright: © 2005 McArthur et al.2005This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are properly credited.
An Immeasurable Crisis? A Criticism of the Millennium Development Goals and Why They Cannot Be Measured
Author's Reply
==== Body
Amir Attaran's Policy Forum [1] raises important points on the poor quality of data for some indicators used to measure progress on the Millennium Development Goals (MDGs), but, sadly, uses these findings to draw the wrong conclusions. The evidence he presents on a small number of indicators is partial, and does not justify his conclusion that the MDGs might become a liability and are doomed to fail. Quite the opposite is the case.
Of course the data on the world's extremely poor people are weak, as is just about every other aspect of efforts vis-a-vis the poor. The rich countries dramatically underinvest and make far too little effort in helping to save the poor from dying of malaria and tuberculosis (TB). It is, therefore, no surprise that developing countries and the international system lack the resources and operational support to measure malaria and TB well. Attaran's criticisms in this regard are justified and have been made by many others before him, including many professionals working for the United Nations (UN) system.
The MDGs are a political commitment made by the 189 countries represented at the 2000 Millennium Summit to halve extreme poverty in its many forms by 2015. The author ignores that such broad outcome goals adopted by world leaders are distinct from the technical question, how to define and measure corresponding indicators, which the UN has been asked to help answer. It is, therefore, inaccurate to blame the UN system for setting goals that are difficult to measure. Goal setting is the prerogative of world leaders, and they have correctly reaffirmed their commitment to the MDGs many times since 2000.
In response, the UN system has set up an active process to review indicators and data on progress toward achieving the MDGs, involving many UN organizations as well as the World Bank and the International Monetary Fund. In recent years, this interagency process has already revised several MDG indicators and issued guidance notes on how data collection can be improved. The author's assertion that the UN “shows a profound disrespect for the scientific process” [1] is wrong and misleading. The UN leadership rightly decided that the heads of state and government convening at the 2005 World Summit should focus their attention on the high-level political decisions needed to strengthen the international framework for security, development, and human rights. On the development side, the greatest priority was to cement the MDGs as operational rather than simply rhetorical targets. The world leaders did not delve into the technical issues of measurement and indicators, but this important work will continue to be addressed by UN statisticians and independent experts. Such experts have indeed been scrutinizing the definition and measurement of these indicators for some time—as did, for example, several of the UN Millennium Project task forces.
Another shortcoming of Attaran's article is that it generalizes incorrectly across the MDGs. It describes some of the toughest measurement challenges (e.g., maternal mortality and malaria), and uses them to paint all the MDGs with the same brush. In addition to the example of child mortality rates cited in the article, several other MDG indicators can be measured quite well. These indicators include anthropometric measures of malnutrition, primary school enrollment, gender parity in education, and access to basic infrastructure services.
An implication of Attaran's argument is that there should be no goals when measurement is imperfect, as it is in many countries with indicators for maternal mortality. Should world leaders, therefore, not set time-bound goals for reducing maternal mortality? This would be wrongheaded for three clear reasons. First, even with incomplete or missing data, dramatic and verifiable improvements in women's health can be achieved by investing in emergency obstetric care and other known, monitorable, and practical interventions to build and sustain primary health systems. The MDGs provide a major political and operational framework for doing this. Second, the MDGs are already promoting strengthened health systems in low-income settings, and those improved systems are key to ensuring the vital registrations that the author rightly recommends for improving the measurement of maternal mortality. Third, the very adoption of the maternal mortality goal (and others) is provoking greatly increased attention to improvements in data collection from the World Health Organization, the Gates Foundation, the World Bank, academia, and others. The MDGs should not be misunderstood as a static, one-off process.
Attaran misleads the reader when he argues that the MDGs have become “all-encompassing” catch-alls for tenuously related interventions. To say that roads and electricity are falsely linked to achieving the MDGs is incorrect, and suggests a lack of understanding of the integrated nature of development processes. Roads and electricity play a critical role in poverty reduction, access to essential health services, reduction of maternal and child mortality, and a host of other channels directly related to success in achieving the MDGs. Therefore, any strategy to achieve the MDGs needs to include these interventions. The UN Millennium Project described these linkages in the most detailed series of studies on practical approaches to achieving the MDGs that has ever been produced.
We hope that Attaran's key message on improved data collection and interpretation is heard. More and better data are certainly needed on the conditions of the world's poorest and most vulnerable people. However, even in countries with poor data systems, enough is known today to start making the practical and measurable investments in education, health, basic infrastructure, and improved environmental management that are vitally needed to cut, and eventually to end, extreme poverty. Crucially, the MDGs provide the unique framework for prompting the international cooperation that is indispensable to helping poor countries escape the poverty trap, and to benchmarking progress en route. No discussion about indicators and measurement—no matter how justified it is—should deflect from the overarching global commitment to the poorest of the poor that world leaders struck at the Millennium Summit in 2000 and reconfirmed at the World Summit in 2005.
Citation: McArthur JW, Sachs JD, Schmidt-Traub G (2005) Response to Amir Attaran. PLoS Med 2(11): e379.
==== Refs
Reference
Attaran A An immeasurable crisis? A criticism of the Millennium Development Goals and why they cannot be measured PLoS Med 2005 2 e318 10.1371/journal.pmed.0020318 16156696
|
16288557
|
PMC1297542
|
CC BY
|
2021-01-05 10:39:19
|
no
|
PLoS Med. 2005 Nov 29; 2(11):e379
|
utf-8
|
PLoS Med
| 2,005 |
10.1371/journal.pmed.0020379
|
oa_comm
|
==== Front
PLoS MedPLoS MedpmedplosmedPLoS Medicine1549-12771549-1676Public Library of Science San Francisco, USA 1628855810.1371/journal.pmed.0020383Correspondence and Other CommunicationsInfectious DiseasesOtherEpidemiology/Public HealthHealth PolicyEpidemiologyHealth PolicyInfectious DiseasesMicrobiologyPublic HealthDid Glycopeptide Use in Animals Result in Hospital Infections of VRE? CorrespondenceMudd Anthony
1
1SouthamptonUnited KingdomE-mail: [email protected]
Competing Interests: AM was a consultant advising the International Federation for Animal Health representing the Global Animal Health companies.
11 2005 29 11 2005 2 11 e383Copyright: © 2005 Anthony Mudd.2005This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are properly credited.
Agricultural Antibiotics and Human Health
==== Body
As one of the persons involved with the development of avoparcin for farm animals, I have followed the discussion on vancomycin-resistant enterococci (VRE) and its potential transfer from animals to humans over the past decade. What a pity that the authors of this PLoS Medicine Policy Forum [1] did not reference a recent review by Wassenaar [2] that comprehensively discussed this topic. In this review, evidence is presented to show that VRE infections in humans have actually increased in the European Union since avoparcin was removed from the market. Other data show that whole-genome typing methods separate clinical VRE strains from animal or nonhospitalized human strains.
The conclusion of the Smith et al. article [1] that a correct decision was made to adopt the EU “precautionary principle” and remove avoparcin from the market is surprising, as this is contrary to the opinion of the independent EU Scientific Committee for Animal Nutrition, and since a quantitative risk analysis, as suggested by the authors, could not conclude a relationship between glycopeptide use in animals and incidence of clinical infection in humans.
Citation: Mudd A (2005) Did glycopeptide use in animals result in hospital infections of VRE? PLoS Med 2(11): e383.
==== Refs
References
Smith DL Dushoff J Morris G Agricultural antibiotics and human health PLoS Med 2005 2 e232 10.1371/journal.pmed.0020232 15984910
Wassenaar TM Use of antimicrobial agents in veterinary medicine and implications for human health Crit Rev Microbiol 2005 31 155 169 16170906
|
16288558
|
PMC1297545
|
CC BY
|
2021-01-05 10:39:18
|
no
|
PLoS Med. 2005 Nov 29; 2(11):e383
|
utf-8
|
PLoS Med
| 2,005 |
10.1371/journal.pmed.0020383
|
oa_comm
|
==== Front
PLoS MedPLoS MedpmedplosmedPLoS Medicine1549-12771549-1676Public Library of Science San Francisco, USA 1628855910.1371/journal.pmed.0020386Correspondence and Other CommunicationsGenetics/Genomics/Gene TherapyOtherScience PolicyEpidemiology/Public HealthStatisticsGeneral MedicineCommunication in Health CareEditorial policies (including conflicts of interest)Medical journalsPower, Reliability, and Heterogeneous Results CorrespondenceShrier Ian
1
1McGill UniversityMontreal, QuebecCanadaE-mail: [email protected]
Competing Interests: The author has declared that no competing interests exist.
11 2005 29 11 2005 2 11 e386Copyright: © 2005 Ian Shrier.2005This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are properly credited.
Why Most Published Research Findings Are False
N/A
==== Body
I want to congratulate John P. A. Ioannidis on his thought-provoking Essay [1]. I have two comments.
In Corollary 1, he suggests that small sample sizes mean smaller power, and implies that larger studies with thousands of subjects are more likely to be true. I think it is important to stress that if the effect size is large (e.g., very small variance that is seen in physiological studies), then adequate power is obtained with small numbers. Furthermore, some would argue that exposing subjects to research risks unnecessarily (e.g., when fewer subjects would yield sufficient power) is unethical. Since the analysis is based on power, we should remember that larger is not always better.
In Corollary 4, Ioannidis argues that greater flexibility in designs, definitions, etc. means the results are less likely to be true. I agree that replication of all aspects of the study is more likely to yield consistent results, but this does not necessarily mean true results. Since we don't know a priori which methodological details are most appropriate (e.g., dose, timing, etc.), heterogeneous results from different designs is an important source of information and can lead to a new, more in-depth understanding of the subject—and sometimes even paradigm shifts. I agree with the accompanying Editorial [2] to the article that we need to distinguish between the validity of the data and the validity of the authors' conclusions.
Citation: Shrier I (2005) Power, reliability, and heterogeneous results. PLoS Med 2(11): e386.
==== Refs
References
Ioannidis JPA Why most published research findings are false PLoS Med 2005 2 e124 10.1371/journal.pmed.0020124 16060722
PLoS Medicine Editors Minimizing mistakes and embracing uncertainty PLoS Med 2005 2 e272 10.1371/journal.pmed.0020272 16120013
|
16288559
|
PMC1297547
|
CC BY
|
2021-01-05 10:39:22
|
no
|
PLoS Med. 2005 Nov 29; 2(11):e386
|
utf-8
|
PLoS Med
| 2,005 |
10.1371/journal.pmed.0020386
|
oa_comm
|
==== Front
PLoS MedPLoS MedpmedplosmedPLoS Medicine1549-12771549-1676Public Library of Science San Francisco, USA 1628856010.1371/journal.pmed.0020387Correspondence and Other CommunicationsInfectious DiseasesPharmacology/Drug DiscoveryDrugs and adverse drug reactionsHealth PolicyInfectious DiseasesInternational healthMedicine in Developing CountriesThe Need of a Neonatal Preparation for Chagas Disease CorrespondenceSosa-Estani Sergio Belizan Jose M Althabe Fernando Rubinstein Aldofo
1
1Institute for Clinical Effectiveness and Health PolicyBuenos AiresArgentinaE-mail: [email protected]
Competing Interests: The authors have declared that no competing interests exist.
11 2005 29 11 2005 2 11 e387Copyright: © 2005 Sosa-Estani et al.2005This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are properly credited.
Designing Drugs for Parasitic Diseases of the Developing World
==== Body
We have read about the efforts and initiatives related to the design of drugs for parasitic diseases in McKerrow's article [1] with interest and expectation. One of the pressing needs in this area is for a neonatal preparation for Chagas disease.
Satisfactory achievements have been made in Argentina in relation to the transmission of the disease by vectors and through blood transfusion [2,3]. Vertical transmission is now the great challenge in eradicating Chagas disease. Around 800–1,300 neonates infected with Trypanosmoma cruzi are born every year in our country [4]. Almost 99% of all births occur in hospital, thus allowing the detection of infants born with parasites immediately after birth. The initiation of treatment of these neonates before they and their mothers leave the hospital is a good strategy to obtain high treatment coverage. The later attendance of mothers with their children to health-care facilities is quite unpredictable and irregular. Also, it is difficult to link information about maternal and neonatal parasitic status obtained at birth with later attendance at other health-care facilities.
It would, therefore, be of great value to have a neonatal preparation for the treatment of Chagas disease. There is currently no neonatal or infant preparation available. Instead, one of the two available adult preparations (nifurtimox or benznidazol) is mashed and diluted at local level in order to be administered to newborns and infants. It is easy to understand the difficulties and uncertainties that these procedures involve.
We hope that in the agenda of the several initiatives mentioned in this article [1] the development of a neonatal preparation for Chagas disease could be considered. It would benefit many infants every year.
Citation: Sosa-Estani S, Belizan JM, Althabe F, Rubinstein A (2005) The need of a neonatal preparation for Chagas disease. PLoS Med 2(11): e387.
==== Refs
References
McKerrow JH Designing drugs for parasitic diseases of the developing world PLoS Med 2005 2 e210 10.1371/journal.pmed.0020210 16120007
Segura EL Cura EN Sosa Estani S Andrade J Lansetti JC Long-term effects of a Nation-wide control program on the seropositivity for Trypanosoma cruzi infection in young men from Argentina Am J Trop Med Hyg 2000 62 353 362 11037777
Segura EL Esquivel ML Salomón O Gómez AO Sosa Estani S Participación comunitaria en el Programa Nacional de Control de la Transmisión de la Enfermedad de Chagas Medicina (B Aires) 1994 54 610 611 7658995
Gurtler RE Segura EL Cohen JE Congenital transmission of Trypanososma cruzi infection in Argentina Emerg Infect Dis 2003 9 29 32 12533278
|
16288560
|
PMC1297548
|
CC BY
|
2021-01-05 11:13:37
|
no
|
PLoS Med. 2005 Nov 29; 2(11):e387
|
utf-8
|
PLoS Med
| 2,005 |
10.1371/journal.pmed.0020387
|
oa_comm
|
==== Front
PLoS MedPLoS MedpmedplosmedPLoS Medicine1549-12771549-1676Public Library of Science San Francisco, USA 1628856110.1371/journal.pmed.0020388Correspondence and Other CommunicationsMedical EthicsClinical trialsEthicsResearch designResearch MethodsPlacebo: Physician, Heal Thyself CorrespondenceKumar Arunachalam Kumar C. Jairaj
1
1Kasturba Medical CollegeMangalore, KarnatakaIndiaE-mail: [email protected]
Competing Interests: The authors have declared that no competing interests exist.
11 2005 29 11 2005 2 11 e388Copyright: © 2005 Kumar and Kumar.2005This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are properly credited.
Deception in Research on the Placebo Effect
==== Body
Kudos to Miller et al. [1] for exposing the seamy side of medical research. No matter how positive the results of drug effectiveness, as demonstrated through against-placebo comparison studies, the fact remains that the ends do not justify the means.
To knowingly and willingly induce a patient to trust in the pharmacotherapeutic effectiveness of a phony drug not only trivializes the patient's intelligence, but also devalues the mentor status of the treating physician.
The role of placebo therapy, though quite dramatic at times, does not sanctify the manner in which large-scale research studies pit one group of falsely guided patients or volunteers against another. The medical and pharmacological world must relegate blind new drug–placebo comparison studies to the back burner. Medical journals, too, have a major role in publishing papers that contain misinformation and that mislead. The psychosomatic pharmacokinetics of drugs can be tested or evaluated through much less dubious means than placebo-based research.
The medical profession has an onus to constantly and continuously present itself as an educated partner of the patient in the treatment of disease and ill health: we are quite positive that not a single volunteer would agree to participate in any placebo–new drug study, if informed of the patently false and fake nature of the research protocols.
There is no denying the fact that placebo therapy has a role in fighting disease, but too much nearly criminal injustice is perpetuated in the form of misinformation in placebo-based research studies. The sooner medical professionals and drug manufacturers proscribe placebo-based research, the better for all.
We congratulate the authors of this Policy Forum [1] for bringing to light this rather dark side of research, and hope that the debate engendered will result in re-establishing the physician as a trusted confidant of the trusting patient.
Citation: Kumar A, Kumar CJ (2005) Placebo: Physician, heal thyself. PLoS Med 2(11): e388.
==== Refs
Reference
Miller FG Wendler D Swartzman LC Deception in research on the placebo effect PLoS Med 2005 2 e262 10.1371/journal.pmed.0020262 16173830
|
16288561
|
PMC1297549
|
CC BY
|
2021-01-05 10:39:17
|
no
|
PLoS Med. 2005 Nov 29; 2(11):e388
|
utf-8
|
PLoS Med
| 2,005 |
10.1371/journal.pmed.0020388
|
oa_comm
|
==== Front
PLoS MedPLoS MedpmedplosmedPLoS Medicine1549-12771549-1676Public Library of Science San Francisco, USA 1628856210.1371/journal.pmed.0020395Correspondence and Other CommunicationsGenetics/Genomics/Gene TherapyOtherScience PolicyEpidemiology/Public HealthStatisticsGeneral MedicineCommunication in Health CareEditorial policies (including conflicts of interest)Medical journalsThe Clinical Interpretation of Research CorrespondencePauker Stephen G
1
1Tufts-New England Medical CenterBoston, MassachusettsUnited States of AmericaE-mail: [email protected]
Competing Interests: The author has declared that no competing interests exist.
11 2005 29 11 2005 2 11 e395Copyright: © 2005 Stephen G. Pauker.2005This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are properly credited.
Why Most Published Research Findings Are False
==== Body
John P. A. Ioannidis emphasizes the central role of prior probabilities [1]. His conclusion rests on the presumed low probability that a hypothesis was true before the study.
Unfortunately, his formulation relates the post-study probability that the study's conclusion is true to the pre-study odds. The results might have been clearer had he also plotted the relation of odds to probability, a curvilinear relationship, assuming the study carried no information. Further, the various graphs are right-truncated at pre-study odds, R, of 1.0 (a probability of 0.5), although his examples go as high as R = 2.0. A positive study must, by definition, increase the likelihood that the hypothesis is true. It might have been clearer had Ioannidis chosen to relate odds to odds or probability to probability; in both cases, a neutral study would produce a straight line along a 45-degree diagonal.
The pre-study to post-study relation can more simply be expressed using the odds-likelihood form of Bayes rule—i.e., the post-study odds equal the pre-study odds multiplied times the likelihood ratio (LR) of the study. Then, the equations for positive predictive value (PPV) become the simple product of R × LR. For a single unbiased study, LR = (1 − β)/α. When incorporating study bias, u, as defined by Ioannidis, LR = (1 − β[1 − u])/(α[1 − u] + u). For a typical study with α = 0.05 and β = 0.2 (i.e., with a power of 0.8), LR = 16. When R is less than 1:16 (a probability of 0.0588), the post-study odds will be less than one—i.e., the study's hypothesis will be more likely false than true.
For non-Bayesians, statistical significance testing presumes uninformative prior probability—i.e., R = 1. Then, LR would merely need to exceed one for the study's conclusions to be more likely true than false. At the common significance levels (α) of 0.05 and 0.01, the requisite study powers would merely need to exceed 0.05 and 0.01 respectively, corresponding to maximum type II error rates (β) of 0.95 and 0.99. Such lax requirements would almost always be met for a published study. Hence, the common belief that the vast majority of studies have valid conclusions would be correct if we can assume that the pre-study odds are truly uninformative. However, as Ioannidis suggests, this is unlikely to be the case.
Two more corollaries might be added. The higher the pre-study odds that the study's hypothesis is true, the lower the requisite power (study size and effect size) required to make the study's findings more likely true than false. When studies are published, the investigator should estimate the pre-study odds and report the LR implied by the observed effect.
From the perspective of an epidemiologist or a statistician, the relevant question is whether the study's hypothesis is true—i.e., is the probability of the hypothesis greater than 0.5? For clinicians and their patients, the relevant question is whether a particular strategy should be followed in an individual patient or a subset of similar patients. That decision (or recommendation to the patient) will depend on the pre-study likelihood of benefit in that patient and on the relative magnitude of benefits and risks of that strategy, if the diagnosis in that patient is uncertain. For many such decisions, the “more likely true than false” criterion may not be the best decision rule. For serious diseases and treatments of only modest risk, post-study probabilities of considerably less than 0.5 may be sufficient to justify treatment [2].
Ioannidis's provocative Essay is a timely call for careful consideration of published studies. The odds-likelihood formulation suggested herein may be helpful in providing a more intuitive model. Clinicians now need to take it to the next step.
Citation: Pauker S (2005) The clinical interpretation of research. PLoS Med 2(11): e395.
==== Refs
References
Ioannidis JPA Why most published research findings are false PLoS Med 2005 2 e124 10.1371/journal.pmed.0020124 16060722
Pauker SG Kassirer JP Therapeutic decision making: A cost-benefit analysis N Eng J Med 1975 293 229 234
|
16288562
|
PMC1297552
|
CC BY
|
2021-01-05 10:39:22
|
no
|
PLoS Med. 2005 Nov 29; 2(11):e395
|
utf-8
|
PLoS Med
| 2,005 |
10.1371/journal.pmed.0020395
|
oa_comm
|
==== Front
PLoS MedPLoS MedpmedplosmedPLoS Medicine1549-12771549-1676Public Library of Science San Francisco, USA 1628856310.1371/journal.pmed.0020396Correspondence and Other CommunicationsGenetics/Genomics/Gene TherapyImmunologyOtherPharmacology/Drug DiscoveryHealth PolicyRespiratory MedicinePsychiatryPublic HealthRespiratory MedicineSmokingSubstance use (including alcohol)Genetic Research on Nicotine Dependence Will Facilitate Public Health CorrespondenceCubells Joe
1
1Emory University School of MedicineAtlanta, GeorgiaUnited States of AmericaE-mail: [email protected]
Competing Interests: JC is currently supported by a career development award from the National Institute on Drug Abuse, and collaborates in several ongoing basic research projects in addiction genetics. The opinions expressed here are solely the author's.
11 2005 29 11 2005 2 11 e396Copyright: © 2005 Joe Cubells.2005This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are properly credited.
Will Nicotine Genetics and a Nicotine Vaccine Prevent Cigarette Smoking and Smoking-Related Diseases?
==== Body
W. D. Hall's cogent Research in Translation article [1] on the folly of attempting to use predictive genetic testing in public-health measures to prevent nicotine dependence is a valuable contribution. Indeed, his arguments against predictive testing can easily be applied to virtually any complex genetic disorder. It is certainly important that we in the medical research community continue to offer such articulate education to clinicians, the press, and society in general.
There is a danger, however, that his arguments will be seized by those who oppose supporting research on the genetics of nicotine dependence and other addictions, to the detriment of public health. For example, Merikangas and Risch [2] have already proposed that addictions and several other complex diseases should be deprived of federal research support in favor of other complex disorders, arguing, “The expensive and laborious tools of molecular genetics [should] be prioritized to those diseases ... that cannot now be treated or prevented with environmental changes [such as] type 1 diabetes, multiple sclerosis, autism and schizophrenia. In contrast, gene hunting for disorders that appear to be highly amenable to environmental modification, such as type 2 diabetes, AIDS, alcohol dependence and nicotine dependence, would have lower priority [for federal research support], even though genes may be involved in their etiology.” Those arguments were picked up by right-wing commentators and trumpeted in high-profile lay outlets such as the New York Times. For example, Humphreys and Satel [3] (the latter a resident scholar at the American Enterprise Institute) cite Merikangas and Risch when they conclude, “Some gene research just isn't worth the money ... [because] major cuts in drug- and alcohol-related harm depend not on genes but on choices by policy makers and individual citizens.” Thus, the myth that addictive behavior is simply a matter of “choice” is made to appear as if it has solid science behind it, when in my view, the only real rationale for opposing genetic research on disorders related to smoking, drinking, overeating, homosexual sex, and other “sinful” behavior derives from the same strain of religious moralism underlying creationism and intelligent design.
Such arguments miss the most important rationale for genetic research on addictions and other environmentally influenced complex disorders. These conditions deserve continued vigorous support from the National Institutes of Health and other sources because genetic research is a powerful tool for pointing us toward new treatments based on improved understanding of the biology of the disorders. Nicotine dependence is an important case in point because current treatments, which consist of psychosocial interventions, and medication therapies such as nicotine replacement and buproprion are, in a word, ineffective: relapse rates following smoking cessation with those strategies, while slightly better than no intervention, usually exceed 80% at one-year follow-up [4].
Genetic research, by providing suggestive evidence for associations to the genes encoding the gamma-amino butyric acid (GABA) B receptor subunit 2 [5], or the cannabinoid-1 receptor [6,7], has already helped light the way toward potentially more effective interventions for millions who struggle to quit smoking, but repeatedly fail. While predictive testing of risk for nicotine dependence based on those genetic findings is quite useless, it remains to be ascertained whether pharmacogenetic profiling will be useful for identifying those most likely to benefit from specific medications (or for that matter, psychosocial interventions), or who would be at risk for harmful side-effects from an otherwise effective drug. While the potential for such profiling has been hyped up in the popular press just as predictive testing has been, we have only to recall the lives saved by understanding the genetic basis of variation in thiopurine methyltransferase activity, in the context of thiopurine chemotherapy for acute lymphoblastic leukemia [8], to convince ourselves of the importance of studying the genetic basis of all common complex diseases, whether partially amenable to environmental prevention or not.
Citation: Cubells J (2005) Genetic research on nicotine dependence will facilitate public health? PLoS Med 2(11): e396.
==== Refs
References
Hall WD Will nicotine genetics and a nicotine vaccine prevent cigarette smoking and smoking-related diseases? PLoS Med 2005 2 e266 10.1371/journal.pmed.0020266 16173831
Merikangas KR Risch N Genomic priorities and public health Science 2003 302 599 601 14576422
Humphreys K Satel S Some gene research just isn't worth the money The New York Times; Section F 2005 January 18 5
Office of the US Surgeon General Reducing tobacco use: A report of the surgeon general 2000 Centers for Disease Control and Prevention, Office on Smoking and Health. Available: http://www.cdc.gov/tobacco/sgr/sgr_2000/index.htm . Accessed 11 October 2005
Beuten J Ma JZ Payne TJ Dupont RT Crews KM Single- and multilocus allelic variants within the GABA(B) receptor subunit 2 (GABAB2) gene are significantly associated with nicotine dependence Am J Hum Genet 2005 76 859 864 15759211
Uhl GR Liu QR Walther D Hess J Naiman D Polysubstance abuse-vulnerability genes: Genome scans for association, using 1,004 subjects and 1,494 single-nucleotide polymorphisms Am J Hum Genet 2001 69 1290 300 11704927
Zhang PW Ishiguro H Ohtsuki T Hess J Carillo F Human cannabinoid receptor. 1: 5′ exons, candidate regulatory regions, polymorphisms, haplotypes and association with polysubstance abuse Mol Psychiatry 2004 9 916 931 15289816
Weinshilboum R Wang L Pharmacogenomics: Bench to bedside Nat Rev Drug Discov 2004 3 739 748 15340384
|
16288563
|
PMC1297553
|
CC BY
|
2021-01-05 10:39:20
|
no
|
PLoS Med. 2005 Nov 29; 2(11):e396
|
utf-8
|
PLoS Med
| 2,005 |
10.1371/journal.pmed.0020396
|
oa_comm
|
==== Front
PLoS MedPLoS MedpmedplosmedPLoS Medicine1549-12771549-1676Public Library of Science San Francisco, USA 1628856410.1371/journal.pmed.0020397Correspondence and Other CommunicationsHealth PolicyMedical EthicsMedical LawMental HealthPsychiatryEthicsHealth PolicyHuman rightsPsychiatryMental Health Care and Mental Health Legislation in Pakistan: No Mercy for Losers CorrespondenceNaqvi Haider Ali
1
1Aga Khan University HospitalKarachiPakistanE-mail: [email protected]
Competing Interests: The author has declared that no competing interests exist.
11 2005 29 11 2005 2 11 e397Copyright: © 2005 Haider Ali Naqvi.2005This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are properly credited.
Psychiatric Health Laws in Pakistan: From Lunacy to Mental Health
==== Body
In the Policy Forum by Gilani et al. [1], the authors have quite vividly described the state of mental health care and legislation in Pakistan. Indeed, the situation is one of low awareness and resources, much like other low-income countries in Southeast Asia. Even at the international level, there is deep-seated societal ambivalence toward the mentally ill. The so-called human rights principles have little material effect on the lives of psychiatric patients, and create double standards in the exercise of choice [2].
The promulgation of the new mental health ordinance has, indeed, been a red-letter day in the history of Pakistani legislation. However, how this document translates into real world action remains to be seen. In Pakistan, 70% of health-care services are provided by the private sector, and this, too, is mostly curative in nature [3]. According to the World Health Report (2004), 100% of health-care payments are an out-of-pocket expense for Pakistanis. Most private health care is unregulated. No hospital in Pakistan has Joint Commission on Accreditation on Healthcare Organization (JCAHO) accreditation. Anecdotal reports on abuse of individuals who are mentally ill are ubiquitous. It is not uncommon to see patients who are mentally ill chained to their beds. There is poor provision of psychotropic medication in government-run hospitals. Contrarily, one sees a cocktail of five medications prescribed by an inadequately trained mental health professional in private practice. Out of 342 registered psychiatrists, hardly 100–150 have adequate training. The Pakistan Medical and Dental Council is the sole body for the proper licensing and credentialing of physicians. The problem lies in the implementation of rules and regulations, rather than their existence. One sees a chain of psychiatric hospitals, claiming to deliver psychiatric care, with no qualified psychiatrists on their panels. There is no legal action taken against these people who blatantly exploit patients with mental illness.
All the major tertiary care centers in Pakistan have allied general medical and anesthesia services, yet the provision of modified electroconvulsive therapy (ECT) is deemed only “preferable” in the new mental health ordinance. Unmodified use of ECT results in serious and potentially life-threatening complications. Similarly, there are other paradoxes in the actual care and legislative protection of people with mental illness. The Federal Authority for Mental Health has played no active role in addressing these glaring inequities since its organization in 2001. Contrarily, there is a risk that it might become a power broker for bureaucracy and ministry officials, rather than serving the real stakeholders.
Essentially, nothing has changed for people with mental illness, except the nomenclature and terminologies. There is still no mercy for people with mental illness in poor and other marginalized communities.
Governmental low health-care spending (less than 1% of gross national product) should be seen in the context of the bigger geopolitical situation. Countries' major spending is on defense and military armaments. This is in a politically volatile environment, with ongoing border conflicts with neighbors. However, there is a need for strong political will from the government, which oversees the implementation of rules and regulations and protects the rights of people with mental illness.
For a comprehensive solution, an active public–private partnership is required. This requires a unified agenda and commitment from both sectors. Any lasting solution has to address the deep-rooted inequities, ethical misconduct, and micro- and macroeconomic issues.
Citation: Naqvi HA (2005) Mental health care and mental health legislation in Pakistan: No mercy for losers. PLoS Med 2(11): e397.
==== Refs
References
Gilani AI Gilani UI Kasi PM Khan MM Psychiatric health laws in Pakistan: From lunacy to mental health PLoS Med 2005 2 11 e317 10.1371/journal.pmed.0020317 16159307
Harding TW Human rights law in the field of mental health: A critical review Acta Psychiatr Scand 2000 101 24 30
Khattak FH Financing of health sector in health economics and planning in Pakistan 1996 Islamabad Ad-Rays publishers 44 61
|
16288564
|
PMC1297554
|
CC BY
|
2021-01-05 10:39:26
|
no
|
PLoS Med. 2005 Nov 29; 2(11):e397
|
utf-8
|
PLoS Med
| 2,005 |
10.1371/journal.pmed.0020397
|
oa_comm
|
==== Front
PLoS MedPLoS MedpmedplosmedPLoS Medicine1549-12771549-1676Public Library of Science San Francisco, USA 10.1371/journal.pmed.0020398Correspondence and Other CommunicationsGenetics/Genomics/Gene TherapyOtherScience PolicyEpidemiology/Public HealthStatisticsGeneral MedicineCommunication in Health CareEditorial policies (including conflicts of interest)Medical journalsAuthor's Reply CorrespondenceIoannidis John P. A
1
1University of Ioannina School of MedicineIoanninaGreeceE-mail: [email protected]
Competing Interests: The author has declared that no competing interests exist.
11 2005 29 11 2005 2 11 e398Copyright: © 2005 John P. A. Ioannidis.2005This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are properly credited.
Why Most Published Research Findings Are False
==== Body
I agree with Ian Shrier [1] that, when the effect size is large, adequate power is obtained with small numbers, and it is unnecessary to aim at very large studies. However, most effect sizes probed with statistical testing seem to be small. I also agree that heterogeneity is useful and can offer valuable insights [2]. Sometimes heterogeneity can show us that there are actually two or more research questions, where we thought there was only one [3]. The danger is when heterogeneity is silenced and dismissed in favor of claiming consistent results and when heterogeneity is exploited to show only the most spectacular results—unfortunately, this is not uncommon.
As Stephen Pauker [4] also points out correctly, it is useful to think about what the post-study odds are that one is aiming for if a study eventually were to get a “positive” result. Some residual uncertainty is unavoidable in any research question, no matter how strong the evidence. We should learn to live with uncertainty. I also agree that often the credibility level is less than 50%, yet decisions still have to be made. I don't see a problem implementing a very safe and very cheap medical intervention, even if the credibility that it is effective is only 20%. However, it is important to understand and acknowledge that this intervention has a credibility of 20%, while another has a credibility of 70%. I have no objection or preference on how exactly this will be calculated and plotted. Likelihood ratios are also a nice equivalent approach to calculate the probabilities or odds.
I agree with Jonathan D. Wren [5] that it is impossible to be 100% certain about the exact pre-study odds of truth for any research, mine included of course. However, I argue that we need to start thinking more seriously and consistently about these pre-study odds. In the single nucleotide polymorphism (SNP) association example, one might argue that 1:10,000 is not the best choice, but I doubt anyone would choose 1:100 [6]. Some fields may, indeed, have a pre-study odds of zero—these are the “null fields” that I discussed [7]. The differences in the range in pre-study odds are huge in current research, and I am afraid that this is almost completely ignored. I also have no objection about the framework concept. It is nice to see multiple lines of evidence converge. In fact “framework” evidence may be used to formulate more accurate pre-study odds. However, we should be cautious about how this framework is interpreted. We need more empirical data on how scientists try to converge various pieces of biological, epidemiological, and clinical information. I suspect that bias to make things fit, even if they don't, is not negligible.
Citation: Ioannidis JPA (2005) Author's reply. PLoS Med 2(11): e398.
==== Refs
References
Shrier I Power, reliability, and heterogeneous results PLoS Med 2005 2 e386 10.1371/journal.pmed.0020386 16288559
Berlin JA Invited commentary: Benefits of heterogeneity in meta-analysis of data from epidemiologic studies Am J Epidemiol 1995 142 383 387 7625402
Lau J Ioannidis JP Schmid CH Summing up evidence: One answer is not always enough Lancet 1998 351 123 127 9439507
Pauker S The clinical interpretation of research PLoS Med 2005 2 e395 10.1371/journal.pmed.0020395 16288562
Wren J Truth, probability, and frameworks PLoS Med 2005 2 e361 10.1371/journal.pmed.0020361 16288554
Yang Q Khoury MJ Friedman JM Little J Flanders WD How many genes underlie the occurrence of common complex diseases in the population? Int J Epidemiol 2005 34 1129 1137 16043441
Ioannidis JPA Why most published research findings are false PLoS Med 2005 2 e124 10.1371/journal.pmed.0020124 16060722
|
0
|
PMC1297555
|
CC BY
|
2021-01-05 10:39:22
|
no
|
PLoS Med. 2005 Nov 29; 2(11):e398
|
utf-8
|
PLoS Med
| 2,005 |
10.1371/journal.pmed.0020398
|
oa_comm
|
==== Front
PLoS MedPLoS MedpmedplosmedPLoS Medicine1549-12771549-1676Public Library of Science San Francisco, USA 1628856610.1371/journal.pmed.0020399PerspectivesGenetics/Genomics/Gene TherapyOtherOpthalmologyOphthalmologyGene TherapyDrugs and Adverse Drug ReactionsRestoring Retinal Function in a Mouse Model of Hereditary Blindness PerspectiveMoore Tony Tony Moore is in the Division of Inherited Eye Disease, Institute of Ophthalmology, University College London, London, United Kingdom. E-mail: [email protected]
Competing Interests: The author declares that no competing interests exist.
11 2005 29 11 2005 2 11 e399Copyright: © 2005 Tony Moore.2005This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are properly credited.
Pharmacological and rAAV Gene Therapy Rescue of Visual Functions in a Blind Mouse Model of Leber Congenital Amaurosis
Moore discusses a new study showing rescue of photoreceptor function using gene and drug therapies in a mouse model of Leber congenital amaurosis.
==== Body
Leber congenital amaurosis (LCA) is a generic term used to describe a heterogeneous group of inherited disorders that result in severe loss of rod and cone photoreceptor function in infancy and early childhood. Although each individual disorder is rare, LCA accounts for 3%–5% of childhood blindness [1].
To date, ten different genes have been implicated in such early onset retinal dystrophies. The genes are expressed in the photoreceptors or the underlying retinal pigment epithelium (RPE), and encode a variety of proteins, including structural proteins, transcription factors, and key components of the phototransduction cascade and the visual cycle.
The Visual Cycle
Light perception in vertebrates is mediated by a group of G protein–coupled receptors called opsins, which are bound to a chromophore (light-absorbing compound), 11-cis-retinal, derived from vitamin A. Absorption of light induces an 11-cis to all-trans isomerisation of the chromophore, and the ensuing conformational change initiates the visual transduction cascade. In order for the activated opsin to participate again in the visual process, the all-trans-retinal must be isomerised back to the 11-cis form via an enzymatic pathway, the visual cycle (Figure 1). This occurs mainly within the RPE.
Figure 1 The Visual Cycle
(Illustration: Giovanni Maki, adapted from [4])
Genes encoding a number of enzymes within the visual pathway have been implicated in LCA, namely, RPE65 (retinoid isomerase), retinal dehydrogenase 12, and lecithin:retinol acyl transferase (LRAT). The genetic mutations lead to a lack of 11-cis-retinal, resulting in severe loss of photoreceptor function and ultimately in cell death. Therapeutic approaches aimed at restoring levels of 11-cis-retinal by either improving enzyme function or using other biochemical strategies hold promise for restoring visual function. The development of good animal models of LCA is an essential first step in developing effective therapies.
Promising Treatments in a Mouse Model
Batten et al., in this issue of PLoS Medicine, report very encouraging results of two different therapeutic approaches in the Lrat knockout (Lrat−/−) mouse [2]. The mutant mice show an early onset rod–cone dystrophy and, as in the human disorder, have disease confined to the eye. The mouse model appears to faithfully reproduce the phenotype seen in humans with Lrat deficiency. In the first approach, the authors used gene therapy with subretinal injection of a recombinant adeno-associated virus carrying the Lrat gene. Treated animals showed high expression of Lrat in the RPE adjacent to the injection site, markedly increased levels of visual pigment, and improvement of rod photoreceptor function measured electrophysiologically. The rescue of retinal function peaked at six weeks and then slowly declined. The reasons for the lack of sustained rescue are unclear.
A second approach was to use the orally administered pro-drugs 9-cis-retinyl acetate and 9-cis-retinyl succinate to bypass the Lrat stage of the visual cycle. (9-cis-retinoids were chosen as they are easier to synthesise and are more stable than 11-cis-forms.) Again, rescue of photoreceptor function could be unequivocally demonstrated. Furthermore, treatment with both gene therapy and synthetic retinoids was more effective than treatment with either alone.
Clinical Implications
What is the relevance of these results for patients with early onset retinal dystrophies and their clinicians? Clearly, there is cause for optimism in that treatment may become available within the next few years. Gene therapy is perhaps the most promising approach. Photoreceptor rescue has now been demonstrated in a number of animal models of inherited retinal disease, including mice and dogs that lack RPE65 [3]. The first gene therapy trial in humans with early onset retinal dystrophies as a result of mutations in RPE65 is due to start within the next two years, and if this is successful, treatment of other disorders such as Lrat deficiency will follow. However, unlike RPE65, there is no large animal model of Lrat deficiency, and patients with Lrat deficiency are extremely rare; these two problems may limit the development of a treatment trial in humans. To date, only three patients with Lrat deficiency have been described, and there is very limited information about the clinical phenotype and natural history.
What about dietary treatment? This has been used with limited success in other rare retinal dystrophies in humans, for example, Refsum disease and gyrate atrophy, where the biochemistry of the disorder is understood. The results of Batten and colleagues' study suggest that Lrat deficiency may also respond to dietary treatment [2]. It is very interesting that 9-cis-retinoids can replace 11-cis-retinal, and partially restore photoreceptor function. However, as the authors caution, there is much more work that needs to be done before such compounds can be used in humans. We will need to know whether 9-cis-retinoids are well tolerated in humans, whether there is any retinal or systemic toxicity, and whether retinal function is improved long term with prevention of cell death. Will the treatment be effective relatively late in the disease process when most patients tend to be diagnosed, or is early treatment essential?
The Next Steps
Batten and colleagues' excellent study highlights the improvements that are being made in our understanding of the disease mechanisms in inherited retinal disease [2]. This knowledge will lead to the development of novel therapies for specific forms of disease. The challenge for clinicians is to be prepared for treatment trials. We need to identify the genetic mutations causing retinal degeneration in our patients, carefully characterise the phenotype and natural history of disease, and identify those patients who may benefit from the new therapeutic approaches.
Citation: Moore T (2005) Restoring retinal function in a mouse model of hereditary blindness. PLoS Med 2(11): e399.
Abbreviations
LCALeber congenital amaurosis
LRATlecithin:retinol acyl transferase
RPEretinal pigment epithelium
==== Refs
References
Rahi JS Cable N Severe visual impairment and blindness in children in the UK Lancet 2003 362 1359 1365 14585637
Batten ML Imanishi Y Tu DC Doan T Zhu L Pharmacological and rAAV gene therapy rescue of visual functions in a blind mouse model of Leber congenital amaurosis PLoS Med 2005 2 e333 10.1371/journal.pmed.0020333 16250670
Acland GM Aguirre GD Ray J Zhang Q Aleman TS Gene therapy restores vision in a canine model of childhood blindness Nat Genet 2001 28 92 95 11326284
Sparrow JR Therapy for macular degeneration: Insights from acne Proc Natl Acad Sci U S A 2003 100 4353 4354 12682280
|
16288566
|
PMC1297556
|
CC BY
|
2021-01-05 10:39:20
|
no
|
PLoS Med. 2005 Nov 29; 2(11):e399
|
utf-8
|
PLoS Med
| 2,005 |
10.1371/journal.pmed.0020399
|
oa_comm
|
==== Front
PLoS MedPLoS MedpmedplosmedPLoS Medicine1549-12771549-1676Public Library of Science San Francisco, USA 1628856710.1371/journal.pmed.0020401PerspectivesCancer BiologyCell BiologyOncologyOncologyMitochondrial DNA Mutations in Cancer PerspectiveZanssen Stefanie *Schon Eric A Stefanie Zanssen is in the Department of Neurology, College of Physicians and Surgeons, Columbia University, New York, New York, United States of America, and Eric A. Schon is in the Departments of Neurology and Genetics and Development, College of Physicians and Surgeons, Columbia University, New York, New York, United States of America.
Competing Interests: The authors declare that no competing interests exist.
*To whom correspondence should be addressed. E-mail: [email protected] 2005 29 11 2005 2 11 e401Copyright: © 2005 Zanssen and Schon.2005This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are properly credited.
A Critical Reassessment of the Role of Mitochondria in Tumorigenesis
One can no longer ignore mitochondria in cancer biology, argue Zanssen and Schon.
==== Body
In 1956, the German biochemist and Nobel Laureate Otto Warburg proposed that cancer is caused by altered metabolism and by deranged energy processing in mitochondria [1]. After falling on deaf ears for decades, his theory has recently enjoyed a resurrection, coinciding with an explosion of new information on the role of mitochondria in energy metabolism and oxygen sensing.
Mitochondrial Enzyme Deficiencies in Inherited Neoplasia
Mitochondrial defects have been associated with severe neurodegenerative disorders [2], and, more recently, with primary hereditary neoplasias. Germline heterozygote mutations in the nucleus-encoded mitochondrial succinate dehydrogenase (SDH) subunits—a tricarboxylic acid cycle (TCA) enzyme, which is also part of the respiratory chain—cause inherited pheochromocytomas and paragangliomas. Mutations in another TCA enzyme, fumarate hydratase (FH), cause cutaneous and uterine leiomyomas, as well as renal cell carcinomas [3]. Mitochondrial proteins may also play a role in the development of the sporadic kidney tumor oncocytoma [4]. Selak et al. has recently shed light on the mysterious connection between tumors and SDH/FH [5]; they showed that the hypoxia-inducible factor (HIF)–mediated signaling pathway, known to be tumorigenic in von Hippel Lindau syndrome, also plays a role in SDH and FH deficiency (Figure 1) [5].
Figure 1 Schematic Model for Tumorigenesis in SDH deficiency
SDH is a component of the OxPhos complexes (I–IV) as well as the TCA cycle. Due to SDH deficiency succinate accumulates intramitochondrially (the same occurs when FH is deficient) and is transported into the cytosol. Here, its elevated concentration inhibits prolyl hydroxylases (PDH1-3), which can no longer degrade HIF-1α even under normoxic conditions. Heterodimerisation of stabilized HIF-1α and HIF-1β now forms HIF, which induces transcription of nuclear genes involved in tumor progression [5].
(Illustration: Giovanni Maki)
Somatic mtDNA Mutations in Sporadic Tumors
Mitochondrial DNA (mtDNA) mutations have also been linked to nonhereditary tumors. Intragenic deletions (see Glossary) [6], missense and chain-terminating point mutations [7], and alterations of homopolymeric sequences [8] have been identified in nearly every type of tumor. Mutations have been described in the hypervariable regions of the mitochondrial D-loop, the section of DNA controlling mtDNA transcription and replication that is most prone to mutation. Mutations have also been described in all 22 tRNAs, both rRNAs, and all 13 of the mtDNA-encoded subunits of the respiratory chain complexes.
Glossary
Chain-terminating point mutation: A point mutation that leads to a premature stop codon
D-loop analysis: Analysis of polymorphisms in the mtDNA control region, known as the D-loop
Ethnic-specific haplotype-defining polymorphism: A polymorphism representing an mtDNA variant found exclusively in a specific population. Haplotypes H, K, J, and U are found in European populations
Homopolymeric sequences: Consecutive runs of the same nucleotide
Intragenic deletion: Loss of a contiguous segment of DNA from a gene
Missense mutation: A point mutation in a codon resulting in the alteration of one encoded amino acid to another
Mitomap database: An online human mitochondrial genome database (http://www.mitomap.org/)
Nuclear mitochondrial pseudogenes: Nuclear DNA sequences homologous to mtDNA sequences, but without any function
ROS: Reactive oxygen species
rRNA: Ribosomal RNA
tRNA: Transfer RNA
These findings pose a key conceptual question: how do these mutations arise and then establish themselves so quickly in the tumor? Many of the (noninherited) mtDNA mutations lead to the mtDNA of the tumor becoming heteroplasmic; i.e., they are a mixture of normal germline mtDNAs found in the noncancerous tissue and mtDNAs containing one or more mutations. Even more surprising than this heteroplasmy is the observation that most mutations are apparently homoplasmic (i.e., there is no evidence of the germline mtDNA haplotype in the tumor), and further, some of those genomes contain not one, but two, three, and even four mutations, all on the same mtDNA circle. How can a cell with mitochondrial haplotype A in noncancerous tissue shift its mitochondrial genotype to haplotype B so rapidly in the adjacent tumor? Mathematical models of heteroplasmic shifting in rapidly dividing tumor cells [9] have been invoked to explain this conundrum, but the mechanism of shifting remains both elusive and controversial.
In this issue of PLoS Medicine, Salas et al. question the very foundation of the observation of haplotype shifting in tumors [10]. They present evidence that much of the literature demonstrating mutations—whether heteroplasmic or homoplasmic—has been laden with technical and conceptual errors that reduce the finding of mtDNA mutations in tumors to the level of technical artifact.
Salas et al. rightly emphasize the importance of the highest standards in mtDNA analysis to produce reliable results in the many investigations on this subject [10]. In summary, they point out that double-checking to avoid sample mix-ups and using stringent procedures to extract either purified mtDNA or total DNA containing the mtDNA will guarantee high-quality templates (avoiding, for example, amplification of nuclear mitochondrial pseudogenes, as in the criticized study of Reddy et al. [11]). They also point out that controls should be ethnicity- and age-matched, and that every mutation should be checked to see whether the mutation is listed in the Mitomap database or is an ethnic-specific haplotype-defining polymorphism. Furthermore, when interpreting results, researchers should assess whether the missense mutations that are being implicated in carcinogenesis fulfill the criteria for pathogenicity (for example, they should occur in highly conserved regions or destroy structurally important motifs). If these criteria are adhered to, Salas et al. claim that the vast majority of mtDNA mutations in tumors will simply disappear [10].
However, although technical errors probably do account for some of the positives found in some of the literature, they cannot account for all of them. Putting aside the issue of mechanism, the question that remains—can mtDNA haplotypes shift rapidly in tumors—is not only still with us, but is almost certainly answered affirmatively.
We would interpret the mtDNA sequencing data from the phylogenetic point of view differently than Salas and colleagues. mtDNA haplotype-specifying polymorphisms [12] arising on a background of a fixed mitochondrial germline haplotype are seen commonly in tumors, for example, in a patient with myelodysplastic syndrome ([MDS], case one in [13]) who had mtDNA germline haplotype U (confirmed by U-specifying polymorphisms at nt 150, 3197, 12308, and 12372). Such polymorphisms are not necessarily due to sequencing errors or contaminations. The typical slow progression of MDS to acute leukemia in this patient made it possible to monitor the shift from “initial” germline homoplasmy to heteroplasmy to “final” homoplasmy of mutated mtDNA in the tumor in different stages of the disease. Figure 2 shows that the haplotype U–specifying positions 12308 and 12372 in the patient's platelets had a “beginning” heteroplasmy in MDS (Figure 2A), which reached more than 50% when the cells were in transformation (Figure 2B), and “ended” at homoplasmy when the acute leukemia arose (Figure 2C). Sample contamination or a shift from haplotype U to H can be excluded here because all other haplotype-specific polymorphisms were homoplasmic for type U, not H. The same observation was made in patient 2 of the cited study from Kirches et al. [14]. This patient had a germline mitochondrial haplotype J, which “shifted” in positions 185, 295, and 16126 back to the phylogenetically older haplotype H, but shifted in position 195 to haplotype W and in position 204 to nowhere.
Figure 2 Mutations C12308T and T12372C Alter Polymorphisms Specifying European Mitochondrial Haplotype U
Mutation load in (A) MDS, (B) MDS in transformation, and (C) acute leukemia.
Although the mechanism of rapid shifting is still unclear, random processes, extensively modeled in computer systems [9], may well explain the presence of homoplasmic mutations in cancer. It may be that certain mitochondrial haplotypes or pathogenic mutations represent a selective advantage in tumor progression, similar to the overrepresentation of haplotype J seen in Leber hereditary optic neuropathy, a maternally inherited mitochondrial disease [15].
Challenges for the Future
Warburg was right. One can no longer ignore mitochondria in cancer biology. Mutations of mtDNA, even driven by random processes during malignant transformation, present an excellent possibility for early tumor detection using, e.g., D-loop analysis of bodily fluids from patients with tumors [16]. (As an aside, direct sequencing of tumor mtDNA, as performed by many groups, is a poor screening technique, as it misses levels of heteroplasmy below approximately 20%; a better method is denaturing high-performance liquid chromatography followed by confirmatory polymerase chain reaction/restriction fragment length polymorphism analysis.)
The proposal that mtDNA mutations and respiratory dysfunction may be linked directly to carcinogenesis via apoptotic or reactive oxygen species (ROS)–mediated pathways is challenging, but urgently needs experimental proof. Questions regarding whether the HIF-mediated pathway is also initiated in hypoxia and mitochondrial deficiency [17], both characteristics of tumors, also need to be clarified. If these pathways are confirmed as being involved in tumorigenesis, metabolic targeting (e.g., blocking the HIF-pathway by administration of α-ketoglutarate) of the mitochondrion in cancer may open up new avenues to antineoplastic therapies and prevention of cancer.
Citation: Zanssen S, Schon EA (2005) Mitochondrial DNA mutations in cancer. PLoS Med 2(11): e401.
Abbreviations
FHfumarate hydratase
HIFhypoxia-inducible factor
MDSmyelodysplastic syndrome
mtDNAmitochondrial DNA
SDHsuccinate dehydrogenase
TCAtricarboxylic acid cycle
==== Refs
References
Warburg O On the origin of cancer cells Science 1956 123 309 314 13298683
DiMauro S Schon EA Mitochondrial respiratory-chain diseases N Engl J Med 2003 348 2656 2668 12826641
Eng C Kiuru M Fernandez MJ Aaltonen LA A role for mitochondrial enzymes in inherited neoplasia and beyond Nat Rev Cancer 2003 3 193 202 12612654
Zanssen S Gunawan B Fuzesi L Warburton D Schon EA Renal oncocytomas with rearrangements involving 11q13 contain breakpoints near CCND1 Cancer Genet Cytogenet 2004 149 120 124 15036887
Selak MA Armour SM MacKenzie ED Boulahbel H Watson DG Succinate links TCA cycle dysfunction to oncogenesis by inhibiting HIF-alpha prolyl hydroxylase Cancer Cell 2005 7 77 85 15652751
Horton TM Petros JA Heddi A Shoffner J Kaufman AE Novel mitochondrial DNA deletion found in a renal cell carcinoma Genes Chromosomes Cancer 1996 15 95 101 8834172
Polyak K Li Y Zhu H Lengauer C Willson JK Somatic mutations of the mitochondrial genome in human colorectal tumours Nat Genet 1998 20 291 293 9806551
Habano W Nakamura S Sugai T Microsatellite instability in the mitochondrial DNA of colorectal carcinomas: Evidence for mismatch repair systems in mitochondrial genome Oncogene 1998 17 1931 1937 9788436
Coller HA Khrapko K Bodyak ND Nekhaeva E Herrero-Jimenez P High frequency of homoplasmic mitochondrial DNA mutations in human tumors can be explained without selection Nat Genet 2001 28 147 150 11381261
Salas A Yao YG Macaulay V Vega A Carracedo Á A critical reassessment of the role of mitochondria in tumorigenesis PLoS Med 2005 2 e296 10.1371/journal.pmed.0020296 16187796
Reddy PL Shetty VT Dutt D York A Dar S Increased incidence of mitochondrial cytochrome c-oxidase gene mutations in patients with myelodysplastic syndromes Br J Haematol 2002 116 564 575 11849212
Torroni A Wallace DC Mitochondrial DNA variation in human populations and implications for detection of mitochondrial DNA mutations of pathological significance J Bioenerg Biomembr 1994 26 261 271 7521328
Linnartz B Anglmayer R Zanssen S Comprehensive scanning of somatic mitochondrial DNA alterations in acute leukemia developing from myelodysplastic syndromes Cancer Res 2004 64 1966 11971 15026331
Kirches E Krause G Warich-Kirches M Weis S Schneider T High frequency of mitochondrial DNA mutations in glioblastoma multiforme identified by direct sequence comparison to blood samples Int J Cancer 2001 93 534 538 11477557
Hofmann S Bezold R Jaksch M Kaufhold P Obermaier-Kusser B Disease relevance of the so-called secondary Leber hereditary optic neuropathy mutations Am J Hum Genet 1997 60 1539 1542 9199577
Fliss MS Usadel H Caballero OL Wu L Buta MR Facile detection of mitochondrial DNA mutations in tumors and bodily fluids Science 2000 287 2017 2019 10720328
Doege K Heine S Jensen I Jelkmann W Metzen E Inhibition of mitochondrial respiration elevates oxygen concentration but leaves regulation of hypoxia-inducible factor (HIF) intact Blood 2005 106 2311 2317 15947089
|
16288567
|
PMC1297557
|
CC BY
|
2021-01-05 11:13:38
|
no
|
PLoS Med. 2005 Nov 29; 2(11):e401
|
utf-8
|
PLoS Med
| 2,005 |
10.1371/journal.pmed.0020401
|
oa_comm
|
==== Front
PLoS MedPLoS MedpmedplosmedPLoS Medicine1549-12771549-1676Public Library of Science San Francisco, USA 10.1371/journal.pmed.0020405Correspondence and Other CommunicationsInfectious DiseasesOtherEpidemiology/Public HealthHealth PolicyObstetrics/GynecologyPediatricsWomen's HealthInfectious DiseasesInternational healthMedicine in Developing CountriesPublic HealthAuthor's Reply CorrespondenceAttaran Amir
1
1University of OttawaOttawa, OntarioCanadaE-mail: [email protected]
Competing Interests: AA has held small contracts or been paid per diem by the World Bank, the United Nations Development Program, and the Roll Back Malaria Partnership in the last five years. None of these agencies were consulted in the development of this manuscript. Research funding was provided exclusively by the Canada Research Chairs program.
11 2005 29 11 2005 2 11 e405Copyright: © 2005 Amir Attaran.2005This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are properly credited.
An Immeasurable Crisis? A Criticism of the Millennium Development Goals and Why They Cannot Be Measured
Response to Amir Attaran
==== Body
I am grateful for the reply of Jeff Sachs, John McArthur, and Guido Schmidt-Traub to my article [1,2]. Their reply, written on behalf of the United Nations (UN) Millennium Project, shows that even the leading thinkers of that organization possess beliefs about the Millennium Development Goals (MDGs) that contradict the factual evidence.
On one issue, the UN Millennium Project team agrees with me: progress on the MDGs is sometimes not well measured, such that it is impossible to know if the goals are on track to being fulfilled by their 2015 deadline.
But even on this issue, we disagree on the extent of the problem. The UN Millennium Project team writes that, because my analysis was limited to only the public health MDGs, I based my conclusions on only “the toughest measurement challenges,” and “generalize[d] incorrectly across the MDGs” [1]. In their view, if one leaves the difficult health MDGs aside, then “several other MDG indicators can be measured quite well” [1].
To determine whether this assertion is correct, I accessed the UN's own “Data Availability Analysis” [3] for 2005, which takes into account every one of the 48 UN-designated MDG indicators (really 65 indicators in 48 categories). For each indicator, the UN's analysis summarizes the percentage of countries possessing measurements taken in two benchmark years: one year near the starting point of the MDGs (usually 1990) and another year nearer to the present (after 1999). Naturally, it is the difference of this pair of benchmark measurements, taken years apart, which proves whether progress is or is not being achieved for a particular MDG indicator.
The disappointing result of the UN's “Data Availability Analysis” is that, more often than not, the requisite pair of benchmark measurements doesn't exist, such that no factual conclusion about progress can be made. In the best case scenario, there are two indicators with paired benchmark measurements in 98% of countries. In the worst case scenario, there are 26 indicators—13 times as many—with paired benchmark measurements in none of the countries. Even the median MDG indicator, which is the one indicator most typical of the bunch, has paired benchmark measurements in just 5% of countries—meaning that the UN does not possess those data for 95% of countries.
In that context, the UN Millennium Project team is definitely wrong to believe that, health indicators aside, the MDGs “can be measured quite well” [1]. The UN's comprehensive analysis of all MDG indicators shows that, for whatever reason, most are not measured well, not even so rarely as twice a decade.
Why is measurement of the MDGs so generally poor? According to the UN Millennium Project team, the answer is money. They write that “developing countries and the international system,” which presumably includes the UN, “lack the resources to measure” the MDGs [1].
However, this belief, too, contradicts the evidence. Concerning the health MDGs, my article recommended expanding the network of demographic surveillance sites (DSS) as the single most efficient way to obtain timely, accurate measurements [2]. According to a recent study of DSS in Tanzania, this costs $0.01 per person, per year [4]. Thus, to institute DSS and good quality MDG measurements for the 4 billion poorest people worldwide would cost perhaps $40 million annually.
In this context, the UN Millennium Project team's argument that the “international system lacks the resources” to effectively measure the health MDGs is without credibility. The sum of $40 million is under 0.1% of the global foreign aid budget (Organization for Economic Cooperation and Development Development Assistance Committee [OECD DAC]).
Without such steps to measure progress on achieving the MDGs, any claims made for them are necessarily conjectural, rather than objective. The UN Millennium Project team writes that, even without measurement, “the MDGs are already promoting strengthened health systems in low-income countries” [1]—but regrettably, they fail to furnish evidence of this. They also write that “the very adoption of the maternal mortality goal … is provoking greatly increased attention to improvements in data collection” [1]. What they fail to mention is that the UN adopted that goal in 1990, and despite 15 years of provoking increased attention, elsewhere the UN Millennium Project team have called the data “unreliable” [5].
I could dispute other unsupported assertions in the reply of the UN Millennium Project team, but choose not to do so because it would distract from this fundamental point: whether to honor the health of the world's poorest or sickest people or to restore the earth's most vulnerable natural environments or to secure human rights for children and women, the UN must demonstrate much greater responsibility than it has in measuring the status of the MDGs. The UN Millennium Project team urges to “cement the MDGs as operational rather than simply rhetorical targets” [1]. I agree this is desirable, and, actually, to place rhetoric ahead of evidence is unethical.
That is why measurement to prove—not just to speculate—on the MDGs' operational progress cannot continue to be neglected, and also why the leading intellectuals of the UN Millennium Project err awfully in their judgment when justifying the neglect to date.
Citation: Attaran A (2005) Author's reply. PLoS Med 2(11): e405.
==== Refs
References
Sachs JD McArthur J Schmidt-Traub G Response to Amir Attaran PLoS Med 2005 2 e379 10.1371/journal.pmed.0020379 16288557
Attaran A An immeasurable crisis? A criticism of the Millennium Development Goals and why they cannot be measured PLoS Med 2005 2 e318 10.1371/journal.pmed.0020318 16156696
United Nations Data availability analysis 2005 New York United Nations Available: http://unstats.un.org/unsd/mi/techgroup/January2005/Series%20update%20status%20query_FC.xls . Accessed 13 October 2005
Rommelmann V Setel PW Hemed Y Angeles G Mponezya H Cost and results of information systems for health and poverty indicators in the United Republic of Tanzania Bull World Health Organ 2005 83 569 577 16184275
United Nations Millennium Project Investing in development: A practical plan to achieve the Millennium Development Goals 2005 New York Earthscan 329
|
0
|
PMC1297559
|
CC BY
|
2021-01-05 10:39:22
|
no
|
PLoS Med. 2005 Nov 29; 2(11):e405
|
utf-8
|
PLoS Med
| 2,005 |
10.1371/journal.pmed.0020405
|
oa_comm
|
==== Front
Arthritis Res TherArthritis Research & Therapy1478-63541478-6362BioMed Central London ar17961627766810.1186/ar1796Research ArticleExpression of ADAM15 in rheumatoid synovium: up-regulation by vascular endothelial growth factor and possible implications for angiogenesis Komiya Koichiro [email protected] Hiroyuki [email protected] Isao [email protected] Satoko [email protected] Yoshinari [email protected] Eiji [email protected] Eiko [email protected] Hideo [email protected] Yoshiaki [email protected] Yasunori [email protected] Department of Pathology, School of Medicine, Keio University, Shinjuku-ku, Tokyo, Japan2 Department of Orthopaedic Surgery, School of Medicine, Keio University, Shinjuku-ku, Tokyo, Japan3 Biopharmaceutical Department, Daiichi Fine Chemical Co. Ltd., Takaoka, Toyama, Japan2005 5 8 2005 7 6 R1158 R1173 28 5 2005 20 6 2005 23 6 2005 27 6 2005 Copyright © 2005 Komiya et al.; licensee BioMed Central Ltd.This is an Open Access article distributed under the terms of the Creative Commons Attribution License , which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
ADAMs (a disintegrin and metalloproteinases) comprise a new gene family of metalloproteinases, and may play roles in cell-cell interaction, cell migration, signal transduction, shedding of membrane-anchored proteins and degradation of extracellular matrix. We screened the mRNA expression of 10 different ADAMs with a putative metalloproteinase motif in synovial tissues from patients with rheumatoid arthritis (RA) or osteoarthritis (OA). Reverse transcription PCR and real-time quantitative PCR analyses indicated that among the ADAMs, ADAM15 mRNA was more frequently expressed in the RA samples and its expression level was significantly 3.8-fold higher in RA than in OA (p < 0.01). In situ hybridization, immunohistochemistry and immunoblotting demonstrated that ADAM15 is expressed in active and precursor forms in the synovial lining cells, endothelial cells of blood vessels and macrophage-like cells in the sublining layer of RA synovium. There was a direct correlation between ADAM15 mRNA expression levels and vascular density in the synovial tissues (r = 0.907, p < 0.001; n = 20). ADAM15 was constitutively expressed in RA synovial fibroblasts and human umbilical vein endothelial cells (HUVECs), and the expression level was increased in HUVECs by treatment with vascular endothelial growth factor (VEGF)165. On the other hand, ADAM15 expression in RA synovial fibroblasts was enhanced with VEGF165 only if vascular endothelial growth factor receptor (VEGFR)-2 expression was induced by treatment with tumor necrosis factor-α, and the expression was blocked with SU1498, a specific inhibitor of VEGFR-2. These data demonstrate that ADAM15 is overexpressed in RA synovium and its expression is up-regulated by the action of VEGF165 through VEGFR-2, and suggest the possibility that ADAM15 is involved in angiogenesis in RA synovium.
==== Body
Introduction
In rheumatoid arthritis (RA), the affected joints develop chronic synovitis that is characterized by hyperplasia of lining cells, infiltration of inflammatory cells and abundant neovascularization. Various factors such as proteinases, growth factors and cytokines are produced in the RA synovium and implicated in the destruction of articular cartilage and subchondral bones, leading to disability of the joints. Among the proteinases, matrix metalloproteinases (MMPs), a gene family of zinc metalloproteinases, are well known to play a major role in the proteolytic degradation of extracellular matrix (ECM) macromolecules of cartilage and bone, which is a key step in joint destruction in RA [1]. Members of a new family of metalloproteinases, the 'a disintegrin and metalloproteinases' (ADAMs), which share structural homology with MMPs and snake venom metalloproteinases, have recently been cloned. ADAMs consist of propeptide, metalloproteinase, disintegrin-like, cysteine-rich, epidermal growth factor-like, transmembrane and cytoplasmic tail domains [1,2]. Members are classified into putative proteinase-type and non-proteinase-type ADAMs according to the different structures of the catalytic site motif in the metalloproteinase domain [1,3]. Although the specific biological functions of ADAMs are not well elucidated at the present time, they may be involved in cell-cell interaction, cell migration, signal transduction, shedding of various membrane-anchored proteins and degradation of ECM components under pathophysiological conditions such as fertilization [4,5], morphogenesis [6,7], angiogenesis [8] and cancer [9]. The expression of ADAM10, ADAM15 and ADAM17 in arthritic cartilage and synovial tissues has been examined [10-12], but there are no reports of systematic analyses of the expression of ADAMs in arthritic joint tissues. In addition, little or no information is available for correlation between the expression and synovial pathology or for regulation mechanism of ADAM expression. Angiogenesis in the synovium during RA begins at the early stage of the disease and is crucial for progression of the synovitis [13]. Vascular endothelial growth factor (VEGF), which has five different isoforms (VEGF121, VEGF145, VEGF165, VEGF189 and VEGF206) is known to play a key role in the angiogenesis in RA synovium [13,14]. All these VEGF isoforms bind to high-affinity receptors, namely VEGFR-1 (fms-like tyrosine kinase; Flt-1) and VEGFR-2 (kinase insert domain-containing receptor; KDR). Neuropilin-1, an isoform-specific co-receptor of VEGFR-2, enhances the bioactivity of VEGF165 by increasing its binding affinity for VEGFR-2 [15]. Interestingly, binding of VEGF to its receptors on endothelial cells enhances not only their proliferation and migration but also production of MMPs [16-18]. In addition, VEGF stimulates other cells such as chondrocytes to induce expression of MMPs [19]. Thus, it might be possible to speculate that VEGF regulates the expression of proteinase-type ADAMs.
In the present study, we examined the expression of 10 different ADAM species with a putative metalloproteinase motif in synovial tissues of RA and osteoarthritis (OA), correlation of ADAM15 expression with synovial pathology, localization of ADAM15 in RA synovium, and the mechanism of regulation of ADAM15 expression in RA synovial fibroblasts and endothelial cells. Our results demonstrate that ADAM15 is expressed in lining cells, endothelial cells of blood vessels and macrophage-like cells in the sublining layer of RA synovium with a direct correlation with vascular density in the synovium, and that the expression of ADAM15 is up-regulated by the action of VEGF165 via VEGFR-2.
Materials and methods
Clinical samples and histology
Synovial tissues were obtained from patients with RA (56 ± 14 years old (mean ± SD); n = 16) or OA (73 ± 6 years old; n = 20) at total knee arthroplasty. Diagnosis of the patients with RA or OA was based on the 1987 revised American Rheumatism Association Criteria for RA [20] and the American Rheumatism Association Criteria for OA [21]. Synovial specimens were fixed with periodate-lysine-paraformaldehyde, and paraffin sections stained with hematoxylin and eosin were analyzed by light microscopy according to our grading system of synovial lining cell hyperplasia, cellular infiltration and fibrosis [22]. For the experimental use of the surgical specimens, written informed consent was obtained from the patients according to the hospital ethical guidelines.
Reverse transcription-PCR
Total RNA was extracted directly from RA (n = 16) and OA (n = 20) synovial tissues and evaluated by using the Agilent 2100 Bioanalyzer (Agilent Technologies, Palo Alto, CA, USA) as descried previously [9]. By using a random hexamer of oligonucleotides (Takara Bio Inc., Otsu, Japan), cDNAs were prepared from total RNA with SuperScript II reverse transcriptase (Life Technologies Inc., Rockville, MD, USA). The reaction product was subjected to reverse transcription (RT)-PCR analysis on the expression of ADAMs 8, 9, 10, 12, 15, 17, 20, 21, 28 and 30, VEGFR-1, VEGFR-2, neuropilin-1 and β-actin for 25–30 cycles. PCR was carried out in 50 μl reaction volume containing 800 nM of each primer, 220 μM of dNTPs and 1 unit of ExTaq DNA polymerase (Takara Bio Inc.). The thermal cycle was 1 minute at 94°C, 1 minute at 62°C for ADAMs 8, 9, 10, 12, 17, 20, 21, 28 and 30, 67°C for ADAM15, 64°C for VEGFR-1, 63°C for VEGFR-2 and neuropilin-1 and 65°C for β-actin, and 1 minute at 72°C, followed by 3 minutes at 72°C for the final extension. The nucleotide sequences of the PCR primers and the expected sizes of the amplified cDNA fragments are shown in Table 1. Aliquots of the PCR products were electrophoresed in 2% agarose gels, and stained with ethidium bromide. For positive controls, total RNA was extracted from cancer cell lines as described previously [9]. The specific amplification of these ADAMs, VEGFRs, neuropilin-1 and β-actin was confirmed by direct sequencing of the PCR products.
Real-time quantitative PCR for ADAM15
The mRNA expression levels of ADAM15 were evaluated in a TaqMan real-time PCR assay using the ABI Prism 7000 sequence detection system (Applied Biosystems, Foster City, CA, USA) according to the manufacturer's protocols. Cycling conditions were 50°C for 2 minutes, 95°C for 10 minutes, and then 40 cycles at 95°C for 15 seconds and 60°C for 1 minute. Primers were designed and selected using Primer Express software (Applied Biosystems). Sequences of the primers and TaqMan probe for ADAM15 were as follows: forward primer, 5'-GGCAATCGAGGCAGCAAAT-3'; reverse primer, 5'-TGGTGGAGATCAGCCCAAAC-3'; and TaqMan probe, 5'-FAM-CAGCTGTCACCCTCGAAAACTTCCTCC-TAMRA-3'. The relative quantification value of ADAM15 was normalized to an endogenous control, 18S ribosomal RNA, after confirming that ADAM15 and ribosomal 18S cDNAs were amplified with the same efficiency according to the manufacturer's protocol. The total gene specificity of the nucleotide sequences chosen for the primers and probe and the absence of DNA polymorphisms were ascertained using BLASTN and Entrez from the National Center for Biotechnology Information web site [23].
In situ hybridization
Paraffin sections of the RA synovial tissues (n = 5) were used for in situ hybridization of ADAM15 according to a modification of our methods used previously [19]. Briefly, the sections were treated with proteinase K (5 μg/ml; Sigma-Aldrich Inc., St Louis, MO, USA) in 10 mM Tris-HCl buffer, pH 8.0, and 1 mM ethylenediaminetetraacetic acid, and endogenous peroxidase was blocked with 0.3% hydrogen peroxide in methanol. Single-stranded sense and anti-sense digoxigenin-labeled RNA probes were generated by in vitro transcription from the cDNA encoding ADAM15, nucleotides 1091 to 1331 (241 bp), using the DIG RNA labeling kit (Roche Diagnostics GmbH, Mannheim, Germany). BLASTN searches were performed to confirm the specificity of the probes. Hybridization with the probes was performed at 42°C for 16 h, and the sections were washed in 2× standard saline citrate/50% formamide, followed by digestion with 10 μg/ml ribonuclease A (Wako Pure Chemical Industries, Osaka, Japan). After washing once in 2× standard saline citrate and twice in 0.2× standard saline citrate and blocking nonspecific binding with blocking solution (DakoCytomation Norden A/S, Glostrup, Denmark), they were incubated with mouse anti-digoxigenin antibody (1/750 dilution; Roche Diagnostics GmbH), and subjected to the following steps using the Catalyzed Signal Amplification System (DakoCytomation Norden A/S) according to the manufacturer's protocol. Counterstaining was performed with hematoxylin.
Characterization of monoclonal antibody against human ADAM15 and immunoblotting
A monoclonal antibody against human ADAM15 was developed using the synthetic peptide corresponding to the amino acid sequence of the cysteine-rich domain (residues 596 to 612, RDLLWETIDVNGTELNC) of human ADAM15 as an antigen according to our methods [24]. Clones were screened by enzyme-linked immunosorbent assay using the peptide, and a clone 240-2C7 was selected as a candidate. The specific reactivity of the antibody was further evaluated by immunoblotting. The cysteine-rich domain of ADAM15 with FLAG-tag was expressed in Escherichia coli DH5α (Takara Bio Inc.) by transfecting with the expression vector pFLAG-ADAM15, which was prepared by inserting a cDNA fragment encoding the human ADAM15 cysteine-rich domain (nucleotides 1887 to 1937) [25] into pFLAG-CTC vector (Sigma-Aldrich Inc.). As a negative control, the vector pFLAG-CTC alone was transfected to DH5α (mock transfectants). Cell lysates were subjected to SDS-PAGE (15% total acrylamide) under reducing conditions. The resolved proteins were then transferred to polyvinylidene difluoride membranes (ATTO, Tokyo, Japan), which were reacted with anti-FLAG antibody (1 μg/ml; Sigma-Aldrich, Inc.), anti-ADAM15 antibody (1 μg/ml; 240-2C7) or non-immune mouse immunoglobulin G (IgG) (1 μg/ml; DakoCytomation Norden A/S) after blocking with 5% skim milk. They were then incubated with horseradish peroxidase-labeled anti-mouse IgG (1/5000 dilution; Amersham Biosciences Corp., Piscataway, NJ, USA). Immunoreactive bands were detected with ECL Western blotting reagents (Amersham Biosciences Corp.). As shown in Fig. 1, two protein bands of 15 kDa and 10 kDa were detected with anti-FLAG antibody in the cell lysates of ADAM15 transfectants (lane 1) but not mock transfectants (lane 2). On the other hand, anti-ADAM15 antibody (240-2C7) selectively reacted with the band of 15 kDa in the ADAM15 transfectants (Fig. 1, lane 3) but not mock transfectants (lane 4). The molecular weight of the immunoreactive 15-kDa band corresponds to that of the cysteine-rich domain of ADAM15 [25]. Importantly, the immunoreactivity of the 15-kDa band was blocked after absorption of the antibody with the antigen peptide (Fig. 1, lanes 5 and 6). Blotting with non-immune mouse IgG showed no reactive bands (data not shown). This indicates that the monoclonal antibody (240-2C7) is monospecific to the cysteine-rich domain of ADAM15.
RA synovial tissues (n = 5) were homogenized on ice in a lysis buffer (50 mM Tris-HCl buffer, pH 7.5, 150 mM NaCl, 10 mM CaCl2 and 0.05% Brij35) containing a cocktail of proteinase inhibitors (Roche Diagnostics, GmbH). Supernatants of the homogenates were subjected to SDS-PAGE (10% total acrylamide) under reducing conditions, transferred onto polyvinylidene difluoride membranes and reacted with anti-ADAM15 antibody (240-2C7; 1 μg/ml) or non-immune mouse IgG (1 μg/ml) after blocking with 5% skim milk. They were then incubated with horseradish peroxidase-labeled anti-mouse IgG and immunoreactive bands were detected with ECL Western blotting reagents as described above.
Immunohistochemistry
Paraffin sections of the RA synovial samples were treated with 0.3% H2O2 and 10% normal goat serum to block endogenous peroxidase and non-specific binding, respectively. As antigen retrieval, the sections were subjected to microwave treatment at 500 W for 10 minutes in 10 mM citrate buffer, pH 6.0. They were then treated with mouse anti-ADAM15 antibody (240-2C7; 20 μg/ml), rabbit anti-VEGFR-2 antibody (Flk-1; 5 μg/ml; Santa Cruz Biotechnology, Santa Cruz, CA, USA), rabbit anti-von Willebrand factor (vWF; 15 μg/ml; DakoCytomation Norden A/S) or mouse anti-CD31 antibody (8 mg/ml; DakoCytomation Norden A/S). After the reaction with goat anti-mouse IgG or goat anti-rabbit IgG conjugated to peroxidase-labeled dextran polymer (no dilution; En Vison+ Mouse or En Vison+ Rabbit; DakoCytomation Norden A/S), the color was developed with 3,3'-diaminobenzidine tetrahydrochloride in 50 mM Tris-HCl buffer, pH 7.6, containing 0.006% H2O2. Counterstaining was performed with hematoxylin. As for a control, sections were reacted by replacing the first antibodies with non-immune mouse IgG or rabbit IgG.
Vascular density
Vascular density in RA and OA synovial tissues was evaluated by the morphometric analysis of RA and OA tissue sections immunostained with anti-CD31 antibody without any clinical information on each sample. Four fields were selected at random and vessels with a distinct lumen were counted to calculate the number of vessels per square millimeter as we described previously [14]. The average vascular density (vessels/mm2) from the fields for each patient was processed for further statistical analysis.
Cell cultures of rheumatoid arthritis synovial fibroblasts and endothelial cells
RA synovial fibroblasts (SFs) were prepared from RA synovial tissues obtained at total knee arthroplasty. The tissues were minced and incubated with 0.04% bacterial collagenase type I (Worthington Biochemical Corp., Freehold, NJ, USA). Isolated cells were seeded in culture dishes and maintained in DMEM (Life Technologies) supplemented with 10% fetal bovine serum (Life Technologies) at 37°C in humidified 5% CO2 in air. After the cells were cultured in confluence, they were trypsinized and reseeded in culture dishes. RA SFs at 5–9 passages were used for experiments. Human umbilical vein endothelial cells (HUVECs 7943; Cambrex Co., East Rutherford, NJ, USA) were grown in medium EBM-2 supplemented with EGM-2 (Cambrex Co.).
In order to examine the expression of VEGFR-2 and ADAM15 and exclude the possibility of contamination of cultured RA SF by endothelial cells, RA SFs at 5–9 passages were seeded on Lab-Tek II chamber slides (Nalge-Nunc International, Naperville, IL, USA) and subjected to immunohistochemistry for VEGFR-2, ADAM15, CD31 and vWF as described above. For a positive control, HUVECs were immunostained with these antibodies. In addition, mRNA expression of CD31 and vWF in cultured RA SFs was examined by RT-PCR using the PCR primers (Table 1).
Stimulation of RA synovial fibroblasts with proinflammatory cytokines and/or growth factors
RA SFs were plated on a 60 mm dish at a density of 3 × 105 cells/dish in DMEM supplemented with 10% fetal bovine serum. The culture media were replaced with serum-free DMEM containing 0.2% lactalbumin hydrolysate and starved for 24 h before they were used in experiments. The cells were treated with tumor necrosis factor-α (TNF-α; Dainippon Pharmatheutical, Osaka, Japan), IL-1α (Dainippon Pharmatheutical), transforming growth factor-β (TGF-β; R&D Systems, Minneapolis, MN, USA; 0, 0.1, 1 or 10 ng/ml) or recombinant VEGF165 (R&D Systems; 0, 1, 10 or 50 ng/ml) for 24 h. For co-stimulation of RA SFs with TNF-α and VEGF165, the cells were first starved with serum-free DMEM containing 0.2% lactalbumin hydrolysate for 24 h, treated with TNF-α (1 or 10 ng/ml) for 24 h, and then stimulated with VEGF165 (40 ng/ml) for 24 h. HUVECs were also stimulated with these cytokines or growth factors for 24 h after being starved with serum-free medium EBM-2 containing 1% bovine serum albumin for 24 h.
To block the signaling of VEGF165, RA SFs that had been stimulated with TNF-α (10 ng/ml) for 24 h were incubated with SU1498 (1 or 10 μM; Calbiochem, San Diego, CA, USA), a selective VEGFR-2 tyrosine kinase inhibitor [26,27], for 30 minutes and then treated with VEGF165 (40 ng/ml) for 24 h. HUVECs were also treated with SU1498 in a similar way except for no treatment with TNF-α. To exclude the possible involvement of VEGFR-1 in ADAM15 expression, RA SFs and HUVECs were stimulated with recombinant human placenta growth factor (PlGF; 1, 10 or 50 ng/ml; R&D Systems), which selectively binds to VEGFR-1 [28].
Statistics
Comparisons between two independent groups were determined by Mann-Whitney U test. Spearman's rank correlation was used for analysis of the relationship between relative ADAM15 mRNA expression and vascular density. P-values less than 0.05 were considered significant.
Results
Screening of mRNA expression of ADAMs and relative expression levels of ADAM15 in RA and OA synovial tissues
The mRNA expression of 10 different ADAMs (8, 9, 10, 12, 15, 17, 20, 21, 28 and 30) with a putative metalloproteinase motif was screened by RT-PCR analysis in eight RA and eight OA synovial samples. ADAM9, ADAM10 and ADAM17 were expressed in more than 88% of RA samples, but their expression was also observed in more than 75% of OA samples (Fig. 2). ADAM12 was expressed in 38% and 13% of RA and OA samples, respectively. ADAMs 8, 20, 21, 28 and 30 were expressed in less than 13% of both RA and OA samples. On the other hand, ADAM15 was intensely expressed in all the RA synovial samples, whereas it was detected in 63% of OA samples (Fig. 2). Because of the more selective expression of ADAM15 in RA than in OA, we focused on ADAM15 for further studies. When the expression was examined in a larger number of RA and OA samples, ADAM15 was detected in 100% of the RA samples (16 of 16 cases) and in 60% of the OA samples (12 of 20 cases) (data not shown). By real-time quantitative PCR analysis, the expression levels (ratio of ADAM15 to ribosomal 18S RNA) were significantly higher in RA samples (0.344 ± 0.276; n = 10) than in OA samples (0.091 ± 0.030; n = 10) (p < 0.01; Fig. 3).
Expression of ADAM15 in RA synovial tissues studied by in situ hybridization, immunohistochemistry and immunoblotting
Cells expressing ADAM15 mRNA were examined by in situ hybridization. Synovial lining cells, endothelial cells of blood vessels and macrophage-like cells in the sublining layer were labeled with the anti-sense RNA probe (Fig. 4a), whereas the sense probe gave only a background signal in these cells (Fig. 4b). Immunohistochemical studies showed that ADAM15 was localized to synovial lining cells, endothelial cells of blood vessels and macrophage-like cells in the sublining layer of the RA synovium (Fig. 5a,b), confirming the findings from in situ hybridization. No staining was obtained with non-immune IgG (Fig. 5f). VEGFR-2 was immunolocalized to some synovial lining cells and endothelial cells of blood vessels in RA samples (Fig. 5c), but vWF and CD31 were almost exclusively localized to endothelial cells (Fig. 5d,e).
When homogenates of RA synovial tissues were subjected to immunoblotting analysis, eight major immunoreactive bands of 100, 76, 66, 47, 41, 38, 34 and 29 kDa were observed (Fig. 6, lane 2). Because the molecular weight of the 100-kDa band is similar to that of the precursor form of ADAM15 [12,25], this band appears to correspond to proADAM15. On the other hand, at least some of the other bands are considered to be active ADAM15 forms containing the metalloproteinase domain because of their positive immunoreactivity to the antibody specific to the cysteine-rich domain and their molecular weights. An immunoreactive band of 58 kDa was a non-specific reaction, because it was also detected with non-immune mouse IgG (Fig. 6, lane 1).
Correlation between ADAM15 expression and vascular density in RA and OA synovial tissues
No definite correlation between relative mRNA expression levels of ADAM15 and the separate or total histological scores of RA and OA synovial tissues was observed (data not shown). Thus, we further evaluated vascular density in the RA and OA synovial tissues, by counting CD31-positive vessels in synovial tissue sections, and compared it with the mRNA expression levels of ADAM15. A linear correlation was found between expression levels and vascular density in RA and OA synovial tissues (r = 0.907, p < 0.001; n = 20; Fig. 7).
Effect of cytokines and growth factors on ADAM15 expression in RA SFs and HUVECs
When the expression of ADAM15 in RA SFs was examined by RT-PCR, it was constitutively expressed regardless of the passage numbers (5 to 9) of the cells (Fig. 8a). Expression was decreased to low levels in a time-dependent manner, however, after starvation with serum-free medium for up to 72 h (Fig. 8b). To test the effect of cytokines and growth factors on ADAM15 expression, RA SFs were stimulated with TNF-α, IL-1α, TGF-β or VEGF165; however, no changes in mRNA expression were found with these factors (Fig. 8c,d). In contrast, VEGF165 appeared to selectively enhance ADAM15 expression in HUVECs (Fig. 8d), whereas TNF-α, IL-1α or TGF-β did not alter the expression (data not shown). Using real-time quantitative PCR, we found that the relative expression levels of ADAM15 mRNA (ratio of ADAM15 to ribosomal 18S) in HUVECs are significantly 2.2-fold higher after treatment with VEGF165 (p < 0.05).
Regulation of VEGFR-1, VEGFR-2 and neuropilin-1 expression by cytokines and growth factors in RA SFs and HUVECs
As previously reported [15,29], HUVECs expressed the three major VEGF receptors (VEGFR-1, VEGFR-2 and neuropilin-1), but RA SFs expressed only neuropilin-1 under unstimulated conditions (Fig. 9a). Because the data that VEGF165 stimulated ADAM15 expression only in HUVECs suggested that the effect is dependent on the expression of VEGF receptors, we tried to induce VEGF receptors by treating RA SFs with cytokines and growth factor and found that TNF-α, but not IL-1α or TGF-β, can induce VEGFR-2 expression without affecting the expression of VEGFR-1 or neuropilin-1 (Fig. 9b).
Stimulation of ADAM15 expression by VEGF165 in VEGFR-2-expressing RA SFs
As TNF-α induced VEGFR-2 expression in RA SFs, we further examined whether VEGF165 enhances ADAM15 expression in VEGFR-2-expressing RA SFs. After stimulation of RA SFs with TNF-α or VEGF165 alone, ADAM15 mRNA expression, which was only weak after starvation of the cells, did not change (Fig. 10a). When the cells were sequentially treated with TNF-α and VEGF165, however, the level of ADAM15 expression appeared to be increased (Fig. 10a). Real-time quantitative PCR analysis demonstrated that the expression levels are significantly 2.2-fold higher in RA SFs treated with TNF-α and VEGF165 compared with the control without treatment (p < 0.05).
VEGFR-2 signaling in VEGF165-stimulated ADAM15 expression in RA SFs
To examine the involvement of VEGFR-2 signaling in the stimulation of ADAM15 expression with TNF-α and VEGF165, RA SFs were incubated with SU1498, a selective VEGFR-2 tyrosine kinase inhibitor, prior to the stimulation with VEGF165 and ADAM15 mRNA expression was examined. ADAM15 expression decreased with 1 μM SU1498 and was completely suppressed with 10 μM SU1498, while VEGFR-2 expression was not affected by the treatment with such concentrations of the inhibitor (Fig. 10b). The enhanced expression of ADAM15 in HUVECs was also inhibited by the treatment with SU1498 (data not shown). PlGF, which selectively binds to VEGFR-1, did not affect the mRNA expression of ADAM15 in either RA SFs or HUVECs (data not shown).
Immunohistochemical demonstration of ADAM15 expression and VEGFR-2 induction by TNF-α in RA SFs
Protein expression of ADAM15 and endothelial cell markers in cultured RA SFs was examined by immunohistochemistry. ADAM15 was immunolocalized to RA SFs and HUVECs (Fig. 11a,e), whereas no staining was observed with non-immune IgG (Fig. 11d,h). On the other hand, although VEGFR-2, vWF and CD31 were all immunostained in HUVECs (VEGFR-2 and vWF, Fig. 11f,g; CD31, data not shown), RA SFs were negative for these endothelial cell markers (vWF, Fig. 11c; VEGFR-2 and CD31, data not shown). When RA SFs were treated with TNF-α for 24 h, however, they were positively immunostained with anti-VEGFR-2 antibody (Fig. 11b). In accordance with the immunohistochemical data, the mRNA expression of CD31 and vWF in untreated RA SFs was not detected by RT-PCR (data not shown).
Discussion
In the present study, we have demonstrated that among the 10 different ADAM species with the putative metalloproteinase motif, ADAM15 is more frequently and intensely expressed in RA synovium than in OA synovium. The mRNA expression patterns of the ADAM species in synovial tissues could be classified into three groups: constitutive expression in both RA and OA samples (ADAM9, ADAM10 and ADAM17); negligible or no expression in RA or OA (ADAM8, ADAM20, ADAM21, ADAM28 and ADAM30); and more selective expression in RA than in OA (ADAM12 and ADAM15). When the expression patterns were compared with those in human astrocytic tumor and normal brain tissues [9], they were different in that more ADAM species, including ADAM9, ADAM10, ADAM15, ADAM17, ADAM20, ADAM21 and ADAM28, are constitutively expressed in brain tumor and normal brain tissues, but similar in that the expression of ADAM8 and ADAM30 is negligible. ADAM12 was selectively overexpressed in the highly malignant glioblastomas and appeared to play a key role in the tumor cell proliferation through shedding of heparin-binding epidermal growth factor [9]. This was not the case in RA synovium, however, because ADAM12 expression was confined to less than 40% of the RA samples and the expression level did not correlate with synovial lining cell hyperplasia.
A study by Bohm and co-workers [12] described the expression of ADAM15 in RA and OA synovial tissues by immunohistochemistry and in situ hybridization, but their study did not quantitatively analyze the expression levels. The present study has provided the first evidence that the mRNA expression level of ADAM15 is significantly 3.8-fold higher in RA than in OA. Our data of in situ hybridization and immunohistochemistry in RA synovium demonstrated that synovial lining cells, endothelial cells of blood vessels and macrophage-like cells in the sublining layer are responsible for the expression of ADAM15. The finding confirms the previous observation that synovial lining cells and macrophage-like cells express ADAM15 [12], but further indicate that endothelial cells, which are positive for CD31 and vWF, express ADAM15 in RA synovium. The expression by RA synovial lining cells and endothelial cells was also supported by our immunohistochemical data with cultured RA SFs and HUVECs. Interestingly, several ADAM15 species with molecular weights ranging from 100 kDa to 29 kDa were immunoblotted with RA synovial tissues. The data suggest that proADAM15 is susceptible to proteolytic cleavages and processed into fragments including active forms in RA synovial tissues.
One of the important findings in the present study is that VEGF165 up-regulates the expression of ADAM15. Overexpression of ADAM species is known in tumor cell lines, for example, ADAM10, ADAM12 and ADAM15 in hematological malignant tumor cell lines [30], and ADAM9, ADAM10, ADAM15 and ADAM17 in prostate cancer cell lines [31]. Although dihydrotestosterone modulates the gene expression of ADAM9, ADAM10 and ADAM17 in an androgen-dependent cell line [32], the expression of the ADAM species in these tumor cell lines may result mainly from the gene regulation associated with transformation of the cells. Phorbol 12-myristate 13-acetate stimulates human monocyte-like cell line THP-1 to enhance expression of ADAM8, ADAM9 and ADAM17, but decreases ADAM15 expression [33]. In addition, platelet-derived growth factor stimulates human vascular smooth muscle cells to overexpress ADAM9 and ADAM15 [34]. In human OA chondrocytes, IL-1 and/or retinoic acid up-regulate ADAM15 and ADAM17, but down-regulate ADAM9 [35]. In the present study, however, we have shown that VEGF165, but not IL-1α, TNF-α or TGF-β, stimulates HUVECs to enhance the gene expression of ADAM15. In addition, our data indicate that TNF-α induces VEGFR-2 expression in RA SFs, and VEGF165 up-regulates ADAM15 expression in the TNF-α-stimulated RA SFs. The specific involvement of VEGFR-2 signaling in the enhanced ADAM15 gene expression was demonstrated by the findings that the stimulation was blocked with SU1498, an inhibitor of VEGFR-2, and that PlGF, which selectively binds to VEGFR-1, had no effect on ADAM15 expression. Thus, these results demonstrate for the first time that VEGF165 is a stimulator of ADAM15 gene expression in cells that express VEGFR-2. Although VEGFR-2 expression was originally thought to be specific to endothelial cells, accumulated evidence indicates that many cell types, such as hematopoietic stem cells [36], megakaryocytes [36], retinal progenitor cells [37] and OA chondrocytes [19], express the receptor. Because RA synovial fluids contain high amounts of TNF-α and VEGF165 [38,39], it is reasonable to think that synovial lining cells are co-stimulated with these factors to induce VEGFR-2 and ADAM15 in RA synovial tissue. In fact, VEGFR-2 and ADAM15 were immunolocalized to RA synovial lining cells in the present study, suggesting the validity of the hypothesis.
The present study has demonstrated that the expression levels of ADAM15 directly correlate with the vascular density of synovial tissues. This suggests the possible involvement of ADAM15 in the angiogenic steps of RA synovial tissues, which include sprouting, migration and proliferation of endothelial cells, and maturation of vessels [40]. The notion that ADAM15 may be implicated in pathological angiogenesis is supported by recent experimental data: ADAM15 is up-regulated on the angiogenic endothelial cell surface [41]; ADAM15 has proteinase activity to digest gelatin and type IV collagen [42], which is essential for endothelial cells to sprout and migrate; ADAM15 interacts with αvβ3 and α5β1 integrins [43], which promote cell migration and proliferation of endothelial cells and smooth muscle cells [44,45]; inhibition of ADAM15's functions by antibodies, antisense oligonucleotide and metalloproteinase inhibitor decreases cell migration [42]; ADAM15-deficient mice show reduced neovascularization in hypoxia-induced proliferative retinopathy compared to wild-type control mice [8]; and the recombinant human disintegrin domain of ADAM15 itself exhibits anti-angiogenic activity, probably by disturbing the interaction between ADAM15 and integrins, leading to inhibition of endothelial cell migration and proliferation [46]. In addition, Ham et al. [47] have reported that ADAM15 is co-localized with vascular endothelial cadherin in the adherens junctions of endothelial cells and this cadherin drives ADAM15 to the cell junctions. Moreover, overexpression of ADAM15 is known to enhance cell-cell interactions [48]. VEGF is highly expressed in RA synovial tissues and believed to play a central role in synovial angiogenesis [49], and it stimulates ADAM15 expression in endothelial cells as shown in the present study. Altogether, these data suggest the possibility that ADAM15 may play a role in maturation of the blood vessels newly formed by VEGF-induced angiogenesis through enhancing endothelial cell-cell interaction in the RA synovium.
The biological significance of the overexpression of ADAM15 in RA synovial lining cells and macrophage-like cells is not clear at present. The fact that RA synovial lining cells express α5β1 and αvβ3 integrins [50-52], however, suggests the possibility that ADAM15 may function as a binding molecule to reinforce cell-cell and/or cell-ECM interactions. On the other hand, because ADAM15 is reported to have proteolytic activities towards ECM components (gelatin and type IV collagen) [42], it is possible that ADAM15 may be involved in the degradation of cartilage ECM at the synovial cartilage junction. In addition, ADAM15 is known to produce soluble forms of CD23 through ectodomain shedding of membrane-bound CD23 on cell membranes of inflammatory cells [53]. Soluble forms of CD23, which are elevated in synovial fluid and sera of RA patients [54-56], promote the production of inflammatory cytokines by macrophages [57] and stimulate monocytes and T cells [58]. Thus, these data suggest that ADAM15 may play a role in the aggravation of inflammation of RA synovitis through production of soluble forms of CD23, which trigger inflammatory process via monokine release and stimulate monocytes and T cells [59,60]. In contrast to the prospective role of ADAM15 in joint destruction, a recent study has demonstrated that aging mice with a targeted disruption of ADAM15 exhibit accelerated development of OA changes in the articular cartilage, and suggested that ADAM15 expressed by chondrocytes has a homeostatic, rather than a destructive, role in the cartilage [61]. Thus, further studies are definitely needed to elucidate the exact role of this unique and multifunctional molecule in RA and OA.
Conclusion
Our results demonstrate that ADAM15 is overexpressed in lining cells, endothelial cells of blood vessels and macrophage-like cells in the sublining layer of RA synovium, with a direct correlation with vascular density in the synovium, and that the expression of ADAM15 in RA SFs is up-regulated by the action of VEGF165 via VEGFR-2. These data suggest that ADAM15 is involved in angiogenesis in the RA synovium.
Abbreviations
ADAM = a disintegrin and metalloproteinase; DMEM = Dulbecco's modified Eagle's medium; ECM = extracellular matrix; HUVEC = human umbilical vein endothelial cells; IL = interleukin; MMP = matrix metalloproteinase; OA = osteoarthritis; PlGF = placenta growth factor; RA = rheumatoid arthritis; RT-PCR = reverse transcription polymerase chain reaction; SF = synovial fibroblast; TGF = transforming growth factor; TNF = tumor necrosis factor; VEGF = vascular endothelial growth factor; VEGFR = vascular endothelial growth factor receptor; vWF = von Willebrand factor.
Competing interests
The authors declare that they have no competing interests.
Authors' contributions
KK carried out critical examinations in this study, especially histopathological analyses, RT-PCR, real-time PCR, in situ hybridization, characterization of monoclonal anti-ADAM15 antibody, immunohistochemistry, immunoblotting, cell cultures, and stimulation of cells, and drafted the manuscript as a part of his doctoral thesis, with the assistance of the coauthors. HE and SO prepared histological specimens and carried out RNA extraction. II participated in the cultures and stimulation of cells. YF participated in the in situ hybridization. EI participated in the vascular density analyses. EO carried out the development of the monoclonal antibody against human ADAM15. YT gave critical suggestions concerning orthopedics and experimental design. HM carried out the clinical studies of each case and performed surgery to obtain synovial tissues with the written informed consent of patients. YO conceived of the study, participated in its design and coordination, and is the corresponding author. All authors read and approved the final manuscript.
Acknowledgements
We are grateful to Drs T Otani, E Nomura, Y Suda, M Kurimura and T Toyoda for providing us with synovial samples. We also thank Miss M Uchiyama for her technical assistance.
Figures and Tables
Figure 1 Characterization of monoclonal antibody against human ADAM15. Lysates of Escherichia coli transfected with the expression vector FLAG-ADAM15 containing a cDNA fragment encoding the cysteine-rich domain of human ADAM15 (lanes 1, 3, 5 and 6) or the pFLAG-CTC vector alone (lanes 2 and 4) were immunoblotted with anti-FLAG antibody (lanes 1 and 2) or anti-ADAM15 antibody (240-2C7) (lanes 3-6) as described in Materials and methods. The absorption study was carried out by incubation of the antibody with 1000-fold excess amounts of the peptide for 16 h at 4°C (lane 6). The arrow indicates the protein band of the cysteine-rich domain reactive with anti-ADAM15 antibody. Note that no staining is present with mock transfectants (lanes 2 and 4), and that the immunoreactive band with anti-ADAM15 antibody (lane 5) completely disappears after reaction with the antibody absorbed with the peptide (lane 6).
Figure 2 mRNA expression of ADAM species in rheumatoid arthritis (RA) and osteoarthritis (OA) synovial samples. Total RNA was extracted from eight RA (lanes 1-8) or eight OA (lanes 9-16) synovial samples, and reverse-transcribed into cDNA, followed by PCR as described in Materials and methods. C, positive controls.
Figure 3 The mRNA expression levels of ADAM15 in rheumatoid arthritis (RA) or osteoarthritis (OA) synovial samples. The relative expression levels (ADAM15:ribosomal 18S ratios) were determined by real-time PCR analysis as described in Materials and methods. Bars indicate the mean values of the 10 samples of RA and OA synovial tissues. Asterisk indicates p < 0.01.
Figure 4 In situ hybridization of ADAM15 in rheumatoid arthritis synovial tissues. Paraffin sections were reacted with digoxigenin-labeled anti-sense or sense RNA probes as described in Materials and methods. Note (a) a positive signal in the synovial lining cells (arrows), endothelial cells (black arrowhead) and macrophage-like cells (white arrowhead) with anti-sense probe, whereas (b) there was only a background signal with the sense probe. Scale bar, 50 μm.
Figure 5 Immunolocalization of ADAM15, VEGFR-2, vWF and CD31 in rheumatoid arthritis (RA) synovial tissues. Paraffin sections were stained with (a) hematoxylin and eosin or immunostained with antibodies against (b) ADAM15, (c) VEGFR-2, (d) vWF or (e) CD31, or (f) non-immune mouse IgG as described in Materials and methods. (b) Note that ADAM15 is expressed in synovial lining cells and endothelial cells of blood vessels in the sublining layer. Immunostained sections were counterstained with hematoxylin. Arrows, synovial lining cells; arrowheads, endothelial cells of blood vessel. Scale bar, 100 μm.
Figure 6 Immunoblotting of ADAM15 in rheumatoid arthritis (RA) synovial tissues. Homogenates of RA synovial tissues were prepared and subjected to immunoblotting with anti-ADAM15 antibody specific to the cysteine-rich domain of ADAM15 (240-2C7) (lane 2) or non-immune mouse IgG (lane 1) as described in Materials and methods. Immunoreactive bands of 100, 76, 66, 47, 41, 38, 34 and 29 kDa are indicated (arrow heads). The 58 kDa protein band is considered to be a non-specific band because it is also detected with non-immune IgG (lanes 1 and 2).
Figure 7 Correlation between ADAM15 mRNA expression levels and vascular density in synovial tissues. Vascular density was determined by the morphometric analysis of the CD31-immunostained sections and correlation was examined by Spearman's rank correlation. Note a direct correlation between the parameters (r = 0.907, p < 0.001; n = 20). Open and closed circles indicate osteoarthritis (OA) and rheumatoid arthritis (RA) synovial samples, respectively.
Figure 8 Effects of passages, starvation, cytokines and growth factors on ADAM15 expression in cultured cells. The mRNA expression of ADAM15 was examined by RT-PCR at 25 cycles as described in Materials and methods. (a) Effect of passages on the mRNA expression of ADAM15 in rheumatoid arthritis (RA) synovial fibroblasts (SFs). Lanes 1 to 5 indicate passages 5, 6, 7, 8 and 9 of RA SFs. (b) Effect of starvation on the mRNA expression of ADAM15 in RA SFs. (c) Effect of tumor necrosis factor (TNF)-α (0, 0.1, 1 and 10 ng/ml), IL-1α (0, 0.1, 1 and 10 ng/ml) or transforming growth factor (TGF)-β (0, 0.1, 1 and 10 ng/ml) on the mRNA expression of ADAM15 in RA SFs after stimulation with these factors for 24 h. (d) Regulation of the mRNA expression of ADAM15 by vascular endothelial growth factor (VEGF)165 in RA SFs and human umbilical vein endothelial cells (HUVECs). Cells were stimulated with VEGF165 (0, 1, 10 and 50 ng/ml) for 24 h. Note that VEGF165 enhances the expression of ADAM15 only in HUVECs.
Figure 9 Effects of cytokines and growth factors on expression of vascular endothelial growth factor receptors (VEGFRs). The mRNA expression of the VEGFRs was examined by RT-PCR at 30 cycles as described in Materials and methods. (a) The expression of VEGFR-1, VEGFR-2 and neuropilin-1 in rheumatoid arthritis (RA) synovial fibroblasts (SFs) of different passages and human umbilical vein endothelial cells (HUVECs). Lanes 1 to 5 correspond to RA SFs of passages 5, 6, 7, 8 and 9, respectively. (b) The mRNA expression of VEGFR-1, VEGFR-2 and neuropilin-1 in RA SFs after 24 h stimulation with tumor necrosis factor (TNF)-α (0, 0.1, 1, 10 and 50 ng/ml), IL-1α (0, 0.1, 1, 10 and 50 ng/ml) or transforming growth factor (TGF)-β (0, 0.01, 0.1, 1 and 10 ng/ml). Note that TNF-α selectively induces the mRNA expression of VEGFR-2 in RA SFs.
Figure 10 Effects of tumor necrosis factor-α, vascular endothelial growth factor165 and SU1498 on ADAM15 expression. (a) Effect of tumor necrosis factor (TNF)-α and vascular endothelial growth factor (VEGF)165 on ADAM15 expression. After starvation, rheumatoid arthritis (RA) synovial fibroblasts (SFs) were treated with 10 ng/ml TNF-α and/or 40 ng/ml VEGF165 for 24 h, and then the expression of ADAM15, vascular endothelial growth factor receptor (VEGFR)-1, VEGFR-2 and neuropilin-1 was examined by RT-PCR at 25 cycles (ADAM15 and β-actin) or 30 cycles (VEGFR-1, VEGFR-2 and neuropilin-1) as described in Materials and methods. Note that treatment with TNF-α induces VEGFR-2 and co-treatment with TNF-α and VEGF165 stimulates the mRNA expression of ADAM15 in RA SFs. (b) Effect of SU1498 on ADAM15 expression. After starvation for 24 h, RA SFs were stimulated with 10 ng/ml TNF-α for 24 h to induce VEGFR-2. Then, the cells were treated with SU1498 (0, 1 and 10 μM) for 30 minutes and stimulated with 40 ng/ml VEGF165 for 24 h. The mRNA expression of ADAM15 was determined by RT-PCR at 25 cycles as described in Materials and methods. Note that the stimulated expression of ADAM15 mRNA is inhibited by the treatment with SU1498, while the expression of VEGFR-2 mRNA is not affected by the treatment.
Figure 11 Immunohistochemistry of ADAM15, vascular endothelial growth factor receptor (VEGFR)-2 and von Willebrand factor (vWF). (a-d) Rheumatoid arthritis (RA) synovial fibroblasts (SFs) and (e-h) human umbilical vein endothelial cells (HUVEC) were cultured on Lab-Tek II chamber slides and immunostained with antibodies against (a,e) ADAM15, (b,f) VEGFR-2 or (c,g) vWF or (d,h) non-immune mouse IgG as described in Materials and methods. Immunostaining of VEGFR-2 in RA SFs (b) was performed with RA SFs that were treated with 10 ng/ml TNF-α for 24 h prior to immunohistochemistry. Note that vWF is not immunostained in RA SFs (c), but VEGFR-2 is expressed in those stimulated with TNF-α (b). Scale bar, 25 μm.
Table 1 Sequences of the primers used for RT-PCR
Primer's name Product Oligonucleotide sequence Size (base pairs) Accession number
ADAM8 Forward 5'-GCCGTCTTCAGGCCTCGGCCCGGGGACTCT-3' 651 NM001109
Reverse 5'-AGGGGCGTTGGCGAGGCACACCGACTGCGG-3'
ADAM9 Forward 5'-GCTGTCTTGCCACAGACCCGGTATGTGGAG-3' 604 HSU41766
Reverse 5'-TGGAATATTAAGAAGGCAGTTTCCTCCTTT-3'
ADAM10 Forward 5'-ATCCAGTCATGTTAAAGCGATTGATACAATTTAC-3' 434 NM001110
Reverse 5'-TCCAAAGTTATGTCCAACTTCGTGAGCAAAAGTAA-3'
ADAM12 Forward 5'-GAGACCCTCAAGGCAACTAAGTATGTGGAG-3' 627 AF023476
Reverse 5'-CGGCAGGTTAAACAGGCACACCCCCATTCC-3'
ADAM15 Forward 5'-CTGGGACAGCGCCACATTCGCCGGAGGCGG-3' 688 HSU41767
Reverse 5'-TCCGCAGAAAGCAGCCATAGGGGGTAGGCT-3'
ADAM17 Forward 5'-AGAGCTGACCCAGATCCCATGAAGAACACG-3' 777 HSU69611
Reverse 5'-GCGTTCTTGAAAACACTCCTGGGCCTTACT-3'
ADAM20 Forward 5'-AAAATAGCACACCAGATGGAGTTGCAATTG-3' 702 AF029899
Reverse 5'-ATTCCCACAGTACTTCAGTCTAAATATATT-3'
ADAM21 Forward 5'-TCTGGCTTGGGGTATTTTTG-3' 500 AF158644
Reverse 5'-TTGGCGTGCTACTTCCTTCT-3'
ADAM28 Forward 5'-GCTGTGATGCTAAGACATGT-3' 871 AF137334
Reverse 5'-TGAACAGCCTTTACCATCTG-3'
ADAM30 Forward 5'-AACCAGGTGCCAACTGTAGC-3' 496 AF171932
Reverse 5'-CCCATGGGTTTCATGGATAG-3'
VEGFR-1 Forward 5'-GATGTTGAGGAAGAGGAGGATT-3' 1146 NM002019
Reverse 5'-AAGCTAGTTTCCTGGGGGTATA-3'
VEGFR-2 Forward 5'-GATGTGGTTCTGAGTCCGTCT-3' 562 NM002253
Reverse 5'-CATGGCTCTGCTTCTCCTTTG-3'
Neuropilin-1 Forward 5'-CAACGATAAATGTGGCGATACT-3' 824 NM003873
Reverse 5'-TATACTGGGAAGAAGCTGTGAT-3'
CD31 Forward 5'-CAACGAGAAAATGTCAGA-3' 259 NM000442
Reverse 5'-GGAGCCTTCCGTTCTAGAGT-3'
vWF Forward 5'-GTTCGTCCTGGAAGGATCGG-3' 697 NM000552
Reverse 5'-CACTGACACCTGAGTGAGAC-3'
β-Actin Forward 5'-TGACGGGGTCACCCACACTGTGCCCATCTA-3' 661 NM001101
Reverse 5'-CTAGAAGCATTTGCGGTGGACGATGGAGGG-3'
ADAM, a disintegrin and metalloproteinase; VEGFR, vascular endothelial growth factor receptor; vWF, von Willebrand factor.
==== Refs
Okada Y Harris ED Jr BR, Genovese MC, Firestein GS, Sargent JS Proteinases and matrix degradation Kelley's Textbook of Rheumatology 2005 7 Philadelphia: Elsevier Saunders 63 81
Primakoff P Myles DG The ADAM gene family: surface proteins with adhesion and protease activity Trends Genet 2000 16 83 87 10652535 10.1016/S0168-9525(99)01926-5
Seals DF Courtneidge SA The ADAMs family of metalloproteases: multidomain proteins with multiple functions Genes Dev 2003 17 7 30 12514095 10.1101/gad.1039703
Cho C Bunch DO Faure JE Goulding EH Eddy EM Primakoff P Myles DG Fertilization defects in sperm from mice lacking fertilin beta Science 1998 281 1857 1859 9743500 10.1126/science.281.5384.1857
Nishimura H Cho C Branciforte DR Myles DG Primakoff P Analysis of loss of adhesive function in sperm lacking cyritestin or fertilin beta Dev Biol 2001 233 204 213 11319869 10.1006/dbio.2001.0166
Alfandari D Cousin H Gaultier A Smith K White JM Darribere T DeSimone DW Xenopus ADAM 13 is a metalloprotease required for cranial neural crest-cell migration Curr Biol 2001 11 918 930 11448768 10.1016/S0960-9822(01)00263-9
Rooke J Pan D Xu T Rubin GM KUZ, a conserved metalloprotease-disintegrin protein with two roles in Drosophila neurogenesis Science 1996 273 1227 1231 8703057
Horiuchi K Weskamp G Lum L Hammes HP Cai H Brodie TA Ludwig T Chiusaroli R Baron R Preissner KT Potential role for ADAM15 in pathological neovascularization in mice Mol Cell Biol 2003 23 5614 5624 12897135 10.1128/MCB.23.16.5614-5624.2003
Kodama T Ikeda E Okada A Ohtsuka T Shimoda M Shiomi T Yoshida K Nakada M Ohuchi E Okada Y ADAM12 is selectively overexpressed in human glioblastomas and is associated with glioblastoma cell proliferation and shedding of heparin-binding epidermal growth factor Am J Pathol 2004 165 1743 1753 15509542
Chubinskaya S Cs-Szabo G Kuettner KE ADAM-10 message is expressed in human articular cartilage J Histochem Cytochem 1998 46 723 729 9603783
Bohm BB Aigner T Gehrsitz A Blobel CP Kalden JR Burkhardt H Up-regulation of MDC15 (metargidin) messenger RNA in human osteoarthritic cartilage Arthritis Rheum 1999 42 1946 1950 10513811 10.1002/1529-0131(199909)42:9<1946::AID-ANR21>3.0.CO;2-E
Bohm BB Aigner T Blobel CP Kalden JR Burkhardt H Highly enhanced expression of the disintegrin metalloproteinase MDC15 (metargidin) in rheumatoid synovial tissue Arthritis Rheum 2001 44 2046 2054 11592366 10.1002/1529-0131(200109)44:9<2046::AID-ART354>3.0.CO;2-3
Koch AE Review: angiogenesis: implications for rheumatoid arthritis Arthritis Rheum 1998 41 951 962 9627005 10.1002/1529-0131(199806)41:6<951::AID-ART2>3.0.CO;2-D
Ikeda M Hosoda Y Hirose S Okada Y Ikeda E Expression of vascular endothelial growth factor isoforms and their receptors Flt-1, KDR, and neuropilin-1 in synovial tissues of rheumatoid arthritis J Pathol 2000 191 426 433 10918218 10.1002/1096-9896(2000)9999:9999<::AID-PATH649>3.0.CO;2-E
Soker S Takashima S Miao HQ Neufeld G Klagsbrun M Neuropilin-1 is expressed by endothelial and tumor cells as an isoform-specific receptor for vascular endothelial growth factor Cell 1998 92 735 745 9529250 10.1016/S0092-8674(00)81402-6
Zucker S Mirza H Conner CE Lorenz AF Drews MH Bahou WF Jesty J Vascular endothelial growth factor induces tissue factor and matrix metalloproteinase production in endothelial cells: conversion of prothrombin to thrombin results in progelatinase A activation and cell proliferation Int J Cancer 1998 75 780 786 9495249 10.1002/(SICI)1097-0215(19980302)75:5<780::AID-IJC19>3.0.CO;2-A
Unemori EN Ferrara N Bauer EA Amento EP Vascular endothelial growth factor induces interstitial collagenase expression in human endothelial cells J Cell Physiol 1992 153 557 562 1447317 10.1002/jcp.1041530317
Lamoreaux WJ Fitzgerald ME Reiner A Hasty KA Charles ST Vascular endothelial growth factor increases release of gelatinase A and decreases release of tissue inhibitor of metalloproteinases by microvascular endothelial cells in vitro Microvasc Res 1998 55 29 42 9473407 10.1006/mvre.1997.2056
Enomoto H Inoki I Komiya K Shiomi T Ikeda E Obata K Matsumoto H Toyama Y Okada Y Vascular endothelial growth factor isoforms and their receptors are expressed in human osteoarthritic cartilage Am J Pathol 2003 162 171 181 12507900
Arnett FC Edworthy SM Bloch DA McShane DJ Fries JF Cooper NS Healey LA Kaplan SR Liang MH Luthra HS The American Rheumatism Association 1987 revised criteria for the classification of rheumatoid arthritis Arthritis Rheum 1988 31 315 324 3358796
Altman R Asch E Bloch D Bole G Borenstein D Brandt K Christy W Cooke TD Greenwald R Hochberg M Development of criteria for the classification and reporting of osteoarthritis. Classification of osteoarthritis of the knee. Diagnostic and Therapeutic Criteria Committee of the American Rheumatism Association Arthritis Rheum 1986 29 1039 1049 3741515
Okada Y Shinmei M Tanaka O Naka K Kimura A Nakanishi I Bayliss MT Iwata K Nagase H Localization of matrix metalloproteinase 3 (stromelysin) in osteoarthritic cartilage and synovium Lab Invest 1992 66 680 690 1602738
National Center for Biotechnology Information
Ueno H Nakamura H Inoue M Imai K Noguchi M Sato H Seiki M Okada Y Expression and tissue localization of membrane-types 1, 2, and 3 matrix metalloproteinases in human invasive breast carcinomas Cancer Res 1997 57 2055 2060 9158005
Kratzschmar J Lum L Blobel CP Metargidin, a membrane-anchored metalloprotease-disintegrin protein with an RGD integrin binding sequence J Biol Chem 1996 271 4593 4596 8617717 10.1074/jbc.271.9.4593
Shen BQ Lee DY Zioncheck TF Vascular endothelial growth factor governs endothelial nitric-oxide synthase expression via a KDR/Flk-1 receptor and a protein kinase C signaling pathway J Biol Chem 1999 274 33057 33063 10551875 10.1074/jbc.274.46.33057
Boguslawski G McGlynn PW Harvey KA Kovala AT SU1498, an inhibitor of vascular endothelial growth factor receptor 2, causes accumulation of phosphorylated ERK kinases and inhibits their activity in vivo and in vitro J Biol Chem 2004 279 5716 5724 14625306 10.1074/jbc.M308625200
Park JE Chen HH Winer J Houck KA Ferrara N Placenta growth factor. Potentiation of vascular endothelial growth factor bioactivity, in vitro and in vivo, and high affinity binding to Flt-1 but not to Flk-1/KDR J Biol Chem 1994 269 25646 25654 7929268
Barleon B Hauser S Schollmann C Weindel K Marme D Yayon A Weich HA Differential expression of the two VEGF receptors flt and KDR in placenta and vascular endothelial cells J Cell Biochem 1994 54 56 66 8126087 10.1002/jcb.240540107
Wu E Croucher PI McKie N Expression of members of the novel membrane linked metalloproteinase family ADAM in cells derived from a range of haematological malignancies Biochem Biophys Res Commun 1997 235 437 442 9199213 10.1006/bbrc.1997.6714
McCulloch DR Harvey M Herington AC The expression of the ADAMs proteases in prostate cancer cell lines and their regulation by dihydrotestosterone Mol Cell Endocrinol 2000 167 11 21 11000516 10.1016/S0303-7207(00)00305-1
Karan D Lin FC Bryan M Ringel J Moniaux N Lin MF Batra SK Expression of ADAMs (a disintegrin and metalloproteases) and TIMP-3 (tissue inhibitor of metalloproteinase-3) in human prostatic adenocarcinomas Int J Oncol 2003 23 1365 1371 14532978
Worley JR Baugh MD Hughes DA Edwards DR Hogan A Sampson MJ Gavrilovic J Metalloproteinase expression in PMA-stimulated THP-1 cells. Effects of peroxisome proliferator-activated receptor-gamma (PPAR gamma) agonists and 9-cis-retinoic acid J Biol Chem 2003 278 51340 51346 14534304 10.1074/jbc.M310865200
Al-Fakhri N Wilhelm J Hahn M Heidt M Hehrlein FW Endisch AM Hupp T Cherian SM Bobryshev YV Lord RS Katz N Increased expression of disintegrin-metalloproteinases ADAM-15 and ADAM-9 following upregulation of integrins alpha5beta1 and alphavbeta3 in atherosclerosis J Cell Biochem 2003 89 808 823 12858346 10.1002/jcb.10550
Flannery CR Little CB Caterson B Hughes CE Effects of culture conditions and exposure to catabolic stimulators (IL-1 and retinoic acid) on the expression of matrix metalloproteinases (MMPs) and disintegrin metalloproteinases (ADAMs) by articular cartilage chondrocytes Matrix Biol 1999 18 225 237 10429942 10.1016/S0945-053X(99)00024-4
Katoh O Tauchi H Kawaishi K Kimura A Satow Y Expression of the vascular endothelial growth factor (VEGF) receptor gene, KDR, in hematopoietic cells and inhibitory effect of VEGF on apoptotic cell death caused by ionizing radiation Cancer Res 1995 55 5687 5692 7585655
Yang K Cepko CL Flk-1, a receptor for vascular endothelial growth factor (VEGF), is expressed by retinal progenitor cells J Neurosci 1996 16 6089 6099 8815891
Koch AE Harlow LA Haines GK Amento EP Unemori EN Wong WL Pope RM Ferrara N Vascular endothelial growth factor. A cytokine modulating endothelial function in rheumatoid arthritis J Immunol 1994 152 4149 4156 7511670
Fava RA Olsen NJ Spencer-Green G Yeo KT Yeo TK Berse B Jackman RW Senger DR Dvorak HF Brown LF Vascular permeability factor/endothelial growth factor (VPF/VEGF): accumulation and expression in human synovial fluids and rheumatoid synovial tissue J Exp Med 1994 180 341 346 8006592 10.1084/jem.180.1.341
Colville-Nash PR Scott DL Angiogenesis and rheumatoid arthritis: pathogenic and therapeutic implications Ann Rheum Dis 1992 51 919 925 1378718
Herren B Raines EW Ross R Expression of a disintegrin-like protein in cultured human vascular cells and in vivo FASEB J 1997 11 173 180 9039960
Martin J Eynstone LV Davies M Williams JD Steadman R The role of ADAM 15 in glomerular mesangial cell migration J Biol Chem 2002 277 33683 33689 12091380 10.1074/jbc.M200988200
Nath D Slocombe PM Stephens PE Warn A Hutchinson GR Yamada KM Docherty AJ Murphy G Interaction of metargidin (ADAM-15) with alphavbeta3 and alpha5beta1 integrins on different haemopoietic cells J Cell Sci 1999 112 579 587 9914169
Moiseeva EP Adhesion receptors of vascular smooth muscle cells and their functions Cardiovasc Res 2001 52 372 386 11738054 10.1016/S0008-6363(01)00399-6
Kim LT Yamada KM The regulation of expression of integrin receptors Proc Soc Exp Biol Med 1997 214 123 131 9034129
Trochon-Joseph V Martel-Renoir D Mir LM Thomaidis A Opolon P Connault E Li H Grenet C Fauvel-Lafeve F Soria J Evidence of antiangiogenic and antimetastatic activities of the recombinant disintegrin domain of metargidin Cancer Res 2004 64 2062 2069 15026344
Ham C Levkau B Raines EW Herren B ADAM15 is an adherens junction molecule whose surface expression can be driven by VE-cadherin Exp Cell Res 2002 279 239 247 12243749 10.1006/excr.2002.5606
Herren B Garton KJ Coats S Bowen-Pope DF Ross R Raines EW ADAM15 overexpression in NIH3T3 cells enhances cell-cell interactions Exp Cell Res 2001 271 152 160 11697891 10.1006/excr.2001.5353
Paleolog EM Angiogenesis in rheumatoid arthritis Arthritis Res 2002 4 Suppl 3 S81 S90 12110126 10.1186/ar575
Rinaldi N Barth T Henne C Mechterscheimer G Moller P Synoviocytes in chronic synovitis in situ and cytokine stimulated synovial cells in vitro neo-express alpha 1, alpha 3 and alpha 5 chains of beta 1 integrins Virchows Arch 1994 425 171 180 7524977 10.1007/BF00230354
Rinaldi N Schwarz-Eywill M Weis D Leppelmann-Jansen P Lukoschek M Keilholz U Barth TF Increased expression of integrins on fibroblast-like synoviocytes from rheumatoid arthritis in vitro correlates with enhanced binding to extracellular matrix proteins Ann Rheum Dis 1997 56 45 51 9059141
Rinaldi N Weis D Brado B Schwarz-Eywill M Lukoschek M Pezzutto A Keilholz U Barth TF Differential expression and functional behaviour of the alpha v and beta 3 integrin subunits in cytokine stimulated fibroblast-like cells derived from synovial tissue of rheumatoid arthritis and osteoarthritis in vitro Ann Rheum Dis 1997 56 729 736 9496152
Fourie AM Coles F Moreno V Karlsson L Catalytic activity of ADAM8, ADAM15, and MDC-L (ADAM28) on synthetic peptide substrates and in ectodomain cleavage of CD23 J Biol Chem 2003 278 30469 30477 12777399 10.1074/jbc.M213157200
Bansal AS Ollier W Marsh MN Pumphrey RS Wilson PB Variations in serum sCD23 in conditions with either enhanced humoral or cell-mediated immunity Immunology 1993 79 285 289 8344706
Chomarat P Briolay J Banchereau J Miossec P Increased production of soluble CD23 in rheumatoid arthritis, and its regulation by interleukin-4 Arthritis Rheum 1993 36 234 242 8431213
Ribbens C Bonnet V Kaiser MJ Andre B Kaye O Franchimont N de Groote D Beguin Y Malaise MG Increased synovial fluid levels of soluble CD23 are associated with an erosive status in rheumatoid arthritis (RA) Clin Exp Immunol 2000 120 194 199 10759783 10.1046/j.1365-2249.2000.01198.x
Bonnefoy JY Plater-Zyberk C Lecoanet-Henchoz S Gauchat JF Aubry JP Graber P A new role for CD23 in inflammation Immunol Today 1996 17 418 420 8854559 10.1016/0167-5699(96)10054-2
Rezzonico R Chicheportiche R Imbert V Dayer JM Engagement of CD11b and CD11c beta2 integrin by antibodies or soluble CD23 induces IL-1beta production on primary human monocytes through mitogen-activated protein kinase-dependent pathways Blood 2000 95 3868 3877 10845922
Armant M Rubio M Delespesse G Sarfati M Soluble CD23 directly activates monocytes to contribute to the antigen-independent stimulation of resting T cells J Immunol 1995 155 4868 4875 7594490
Lecoanet-Henchoz S Plater-Zyberk C Graber P Gretener D Aubry JP Conrad DH Bonnefoy JY Mouse CD23 regulates monocyte activation through an interaction with the adhesion molecule CD11b/CD18 Eur J Immunol 1997 27 2290 2294 9341771
Bohm BB Aigner T Roy B Brodie TA Blobel CP Burkhardt H Homeostatic effects of the metalloproteinase disintegrin ADAM15 in degenerative cartilage remodeling Arthritis Rheum 2005 52 1100 1109 15818704 10.1002/art.20974
|
16277668
|
PMC1297561
|
CC BY
|
2021-01-04 16:02:40
|
no
|
Arthritis Res Ther. 2005 Aug 5; 7(6):R1158-R1173
|
utf-8
|
Arthritis Res Ther
| 2,005 |
10.1186/ar1796
|
oa_comm
|
==== Front
Arthritis Res TherArthritis Research & Therapy1478-63541478-6362BioMed Central London ar18011627766910.1186/ar1801Research ArticleCopper chelation with tetrathiomolybdate suppresses adjuvant-induced arthritis and inflammation-associated cachexia in rats Omoto Atsushi [email protected] Yutaka [email protected] Igor [email protected] Yasunori [email protected] Mizuho [email protected] Hidetaka [email protected] Makoto [email protected] Makie [email protected] Masataka [email protected] Rikio [email protected] Toshikazu [email protected] Hajime [email protected] Inflammation and Immunology, Graduate School of Medical Science, Kyoto Prefectural University of Medicine, Kyoto, Japan2 Center for Molecular Medicine, Maine Medical Center Research Institute, Scarborough, Maine, USA3 Department of Urology, Osaka City Graduate School of Medicine, Osaka, Japan4 Division of Rheumatology and Clinical Immunology, Department of Internal Medicine, Hyogo College of Medicine, Nishinomiya, Japan2005 8 8 2005 7 6 R1174 R1182 15 3 2005 11 4 2005 9 6 2005 7 7 2005 Copyright © 2005 Omoto et al.; licensee BioMed Central Ltd.This is an Open Access article distributed under the terms of the Creative Commons Attribution License (), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
Tetrathiomolybdate (TM), a drug developed for Wilson's disease, produces an anti-angiogenic and anti-inflammatory effect by reducing systemic copper levels. TM therapy has proved effective in inhibiting the growth of tumors in animal tumor models and in cancer patients. We have hypothesized that TM may be used for the therapy of rheumatoid arthritis and have examined the efficacy of TM on adjuvant-induced arthritis in the rat, which is a model of acute inflammatory arthritis and inflammatory cachexia. TM delayed the onset of and suppressed the severity of clinical arthritis on both paw volume and the arthritis score. Histological examination demonstrated that TM significantly reduces the synovial hyperplasia and inflammatory cell invasion in joint tissues. Interestingly, TM can inhibit the expression of vascular endothelial growth factor in serum synovial tissues, especially in endothelial cells and macrophages. Moreover, the extent of pannus formation, which leads to bone destruction, is correlated with the content of vascular endothelial growth factor in the serum. There was no mortality in TM-treated rat abnormalities. TM also suppressed inflammatory cachexia. We suggest that copper deficiency induced by TM is a potent approach both to inhibit the progression of rheumatoid arthritis with minimal adverse effects and to improve the well-being of rheumatoid arthritis patients.
==== Body
Introduction
Rheumatoid arthritis (RA) is a chronic, destructive inflammatory polyarticular joint disease. It is characterized by massive synovial proliferation and subintimal infiltration of inflammatory cells, which along with angiogenesis leads to the formation of a very aggressive tissue called pannus [1,2]. Expansion of the pannus induces bone erosion and cartilage thinning, leading to the loss of joint function. The rheumatoid pannus can thus be considered a local tumor. One of the earliest phenomena observed in RA is synovial neovascular formation delivering nutrients and oxygen to this proliferating pannus [3]. It has been demonstrated that angiogenesis inhibitors can inhibit the growth of pannus in animal arthritis models [4]. Vascular endothelial growth factor (VEGF) plays a pivotal role in the pathogenesis of RA [3,5,6]. Immunohistochemical and in situ hybridization studies indicate that VEGF is strongly expressed in subsynovial macrophages, in fibroblasts surrounding microvessels, in vascular smooth muscle cells, and in synoviocytes [7-9]. VEGF expression is activated at the very early stages of RA, and it continues throughout the course of the disease [10,11]. The VEGF level in synovial fluid and tissues correlates with the clinical severity of RA and with the degree of joint destruction [10]. Moreover, VEGF mediates the recruitment, chemotaxis, and proliferation of osteoclast precursor macrophages, leading to bone destruction [11,12].
RA is also characterized by increased production of the inflammatory cytokines tumor necrosis factor alpha (TNF-α) [1], IL-1α [13], IL-1β [1], and fibroblast growth factor (FGF) 1 [14]. TNF-α appears to be a key mediator in the disease process, and IL-1β plays a permissive role by acting to shift the whole-body protein metabolism towards net catabolism, to elevate resting energy expenditure, and to increase joint pain and stiffness [15]. Treatment with antibodies against TNF-α, IL-1α, and IL-1β attenuated RA in the experimental mouse model [16]. FGF-1 is important for the growth of synoviocytes in the course of RA [17].
Rheumatoid cachexia was first described more than a century ago [18]. However, it has not been recognized as a common problem among patients with RA until relatively recently. Rheumatoid cachexia leads to muscle weakness, osteoporosis, and a loss of functional capacity. It also increases susceptibility to infection [19], and is believed to accelerate morbidity and mortality in RA [15].
Copper is an essential trace element that acts as a cofactor for a variety of enzymes by virtue of its ability to accept and donate electrons under physiologic conditions [20]. Additionally, copper ions have recently been demonstrated to be required for the assembly of multiprotein release complexes in the process of stress-induced nonclassical release of FGF-1 and IL-1α [21-23]. These two proteins lack signal sequences in their primary structures, and cannot be released through the classical endoplasmic reticulum-Golgi pathway. Their nonclassical export involves copper-dependent association with a small calcium-binding protein, S100A13.
Tetrathiomolybdate (TM), which forms a stable tripartite complex with copper and protein, is a copper-lowering agent that has been evaluated extensively in the treatment of Wilson's disease [24]. TM treatment decreases serum copper levels and attenuates angiogenesis and tumor growth in animal tumor models [25-27]. The hypothesis underlying this approach is that one or more copper-containing or copper-binding angiogenic proteins (e.g. VEGF, FGF-1, FGF-2, angiogenin, angiotropin, or others) require higher levels of copper to be active than are required for basic cellular needs [28]. In fact, the antitumor activity of TM was evaluated in patients with advanced kidney cancer in a phase II trial [29].
Alternatively, the in vivo effects of TM may be explained by its ability to block the release of FGF-1 and IL-1α, both known as potent proangiogenic and proinflammatory polypeptides. Indeed, the inhibition of restenosis by TM in the model of damaged rat carotid artery was accompanied by the downregulation of FGF-1 and IL-1α levels in the vessel wall [22]. Additionally, copper is known to play an important role in the development and maintenance of the immune system [30]. Some reports revealed that the possibility of inhibiting both fibrotic response and inflammatory response by copper chelation is due to the suppression of transforming growth factor beta and TNF-α production [31,32].
The serum copper level in RA patients has been reported to be high [33], and the IL-1β and TNF-α serum content might correlate with the serum copper level in RA patients [34]. In addition, D-penicillamine (another anticopper agent) has been used as the therapy for RA for many years. A previous study suggested that D-penicillamine might regress rheumatoid synovial hyperplasia via Fas-mediated apoptosis, but the mechanism of the effect of D-penicillamine is still unknown [35]. In animal studies, TM is a more fast-acting, more potent, copper chelating agent than D-penicillamine [36]. We hypothesized that the anticopper drug TM can be useful for the treatment of RA through inhibition of proangiogenic and proinflammatory cytokines. We examined whether TM has the potency to suppress chronic inflammation, pannus formation, and angiogenesis in the course of adjuvant-induced arthritis (AIA) in female Lewis rats. We also examined whether copper chelation with TM reduces the production of VEGF in serum and synovium of AIA rats and suppresses inflammatory cachexia in AIA rats.
Materials and methods
AIA in rats
Eight-week-old female Lewis rats were obtained from Charles River Japan (Yokohama, Japan). Complete Freund's adjuvant was prepared by suspending heat-killed Mycobacterium butyricum (Difco Laboratories, Detroit, MI, USA) in liquid paraffin (Merck & Co., Whitehouse Station, NJ, USA) (a kind gift from Nippon Shinyaku, Kyoto, Japan) at 12 mg/ml. Complete Freund's adjuvant-induced arthritis was stimulated by injection of 50 μl Complete Freund's adjuvant emulsion intradermally at the base of the tail, as described previously [37-39]. This experimental procedure was designed and carried out according to the institutional rules and regulations of the animal research of Kyoto Prefectural University of Medicine.
Administration of TM
TM treatment commenced 2 weeks before immunization; TM (10 mg/kg) was given in 3 ml water once daily by means of intragastric gavage until 10 days after the onset of arthritis. Deionized water was given to control rats. The animals were housed four to a cage at 21°C in a 12-hour light/dark cycle. The rats were killed on day 17 after immunization under anesthesia with sodium pentobarbital.
Evaluation of arthritis
From day 7 after immunization (onset of arthritis), rats were examined every 2 days for three clinical parameters: paw volume, arthritis score, and body weight. The footpad volume was measured with a water replacement plethysmometer (Unicom Japan, Tokyo, Japan). For clinical evaluation of AIA, the mid-forepaw, the wrist, the joints of the finger, the midfoot, the ankle, and the joints of the digits were scored on a 0–4 scale: 0, normal; 1, minimal swelling; 2, medium swelling; 3, severe swelling; 4, severe and nonweight-bearing arthritis. Each limb was graded, resulting in a maximal clinical score of 48 per animal.
Copper status
In the course of TM therapy, the copper status cannot be assessed by direct measurement of serum copper. The accumulation of a tripartite complex of TM, copper, and albumin turns over slowly. The serum copper is therefore increased even though the availability of copper is decreased. Serum ceruloplasmin is a good surrogate marker of copper status because the liver secretes this copper-containing protein into the blood at a rate dependent on copper availability [40].
We monitored the copper status by assaying serum ceruloplasmin in blood from the tail vein. The ceruloplasmin measurements were made by nephelometry (differential light scattering from a colored or turbid case solution with respect to a control solution) using an automated system and reagents available commercially (Dade Behring Inc., Deerfield, IL, USA) when TM-treated rats were immunized.
Histological examination
After euthanasia on day 17, the hindpaws were amputated above the knee joint and were fixed in 7.4% formaldehyde solution. The paws were then decalcified, embedded in paraffin, and sectioned in a mid-sagittal plane. The sections of articulation of the tarsal joints were stained with hematoxylin and eosin, and were examined microscopically. We also performed the hematoxylin and eosin staining of tissue specimens of the liver and the kidney.
Two blinded observers evaluated cartilage and bone destruction by pannus formation, mononuclear cell infiltration, and vascularity in synovial tissues in each preparation on two separate occasions, using the following scoring system [41]: mononuclear cell infiltration (0, no infiltration; 1, mild infiltration; 2, moderate infiltration; 3, severe infiltration); cartilage and bone destruction by pannus formation (0, no change; 1, mild change [pannus invasion within cartilage]; 2, moderate change [pannus invasion into cartilage/subchondral bone]; 3, severe change [pannus invasion into the subchondral bone]); and vascularity (0, almost no blood vessels; 1, a few blood vessels; 2, some blood vessels; 3, many blood vessels).
Immunostaining
VEGF, CD11b, and von Willebrand factor (vWF) chain antigens were detected by the use of saturating amounts of antibodies against VEGF, CD11b, and vWF in combinations with immunoperoxidase staining with a Vectastain ABC kit (Vector Laboratories, Burlingame, CA, USA) according to the manufacturer's protocol [14]. The sections were deparaffinized with xylene and graded ethanol and were immersed in 0.3% peroxidase in 90% methanol for 45 min in order to exhaust endogenous peroxidase. They were preincubated with 0.2% bovine serum albumin in PBS for 20 min and with diluted normal horse serum (1:66.7), or normal goat serum (1:66.7) for 30 min followed by incubation with 5 μg/ml anti-VEGF monoclonal IgG antibody (sc-7269: Santa Cruz Biotechnology, Santa Cruz, CA, USA), 10 μg/ml anti-CD11b monoclonal antibody (Serotec Ltd, Kidlington, UK), 10 μg/ml anti monoclonal vWF antibody (Sigma, St Louis, MO, USA), 5 μg/ml purified rabbit IgG (Vector Laboratories), or 5 μg/ml purified mouse IgG (Vector Laboratories) for 16 hours in a humid chamber at 4°C. After washing with PBS, the sections were incubated with biotinylated horse anti-mouse IgG (Vector Laboratories) or goat anti-rabbit IgG (Vector Laboratories) for 30 min. Then, after again washing with PBS, the sections were incubated with avidin and further incubated with biotinylated horseradish peroxidase complex, each for 45 min. Finally, the sections were washed with PBS for 10 min and developed by immersing in a solution of 0.05% (w/v) 3,3'-diaminobenzidine tetrahydrochloride (Sigma Chemical, St Louis, MO, USA), and 0.01% hydrogen peroxide in 0.05 M Tris (pH 7.4) for 2 min. The sections were then counterstained with hematoxylin for 2 min, dehydrated with graded ethanol and xylene for 1 min, respectively, and finally coverslips were mounted.
For both tissue specimens from TM-treated rats and control rats, the extent and intensity of staining with anti-VEGF antibody in synovial lining cells, macrophages, endothelial cells, and fibroblasts were graded on a scale of 0–3+ by two blinded observers on two separate occasions using coded slides as previously described [37]. A 3+ grade implies maximally intense staining, whereas 0 implies no staining.
CD11b immunostaining in monocytes was used to evaluate mononuclear cell infiltration for each of tissue specimens. vWF immunostaining in endothelial cells was used to evaluate vascularity for each of the tissue specimens.
Measurement of body weight of AIA rats
From day 7 after immunization (onset of arthritis), the body weight of AIA rats was examined every 2 days. We also measured body weights of rats at the first administration of TM or of deionized water. At this initial time point, the differences between the two groups were not significant. We also measured body weights of 10-week-old female Lewis rats (n = 10), which were not immunized, for a period of 14 days as a negative control.
Measurement of VEGF production in rats
When rats were sacrificed, the serum was collected. We measured the concentrations of VEGF in the serum using the VEGF ELISA kit (Biosource International, Camarillo, CA, USA).
Statistical analysis
Two-way analysis of variance was used to test the statistical significance of differences between a TM-treated group and a control group for the analysis of hindlimb paw volume, body weight, and clinical score of arthritis. The Mann–Whitney U test to compare nonparametric data for statistical significance was applied for the analysis of histological scores. A nonpaired t test was used for the analysis of the serum concentration of VEGF. The Spearman coefficient of correlation was used to examine the correlation between the extent of pannus formation and the VEGF level in the serum.
Results
Oral administration of TTM attenuates AIA
To explore the effect of TM on AIA rats, TM (10 mg/kg) was administered daily starting from 14 days before immunization. Inflammatory polyarthritis was induced in all nontreated immunized rats and it occurred on day 8 after immunization (day 7 in Materials and methods). TM administration delayed the onset and suppressed severity of clinical arthritis comparative to control rats fed with deionized water, as demonstrated by both the paw volume (Figs 1a and 2) (P < 0.0001) and the arthritis score (Fig. 1b) (P < 0.0001). Especially significant TM effects were observed at days 11–17 after immunization. These data suggest that oral TM administration inhibits the onset of and reduces the severity of arthritis in AIA rats. The body weight of TM-treated rats was significantly increased compared with control AIA rats (Fig. 1c) (P < 0.0001). We also measured the body weights of 10-week-old female Lewis rats (n = 10), which were not immunized, for a period of 14 days. The results show that the body weight of control AIA rats was significantly decreased compared with that of nonimmunized normal rats, and the body weight gain of TM-treated AIA rats was almost similar to that of nonimmunized normal rats (data not shown). These data suggest that oral TM administration prevents inflammatory body weight loss in AIA rats.
Histological effects of TM in the foot joint of AIA rats
At day 17, histological study of foot joints (shown in Fig. 2) in TM-treated rats revealed that the infiltration of mononuclear cells, the formation of pannus in synovial tissues, and the extent of vascularity were significantly decreased compared with control rats (P < 0.01; Fig. 3) (data of immunostaining of CD11b and vWF not shown). These data suggest that TM exhibits anti-inflammatory and anti-angiogenic effects, and inhibits the growth of synoviocytes in AIA.
VEGF expression in synovial tissues in AIA rats
Immunohistochemistry was used to examine the expression and localization of VEGF in AIA rats (Fig. 4). We found markedly enhanced expression of VEGF in endothelial cells (immunohistochemical score 2.1 ± 0.5), and found moderate expression in fibroblasts (1.7 ± 0.6) and macrophages (1.6 ± 0.9) of immunized rats. In control nonimmunized rats (n = 6), the expression of VEGF was 1.1 ± 0.5 in endothelial cells, was 0.6 ± 0.3 in fibroblasts, and was 0.1 ± 0.0 in macrophages. In TM-treated immunized rats, the localization of VEGF in synovial tissues was similar to that in control immunized rats. However, the immunohistochemical score of VEGF in TM-treated immunized rats in endothelial cells (1.1 ± 0.5, P < 0.01), in macrophages (0.7 ± 0.7, P < 0.01), and in fibroblasts (1.2 ± 0.5, P < 0.05) was significantly lower than in control immunized rats. Control immunostaining with normal mouse serum was completely negative in all animals.
TM effect on VEGF production in AIA rats
To examine the effect of TM upon VEGF production, we measured the VEGF level in the serum of TM-treated rats and control AIA rats (Fig. 5a). The production of VEGF in the serum was significantly suppressed by TM treatment (755.0 ± 354.7 pg/ml versus 1912.5 ± 800.3 pg/ml in TM-untreated animals, P = 0.0038). Interestingly, the extent of pannus formation significantly correlated with the production of VEGF in the serum (Fig. 5b).
Side effects of oral administration of TM
There was no mortality in TM-treated rats. The liver and kidney histological examination in TM-treated rats did not show any abnormalities, and we could not detect any pathological changes when comparing them with TM-untreated rats.
Discussion
We demonstrated that TM therapy has a strong protective effect against progression of AIA in rats. Usually AIA is accompanied by weight loss because of inflammatory response, but TM prevents this. TM also significantly reduces the level of VEGF in the serum and inhibits the expression of VEGF in synovial tissues, especially in endothelial cells and macrophages. Joint and bone destruction due to arthritis are markedly suppressed by TM, and the extent of bone destruction is significantly correlated with the production of VEGF in the serum.
An increase of serum copper and ceruloplasmin concentrations has been demonstrated in RA patients [34,42]. In RA these parameters are measures of disease, and they do not depend on dietary factors [43]. Acute or chronic inflammatory processes cause an accumulation of zinc and copper in many organs, particularly in the inflamed areas [44]. Additionally, a number of biologically active extracellular polypeptides, including cytokines and angiogenic factors, which participate in the pathogenesis and development of inflammatory processes, are known to be involved in trace metal metabolism. Copper plays an important role in development and maintenance of the immune system [30]. IL-1α and FGF-1 are Cu2+-binding proteins. The stress-induced IL-1α and FGF-1 release pathways in murine NIH 3T3 cells and human U937 cells are sensitive to TM treatment [45]. The presence of copper in cell cultures is essential for T-cell proliferation induced by macrophages or by macrophage-mediated cytokines [30]. A recent study revealed that IL-1β and TNF-α levels significantly correlate with serum copper concentrations [34]. In our study, in vivo, copper chelation with TM strongly repressed acute inflammation and onset of AIA through inhibition of mononuclear cell infiltration, and pannus formation.
TM has been shown to be a potent anti-angiogenic and anti-metastatic compound in tumors. In a phase II clinical trial for advanced renal cell carcinoma, patients rendered copper deficient with TM therapy had a significantly decreased serum content of proangiogenic mediators, VEGF, FGF-2, IL-6, and IL-8 [29]. In RA patients, proangiogenic factors such as VEGF play an important role in pathological angiogenesis, and the other factors such as IL-6 and IL-8 are considered to have additional effects on its development. VEGF is also a key player in pannus development, acting through the VEGF receptor I signaling pathway. The blockade of the VEGF receptor I suppresses joint destruction in the K/BxN model of RA [46], and serum concentrations of VEGF are elevated in RA patients and correlate with disease activity [47]. In our study, the onset of AIA in rats is delayed, and its severity is suppressed by TM administration through the inhibition of pannus formation and angiogenesis. The anti-arthritic effect of TM might therefore result from the inhibition of VEGF production by synovial tissues. We also examined the efficacy of TM administration starting from day 7 after immunization (the onset of the disease) for AIA in rats; in this case, TM had only a mild anti-arthritic effect. Apparently, to efficiently attenuate arthritis, copper depletion needs to be achieved by the moment of its onset.
TM was well tolerated in patients with advanced kidney cancer in a phase II trial, with dose reductions most commonly occurring for grade 3–4 granulocytopenia of short duration not associated with febrile episodes [29]. The principal features observed in severe copper deficiency are anemia, neutropenia, and osteoporosis. TM was remarkably nontoxic when ceruloplasmin was lowered to 10–20% of baseline levels for up to 17 months of treatment, and the only drug-related toxicity observed was mild anemia, which was easily reversible with adjustment of the TM dose to bring the ceruloplasmin level to the desired target [40]. But various side effects may occur when ceruloplasmin is reduced below 5 mg/dl, such as bone marrow suppression, diarrhea, and arrhythmia. We assayed the serum ceruloplasmin level as a surrogate marker of copper status, and kept it in a range between 5 and 10 mg/dl in TM-treated rats. As the histological examination of the liver and kidney demonstrated, no significant adverse effects were observed. This means that the extent of copper chelation in this study was sufficient and not excessive.
We have demonstrated that TM administration prevents cachexia, which is associated with RA. It is known that AIA in rats is a useful model of inflammatory cachexia that mimics the human pathophysiology in important ways, and is consistent with cytokine-driven cachexia in chronic inflammatory arthritis [48]. We found that the body weight of control AIA rats was significantly decreased compared with that of nonimmunized normal rats, and the body weight gain of TM-treated AIA rats was almost similar to that of nonimmunized normal rats. Rheumatoid cachexia is characterized by altered energy and protein metabolism (reduced total energy expenditure, increased resting energy expenditure, and increased whole-body protein catabolism) and increased inflammatory cytokine production [15].
Leptin is a peptide hormone-regulating body weight, and it exhibits a variety of other effects including the regulation of the endocrine system, reproduction, and immunity [49,50]. The severity of antigen-induced arthritis is decreased in leptin-deficient ob/ob mice [49]. Furthermore, serum leptin levels in patients with RA are significantly higher than those in control patients, and leptin stimulates proinflammatory cytokine production in monocytes and macrophages in vitro [51]. Moreover, it is reported that TM therapy resulted in significantly reduced body-weight loss caused by bleomycin-induced pulmonary fibrosis in mice [31].
These findings suggest that copper chelation by TM may not only suppress joint destruction, but also may influence energy and protein metabolism in the course of RA through the upregulation of the adipocytokine leptin.
Conclusion
TM therapy had a strong protective effect against progression of adjuvant arthritis in rats and inflammatory cachexia with minimal adverse effects. TM also significantly reduced the content of VEGF in the serum and synovial tissues. Joint and bone destruction due to arthritis was markedly suppressed with TM. The extent of bone destruction was correlated with the production of VEGF in the serum.
TM is a potential therapeutic candidate for the treatment of angiogenic and inflammatory diseases in which the serum copper and VEGF levels are elevated. Additionally, the efficacy of TM against inflammatory cachexia may be useful for improving the well-being of RA patents.
Abbreviations
AIA = adjuvant-induced arthritis; ELISA = enzyme-linked immunosorbent assay; FGF = fibroblast growth factor; IL = interleukin; PBS = phosphate-buffered saline; RA = rheumatoid arthritis; TM = tetrathiomolybdate; TNF-α = tumor necrosis factor alpha; VEGF = vascular endothelial growth factor; vWF = von Willebrand factor.
Competing interests
The author(s) declare that they have no competing interests.
Authors' contributions
AO conceived of the study, participated in its design, and performed all the experiments. YK conceived of the study, participated in the design of and the coordination of the study, and participated in the interpretation of the results. IP and HS participated in the design of the animal study. TY, YT, and RY participated in the immunohistochemistry and performed the interpretation of the results. MK and HI performed the animal study. MW, MK, and MY participated in the immunoassay. All authors read and approved the final manuscript.
Acknowledgements
IP acknowledges the support of NIH grants HL35627, HL32348 and RR15555.
Figures and Tables
Figure 1 Suppression of clinical arthritis by tetrathiomolybdate (TM) treatment in adjuvant-induced arthritis rats. TM was orally administered daily to female Lewis rats, from 2 weeks before immunization to day 17. TM-treated rats, 10 mg/kg/day (n = 8); control rats (n = 8). (a) Arthritis score, (b) paw volume, and (c) mean body weight were assessed every second day after onset of arthritis.
Figure 2 Morphological features and histopathological aspects of the hindlimb in AIA rats. (a) Control rats and (b) rats treated with tetrathiomolybdate (TM). Joint swelling, redness, and edema of the foot in AIA was clearly reduced with TM administration at day 17 after immunization. Histopathological studies using hematoxylin and eosin staining of the foot joint also revealed (c) a marked decrease of synovial inflammatory cell infiltrate and synovial lining hyperplasia compared with (d) control rats. (c), (d) Original magnification × 40. AIA, adjuvant-induced arthritis; MC, monocyte; SL, synovial lining cell; OC, osteoclast.
Figure 3 Histopathological scores of the hindlimb in AIA rats fed with TM or deionized water. Mononuclear cell infiltration, pannus invasion into the cartilage and bone, and vascularity were measured by microscopic examination scores of the sections on two separate occasions (see Histological examination). We measured the scores of 16 hindlimbs (both hindlimbs of each rat). AIA, adjuvant-induced arthritis; TM, tetrathiomolybdate.
Figure 4 Immnunostaining for VEGF in synovial tissue in AIA rats. (a),(c), (e) AIA rats treated with deionized water and (b),(d) tetrathiomolybdate (TM). Immunohistochemical staining was performed with the Vecto Stain avidin–biotin peroxidase complex kit (Vector Laboratories, Burlingame, CA, USA). Synovial tissues sections were stained with (a)–(d) mouse anti-VEGF monoclonal IgG antibodies (1:200 dilution, in PBS; Santa Cruz Biotechnology, Santa Cruz California, USA) and (e) a normal mouse IgG (1:200 dilution, in PBS). Positive immunostaining was indicated by brownish deposits. The counterstain was an aqueous solution of hematoxylin. (a), (b) Original magnification × 40; (c)–(e) original magnification × 400. AIA, adjuvant-induced arthritis; ET, endothelial cell; SL, synovial lining cell.
Figure 5 VEGF levels in the serum are correlated with the extent of arthritis in rats. (a) VEGF levels in the serum are higher in control rats than in tetrathiomolybdate (TM)-treated rats (* P < 0.01). VEGF was analyzed by ELISA, as described in Materials and methods. Data represent the mean ± standard deviation. (b) A significant positive relationship between VEGF levels in the serum and the severity of pannus formation in hindlimbs of adjuvant-induced arthritis rats, which includes both the control group and the TM-treated group. VEGF, vascular endothelial growth factor.
==== Refs
Feldmann M Brennan FM Maini RN Rheumatoid arthritis Cell 1996 85 307 310 8616886 10.1016/S0092-8674(00)81109-5
Koch AE Angiogenesis: implication of rheumatoid arthritis Arthritis Rheum 1998 41 951 962 9627005 10.1002/1529-0131(199806)41:6<951::AID-ART2>3.0.CO;2-D
Folkman J Angiogenesis in cancer, vascular, rheumatoid and other disease Nat Med 1995 1 27 31 7584949 10.1038/nm0195-27
Ferrara N Alitalo K Clinical applications of angiogenic growth factors and their inhibitors Nat Med 1999 5 1359 1364 10581076 10.1038/70928
Firestein GS Invasive fibroblast-like synoviocytes in rheumatoid arthritis: passive responders or transformed aggressors? Arthritis Rheum 1996 39 1781 1790 8912499
Koch AE Harlow LA Haines GK Amento EP Unemori EN Wong WL Pope RM Ferrara N Vascular endothelial growth factor. A cytokine modulating endothelial function in rheumatoid arthritis J Immunol 1994 152 4149 4156 7511670
Ballara S Taylor PC Reusch P Marme D Feldmann M Maini RN Paleolog EM Raised serum vascular endothelial growth factor levels are associated with destructive change in inflammatory arthritis Arthritis Rheum 2001 44 2055 2064 11592367 10.1002/1529-0131(200109)44:9<2055::AID-ART355>3.0.CO;2-2
Scola MP Imagawa T Boivin GP Giannini EH Glass DN Hirsch R Grom AA Expression of angiogenic factors in juvenile rheumatoid arthritis: correlation with revascularization of human synovium engrafted into SCID mice Arthritis Rheum 2001 44 794 801 11315918 10.1002/1529-0131(200104)44:4<794::AID-ANR135>3.0.CO;2-7
Weber AJ De Bandt M Gaudry M Immunohistochemical analysis of vascular endothelial growth factor expression in severe and destructive rheumatoid arthritis J Rheumatol 2000 27 2284 2286 10990256
Hitchon C Wong K Ma G Reed J Lyttle D El-Gabalawy H Hypoxia-induced production of stromal cell-derived factor 1 (CXCL12) and endothelial growth factor by synovial fibroblasts Arthritis Rheum 2002 46 2587 2597 12384916 10.1002/art.10520
Henriksen K Karsdal M Delaisse JM Engsig MT RANKL and vascular endothelial growth factor (VEGF) induce osteoclast chemotaxis through an ERK1/2-dependent mechanism J Biol Chem 2003 278 48745 48753 14506249 10.1074/jbc.M309193200
Matsumoto Y Tanaka K Hirata G Hanada M Matsuda S Shuto T Iwamoto Y Possible involvement of the vascular endothelial growth factor-Flt-1-focal adhesion kinase pathway in chemotaxis and cell proliferation of osteoclast precursor cells in arthritic joints J Immunol 2002 168 5824 5831 12023386
Deleuran BW Chu CQ Field M Brennan FM Katsikis P Feldmann M Maini RN Localization of interleukin-1 alpha, type 1 interleukin-1 receptor and interleukin-1 receptor antagonist in the synovial membrane and cartilage/pannus junction in rheumatoid arthritis Br J Rheumatol 1992 31 801 809 1458286
Sano H Forough R Maier JA Case JP Jackson A Engleka K Maciag T Wilder RL Detection of high levels of heparin binding growth factor-1 (acidic fibroblast growth factor) in inflammatory arthritic joints J Cell Biol 1990 110 1417 1426 1691192 10.1083/jcb.110.4.1417
Rall LC Roubenoff R Rheumatoid cachexia: metabolic abnormalities, mechanism and interventions Rheumatology (Oxford) 2004 43 1219 1223 15292530 10.1093/rheumatology/keh321
Joosten LA Helsen MM van de Loo FA van den Berg WB Anticytokine treatment of established type II collagen-induced arthritis in DBA/1 mice. A comparative study using anti-TNF alpha, anti-IL-1 alpha/beta, and IL-1Ra Arthritis Rheum 1996 39 797 809 8639177
Thomas JW Thieu TH Byrd VM Miller GG Acidic fibroblast growth factor in synovial cells Arthritis Rheum 2000 43 2152 2159 11037874 10.1002/1529-0131(200010)43:10<2152::AID-ANR2>3.3.CO;2-I
Paget J Nervous mimicry of organic diseases Lancet 1873 102 727 729 10.1016/S0140-6736(02)69188-3
Roubenoff R Kehayias JJ The meaning and measurement of lean body mass Nutr Rev 1991 49 163 175 2046978
Schilsky ML Diagnosis and treatment of Wilson's disease Pediatr Transplant 2002 6 15 19 11906637 10.1034/j.1399-3046.2002.1r069.x
Landriscina M Bagala C Mandinova A Soldi R Micucci I Bellum S Prudovsky I Maciag T Copper induces the assembly of a multiprotein aggregate implicated in the release of fibroblast growth factor 1 in response to stress J Biol Chem 2001 276 25549 25557 11432880 10.1074/jbc.M102925200
Mandinov L Mandinova A Kyurkchiev S Kyurkchiev D Kehayov I Kolev V Soldi R Bagala C de Muinck ED Lindner V Copper chelation represses the vascular response to injury Proc Natl Acad Sci USA 2003 100 6700 6705 12754378 10.1073/pnas.1231994100
Prudovsky I Mandinova A Soldi R Bagala C Graziani I Landriscina M Tarantini F Duarte M Bellum S Doherty H Maciag T The non-classical export routes: FGF1 and IL-1alpha point the way J Cell Sci 2003 116 4871 4881 14625381 10.1242/jcs.00872
Brewer GJ Johnson V Dick RD Kluin KJ Fink JK Brunberg JA A treatment of Wilson's disease with ammonium tetrathiomolybdate. II. Initial therapy in 33 neurologically affected patients and follow-up with zinc therapy Arch Neurol 1996 53 1017 1025 8859064
Brewer GJ Merajver SD Cancer therapy with tetrathiomolybdate: antiangiogenesis by lowering body copper – a review Integr Cancer Ther 2002 1 327 337 14664727 10.1177/1534735402238185
Khan MK Miller MW Taylor J Gill NK Dick RD Van Golen K Brewer GJ Merajver SD Radiotherapy and antiangiogenic TM in lung cancer Neoplasia 2002 4 164 170 11896571 10.1038/sj.neo.7900218
Pan Q Kleer CG van Golen KL Irani J Bottema KM Bias C De Carvalho M Mesri EA Robins DM Dick RD Copper deficiency induced by tetrathiomolybdate suppresses tumor growth and angiogenesis Cancer Res 2002 62 4854 4859 12208730
Sen CK Khanna S Venojarvi M Trikha P Ellison EC Hunt TK Roy S Copper-induced vascular endothelial growth factor expression and wound healing Am J Physiol Heart Circ Physiol 2002 282 H1821 H1827 11959648
Redman BG Esper P Pan Q Dunn RL Hussain HK Chenevert T Brewer GJ Merajver SD Phase II trial of tetrathiomolybdate in patients with advanced kidney cancer Clin Cancer Res 2003 9 1666 1672 12738719
Percival SS Copper and immunity Am J Clin Nutr 1998 67 1064 1068
Brewer GJ Ullenbruch MR Dick R Olivarez L Phan SH Tetrathiomolybdate therapy protects against bleomycin-pulmonary fibrosis in mice J Lab Clin Med 2003 141 210 216 12624602 10.1067/mlc.2003.20
Marikovsky M Ziv V Nevo N Harris-Cerruti C Mahler O Cu/Zn superoxide plays important role in immune response J Immunol 2003 170 2993 3001 12626552
Honkanen V Konttinen YT Sorsa T Hukkanen M Kemppinen P Santavirta S Saari H Westermarck T Serum zinc, copper and selenium in rheumatoid arthritis J Trace Elem Electrolytes Health Dis 1991 5 261 263 1822334
Zoli A Altomonte L Caricchio R Galossi A Mirone L Ruffini MP Magaro M Serum zinc and copper in active rheumatoid arthritis: correlation with interleukin 1 beta and tumour necrosis factor alpha Clin Rheumatol 1998 17 378 382 9805181 10.1007/BF01450895
Harada S Sugiyama E Taki H Shinoda K Fujita T Maruyama M Kobayashi M D-penicillamine cooperates with copper sulfate to enhance the surface expression of functional Fas antigen in rheumatoid synovial fibroblasts via the generation of hydrogen peroxide Clin Exp Rheumatol 2002 20 469 476 12175101
Brewer GJ Tetrathiomolybdate anticopper therapy for Wilson's disease inhibits angiogenesis, fibrosis and inflammation J Cell Mol Med 2003 7 11 20 12767257
Sano H Hla T Maier JA Crofford LJ Case JP Maciag T Wilder RL In vivo cyclooxygenase expression in synovial tissues of patients with rheumatoid arthritis and osteoarthritis and rats with adjuvant streptococcal cell wall arthritis J Clin Invest 1992 89 97 108 1729286
Kawahito Y Cannon GW Gulko PS Remmers EF Longman RE Reese VR Wang J Griffiths MM Wilder RL Localization of quantitative trait loci regulating adjuvant-induced arthritis in rats: evidence for genetic factors common to multiple autoimmune diseases J Immunol 1998 161 4411 4419 9780220
Kawahito Y Kondo M Tsubouchi Y Hashiramoto A Bishop-Bailey D Inoue K Kohno M Yamada R Hla T Sano H 15-deoxy-delta(12,14)-PGJ2 induces synoviovyte apoptosis and suppresses adjuvant-induced arthritis in rats J Clin Invest 2000 106 189 197 10903334
Brewer GJ Dick RD Grover DK LeClaire V Tseng M Wicha M Pienta K Redman BG Jahan T Sondak VK Treatment of metastatic cancer with tetrathiomolybdate, an anticopper, and antiangiogenic agent: phase I study Clin Cancer Res 2000 6 1 10 10656425
Taniguchi K Kohsaka H Inoue N Terada Y Ito H Hirokawa K Miyasaka N Induction of the p16INK4a senescence gene as a new therapeutic strategy for the treatment of rheumatoid arthritis Nat Med 1999 5 760 767 10395320 10.1038/10480
Milanino R Frigo A Bambara LM Marrella M Moretti U Pasqualicchio M Biasi D Gasperini R Mainenti L Velo GP Copper and zinc status in rheumatoid arthritis: studies of plasma, erythrocytes and urine, and their relationship to disease activity markers and pharmacological treatment Clin Exp Rheumatol 1993 11 271 281 8353981
Honkanen VE Lamberg-Allardt CH Vesterinen MK Lehto JH Westermarck TW Metsa-Ketela TK Mussalo-Rauhamaa MH Konttinen YT Plasma zinc and copper concentrations in rheumatoid arthritis: influence of dietary factors and disease activity Am J Clin Nutr 1991 54 1082 1086 1957825
Milanino R Moretti U Concari E Marrella M Velo GP Copper and zinc status in rats and acute inflammation: focus on the inflamed area Agents Actions 1988 24 356 364 3177096
Mandinova A Soldi R Graziani I Bagala C Bellum S Landriscina M Tarantini F Prudovsky I Maciag T S100A13 mediates the copper-dependent stress-induced release of IL-1alpha from both human U937 and murine NIH 3T3 cells J Cell Sci 2003 116 2687 2696 12746488 10.1242/jcs.00471
De Bandt M Ben Mahdi MH Ollivier V Grossin M Dupuis M Gaudry M Bohlen P Lipson KE Rice A Wu Y Blockade of vascular endothelial growth factor receptor I (VEGF-RI), but not VEGF-RII, suppresses joint destruction in the K/BxN model of rheumatoid arthritis J Immunol 2003 171 4853 4859 14568965
Taylor PC Serum vascular markers and vascular imaging in assessment of rheumatoid arthritis disease activity and response to therapy Rheumatology (Oxford) 2005 44 721 728 15644394 10.1093/rheumatology/keh524
Roubenoff R Freeman LM Smith DE Abad LW Dinarello CA Kehayias JJ Adjuvant arthritis as a model of inflammatory cachexia Arthritis Rheum 1997 40 534 539 9082942
Ahima RS Flier JS Leptin Annu Rev Physiol 2000 62 413 437 10845097 10.1146/annurev.physiol.62.1.413
Fantuzzi G Faggioni R Leptin in the regulation of immunity, inflammation, and hematopoiesis J Leukoc Biol 2000 68 437 446 11037963
Zarkesh-Esfahani H Pockley G Metcalfe RA Bidlingmaier M Wu Z Ajami A Weetman AP Strasburger CJ Ross RJ High-dose leptin activates human leukocytes via receptor expression on monocytes J Immunol 2001 167 4593 4599 11591788
|
16277669
|
PMC1297562
|
CC BY
|
2021-01-04 16:02:40
|
no
|
Arthritis Res Ther. 2005 Aug 8; 7(6):R1174-R1182
|
utf-8
|
Arthritis Res Ther
| 2,005 |
10.1186/ar1801
|
oa_comm
|
==== Front
Arthritis Res TherArthritis Research & Therapy1478-63541478-6362BioMed Central London ar18021627767010.1186/ar1802Research ArticleA functional variant of Fcγ receptor IIIA is associated with rheumatoid arthritis in individuals who are positive for anti-glucose-6-phosphate isomerase antibodies Matsumoto Isao [email protected] Hua [email protected] Yoshifumi [email protected] Taichi [email protected] Takanori [email protected] Yuko [email protected] Daisuke [email protected] Satoshi [email protected] Akito [email protected] Takayuki [email protected] Clinical Immunology, University of Tsukuba, University of Tsukuba, Ibaraki, Japan2 PRESTO, Japan Science and Technology Agency, Saitama, Japan2005 11 8 2005 7 6 R1183 R1188 4 2 2005 15 3 2005 4 7 2005 19 7 2005 Copyright © 2005 Matsumoto et al.; licensee BioMed Central Ltd.This is an Open Access article distributed under the terms of the Creative Commons Attribution License (), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
Anti-glucose-6-phosphate isomerase (GPI) antibodies are known to be arthritogenic autoantibodies in K/B×N mice, although some groups have reported that few healthy humans retain these antibodies. The expression of Fcγ receptors (FcγRs) is genetically regulated and has strong implications for the development of experimental arthritis. The interaction between immune complexes and FcγRs might therefore be involved in the pathogenesis of some arthritic conditions. To explore the relationship between functional polymorphisms in FcγRs (FCGR3A-158V/F and FCGR2A-131H/R) and arthritis in individuals positive for anti-GPI antibodies, we evaluated these individuals with respect to FCGR genotype. Genotyping for FCGR3A-158V/F and FCGR2A-131H/R was performed by PCR amplification of the polymorphic site, followed by site specific restriction digestion using the genome of 187 Japanese patients with rheumatoid arthritis (including 23 who were anti-GPI antibody positive) and 158 Japanese healthy individuals (including nine who were anti-GPI antibody positive). We report here on the association of FCGR3A-158V/F functional polymorphism with anti-GPI antibody positive status. Eight out of nine healthy individuals who were positive for anti-GPI antibodies possessed the homozygous, low affinity genotype FCGR3A-158F (odds ratio = 0.09, 95% confidence interval 0.01–0.89; P = 0.0199), and probably were 'protected' from arthritogenic antibodies. Moreover, among those who were homozygous for the high affinity genotype FCGR3A-158V/V, there were clear differences in anti-human and anti-rabbit GPI titres between patients with rheumatoid arthritis and healthy subjects (P = 0.0027 and P = 0.0015, respectively). Our findings provide a molecular model of the genetic regulation of autoantibody-induced arthritis by allele-specific affinity of the FcγRs.
==== Body
Introduction
Rheumatoid arthritis (RA) is a heterogeneous autoimmune disease that is characterized by chronic inflammatory polyarthritis [1]. One of the characteristic features of RA is the expression of several autoantibodies. The presence of such autoantibodies (e.g. rheumatoid factor [RF]), identified by screening, is commonly used as a diagnostic marker, although the pathogenic role played by autoantibodies in RA remains a mystery.
Fcγ receptors (FcγRs) play a pivotal role in the reaction between immune complex and myeloid cells. Three FcγR types have been identified in mice and humans (FcγRI, FcγRII and FcγRIII). In mouse arthritis models, FcγRIII deficient hosts exhibit resistance to collagen type II induced arthritis and anti-glucose-6-phosphate isomerase (GPI) antibody induced arthritis [2,3], suggesting that FcγRIII is indispensible in autoantibody dependent arthritis. In humans FcγRs are encoded by eight genes, and the genes encoding the low affinity FcγRs (FCGR2A, FCGR3A, FCGR2C, FCGR3B and FCGR2B) are located within a gene cluster on chromosome 1q22-23. Of these FcγRs, FcγRIIIa and FcγRIIa are known to be stimulatory receptors. Various genetic polymorphisms of these receptors were reported to be associated with several autoimmune diseases [4,5], one of which is a polymorphism in FCGR3A, with either a phenylalanine (F) or a valine (V) at amino acid position 158 [6,7]. Moreover, based on findings from a co-crystalization study with IgG1 and FcγRIIIa [8], this residue directly interacts with the lower hinge region of IgG1, suggesting strong binding between IgG1 and FcγRIIIa-158V on both natural killer cells and macrophages. For FCGR2A genes, a polymorphism at position 131 (with either histidine [H] or arginine [R]) alters the ability of the receptor to bind to certain IgG subclasses [9,10].
In RA patients, FCG3A-158V/F polymorphisms were reported to be frequent in UK Caucasian, North Indian and Pakistani individuals [11,12], but not in Japanese, Spanish and French individuals [13-15]. The reason for these differences between populations is unknown, although it is possible that they might depend on the prevalence in these populations of patients with autoantibody related forms of RA, in particular the prevalence of those who have pathogenic autoantibodies that directly interact with FcγRs (especially FcγRIIIa).
Anti-GPI antibodies are candidate arthritogenic antibodies. In K/B×N mice, polyclonal or two monoclonal anti-GPI antibodies induced arthritis in several strains of mice [16]. Moreover, FcγRIII deficient mice were resistant to anti-GPI antibody induced arthritis [3]. Another recent report [17] also confirmed that immune complex and FcγRIII are essential initiators of arthritis through sequential activation of effector cells, thus giving antibodies access into the joint. In human RA, anti-GPI antibodies have frequently been detected in patients with aggressive forms of arthritis [18,19], and their levels correlated significantly with extra-articular manifestations such as rheumatoid nodules, rheumatoid vasculitis and Felty's syndrome [20]. Moreover, a modest association of homozygosity for the FCGR3A-158V allele with RA in the nodular phenotype was suggested by Morgan and coworkers [11], suggesting the presence of a link between anti-GPI antibodies and FCGR3A allele. However, whether anti-GPI antibody positive status correlates with RA is a matter of controversy [18-22]. In our assay few healthy individuals retained anti-GPI antibodies; however, we do not know whether these protective phenotypes are associated with certain human gene polymorphisms.
In order to determine the relationship between functional polymorphisms of FCGR and possible arthritogenic anti-GPI antibodies in human conditions, we examined the correlation of these polymorphisms with anti-GPI positivity.
Materials and methods
Patients
The study was approved by the local ethics review committee and written informed consent was obtained from all participants. Blood samples were collected from 187 Japanese patients with RA (mean age 46 ± 17 years; 33 females; mean disease duration 12.9 years [range 1–46 years]) including four with vasculitis and three with Felty's syndrome. These patients, randomly selected from among patients visiting the clinic, were followed at University of Tsukuba Hospital. The diagnosis of RA was based on the criteria presented by the American College of Rheumatology [23]. In addition, 158 Japanese volunteers (mean age 30 ± 9 years; 105 females) were recruited from our institute to serve as a healthy comparison group. All healthy individuals were free of rheumatic disease symptoms, and derived from the same geographic locations.
Enzyme-linked immunosorbent assay for GPI
In order to select anti-GPI antibody positive patients, we used recombinant human GPI (described in detail previously [18]) or rabbit muscle GPI (Sigma, St Louis, MO, USA). Both antigens were used at 5 μg/ml (diluted in phosphate-buffered saline [PBS]) to coat microtitre plates (12 hours, 4°C). After washing twice with washing buffer (0.05% Tween 20 in PBS), Block Ace (diluted 1/4 in 1 × PBS; Dainippon Pharmaceuticals, Osaka, Japan) was used for saturation (30 min at 37°C). After two washes, sera (diluted 1/50) were added and the plates were incubated for 12 hours at 4°C. After washing, alkaline phosphatase (AP)-conjugated anti-human IgG (Fc fragment specific; Jackson Immuno Research, West Grove, PA, USA) was added to the plate (dilution 1/1000, for 1 hour at room temperature). After three washes, colour was developed with AP reaction solution (containing 9.6% diethanol amine, 0.25 mmol/l MgCl2; pH 9.8) with AP substrate tablets (Sigma; one AP tablet per 5 ml AP reaction solution). Plates were incubated for 1 hour at room temperature, and the optical density (OD) was measured by plate spectrophotometry at 405 nm. Determinations were performed in triplicate and standardized between experiments by reference to a highly positive human anti-GPI serum. The primary reading was processed by subtracting OD readings of control wells (coated with gluthathione-S-transferase (GST) and Block Ace for recombinant GPI–GST and rabbit GPI, respectively). The cutoff OD was calculated from the ELISA reactions of 158 healthy Japanese donors. Those who were double positive to both antigens were considered anti-GPI antibody positive. Because we used two antigens for the discrimination, the cutoff OD (mean value + 1 standard deviation) was 0.98 for human recombinant GPI and 0.64 for rabbit native GPI.
Genomic DNA was isolated from 0.5 ml anticoagulated peripheral blood, from 187 RA patients and 158 healthy individuals, by using DNA QuickII DNA purification kit (Dainippon Pharmaceuticals, Osaka, Japan). FcγR polymorphisms (FCGR3A-158V/F) were identified, as described by Koene and coworkers [6], using a nested PCR followed by allele specific restriction enzyme digestion. For homozygous FcγRIIIA-158F patients only one undigested band (94 bp) was visible. Three bands (94 bp, 61 bp and 33 bp) were seen in heterozygous individuals, whereas for homozygous FcγRIIIA-158V patients only two digested bands (61 bp and 33 bp) were detected (Fig. 1a). These genotyping findings were confirmed by direct sequencing in some individuals.
FcγRIIA-131H/R genotyping
Genotyping of FcγRIIA-131H/R also consisted of PCR followed by an allele specific restriction enzyme digestion, in accordance with the method reported by Jiang and coworkers [24]. The FCGR2A-131H and FCGR2A-131R alleles were visualized as 337 bp and 316 bp DNA fragments, respectively (Fig. 1b). These genotyping findings were confirmed by direct sequencing in some individuals.
Statistical analysis
The data were analyzed using the Student's t-test and the χ2 test, and Fisher's exact test was used when expected frequencies were lower than 5. We used Mann–Whitney U-test to evaluate the distribution of anti-GPI antibodies in FcγRIIIA-158V/V RA patients and healthy individuals. P < 0.05 was considered statistically significant.
Results
Our ELISA assay is highly specific because we used recombinant bacterial human GPI and native rabbit GPI, and double positivity for the two antibodies correlated significantly with the results of western blotting to GPI [18]. Because two GPI antigens were used for discrimination, the cutoff value of the OD was the mean value + one standard deviation from 158 healthy individuals, estimated using ELISA. Those who were positive for both antibodies were considered to be anti-GPI antibody positive. Using these definitions, 23 (12.3%) RA patients were anti-GPI antibody positive, and nine (5.7%) healthy individuals were anti-GPI antibody positive (Fig. 2). Statistical analysis revealed a significant difference in anti-GPI antibody positivity between RA patients and healthy individuals (χ2 = 4.438, with one degree of freedom; P = 0.0352).
To analyze whether functional FCGR polymorphisms were correlated with anti-GPI antibody positive and negative individuals, we performed FCGR genotyping. FCGR3A and FCGR2A genotypes in the control group were in Hardy–Weinberg equilibrium. The FCGR3A-158V allele (high affinity genotype) was more frequently identified in patients with RA than in healthy individuals within the anti-GPI antibody positive population (χ2 = 0.012, with one degree of freedom; P = 0.012; Tables 1 and 2). In addition, these differences were evident when individuals were categorized according to the presence or absence of these genotypes: 56.5% of patients with RA were homozygous or heterozygous with respect to FCGR3A-158V, as compared with 11.1% of healthy individuals; and 43.5% of patients with RA were homozygous with respect to FCGR3A-158F, as compared with 88.9% of healthy individuals (χ2 = 5.42 with one degree of freedom; P < 0.02; Tables 1 and 2). Comparison of FCGR3A-158V allele frequency between RA patients and healthy individuals revealed no statistically significant difference: 48.7% of patients with RA were homozygous or heterozygous with respect to FCGR3A-158V, as compared with 42.4% of healthy individuals; and 51.3% of patients with RA were homozygous with respect to FCGR3A-158F, as compared with 57.6% of healthy individuals (χ2 = 1.04 with one degree of freedom; P = 0.245; Table 1).
Next, FCGR2A genotyping was conducted in the same cohort (Table 1). In contrast to FCGR3A, the frequency of the FCGR2A-131H allele (high affinity genotype) was not significantly different between the two groups within the anti-GPI antibody positive population (χ2 = 0.862 with one degree of freedom; P = 0.35; Tables 1 and 2). These differences were also not evident when individuals were categorized according to the presence or absence of these genotypes (P = 0.19; Tables 1 and 3).
We also analyzed the association between FcγR and other related autoantibodies such as RF. There was no difference between RF positive and RF negative populations of RA patients (P = 0.82 and P = 0.4 for FCGR3A and FCGR2A, respectively; Table 4).
Finally, in order to identify the relationship between FCGR3A-158V allele and anti-GPI antibodies more clearly, we focused on individuals who were homozygous for the high affinity FCGR3A-158V/V genotype (14 RA patients and eight healthy individuals) and compared their anti-GPI antibody titres. Surprisingly, both anti-human GPI antibodies and anti-rabbit GPI antibodies were significantly elevated in the RA group (P = 0.0027 and P = 0.0015 for anti-human GPI antibodies and anti-rabbit GPI antibodies, respectively, by Mann–Whitney U-test; Fig. 3). This suggests that anti-GPI antibody positivity might predispose individuals with the FCGR3A-158V/V genotype to arthritis.
Discussion
Several studies have indicated that anti-GPI antibodies are potential arthritogenic antibodies [18-20] because they were frequently detected in patients with severe forms of RA. Because high titres of these antibodies (IgG, not IgM) were also detected in healthy individuals, the arthritogenicity of these antibodies should be due to modulation – by the low affinity genotype of FcγRs – of the bypass between immune complex and FcγR bearing cells. In a GPI immunized mouse model severe arthritis occurred only in DBA/1 mice, although the production of anti-GPI antibodies was almost equal in arthritis susceptible and resistant mouse strains [25]. Thus, the incidence of arthritis might depend on certain genetic factors such as FcγR. Anti-GPI antibody positive individuals express several GPI variant mRNAs in peripheral blood monocytes [26]. This observation supports the notion that the presence of GPI variants is necessary to produce anti-GPI autoantibodies, and that genetic factors such as FcγRIIIA are important in the development of arthritis. Based on this conclusion, it is conceivable that the production of anti-GPI antibodies does not occur as a 'result' of joint destruction.
Our results do not indicate that individual polymorphisms in the FCGR3A and FCGR2A genes play roles in susceptibility to RA. Despite the lack of association with individual FCGR polymorphisms in the whole cohort, our studies suggest that FCGR3A-158V/F polymorphisms play a crucial role in RA among those individuals who are positive for anti-GPI antibodies (Tables 2 and 3). Moreover, focusing on FCGR3A-158V/V homozygous individuals, anti-GPI antibodies were clearly evident in patients with RA. These findings suggest that anti-GPI antibodies might have arthritogenic potential in individuals homozygous for FCGR3A-158V/V.
Conclusion
Our findings show that FCGR3A-158V/F functional polymorphisms were associated with RA among anti-GPI antibody positive individuals. This is the first report on possible mechanisms of arthritic diseases; they are tightly regulated by some genes, especially by FcγR genotype, as well as by production of arthritogenic autoantibodies.
Abbreviations
AP = alkaline phosphatase; bp = base pairs; ELISA = enzyme-linked immunosorbent assay; FcγR = Fcγ receptor; GPI = glucose-6-phosphate isomerase; GST = gluthathione-S-transferase; OD = optical density; PBS = phosphate-buffered saline; PCR = polymerase chain reaction; RA = rheumatoid arthritis; RF = rheumatoid factor.
Competing interests
The author(s) declare that they have no competing interests.
Authors' contributions
IM wrote the manuscript and conceived the study. HZ performed FcγR genotyping and coordinated the statistical analysis. YM, TY and YK performed GPI ELISA. TH participated in clinical assessment. TS participated in the full design and coordination of the study, and DG, SI and AT participated in writing the discussion.
Acknowledgements
This work was supported in part by the Japanese Ministry of Science and Culture (IM, TS). IM is also a recipient of a fellowship from the Japan Intractable Diseases Research Foundation, Uehara Memorial Foundation, and Japan Rheumatoid Foundation.
Figures and Tables
Figure 1 PCR-RFLP analysis of the FCGR3A and FCGR2A genes. cDNA was amplified with primers and restriction digested using appropriate enzymes. Digested PCR products were visualized with ethidium bromide. (a) FCGR3A gene and (b) FCGR2A gene. ND, nondigested PCR product; RE, restriction enzyme.
Figure 2 Population of anti-GPI antibody positive individuals, and FCGR3A and FCGR2A genotypes. The study included 187 patients with rheumatoid arthritis and 158 healthy Japanese individuals. The horizontal and vertical dotted lines represent the cutoff optical density values calculated from ELISA reactions of 158 healthy individuals for human recombinant GPI and rabbit native GPI, respectively. Individuals positive for both antibodies were considered anti-GPI antibody positive. Numbers in each graph represent the proportions of individuals positive for anti-GPI antibodies relative to the total number of individuals in that group. GPI, glucose-6-phosphate isomerase; HS, healthy subjects; RA, rheumatoid arthritis.
Figure 3 Higher titres of anti-human and anti-rabbit GPI antibodies in FCGR3A-158V/V RA patients versus healthy individuals. In individuals homozygous for the FCGR3A high affinity V/V genotype (14 RA patients and 8 healthy individuals), both anti-human GPI antibodies and anti-rabbit GPI antibodies were significantly elevated in the RA group (P = 0.0027 and P = 0.0015 for anti-human GPI antibodies and anti-rabbit GPI antibodies, respectively, by Mann–Whitney U-test). GPI, glucose-6-phosphate isomerase; RA, rheumatoid arthritis.
Table 1 Frequencies of FCGR3A and FCGR2A genotypes in patients with RA and positive and negative for anti-GPI antibodies
FCGR3A-158 FCGR2A-131
FF low F/V VV high HH high H/R RR low
GPI+ RA (n = 23) 10 (43.5) 9 (39.1) 4 (17.4) 16 (69.6) 6 (26.1) 1 (4.3)
GPI- RA (n = 164) 86 (52.4) 68 (41.5) 10 (6.1) 128 (78) 29 (17.7) 7 (4.3)
GPI+ Control (n = 9) 8(88.9) 1 (11.1) 0 (0) 4 (44.4) 5 (55.6) 0 (0)
GPI- Control (n = 149) 83 (55.7) 58 (38.9) 8 (5.4) 109 (73.2) 40 (26.8) 0 (0)
Data are expressed as number (percentage) of individuals. GPI, glucose-6-phosphate isomerase; high, high affinity genotype; low, low affinity genotype; RA, rheumatoid arthritis.
Table 2 Alleic skewing of FCGR3A and FCGR2A in anti-GPI antibody positive healthy individuals
Polymorphism Allele RA GPI+ (n = 46) Healthy GPI+ (n = 18) P (χ2) P (Fisher's) OR (95% CI)
FCGR3A-158 F 29 17 0.012 0.013 0.10 (0.01–0.82)
V 17 1
FCGR2A-131 H 38 13 0.35 0.4902 1.83 (0.51–6.59)
R 8 5
P values are given for RA versus healthy individuals using a 2×2 contingency table. CI, confidence interval; Fisher's, Fisher's probability test; OR, odds ratio; RA, rheumatoid arthritis.
Table 3 Genotype skewing of FCGR3A and FCGR2A gene polymorphisms in anti-GPI antibody positive healthy individuals
Polymorphism Genotype RA GPI+ (n = 23) Healthy GPI+ (n = 9) P (χ2) P (Fisher's) OR (95% CI)
FCGR3A-158 FF 10 (43.5%) 8 (88.9%) 0.019 0.044 0.09 (0.01–0.89)
FV/VV 13(56.5%) 1 (11.1%)
FCGR2A-131 HH 16 (69.6%) 4(44.4%) 0.19 0.24 2.86 (0.58–13.96)
HR/RR 7 (30.4%) 5 (55.6%)
P values are given for RA versus healthy individuals using a 2×2 contingency table. CI, confidence interval; Fisher's, Fisher's probability test; OR, odds ratio; RA, rheumatoid arthritis.
Table 4 FCGR3A and FCGR2A genotypes in rheumatoid factor positive and negative RA patients
Polymorphism Genotype RA RF+ (n = 130) RA RF- (n = 57) P (χ2) OR (95% CI)
FCGR3A-158 FF 66 (50.8%) 30(52.6%) 0.82 0.93 (0.50–1.73)
FV/VV 64(49.2%) 27 (47.4%)
FCGR2A-131 HH 103 (79.2%) 42(73.7%) 0.4 1.36 (0.66–2.82)
HR/RR 27 (20.8%) 15 (26.3%)
P values are given for RA RF+ versus RA RF- using a 2×2 contingency table. CI, 95% confidence interval; OR, odds ratio; RA, rheumatoid arthritis; RF, rheumatoid factor.
==== Refs
Firestein GS Evolving concepts of rheumatoid arthritis Nature 2003 423 356 361 12748655 10.1038/nature01661
Diaz de Stahl T Andren M Martinsson P Verbeek JS Kleinau S Expression of FcgammaRIII is required for development of collagen-induced arthritis Eur J Immunol 2002 32 2915 2922 12355445 10.1002/1521-4141(2002010)32:10<2915::AID-IMMU2915>3.0.CO;2-4
Ji H Ohmura K Mahmood U Lee DM Hofhuis FM Boackle SA Takahashi K Holers VM Walport M Gerard C Arthritis critically dependent on innate immune system players Immunity 2002 16 157 168 11869678 10.1016/S1074-7613(02)00275-3
Dijstelbloem HM Scheepers RH Oost WW Stegeman CA van der Pol WL Sluiter WJ Kallenberg CG van de Winkel JG Tervaert JW Fcgamma receptor polymorphisms in Wegener's granulomatosis: risk factors for disease relapse Arthritis Rheum 1999 42 1823 1827 10513795 10.1002/1529-0131(199909)42:9<1823::AID-ANR5>3.0.CO;2-X
Myhr KM Raknes G Nyland H Vedeler C Immunoglobulin G Fc-receptor (FcgammaR) IIA and IIIB polymorphisms related to disability in MS Neurology 1999 52 1771 1776 10371522
Koene HR Kleijer M Algra J Roos D von dem Borne AE de Haas M Fc gammaRIIIa-158V/F polymorphism influences the binding of IgG by natural killer cell Fc gammaRIIIa, independently of the Fc gammaRIIIa-48L/R/H phenotype Blood 1997 90 1109 1114 9242542
Wu J Edberg JC Redecha PB Bansal V Guyre PM Coleman K Salmon JE Kimberly RP A novel polymorphism of FcgammaRIIIa (CD16) alters receptor function and predisposes to autoimmune disease J Clin Invest 1997 100 1059 1070 9276722
Sondermann P Huber R Oosthuizen V Jacob U The 3.2-A crystal structure of the human IgG1 Fc fragment-Fc gammaRIII complex Nature 2000 406 267 273 10917521 10.1038/35018508
Warmerdam PA van de Winkel JG Vlug A Westerdaal NA Capel PJ A single amino acid in the second Ig-like domain of the human Fc gamma receptor II is critical for human IgG2 binding J Immunol 1991 147 1338 1343 1831223
Parren PW Warmerdam PA Boeije LC Arts J Westerdaal NA Vlug A Capel PJ Aarden LA van de Winkel JG On the interaction of IgG subclasses with the low affinity Fc gamma RIIa (CD32) on human monocytes, neutrophils, and platelets. Analysis of a functional polymorphism to human IgG2 J Clin Invest 1992 90 1537 1546 1401085
Morgan AW Griffiths B Ponchel F Montague BM Ali M Gardner PP Gooi HC Situnayake RD Markham AF Emery P Isaacs JD Fcgamma receptor type IIIA is associated with rheumatoid arthritis in two distinct ethnic groups Arthritis Rheum 2000 43 2328 2334 11037893 10.1002/1529-0131(200010)43:10<2328::AID-ANR21>3.0.CO;2-Z
Morgan AW Keyte VH Babbage SJ Robinson JI Ponchel F Barrett JH Bhakta BB Bingham SJ Buch MH Conaghan PG FcgammaRIIIA-158V and rheumatoid arthritis: a confirmation study Rheumatology (Oxford) 2003 42 528 533 12649399 10.1093/rheumatology/keg169
Kyogoku C Tsuchiya N Matsuta K Tokunaga K Studies on the association of Fc gamma receptor IIA, IIB, IIIA and IIIB polymorphisms with rheumatoid arthritis in the Japanese: evidence for a genetic interaction between HLA-DRB1 and FCGR3A Genes Immun 2002 3 488 493 12486608 10.1038/sj.gene.6363921
Radstake TR Petit E Pierlot C van de Putte LB Cornelis F Barrera P Role of Fcgamma receptors IIA, IIIA, and IIIB in susceptibility to rheumatoid arthritis J Rheumatol 2003 30 926 933 12734884
Nieto A Caliz R Pascual M Mataran L Garcia S Martin J Involvement of Fcgamma receptor IIIA genotypes in susceptibility to rheumatoid arthritis Arthritis Rheum 2000 43 735 739 10765917 10.1002/1529-0131(200004)43:4<735::AID-ANR3>3.0.CO;2-Q
Matsumoto I Staub A Benoist C Mathis D Arthritis provoked by linked T and B cell recognition of a glycolytic enzyme Science 1999 286 1732 1735 10576739 10.1126/science.286.5445.1732
Wipke BT Wang Z Kim J McCarthy TJ Allen PM Dynamic visualization of a joint-specific autoimmune response through positron emission tomography Nat Immunol 2002 3 366 372 11896393 10.1038/ni775
Matsumoto I Lee DM Goldbach-Mansky R Sumida T Hitchon CA Schur PH Anderson RJ Coblyn JS Weinblatt ME Brenner M Low prevalence of antibodies to glucose-6-phosphate isomerase in patients with rheumatoid arthritis and a spectrum of other chronic autoimmune disorders Arthritis Rheum 2003 48 944 954 12687536 10.1002/art.10898
Schaller M Burton DR Ditzel HJ Autoantibodies to GPI in rheumatoid arthritis: linkage between an animal model and human disease Nat Immunol 2001 2 746 753 11477412 10.1038/90696
van Gaalen FA Toes RE Ditzel HJ Schaller M Breedveld FC Verweij CL Huizinga TW Association of autoantibodies to glucose-6-phosphate isomerase with extraarticular complications in rheumatoid arthritis Arthritis Rheum 2004 50 395 399 14872481 10.1002/art.20028
Kassahn D Kolb C Solomon S Bochtler P Illges H Few human autoimmune sera detect GPI Nat Immunol 2002 3 411 412 11976712 10.1038/ni0502-411b
Schubert D Schmidt M Zaiss D Jungblut PR Kamradt T Autoantibodies to GPI and creatine kinase in RA Nat Immunol 2002 3 411 11976711 10.1038/ni0502-411a
Arnett FC Edworthy SM Bloch DA McShane DJ Fries JF Cooper NS Healey LA Kaplan SR Liang MH Luthra HS The American Rheumatism Association 1987 revised criteria for the classification of rheumatoid arthritis Arthritis Rheum 1988 31 315 324 3358796
Jiang XM Arepally G Poncz M McKenzie SE Rapid detection of the Fc gamma RIIA-H/R 131 ligand-binding polymorphism using an allele-specific restriction enzyme digestion (ASRED) J Immunol Methods 1996 199 55 59 8960098 10.1016/S0022-1759(96)00164-0
Schubert D Maier B Morawietz L Krenn V Kamradt T Immunization with glucose-6-phosphate isomerase induces T cell-dependent peripheral polyarthritis in genetically unaltered mice J Immunol 2004 172 4503 4509 15034067
Muraki Y Matsumoto I Chino Y Hayashi T Suzuki E Goto D Ito S Murata H Tsutsumi A Sumida T Glucose-6-phosphate isomerase variants play a key role in the generation of anti-GPI antibodies: possible mechanism of autoantibody production Biochem Biophys Res Commun 2004 323 518 522 15369782 10.1016/j.bbrc.2004.08.123
|
16277670
|
PMC1297563
|
CC BY
|
2021-01-04 16:02:40
|
no
|
Arthritis Res Ther. 2005 Aug 11; 7(6):R1183-R1188
|
utf-8
|
Arthritis Res Ther
| 2,005 |
10.1186/ar1802
|
oa_comm
|
==== Front
Arthritis Res TherArthritis Research & Therapy1478-63541478-6362BioMed Central London ar18041627767110.1186/ar1804Research ArticleThe contact-mediated response of peripheral-blood monocytes to preactivated T cells is suppressed by serum factors in rheumatoid arthritis Rossol Manuela [email protected]äuser Sylke [email protected] Roger [email protected]äntzschel Holm [email protected] Sunna 3Wagner Ulf [email protected] Department of Medicine IV, University of Leipzig, Leipzig, Germany2 Department of Orthopedics, University of Leipzig, Leipzig, Germany3 Department of Immunobiology, Institute of Zoology, University of Leipzig, Leipzig, Germany2005 17 8 2005 7 6 R1189 R1199 3 3 2005 17 3 2005 16 7 2005 20 7 2005 Copyright © 2005 Rossol et al.; licensee BioMed Central Ltd.This is an Open Access article distributed under the terms of the Creative Commons Attribution License (), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
Stimulation of monocytes/macrophages after cell contact with preactivated T cells has been suggested to contribute to the excessive TNF-α production in rheumatoid arthritis (RA). In this study, T cell-contact-dependent TNF-α production by peripheral-blood monocytes in vitro was investigated and found to be significantly lower in treated and untreated patients with RA than in healthy controls. This suppression was not due to a general deficiency of monocytes to respond, because responses to lipopolysaccharide were comparable in patients and controls. In agreement with the pivotal role of TNF-α in RA, T cell-dependent induction of TNF-α in synovial macrophages was fivefold to tenfold higher than in peripheral-blood monocytes from either patients or controls. The decreased response of peripheral-blood monocytes from patients with RA was found to be mediated by inhibitory serum factors, because the addition of patient sera to monocytes from healthy controls suppressed TNF-α response in the co-culture assay. Preincubation of monocytes from healthy controls with RA serum was sufficient to suppress the subsequent TNF-α response in T cell co-cultures, indicating that inhibitory factors do indeed bind to monocyte surfaces, which might represent a regulatory counter-action of the immune system to the long-standing and consuming autoimmune process in RA. There are some indications that apolipoprotein A-1 might be part of this regulatory system.
==== Body
Introduction
Cytokine production by monocytes/macrophages at sites of active inflammation is an important mechanism in the initiation and perpetuation of various chronic autoimmune diseases including type I diabetes mellitus, multiple sclerosis and rheumatoid arthritis (RA). The signals triggering this proinflammatory cytokine secretion in vivo are not completely understood. Although bacterial endotoxins such as lipopolysaccharide (LPS) and other microbial products are major stimuli of monocyte activation in infectious diseases, they are not considered to be relevant stimuli in autoimmune settings. So far, the most powerful known pathway inducing monocyte cytokine secretion in vivo in non-infectious situations is the direct cellular interaction with preactivated T cells [1].
The cell surface molecules involved in this T cell-dependent monocyte activation have been extensively investigated. T cell surface molecules, some of them upregulated on activation, such as CD69 [2,3], CD23 [4,5], integrins [6], CD40-CD40 ligand [7], LAG-3 [8], CD45 [9], LFA-1 and ICAM-1 [10] and membrane-bound cytokines [11] have all been implicated in this activation.
In RA, elevated levels of monocytic cytokines such as tumour necrosis factor (TNF)-α and IL-1β are present in the synovial membrane of diseased joints. Their relevance to disease pathogenesis has been highlighted by the clinical success of therapeutic strategies neutralizing TNF-α or IL-1β [12-14]. CD4+ T cells, in contrast, have initially been implied in the pathogenesis of the disease because of the association of related HLA DRB1 alleles with susceptibility to and severity of the disease, and have subsequently been found to exhibit numerous pathological features such as oligoclonal expansions, contraction of T cell receptor repertoire, shortened telomere fragments indicative of increased replicative history, and distinct pathological phenotypes [15-19]. The traditionally described paucity of cytokines of T cell origin in the RA synovial membrane, which has been considered an argument against the involvement of T cells in the pathogenesis of the disease, has been put into perspective by the more recent recognition of high levels of IL-15 [20], IL-17 and, although at low levels, IFN-γ [21] in rheumatoid joints.
The monocytic production of TNF-α and IL-1 in RA synovial membranes seems to be independent of T cell cytokines. It has therefore been suggested that the direct interaction of activated T cells with monocytes, rather than the T cell-based production of cytokines, is a major stimulus of the excessive levels of TNF-α and IL-1β in RA. In addition, monocytes have been shown to be able to produce matrix metalloproteinases after contact with T helper type 2 (Th2) cells [22], which further implicates this interaction in the breakdown of extracellular matrix and subsequent joint destruction in RA.
To investigate cell-contact-mediated activation of monocytes by preactivated T cells, monocytic cell lines or monocytes from healthy donors have been primarily used in co-culture with T cells from healthy donors [8,11,23-26]. Disease-relevant CD4+ T cells isolated from the synovial membrane of patients with RA were also used as stimulators and found to be potent inducers of cytokine production in monocytic THP-1 cells [27], in monocytes from healthy donors [10,28] and in mononuclear cells isolated from synovial membranes of patients with RA [29]. So far, peripheral-blood monocytes of patients with RA have not been analysed for T cell-dependent cytokine secretion, although the involvement of circulating monocytes in the disease process has been suggested [30-32]. Here we show that the T cell-dependent response of monocytes is suppressed in RA and that serum factors contribute to this inhibition, most probably by coating monocyte cell surfaces.
Materials and methods
Patients and controls
Twenty patients with RA as defined by the 1987 revised criteria of the American College of Rheumatology [33] were enrolled into the initial study. The study design was approved by the University of Leipzig's Ethics Committee, and informed consent was obtained from each patient before study enrolment. Sixteen of the patients had rheumatoid factor (RF) IgM seropositive disease, and 15 patients expressed the RA-associated shared epitope (on either a DRB1*01 or a DRB1*04 allele).
The median age of the patients was 59 years (range 22 to 74). For each patient, an age-matched control was selected from healthy volunteers. Clinical parameters documented at the time of presentation to the outpatient department included tender and swollen joint count, and concentrations of C-reactive protein (CRP) and of RF IgM.
Six patients had recent-onset RA (disease duration less than 2 years) and had not received therapy with disease-modifying anti-rheumatic drugs (DMARD) before inclusion in the study. The median disease duration for the 14 patients receiving DMARD was 9.5 years (interquartile range 5.0 to 11.0). Current treatment regimens included methotrexate alone (n = 5) or in combination with cyclosporine A (n = 1) or hydroxychloquine (n = 1), intramuscular gold injections (n = 1), leflunomide (n = 1) or TNF-α-blocking agents (n = 5).
For analysis of the T cell-dependent response of synovial macrophages, synovial biopsy specimens were obtained from six patients with RA and active synovitis who underwent synovectomy in the Department of Orthopedics at Leipzig University. Synovectomized joints were elbow joints (n = 2), metacarpophalangeal joints (n = 1), ankle joints (n = 2), knee joint (n = 1) and wrist (n = 1). Five of the patients had RF IgM-positive disease; the median CRP was 14.6 mg/l. Three of the patients received no immunosuppressive therapy, whereas two patients were treated with methotrexate and one was treated with etanercept.
To explore the influence of clinical parameters of the disease on inhibitory serum activity, two additional patient cohorts were recruited. Twenty patients with non-active disease (CRP below 10 mg/l and not more than swollen joints) were identified from a pre-existing institutional serum bank and compared with 20 patients with intermediate to high disease activity (mean CRP 35.6, median swollen joint count 14, median tender joint count 18). As control groups with systemic inflammatory disease, sera from 9 patients with ankylosing spondylitis and from 9 patients with psoriatic arthritis were used.
Isolation of monocytes
Peripheral-blood mononuclear cells (PBMCs) were obtained by Ficoll®-Paque (Pharmacia Biotech, Freiburg, Germany) density-gradient centrifugation [34]. After repeated washing in PBS containing EDTA, monocytes were isolated by negative magnetic depletion with hapten-conjugated CD3, CD7, CD19, CD45RA, CD56 and anti-IgE antibodies (MACS; Miltenyi Biotech, Bergisch Gladbach, Germany) and a magnetic cell separator (MACS) in accordance with the manufacturer's protocol. The cell preparations were more than 95% monocytes as determined by morphology and immunofluorescence staining with a monoclonal antibody against CD14 (BL-M/G14).
To obtain larger amounts of monocytes, PBMCs were separated by counterflow centrifugation with the J6-MC elutriator system (Beckmann Instruments, Palo Alto, CA, USA) as described previously [35]. The cell preparations were more than 90% monocytes as determined by morphology and immunofluorescence staining with a monoclonal antibody against CD14 (BL-M/G14). In the co-culture assays described below, the response of monocytes separated by this technique was indistinguishable from that of monocytes obtained by immunomagnetic separation (data not shown).
Separation of human synovial macrophages from patients with RA
Synovial tissue specimens were cut into 2 to 4 mm3 pieces and washed once in complete medium (RPMI 1640, 10% FCS, 200 μM L-glutamine, 100 U/ml penicillin, 100 μg/ml streptomycin). Then, 1 cm3 of tissue was incubated in 10 ml of digestion solution (0.05 M HEPES buffer, 3 mg/ml type 1A collagenase, 1 mg/ml hyaluronidase, 0.1 mg/ml type IV deoxyribonuclease I in RPMI 1640) at 37°C for 30 to 45 min [36]. Tissue residues were removed, and the resulting single cell suspension was washed twice.
Synovial macrophages were isolated by positive magnetic separation with CD14-conjugated magnetic beads (MACS; Miltenyi Biotech) and a magnetic cell separator (MACS) in accordance with the manufacturer's protocol.
Separation and stimulation of human T cells
Human T cells were isolated by counterflow elutriation from PBMC as described above. T cells (106/ml) were cultured in RPMI 1640 supplemented with 5% human AB serum, 2 mM glutamine, 100 U/ml penicillin and 100 μg/ml streptomycin in culture flasks (Techno Plastic Products AG, Trasadingen, Switzerland) at 37°C and 5% CO2. To stimulate T cells, culture flasks were coated with 3.3 μg/ml monoclonal anti-CD3ε antibodies (R & D Systems Inc., Minneapolis, MN, USA) and soluble monoclonal anti-CD28 antibodies (BD Biosciences Pharmingen, San Diego, CA, USA) were added to the medium at a concentration of 0.8 μg/ml. After stimulation and incubation for 2 days, the cultures contained more than 90% CD3+ T cells as determined by flow cytometric analysis with a monoclonal antibody against CD3. Cells were then washed three times with PBS, fixed for 1 min with 0.05% glutaraldehyde [11] and washed again three times with PBS. Fixed T cells were stored for up to 2 weeks at 4°C.
This method of cell fixation was shown to inhibit blast transformation and TNF-α and IL-2 production in response to phorbol 12-myristate 13-acetate and ionomycin (data not shown).
Stimulation of human monocytes with T cells
Monocytes (1.5 × 106/ml) were cultured in RPMI 1640 supplemented with 10% FCS, 2 mM glutamine, 100 U/ml penicillin and 100 μg/ml streptomycin in 96-well culture plates (Techno Plastic Products AG) at 37°C and 5% CO2. Fixed T cells were added at a T cell : monocyte ratio of 7:1 (or any other indicated ratio) and cells were incubated together for 24 hours (or any other indicated time). LPS (Escherichia coli O55:B5, 100 ng/ml) was used as a positive control for monocyte cytokine production. In some experiments, semi-permeable Anopore membrane inserts (0.02 μm pore size; Nunc Life Technologies) were fitted into the culture wells to separate the monocytes (lower chamber) physically from the T cells (upper chamber). After incubation, supernatants (200 μl per well, two wells per condition) were harvested and stored at -140°C until cytokine concentration was determined.
Data presented in this work show interactions of T cells with monocytes from different donors. Similar results were obtained when T cells were incubated with autologous monocytes (data not shown).
Analysis of inhibitory effects of serum samples in co-culture assays
In some cell–cell contact experiments, monocytes from healthy donors were incubated with human sera either from healthy individuals or from patients with RA. In these experiments, FCS was replaced by 10% heat-inactivated, heterologous human serum matched for blood types. Sera were incubated at 56°C for 30 min, at 70°C for 10 min or at 95°C for 2 min before their addition to the co-culture assay. When preincubation of monocytes (106/ml) with human sera was performed, the monocytes were incubated in RPMI 1640 with 50% heat-inactivated (56°C for 30 minutes) serum at 20°C for 30 min. Monocytes were subsequently washed three times in PBS before proceeding to the standard co-culture assay as described above.
Cytokine measurement
TNF-α concentrations were determined with a commercially available enzmye immunoassay (Beckman Coulter, Coulter-Immunotech, Krefeld, Germany) in accordance with the manufacturer's protocol.
Apolipoprotein A-1 measurement
The serum concentration of apolipoprotein A-1 (ApoA-1) was determined by nephelometry by using a commercially available test kit (N antisera against human ApoA-1; Dade Behring, Liederbach, Germany).
Statistical analysis
For statistical analysis, the software package Sigma Stat (SPSS Inc., Chicago, IL, USA) was used. Before all comparisons, a normality test (Kolmogorov–Smirnov test with Lilliefors' correction) was performed. Student's t-test or the Mann–Whitney rank sum test were used for comparisons where appropriate. To compare cytokine production in the patient population with that of age-matched controls, pairwise comparisons were performed.
Results
T cell-dependent induction of TNF-α production in monocytes
To explore the requirements and dynamics of T cell-dependent monocyte stimulation, an in vitro co-culture system using glutaraldehyde-fixed T cells as stimulator cells was established as described previously [11]. As a positive control, LPS, a potent stimulator of monocytes, was used in all experiments.
The addition of fixed, heterologous T cells preactivated by immobilized anti-CD3 and anti-CD28 antibodies against CD14+ peripheral-blood monocytes induced monocyte TNF production in a manner dependent on T cell concentration (Fig. 1a). Maximum stimulation of TNF-α production occurred at T cell : monocyte ratios between 5:1 and 8:1. Contact of monocytes with resting T cells even at the highest T cell : monocyte ratio did not lead to significant TNF-α production in those experiments (data not shown), and induced only a modest but not statistically significant increase in a later set of co-culture experiments (Fig. 1b). All subsequent experiments were performed at a T cell : monocyte ratio of 7:1, which was also used in several previously published studies with this experimental system [23,25,37]. To analyse the influence of the T cell origin on the monocyte response mediated by T cell contact, autologous and heterologous co-cultures were performed. Contact of monocytes with autologous or heterologous preactivated T cells led to the same amount of TNF-α (data not shown), so a heterologous co-culture system was used for subsequent experiments.
To exclude the influence of soluble mediators released by the fixed T cells, transwell experiments using a tissue culture plate insert with a microporous membrane with a pore size of 0.02 μm were performed. Monocytes placed in the bottom chamber of the transwell system, which had no physical contact with the prestimulated T cells present in the top chamber, failed to produce detectable amounts of TNF-α, indicating that cell-contact-dependent stimuli were necessary for monocyte activation (Fig. 1b). When the time course of monocyte TNF-α production after cell contact with preactivated T cells was analysed, a distinct kinetic profile comparable to that seen after stimulation with LPS was observed (Fig. 1c).
T cell-induced TNF-α production of peripheral monocytes from patients with RA is decreased compared with TNF-α production in healthy donors
Although T cell–monocyte interaction has been proposed to contribute to the abundant TNF-α production seen in this disease, the role of peripheral monocytes from patients with RA in this interaction has not yet been investigated. To address this issue, CD14+ cells were isolated from the peripheral circulation of patients with RA and from age-matched controls by negative immunomagnetic separation and were subsequently used in the co-culture assay.
As seen in Fig. 2a, co-incubation of monocytes of patients with RA with preactivated fixed T cells resulted in significantly lower levels of TNF-α than in the controls. To exclude a global defect of monocyte cytokine production in RA, monocytes from patients with RA were stimulated with LPS as a positive control. In contrast to the T cell-induced TNF-α response, no difference in LPS-induced TNF-α production by monocytes was found between patients with RA and healthy controls.
In view of this unexpected finding, and with regard to the pivotal role of TNF-α in synovitic joints in RA, CD14+ cells were separated with CD14+ MicroBeads from synovial membrane biopsies from patients with RA, and tested for their capacity to produce TNF-α in the co-culture system. In contrast to the peripheral-blood monocytes from patients with RA, synovial mononuclear cells were found to be highly preactivated and to produce large amounts of TNF-α in the presence of resting T cells. In addition, they were found to increase their TNF-α production further after co-culture with preactivated fixed T cells and after stimulation with LPS (Fig. 2b). In parallel experiments, the influence of cell separation by CD14+ MicroBeads and by negative immunomagnetic purification was compared. No significant differences between the two separation techniques with regard to the TNF-α response of synovial monocytes/macrophages were detectable. Most notably, however, the concentrations of TNF-α elicited in synovial monocytes/macrophages by T cell contact were fivefold to tenfold higher than those of peripheral-blood monocytes from either healthy donors or patients with RA under comparable experimental conditions. This result indicates that synovial cell populations are indeed likely to be the major source of the increased TNF-α load in RA, whereas peripheral monocytes are not.
To assess the influence of immunosuppressive therapy on T cell-dependent monocyte TNF-α production, patients were stratified into two groups: one of patients who were receiving DMARD therapy at the time of the study (n = 14) and one of patients with recent-onset RA who had not received steroid or DMARD medication before study enrolment (n = 6). T cell-dependent TNF-α production by peripheral-blood monocytes in the six patients with recent-onset RA was not significantly different (635 ± 210 pg/ml, n = 6) from that of patients who had received therapy (835 ± 233 pg/ml, n = 14).
When TNF-α production by monocytes from untreated patients with RA, who were age-matched with healthy controls, was analysed, monocytes from untreated patients produced significantly less TNF-α in a T cell-dependent manner (635 ± 210 pg/ml, n = 6) than those from controls (1,648 ± 398 pg/ml, n = 6, P = 0.048). Fig. 2c depicts the results from patients treated with methotrexate (n = 5), untreated patients with RA (n = 6) and age-matched healthy controls (n = 6).
Analysis of clinical and laboratory parameters of all patients tested did not reveal any association between T cell-dependent cytokine production and disease activity, RF status, immunogenetic markers, disease duration or the patient's age. Similarly, no age-dependent decline of cytokine production was observed in the age-matched control group.
RA sera inhibit T cell-dependent TNF-α production by monocytes from healthy donors
Since monocytes from patients with RA were able to respond to LPS stimulation similarly to monocytes from healthy controls, we proposed that their diminished response in the co-culture assay was due to a disease-specific inhibitory mechanism present in the systemic circulation of patients with RA.
To test this hypothesis, serum exchange experiments were performed, in which monocytes from healthy donors were incubated with prestimulated T cells in the presence of 10% serum from either patients with RA or healthy controls. To avoid monocyte stimulation through blood type antigen-specific antibodies in the RA sera, patients and controls were matched for blood groups.
The addition of sera from healthy controls to the co-cultures was found to inhibit T cell induced production of TNF-α when compared with control cultures containing culture medium supplemented with FCS (Fig. 3a). This is in line with previous observations reporting the inhibition of T cell-dependent monocyte activation by serum from healthy controls [25].
When sera from patients with RA were added, they also suppressed TNF-α production (Fig. 3a). In comparison with the control sera, this inhibition was found to be significantly stronger. A difference between RA sera and control sera was also seen in experiments in which monocytes were preincubated with either sera and then co-cultured with T cells in the standard FCS-supplemented culture medium (Fig. 3b). Whereas preincubation of monocytes in RA sera was found to induce the inhibitory effect, preincubation with healthy sera was not sufficient to influence TNF-α production (Fig. 3b). In these experiments, we used sera from patients with active RA that had previously been shown to exhibit a pronounced inhibitory activity.
To analyse the influence of clinical parameters of disease activity on the inhibitory activity exhibited by the sera from patients with RA, sera from 20 patients with non-active disease were compared with sera from patients with high disease activity (for definition of active and non-active RA see the Patients and methods section). The results in Fig. 4a indicate that the inhibitory effect of RA sera is evident only in patients with active disease. To explore the disease specificity of this inhibitory activity further, sera from patients with ankylosing spondylitis and with psoriatic arthritis were used as two control groups with chronic, inflammatory autoimmune diseases. Figure 4b shows that sera from the disease controls did not inhibit monocyte cytokine production mediated by T cell contact and did not differ from those from healthy individuals.
The selected RA sera with pronounced inhibitory activity, which were used in the preincubation experiments, were also used for experiments determining the heat resistance of the inhibitory activity. The sera were incubated at different temperatures and added to the standard co-culture assay. As seen in Fig. 5, heating the sera to 56°C (which was routinely used in the previous experiments) preserved the inhibitory activity, whereas increasing the temperature to 70 or 95°C resulted in a loss of that inhibitory activity. Thus, the inhibitory activity is due to heat-labile factors.
Direct inhibition of T cell-dependent monocyte activation has been described previously for the serum protein ApoA-1. To analyse the contribution of ApoA-1 to the inhibitory activity found in sera from patients with RA, ApoA-1 concentrations were determined in sera from patients with active and non-active RA and from controls. The results depicted in Fig. 6a show that ApoA-1 concentrations in sera from patients with active RA are not different from those of the controls, but are significantly decreased in patients with non-active disease. However, when the inhibitory activity of each serum from patients with active disease was plotted against ApoA-1 concentrations, a significant correlation became apparent, because the inhibitory effect increased with increasing ApoA-1 concentration (correlation coefficient R = -0.527, P = 0.016; Fig. 6b). No correlation between ApoA-1 concentration and inhibitory activity was found in the sera from controls and from patients with non-active disease (data not shown).
Discussion
Direct cell–cell contact of monocytes/macrophages with preactivated T lymphocytes leads to the secretion of high levels of proinflammatory cytokines and has been implied in the disturbed cytokine balance seen in RA [21,29,38]. As shown here and described previously, the most likely candidates responsible for the T cell-dependent TNF-α production are synovial macrophages from patients with RA. However, in some studies an involvement of peripheral-blood monocytes in the process of this chronic disease has also been suggested [30-32]. This raises the question of the contribution of monocytes to the increased TNF-α load seen in patients with RA. To address this we used monocytes from patients with RA in the co-culture system, because previous studies have examined monocytes from healthy donors only. The finding that monocytes from patients with RA produced less TNF-α than monocytes from controls was unexpected in view of the proposed contribution of this interaction to the excessive TNF-α levels observed in this disease. Monocytes from patients with RA and controls did not produce any cytokines in the absence of additional stimuli, indicating that peripheral-blood monocytes were not preactivated and that the cell separation techniques used did not lead to artificial ex vivo stimulation of the monocytes.
The induction of TNF-α unequivocally required direct cell contact of the monocytes with T cells, which excludes the possibility that the stimuli of cytokine production are soluble mediators released from the fixed T lymphocytes. Comparing the TNF-α levels produced by synovial monocytes/macrophages with those produced by peripheral monocytes clearly shows that monocytes are not major contributors to the TNF-α load in RA.
The experiments presented here also indicate that the reduced production of TNF-α is not an intrinsic feature of monocytes from patients with RA, because the monocytes are capable of a full TNF-α response to stimulation with LPS. This is in agreement with previous reports about the LPS response of monocytes from patients with RA [39,40], although the evidence is somewhat controversial [41,42]. Furthermore, when measuring other cytokines such as IL-1β, IL-8 and IL-10, we observed that monocytes from patients with RA responded equally well to LPS as did monocytes from controls (data not shown).
Because monocytes from patients with RA are fully capable of producing cytokines, the most likely explanation for the suppression of the T cell-dependent TNF-α response is the presence of regulatory serum factors. The serum protein ApoA-1 has been shown to act as a regulator of cytokine production [25]. The authors found that autologous serum from healthy controls was able to inhibit T cell-dependent TNF-α production in monocytes. They identified ApoA-1 as the molecule blocking the contact-mediated activation of monocytes. ApoA-1 is regarded as a 'negative' acute-phase protein and has been described as being present only in reduced levels in sera from patients with RA [43-45] and juvenile idiopathic arthritis [46], which makes it an unlikely candidate for systemic counter-regulation of cytokine production in RA. High levels of ApoA-1 have been found in the local synovitic environment in RA, where the molecule seems to act as an inhibitory regulator of cytokine production [47]. In its absence, one would expect cytokines to reach extremely high levels.
The present results confirm that patients with RA do not have an increased serum concentration of ApoA-1 compared with that of healthy controls. Consequently, the strong inhibitory activity of RA sera cannot be explained by an increased ApoA-1 concentration alone, although the result of a significant correlation between ApoA-1 concentration and inhibitory serum activity is remarkable.
The results are best explained by additional inhibitory factors that seem to be present in RA sera and seem to bind to the monocyte cell surface, as indicated by two indirect lines of evidence. First, monocytes from patients with RA cultured in FCS produced less TNF-α in response to activated T cells than those from controls (Fig. 4), which indicates that the soluble factor is transferred from the in vivo situation to the co-culture assay. The most likely mode of transfer of the factors would be in cell-bound form on the surface of the monocytes. Second, preincubation of monocytes from healthy controls in RA sera was sufficient to inhibit TNF-α production, and extensive washing did not abolish the inhibitory effect. Again, 'coating' of the monocyte surfaces by the suggested inhibitory factors, which prevents the full interaction between monocytes and T cells, might account for the reduced TNF-α production.
The inhibitory factor(s) described here seem to be specific for RA, and are most pronounced in patients with active disease. It can be proposed that, with increasing disease activity of RA, ApoA-1 (in addition to other factors) becomes upregulated and thus contributes to the downregulation of contact-mediated TNF-α production by monocytes.
Conclusion
The data presented provide evidence for the existence of inhibitory, heat-labile factors in the serum of patients with RA, which downregulate the activation of peripheral-blood monocytes brought about by T cell contact. The possible physiological role of this regulatory mechanism and the specific molecules mediating the suppression remain to be determined.
Abbreviations
ApoA-1 = apolipoprotein A-1; CRP = C-reactive protein; DMARD = disease-modifying anti-rheumatic drugs; FCS = fetal calf serum; IFN = interferon; IL = interleukin; LPS = lipopolysaccharide; PBMCs, peripheral-blood mononuclear cells; PBS = phosphate-buffered saline; RA = rheumatoid arthritis; RF = rheumatoid factor; TNF = tumour necrosis factor.
Competing interests
The author(s) declare that they have no competing interests.
Authors' contributions
MR was responsible for most of the experiments and data analysis as well as drafting the manuscript. SK and RS participated in the collection of the samples and in the interpretation of the results. HH supervised the collection of the samples as well as the design of the study. SH participated in the design of the study and in its coordination, and participated in the interpretation of the results. UW designed the study, participated in its coordination, participated in the interpretation of the results, and drafted the manuscript. All authors read and approved the final manuscript.
Acknowledgements
The work presented here was supported by grants from the German Ministry for Education and Science (Interdisziplinäres Zentrum für Klinische Forschung Leipzig, Teilprojekt A 15, and Kompetenznetzwerk Rheuma, Entzündlich-rheumatische Systemerkrankungen, Teilprojekt C2.7)
Figures and Tables
Figure 1 T cell induced production of TNF-α by monocytes from healthy donors. (a) Fixed stimulated T cells induce tumour necrosis factor (TNF)-α production by monocytes in a dose-dependent manner. Peripheral-blood T cells (106/ml) were cultured for 48 hours in the presence of immobilized anti-CD3 (3.3 μg/ml) and soluble anti-CD28 (0.8 μg/ml) antibodies. Stimulated T cells were fixed and incubated for 24 hours with freshly isolated monocytes (1.5 × 106/ml) at the indicated ratios. Values are means ± SEM from four different experiments. (b) Cell–cell contact is necessary for T cell-induced production of TNF-α in monocytes. Peripheral-blood T cells (106/ml) were cultured for 48 hours in the presence or absence of immobilized anti-CD3 (3.3 μg/ml) and soluble anti-CD28 (0.8 μg/ml) antibodies. Stimulated (sTc) and resting (rTc) T cells were fixed and incubated for 24 hours with freshly isolated monocytes (1.5 × 106/ml) at a ratio of 7:1 in a transwell system as described in the Materials and methods section. T cells and monocytes were physically separated by the semi-permeable membrane (with insert) or had direct cell–cell contact (without insert). Lipopolysaccharide (LPS; 100 ng/ml) was used as a positive control. Values are means ± SEM for experiments with three different donors. (c) T cell-induced TNF-α production by monocytes is time-dependent. Peripheral-blood T cells (106/ml) were cultured for 48 hours in the presence or absence of immobilized anti-CD3 (3.3 μg/ml) and soluble anti-CD28 (0.8 μg/ml) antibodies. Stimulated (sTc) and resting (rTc) T cells were fixed and incubated for the indicated durations with freshly isolated monocytes (1.5 × 106/ml) at a ratio of 7:1. Values are means ± SEM for experiments with three different donors. Levels of significance: * P < 0.05, *** P < 0.001.
Figure 2 Decreased T cell induced production of TNF-α in monocytes from RA patients. (a) T cell-induced tumour necrosis factor (TNF)-α secretion by monocytes from patients with rheumatoid arthritis (RA) is decreased in comparison with those from healthy controls. Peripheral-blood T cells (106/ml) were cultured for 48 hours in the presence or absence of immobilized anti-CD3 (3.3 μg/ml) and soluble anti-CD28 (0.8 μg/ml) antibodies. Stimulated (sTc) and resting (rTc) T cells were fixed and incubated with freshly isolated monocytes (1.5 × 106/ml) at a ratio of 7:1. After 24 hours, the concentration of TNF-α was measured in the supernatant. Lipopolysaccharide (LPS; 100 ng/dl) was used as a positive control. Data are means ± SEM of values from 20 patients with RA and 20 age-matched controls. (b) Macrophages separated from the synovial membrane of patients with RA produce large amounts of TNF-α after contact with preactivated T cells. Stimulated (sTc) and resting (rTc) T cells were fixed and incubated with freshly isolated synovial macrophages from patients with RA (1.5 × 106/ml) at a ratio of 7:1. After 24 hours, the concentration of TNF-α was measured in the supernatant. LPS (100 ng/ml) was used as a positive control. Data are means ± SEM of values from six independent experiments. Level of significance: not significant. (c) Decrease in T cell-induced TNF-α secretion by monocytes from patients with RA is independent of previous and current treatments. The graph depicts the T cell-induced TNF-α production by monocytes of patients with RA either receiving methotrexate (MTX; n = 5) or not receiving immunosuppressive treatment (untreated; n = 6). For comparison, results for six age-matched controls are given (P < 0.05, significant difference compared with untreated patients).
Figure 3 Sera from RA patients inhibit T cell-induced TNF-α production by monocytes from healthy donors. (a) Peripheral-blood T cells (106/ml) were cultured for 48 hours in the presence of immobilized anti-CD3 (3.3 μg/ml) and soluble anti-CD28 (0.8 μg/ml) antibodies. The cells were fixed and incubated for 24 hours with freshly isolated monocytes (1.5 × 106/ml) at a ratio of 7:1. The co-incubation assay was performed in the presence of 10% FCS, 10% serum from patients (serum [RA]) or from healthy donors (serum [control]). All data are expressed as percentages of TNF-α produced in the 10% FCS containing co-culture system (100%). Data are means ± SEM for 10 independent experiments; levels of significance are as indicated. (b) Monocytes from healthy donors were preincubated with 50% control sera (nine donors) or RA sera (six sera from patients with RA, which were previously found to inhibit T cell-dependent monocyte activation) and then washed thoroughly three times. Co-culture experiments were performed as described in the text in medium supplemented with 10% FCS. All data are expressed as percentages of TNF-α produced in the co-culture system containing 10% FCS. Data are means ± SEM for four independent experiments; levels of significance are as indicated.
Figure 4 Inhibition of T cell-induced TNF-α production by monocytes is specific for active RA. (a) Preactivated peripheral-blood T cells (106/ml; see Materials and methods) were fixed and incubated for 24 hours with freshly isolated peripheral-blood monocytes (1.5 × 106/ml) at a ratio of 7:1. The co-incubation assay was performed in the presence of 10% FCS, 10% serum from healthy donors (serum [control], n = 10), from patients with active RA (serum [aRA], n = 20) or from patients with non-active RA (serum [nRA], n = 20). All data are expressed as percentages of TNF-α produced in the co-culture system containing 10% FCS. Data are means ± SEM; levels of significance are as indicated. (b) Co-incubation assays (see (a)) were performed in the presence of 10% FCS, 10% serum from healthy donors (serum [control], n = 10)), serum from patients with psoriatic arthritis (serum [PsA], n = 9) or from patients with ankylosing spondylitis (serum [aSp], n = 9). All data are expressed as percentages of TNF-α produced in the co-culture system containing 10% FCS. Data are means ± SEM; levels of significance are as indicated.
Figure 5 Inhibitory activity of rheumatoid arthritis (RA) sera is not heat stable. Six sera from patients with RA, which previously were found to strongly inhibit T cell-dependent monocyte activation were incubated at 56°C for 30 min, at 70°C for 10 min or at 95°C for 2 min. Co-culture experiments were performed in the presence of 10% FCS or 10% RA sera. All data are expressed as percentages of tumour necrosis factor-α produced in the co-culture system containing 10% FCS. Data are means ± SEM from six independent experiments. P = 0.002 (significant difference from the 100% values).
Figure 6 ApoA-1 concentrations are decreased in non-active RA but correlate with the inhibitory serum activity in active RA. (a) Box plot depicting ApoA-1 concentrations in sera from healthy controls (n = 20), in patients with active RA (n = 20) and in patients with non-active disease (n = 20). Levels of significance are given; n.s., not significant. (b) Scatter plot depicting the correlation between ApoA-1 concentrations and inhibitory serum activity in sera from patients with active RA. Each data point represents the ApoA-1 concentration in relation to the inhibitory activity. Tumour necrosis factor (TNF)-α production in the co-culture system containing 10% FCS is set to 100%. The regression line illustrates the negative correlation between the two parameters (correlation coefficient R = -0.527; P = 0.016).
==== Refs
Burger D Cell contact-mediated signaling of monocytes by stimulated T cells: a major pathway for cytokine induction Eur Cytokine Netw 2000 11 346 353 11022117
Isler P Vey E Zhang JH Dayer JM Cell surface glycoproteins expressed on activated human T cells induce production of interleukin-1 beta by monocytic cells: a possible role of CD69 Eur Cytokine Netw 1993 4 15 23 8490101
Manie S Kubar J Limouse M Ferrua B Ticchioni M Breittmayer JP Peyron JF Schaffar L Rossi B CD3-stimulated Jurkat T cells mediate IL-1 beta production in monocytic THP-1 cells. Role of LFA-1 molecule and participation of CD69 T cell antigen Eur Cytokine Netw 1993 4 7 13 8490106
Lecoanet-Henchoz S Gauchat JF Aubry JP Graber P Life P Paul-Eugene N Ferrua B Corbi AL Dugas B Plater-Zyberk C CD23 regulates monocyte activation through a novel interaction with the adhesion molecules CD11b-CD18 and CD11c-CD18 Immunity 1995 3 119 125 7621072 10.1016/1074-7613(95)90164-7
Armant M Rubio M Delespesse G Sarfati M Soluble CD23 directly activates monocytes to contribute to the antigen-independent stimulation of resting T cells J Immunol 1995 155 4868 4875 7594490
Rezzonico R Chicheportiche R Imbert V Dayer JM Engagement of CD11b and CD11c beta2 integrin by antibodies or soluble CD23 induces IL-1beta production on primary human monocytes through mitogen-activated protein kinase-dependent pathways Blood 2000 95 3868 3877 10845922
Foey AD Feldmann M Brennan FM CD40 ligation induces macrophage IL-10 and TNF-alpha production: differential use of the PI3K and p42/44 MAPK-pathways Cytokine 2001 16 131 142 11792123 10.1006/cyto.2001.0954
Avice MN Sarfati M Triebel F Delespesse G Demeure CE Lymphocyte activation gene-3, a MHC class II ligand expressed on activated T cells, stimulates TNF-alpha and IL-12 production by monocytes and dendritic cells J Immunol 1999 162 2748 2753 10072520
Hayes AL Smith C Foxwell BM Brennan FM CD45-induced tumor necrosis factor α production in monocytes is phosphatidylinositol 3-kinase-dependent and nuclear factor-κB-independent J Biol Chem 1999 274 33455 33461 10559228 10.1074/jbc.274.47.33455
McInnes IB Leung BP Sturrock RD Field M Liew FY Interleukin-15 mediates T cell-dependent regulation of tumor necrosis factor-α production in rheumatoid arthritis Nat Med 1997 3 189 195 9018238 10.1038/nm0297-189
Parry SL Sebbag M Feldmann M Brennan FM Contact with T cells modulates monocyte IL-10 production: role of T cell membrane TNF-alpha J Immunol 1997 158 3673 3681 9103430
Elliott MJ Maini RN Feldmann M Kalden JR Antoni C Smolen JS Leeb B Breedveld FC Macfarlane JD Bijl H Randomised double-blind comparison of chimeric monoclonal antibody to tumour necrosis factor alpha (cA2) versus placebo in rheumatoid arthritis Lancet 1994 344 1105 1110 7934491 10.1016/S0140-6736(94)90628-9
Moreland LW Baumgartner SW Schiff MH Tindall EA Fleischmann RM Weaver AL Ettlinger RE Cohen S Koopman WJ Mohler K Treatment of rheumatoid arthritis with a recombinant human tumor necrosis factor receptor (p75)-Fc fusion protein N Engl J Med 1997 337 141 147 9219699 10.1056/NEJM199707173370301
Feldmann M Development of anti-TNF therapy for rheumatoid arthritis Nat Rev Immunol 2002 2 364 371 12033742 10.1038/nri802
Goronzy JJ Bartz-Bazzanella P Hu W Jendro MC Walser-Kuntz DR Weyand CM Dominant clonotypes in the repertoire of peripheral CD4+ T cells in rheumatoid arthritis J Clin Invest 1994 94 2068 2076 7962553
Walser-Kuntz DR Weyand CM Weaver AJ O'Fallon WM Goronzy JJ Mechanisms underlying the formation of the T cell receptor repertoire in rheumatoid arthritis Immunity 1995 2 597 605 7796293 10.1016/1074-7613(95)90004-7
Wagner UG Koetz K Weyand CM Goronzy JJ Perturbation of the T cell repertoire in rheumatoid arthritis Proc Natl Acad Sci USA 1998 95 14447 14452 9826720 10.1073/pnas.95.24.14447
Koetz K Bryl E Spickschen K O'Fallon WM Goronzy JJ Weyand CM T cell homeostasis in patients with rheumatoid arthritis Proc Natl Acad Sci USA 2000 97 9203 9208 10922071 10.1073/pnas.97.16.9203
Wagner U Pierer M Kaltenhauser S Wilke B Seidel W Arnold S Hantzschel H Clonally expanded CD4+CD28null T cells in rheumatoid arthritis use distinct combinations of T cell receptor BV and BJ elements Eur J Immunol 2003 33 79 84 12594835 10.1002/immu.200390010
McInnes IB al Mughales J Field M Leung BP Huang FP Dixon R Sturrock RD Wilkinson PC Liew FY The role of interleukin-15 in T-cell migration and activation in rheumatoid arthritis Nat Med 1996 2 175 182 8574962 10.1038/nm0296-175
McInnes IB Leung BP Liew FY Cell–cell interactions in synovitis. Interactions between T lymphocytes and synovial cells Arthritis Res 2000 2 374 378 11094451 10.1186/ar115
Chizzolini C Rezzonico R De Luca C Burger D Dayer JM Th2 cell membrane factors in association with IL-4 enhance matrix metalloproteinase-1 (MMP-1) while decreasing MMP-9 production by granulocyte-macrophage colony-stimulating factor-differentiated human monocytes J Immunol 2000 164 5952 5960 10820278
Sebbag M Parry SL Brennan FM Feldmann M Cytokine stimulation of T lymphocytes regulates their capacity to induce monocyte production of tumor necrosis factor-alpha, but not interleukin-10: possible relevance to pathophysiology of rheumatoid arthritis Eur J Immunol 1997 27 624 632 9079801
Foey AD Parry SL Williams LM Feldmann M Foxwell BM Brennan FM Regulation of monocyte IL-10 synthesis by endogenous IL-1 and TNF-alpha: role of the p38 and p42/44 mitogen-activated protein kinases J Immunol 1998 160 920 928 9551930
Hyka N Dayer JM Modoux C Kohno T Edwards CK 3rdRoux-Lombard P Burger D Apolipoprotein A-I inhibits the production of interleukin-1beta and tumor necrosis factor-alpha by blocking contact-mediated activation of monocytes by T lymphocytes Blood 2001 97 2381 2389 11290601 10.1182/blood.V97.8.2381
Vey E Dayer JM Burger D Direct contact with stimulated T cells induces the expression of IL-1beta and IL-1 receptor antagonist in human monocytes. Involvement of serine/threonine phosphatases in differential regulation Cytokine 1997 9 480 487 9237810 10.1006/cyto.1997.0191
Miltenburg AM Lacraz S Welgus HG Dayer JM Immobilized anti-CD3 antibody activates T cell clones to induce the production of interstitial collagenase, but not tissue inhibitor of metalloproteinases, in monocytic THP-1 cells and dermal fibroblasts J Immunol 1995 154 2655 2667 7876539
Brennan FM Hayes AL Ciesielski CJ Green P Foxwell BM Feldmann M Evidence that rheumatoid arthritis synovial T cells are similar to cytokine-activated T cells: involvement of phosphatidylinositol 3-kinase and nuclear factor κB pathways in tumor necrosis factor α production in rheumatoid arthritis Arthritis Rheum 2002 46 31 41 11822409 10.1002/1529-0131(200201)46:1<31::AID-ART10029>3.0.CO;2-5
Foey A Green P Foxwell B Feldmann M Brennan F Cytokine-stimulated T cells induce macrophage IL-10 production dependent on phosphatidylinositol 3-kinase and p70S6K: implications for rheumatoid arthritis Arthritis Res 2002 4 64 70 11879539 10.1186/ar385
Stuhlmuller B Ungethum U Scholze S Martinez L Backhaus M Kraetsch HG Kinne RW Burmester GR Identification of known and novel genes in activated monocytes from patients with rheumatoid arthritis Arthritis Rheum 2000 43 775 790 10765922 10.1002/1529-0131(200004)43:4<775::AID-ANR8>3.0.CO;2-7
Liote F Boval-Boizard B Weill D Kuntz D Wautier JL Blood monocyte activation in rheumatoid arthritis: increased monocyte adhesiveness, integrin expression, and cytokine release Clin Exp Immunol 1996 106 13 19 8870692 10.1046/j.1365-2249.1996.d01-820.x
Shinohara S Hirohata S Inoue T Ito K Phenotypic analysis of peripheral blood monocytes isolated from patients with rheumatoid arthritis J Rheumatol 1992 19 211 215 1378495
Arnett FC Edworthy SM Bloch DA McShane DJ Fries JF Cooper NS Healey LA Kaplan SR Liang MH Luthra HS The American Rheumatism Association 1987 revised criteria for the classification of rheumatoid arthritis Arthritis Rheum 1988 31 315 324 3358796
Ulmer AJ Scholz W Ernst M Brandt E Flad HD Isolation and subfractionation of human peripheral blood mononuclear cells (PBMC) by density gradient centrifugation on Percoll Immunobiology 1984 166 238 250 6329947
Grage-Griebenow E Lorenzen D Fetting R Flad HD Ernst M Phenotypical and functional characterization of Fc gamma receptor I (CD64)-negative monocytes, a minor human monocyte subpopulation with high accessory and antiviral activity Eur J Immunol 1993 23 3126 3135 7504990
Wagner UG Kurtin PJ Wahner A Brackertz M Berry DJ Goronzy JJ Weyand CM The role of CD8+ CD40L+ T cells in the formation of germinal centers in rheumatoid synovitis J Immunol 1998 161 6390 6397 9834130
Vey E Zhang JH Dayer JM IFN-γ and 1,25(OH)2D3 induce on THP-1 cells distinct patterns of cell surface antigen expression, cytokine production, and responsiveness to contact with activated T cells J Immunol 1992 149 2040 2046 1517568
Burger D Dayer JM The role of human T-lymphocyte–monocyte contact in inflammation and tissue destruction Arthritis Res 2002 4 Suppl 3 S169 S176 12110136 10.1186/ar558
Leirisalo-Repo M Paimela L Jaattela M Koskimies S Repo H Production of TNF by monocytes of patients with early rheumatoid arthritis is increased Scand J Rheumatol 1995 24 366 371 8610221
Swaak AJ van den Brink HG Aarden LA Cytokine production in whole blood cell cultures of patients with rheumatoid arthritis Ann Rheum Dis 1997 56 693 695 9462175
Fabris M Tolusso B Di Poi E Tomietto P Sacco S Gremese E Ferraccioli G Mononuclear cell response to lipopolysaccharide in patients with rheumatoid arthritis: relationship with tristetraprolin expression J Rheumatol 2005 32 998 1005 15940758
Forrest CM Harman G McMillan RB Stoy N Stone TW Darlington LG Modulation of cytokine release by purine receptors in patients with rheumatoid arthritis Clin Exp Rheumatol 2005 23 89 92 15789893
Park YB Lee SK Lee WK Suh CH Lee CW Lee CH Song CH Lee J Lipid profiles in untreated patients with rheumatoid arthritis J Rheumatol 1999 26 1701 1704 10451065
Lakatos J Harsagyi A Serum total, HDL, LDL cholesterol, and triglyceride levels in patients with rheumatoid arthritis Clin Biochem 1988 21 93 96 3390902 10.1016/S0009-9120(88)80094-8
Doherty NS Littman BH Reilly K Swindell AC Buss JM Anderson NL Analysis of changes in acute-phase plasma proteins in an acute inflammatory response and in rheumatoid arthritis using two-dimensional gel electrophoresis Electrophoresis 1998 19 355 363 9548303 10.1002/elps.1150190234
Tselepis AD Elisaf M Besis S Karabina SA Chapman MJ Siamopoulou A Association of the inflammatory state in active juvenile rheumatoid arthritis with hypo-high-density lipoproteinemia and reduced lipoprotein-associated platelet-activating factor acetylhydrolase activity Arthritis Rheum 1999 42 373 383 10025933 10.1002/1529-0131(199902)42:2<373::AID-ANR21>3.0.CO;2-3
Bresnihan B Gogarty M Fitzgerald O Dayer JM Burger D Apolipoprotein A-I infiltration in rheumatoid arthritis synovial tissue: a control mechanism of cytokine production? Arthritis Res Ther 2004 6 R563 R566 10.1186/ar1443
|
16277671
|
PMC1297564
|
CC BY
|
2021-01-04 16:02:40
|
no
|
Arthritis Res Ther. 2005 Aug 17; 7(6):R1189-R1199
|
utf-8
|
Arthritis Res Ther
| 2,005 |
10.1186/ar1804
|
oa_comm
|
==== Front
Arthritis Res TherArthritis Research & Therapy1478-63541478-6362BioMed Central London ar18061627767310.1186/ar1806Research ArticlePro-inflammatory properties of stromal cell-derived factor-1 (CXCL12) in collagen-induced arthritis De Klerck Bert [email protected] Lies [email protected] Sigrid [email protected] Hilde [email protected] Yves [email protected] Kurt [email protected] Gary [email protected] Alfons [email protected] Dominique [email protected] Patrick [email protected] Laboratory of Immunobiology, Rega Institute, Katholieke Universiteit Leuven, Leuven, Belgium2 Laboratory of Virology and Chemotherapy, Rega Institute, Katholieke Universiteit Leuven, Leuven, Belgium3 AnorMED, Langley, British Columbia, Canada2005 25 8 2005 7 6 R1208 R1220 25 3 2005 9 5 2005 14 7 2005 29 7 2005 Copyright © 2005 De Klerck et al.; licensee BioMed Central Ltd.This is an Open Access article distributed under the terms of the Creative Commons Attribution License (), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
CXCL12 (stromal cell-derived factor 1) is a unique biological ligand for the chemokine receptor CXCR4. We previously reported that treatment with a specific CXCR4 antagonist, AMD3100, exerts a beneficial effect on the development of collagen-induced arthritis (CIA) in the highly susceptible IFN-γ receptor-deficient (IFN-γR KO) mouse. We concluded that CXCL12 plays a central role in the pathogenesis of CIA in IFN-γR KO mice by promoting delayed type hypersensitivity against the auto-antigen and by interfering with chemotaxis of CXCR4+ cells to the inflamed joints. Here, we investigated whether AMD3100 can likewise inhibit CIA in wild-type mice and analysed the underlying mechanism. Parenteral treatment with the drug at the time of onset of arthritis reduced disease incidence and modestly inhibited severity in affected mice. This beneficial effect was associated with reduced serum concentrations of IL-6. AMD3100 did not affect anti-collagen type II antibodies and, in contrast with its action in IFN-γR KO mice, did not inhibit the delayed type hypersensitivity response against collagen type II, suggesting that the beneficial effect cannot be explained by inhibition of humoral or cellular autoimmune responses. AMD3100 inhibited the in vitro chemotactic effect of CXCL12 on splenocytes, as well as in vivo leukocyte infiltration in CXCL12-containing subcutaneous air pouches. We also demonstrate that, in addition to its effect on cell infiltration, CXCL12 potentiates receptor activator of NF-κB ligand-induced osteoclast differentiation from splenocytes and increases the calcium phosphate-resorbing capacity of these osteoclasts, both processes being potently counteracted by AMD3100. Our observations indicate that CXCL12 acts as a pro-inflammatory factor in the pathogenesis of autoimmune arthritis by attracting inflammatory cells to joints and by stimulating the differentiation and activation of osteoclasts.
==== Body
Introduction
Among chemokines, CXCL12 (formerly stromal cell-derived factor 1) is unique in that it binds to one single chemokine receptor, CXCR4, which itself is recognized by no other chemokines [1-3]. CXCL12 is produced physiologically in various tissues and its receptor CXCR4 is also expressed on various haematopoietic and non-haematopoietic cells. By binding to heparan sulphate proteoglycans, secreted CXCL12 can adhere to certain cells such as bone marrow stromal cells. Through this mechanism, CXCL12-CXCR4 interaction plays an important role in homing of myeloid and lymphoid cells to specific sites in bone marrow or secondary lymphoid organs. CXCR4 also acts as an important co-receptor for HIV entry into CD4+ human lymphocytes [4]. Like other members of the chemokine family, CXCL12 may play a role in inflammatory diseases. Specifically, there is increasing evidence that CXCL12 plays a crucial role in patients with rheumatoid arthritis (RA). In RA patients, abnormally high concentrations of CXCL12 in synovial fluid and overexpression of CXCL12 in synovial cells have been found [5-8]. Moreover, CXCR4+ leukocytes in synovia were found to be significantly more abundant [7]. Evidence also points to a role for CXCL12 in positioning CXCR4+ T and B cells to distinct synovial microdomains as well as in retaining these cells within the inflamed synovial tissue [9]. CXCL12 induces migration of monocytes into human arthritic synovium transplanted into severe combined immunodeficiency (SCID) mice [10]. In addition to exerting these effects on cell migration, CXCL12 also induces angiogenesis during RA development [8] and stimulates chondrocytes to release matrix metalloprotease 3 (MMP3), a matrix-degrading enzyme involved in cartilage destruction [5].
Availability of specific inhibitors of the CXCL12-CXCR4 interaction has allowed the demonstration of the involvement of CXCL12 in experimental animal diseases. One such inhibitor is the bicyclam drug AMD3100, originally discovered as an anti-HIV compound and which specifically interacts with CXCR4 [11,12]. We found that AMD3100 reduces the severity of collagen-induced arthritis (CIA) in mice, a model for RA in man. The study was done on IFN-γ knock-out (IFN-γR KO) DBA/1 mice, which are more susceptible to CIA than wild-type mice [13]. Reduced severity of arthritis was associated with a significant reduction in the delayed type of hypersensitivity (DTH) response to the auto-antigen collagen type II (CII). The majority of leukocytes harvested from inflamed joints of arthritic IFN-γR KO mice were found to be CD11b+, and AMD3100 was demonstrated to interfere with the chemotaxis induced in vitro by CXCL12 on purified CD11b+ splenocytes. We concluded that CXCL12 contributes to the pathogenesis of CIA in these mutant mice by promoting DTH and by interfering with migration of CD11b+ cells into joint tissues.
A major difference in the pathogenesis of CIA between IFN-γR KO and wild-type mice is the presence of more extensive extramedullary myelopoiesis in IFN-γR KO mice, leading to an expansion of CD11b+ cells that can act as DTH and arthritogenic effectors [14-16]. Thus, in IFN-γR KO mice, the balance between cellular (DTH) and humoral autoimmune responses seems to be shifted towards DTH, and this bias may in part explain the beneficial effects of AMD3100 in IFN-γR KO mice. We have tested this hypothesis in the present study. We investigated to what extent AMD3100 affects CIA in wild-type mice and, if so, which mechanisms are involved. We found that AMD3100 does inhibit the disease but that, in contrast to IFN-γR KO mice, this was not associated with reduction in DTH reactivity against CII. We show that, aside from inhibiting chemotaxis in vitro, AMD3100 also inhibits the CXCL12-elicited cell migration into subcutaneous air pouches in vivo. In addition, we found CXCL12 to be able to enhance receptor activator of NF-κB ligand (RANKL)-induced osteoclast differentiation from splenocytes and to increase osteoclast activity, two effects that were counteracted by AMD3100.
Materials and methods
Induction of collagen-induced arthritis
Mice of the DBA/1 strain were bred in the Experimental Animal Centre of the Katholieke Universiteit Leuven (Leuven, Belgium). The experiments were performed in 8- to 12-week-old male mice that were age-matched within each experiment.
CII from chicken sternal cartilage (Sigma-Aldrich Co., St Louis, MO, USA) was dissolved at 2 mg/ml in PBS containing 0.1 M acetic acid by stirring overnight at 6°C. The CII solution was emulsified with an equal volume of complete Freund's adjuvant (CFA; Difco Laboratories, Detroit, MI, USA) with added heat-killed Mycobacterium butyricum (Difco), reaching a final Mycobacterium content of 750 μg/ml emulsion. Mice were injected intradermally with 100 μl emulsion at the base of the tail on day 0.
Mice were examined daily for signs of arthritis. The disease severity was recorded for each limb, as described in [17]: score 0, normal; score 1, redness and/or swelling in one joint; score 2, redness and/or swelling in more than one joint; score 3, redness and/or swelling in the entire paw; score 4, deformity and/or ankylosis.
All animal experiments were approved by the local ethical committee (University of Leuven).
Treatment with AMD3100
AMD3100 was provided by AnorMED (Langley, British Columbia, Canada). For the treatment with AMD3100, Alzet osmotic minipumps model 2002 (DURECT corporation, Cupertino, CA, USA) were subcutaneously implanted at the dorsolateral part of the body. During the procedure, the mice were anaesthetized with a solution of PBS containing 0.2% (v/v) Rompun (Bayer, Brussels, Belgium) and 1% (v/v) Ketalar (Parke-Davis, Zaventem, Belgium). The minipumps delivered AMD3100 at a constant rate of 600 μg/day for 14 days.
Histology
Fore and hind limbs (ankles and interphalanges) were fixed in 10% formalin and decalcified with formic acid. Paraffin sections were haematoxylin stained. Severity of arthritis was evaluated blindly using three parameters: infiltration of mono- and polymorphonuclear cells; hyperplasia of the synovium; and bone destruction. Each parameter was scored on a scale from 0 to 3: score 0, absent; score 1, weak; score 2, moderate; score 3, severe.
Serum anti-collagen type II ELISA
Individual sera were tested for the amount of anti-CII antibody by ELISA, as described previously [17]. Briefly, ELISA plates (Maxisorp, Nunc, Wiesenbaden, Germany) were coated overnight with chicken CII (1μg/ml; 100 μl/well; Sigma-Aldrich Co, St Louis, MO, USA) in coating buffer (50 mM Tris-HCL, pH 8.5; 0.154 mM NaCl) followed by a 2 h incubation with blocking buffer (50 mM Tris-HCl, pH 7.4; 154 mM NaCl and 0.1% (w/v) casein). Serial twofold dilutions of the sera and the standard were incubated overnight in assay buffer (50 mM Tris-HCl; pH 7.4; 154 mM NaCl and 0.5% Tween-20). The quantification of total IgG was done by ELISA making use of a standard with known IgG concentration. For determination of the IgG2a, IgG2b and IgG1 antibody concentrations, a standard of arbitrary U/ml was used (standard = 1,000 U/ml). Plates were then incubated for 2 h with biotinylated rat antibody to mouse total IgG, IgG2a, IgG2b or IgG1 (Zymed Laboratories, San Francisco, CA, USA). Plates were washed and incubated for 1 h with streptavidin-peroxidase. Finally, the substrate 3,3'-5,5'-tetramethyl-benzidine (Sigma-Aldrich Co.) in reaction buffer (100 mM sodium acetate/citric acid, pH 4.9) was added. Reaction was stopped using 50 μl H2SO4 2 M and absorbance was determined at 450 nm.
Delayed-type hypersensitivity experiments
For evaluation of DTH reactivity, CII/CFA-immunized mice were subcutaneously injected with 10 μg of CII/20 μl PBS in the right ear and with 20 μl PBS in the left ear. DTH response was calculated as the percentage swelling (the difference between the increase of thickness of the right and the left ear, divided by the thickness of the ear before challenge, multiplied by 100).
Assays for in vivo leukocyte migration and for in vitro chemotaxis
For the in vivo assay, mice were treated with AMD3100 or PBS as described above. The assay was performed on the last day of the treatment. Six days before, mice were subcutaneously injected at the dorsolateral site of the body with 2.5 ml of sterile air, creating a subcutaneous air pouch. At day three before the assay, injection with 2.5 ml sterile air was repeated at the same location. The chemotactic assay was performed by injecting 1 ml 0.9% (w/v) NaCl/CXCL12 2 μg or 0.9% (w/v) NaCl alone into the air pouch (human CXCL12 was provided by Dr I Clark-Lewis, University of British Columbia, Vancouver, BC, Canada). Two hours later, cells were washed out of the air pouch by 2 ml PBS/FCS 2% (v/v) and cells were immediately counted with a light microscope in the Burker chamber.
In vitro chemotactic assays were performed at day 21 post immunization. Spleens were isolated and passed through cell strainers to obtain a single cell suspension. Erythrocytes were removed by lysis with NH4Cl (0.83% (w/v) in 0.01 M Tris-HCl, pH 7.2; two consecutive incubations of 5 and 3 min, 37°C). Splenocytes of three mice were pooled and incubated with AMD3100 at different concentrations in assay buffer (HBSS, 20 mM Hepes, 0.2% (w/v) BSA, pH 7.2). Transwell filter membranes (5 μm pore; Costar, Boston, MA, USA) were placed in the wells of a 24-well plate, each containing 600 μl buffer with or without CXCL12 at a concentration of 100 ng/ml (human CXCL12 was provided by Dr I Clark-Lewis). 106 cells were loaded on each Transwell filter. The plate was then incubated for 3.5 h at 37°C, whereupon the filter inserts were carefully removed. The migrated cells were collected and counted in a flow cytometer (FACScalibur; Becton Dickinson, San Jose, CA, USA) as described [18-20]. The number of cells is represented as the number of counts registered during a two-minute acquisition (number of cells/2 minutes).
Migrated cells were incubated with anti-CD16/CD32 Fc-blocking antibodies (BD Biosciences Pharmingen, San Diego, CA, USA) and washed with PBS. After washing, the cells were stained for 30 minutes with anti-CD4-PE, anti-CD8-FITC, anti-CD19-PE or anti-CD11b-FITC (BD Biosciences Pharmingen). Cells were washed, fixed with 0.37% formaldehyde in PBS, and analysed by a FACScalibur flow cytometer (Becton Dickinson).
The chemotactic index was calculated as the number of migrated cells obtained with 100 ng/ml CXCL12 divided by the number of cells in the negative control without CXCL12.
Flow cytometric analysis of cells from joint cavities
Cells from joint cavities were obtained by inserting a 25-gauge needle into the ankle joint. Cold PBS (800 μl) was injected into the joint cavity. Fluid exiting spontaneously from the opening was collected and was only used when it was found to contain <5% of erythrocytes. Cells were washed and resuspended in cold PBS. Cells were incubated with anti-CD16/anti-CD32 Fc-receptor-blocking antibodies (BD Biosciences Pharmingen). After washing, the cells were stained for 30 minutes with anti-CD11b-FITC and anti-CXCR4-PE or isotype control rat IgG2b (BD Biosciences Pharmingen). Cells were washed, fixed with 0.37% formaldehyde in PBS, and analysed by a FACScalibur flow cytometer (Becton Dickinson).
Polymerase chain reaction
Synovial tissues from the ankle joints were carefully isolated under a stereomicroscope. Total RNA was extracted with Trizol reagent (Invitrogen, Paisley, Scotland, UK), in accordance with the manufacturer's instructions. cDNA was obtained by reverse transcription with a commercially available kit (Thermoscript; Invitrogen) with oligo(dT)20 as primer.
For PCR reactions we used a TaqMan® Assays-on-Demand™ Gene Expression Product from Applied Biosystems (Foster City, CA, USA; assay ID Mm00445552_m1). Expression levels of the gene were normalized for 18S RNA expression.
Cytokine detection in serum and cultured medium
Control-treated and AMD3100-treated mice were bled both before and 6 h after intraperitoneal injection with 10 μg anti-CD3. Sera were collected and pooled. This allowed us to determine the concentrations of the following cytokines: IL-1β, IL-2, IL-4, IL-6, IL-10, IL-12, tumour necrosis factor-α and IFN-γ.
Spleens of three mice were isolated on day 21 after immunization and were passed through cell strainers to obtain a single cell suspension. Erythrocytes were removed by lysis with NH4Cl (0.83% (w/v) in 0.01 M Tris-HCl, pH 7.2; two consecutive incubations of 5 and 3 minutes, 37°C). Splenocytes of the mice were pooled and cultured in a 96-well plate. 105 cells were cultured in one well in Roswell Park Memorial Institute (RPMI) medium alone, RPMI with mouse CXCL12 (0.1 μg/ml) (PeproTech, London, UK), or RPMI with mouse CXCL12 and AMD3100 (25 μg/ml). Supernatant was collected after 48 h. Detection of cytokine concentrations in serum and cultured medium was done with the Endogen SearchLight™ array (Pierce Boston Technology, Woburn, MA, USA).
In vitro induction of osteoclast formation by splenocytes
Spleens were isolated on day 21 after immunization and were passed through cell strainers to obtain a single cell suspension. Erythrocytes were removed by lysis with NH4Cl (0.83% (w/v) in 0.01 M Tris-HCl, pH 7.2; two consecutive incubations of 5 and 3 minutes, 37°C). Leukocytes from the blood were obtained by lysis of red blood cells by two incubations (5 and 3 minutes at 37°C) with NH4Cl solution (0.083% (w/v) in 0.01 M Tris-HCl; pH 7.2). Remaining cells were washed two times with ice-cold PBS.
Splenocytes were suspended in Minimal Essential Medium alpha Medium (α-MEM) containing 10% (v/v) FCS (GIBCO, Invitrogen corporation, Paisley, Scotland, UK). Cells (2.5 × 105) in a total volume of 400 μl were seeded in chamber slides (LAB-TEK Brand Products, Nalge Nunc International, Naperville, IL, USA). Cells were incubated with macrophage colony stimulating factor (M-CSF; 20 ng/ml) + CXCL12 (0.1 or 0.5 μg/ml; AnorMED), with M-CSF + RANKL (100 ng/ml) + CXCL12 or with M-CSF + RANKL + CXCL12 + AMD3100 (25 μg/ml; AnorMED). M-CSF and RANKL were obtained from R&D Systems Europe (Abingdon, UK). On day 4, supernatants were removed and cultures were provided with fresh media and stimuli. On day 7, media were removed and cells were stained for the presence of tartrate-resistant acid phosphatase (TRAP) (described below).
Pit-forming assay
Splenocyte suspensions were obtained as described above and resuspended in α-MEM containing 10% (v/v) FCS (GIBCO, Invitrogen Corporation). 106 cells were cultured for 5 days with M-CSF (20 ng/ml) and RANKL (100 ng/ml), both from R&D systems Europe, on transparent quartz slides coated with a calcium phosphate film (BioCoat Osteologic Discs; BD Biosciences Pharmingen). On day 6, media were removed and replaced with media containing M-CSF, M-CSF + CXCL12 (0.5 μg/ml), M-CSF + CXCL12 + AMD3100 (25 μg/ml) or M-CSF + AMD3100. Another two days later, cells were removed and resorption pits were quantified using a Leitz DM RBE microscope equipped with a colour video camera (Optronics Engineering, Goleta, CA, USA) and attached to a computer-aided image analysing system (Bioquant, R&M Biometrics, Nashville, TN, USA). Quantification and size determination of the pits was performed at a magnification of ×20 in 15 areas of constant size, positioned adjacent to one another and spanning the whole quartz slide. In all slides, the minimum threshold of a pit surface area was set to 50 μm2. Upon thresholding, the number and square surface of plaques are determined automatically.
Results
Inhibition of collagen-induced arthritis by AMD3100 in DBA/1 wild-type mice
In a first experiment, DBA/1 mice were immunized with CII in CFA. The symptoms of arthritis started to appear on day 27; 4 mice out of 22 showed redness and/or swelling in one of their joints. On that day, mice were divided in two subgroups, matched for incidence and average clinical score. In one group, mice were implanted with osmotic minipumps releasing AMD3100 at a constant rate of 600 μg/day. Mice in the other group were implanted with pumps delivering PBS. From previous experience and according to the manufacturer's specification sheet, the minipumps are known to be active for two weeks. Mice were scored six times a week for symptoms of arthritis. Cumulative incidence and mean scores of arthritis in both groups during the experiment are shown in Fig. 1a,b. The cumulative incidence of arthritis rapidly increased in the control mice, but remained stable in the AMD3100-treated animals (Fig. 1a). In fact, after initiation of treatment, 7 out of 10 mice in the PBS group developed arthritis within 3 days, whereas in the AMD3100-treated group only a single mouse out of 8 developed symptoms 13 days after initiation of the treatment. Correspondingly, the mean arthritic group score gradually increased in controls, but not in AMD3100-treated animals (Fig. 1b). The beneficial effect of AMD3100 in CIA was confirmed in two additional experiments. The data of the three experiments are summarized in Table 1: during the treatment, in total only 2 out of 20 AMD3100-treated mice developed signs of arthritis against 16 out of 22 controls.
Considering all arthritic mice, the average clinical score of arthritic mice was lower in the treated mice, although the difference compared with that in the arthritic control mice was not statistically significant. Thus, the significantly lower average scores reached in the AMD3100-treated group, when all mice are considered, reflects mainly the lower incidence in this group. In addition, evaluation of disease progression in each of the individual mice having arthritis signs at the initiation of treatment revealed lower percent increases in disease scores in the AMD3100- than in the PBS-treated group (Fig. 1c), suggesting that AMD3100 can also exert a beneficial effect on evolving arthritis. These in vivo results show that AMD3100 treatment of CIA initiated at first appearance of symptoms is effective against the development and progression of the disease.
Reduced histological symptoms of arthritis in AMD3100-treated mice
To ascertain that the protective effect of AMD3100 with respect to the clinical symptoms of arthritis was also manifest at the histological level, five mice from each group (Table 1, experiment 1) were sacrificed for histological examination of the joints. These mice were selected such that their mean clinical scores corresponded to the average score of the entire group.
Haematoxylin-stained sections showed that the absence of redness and swelling in AMD3100-treated mice corresponded with the absence of infiltration of immunocompetent cells and tissue destruction (Fig. 1d). Histological examination of joint sections of AMD3100-treated mice that did show clinical symptoms of arthritis revealed a weak hyperplasia and infiltration of mono- and polymorphonuclear cells in the synovium (Fig. 1e). Sections of arthritic PBS-treated mice showed a moderate to severe infiltration, hyperplasia of the synovium and bone destruction (Fig. 1f).
AMD3100 does not interfere with humoral or cellular responses to collagen type II
The pathogenesis of CIA is generally considered to depend on both humoral and cellular immunity against CII. To see whether inhibition of CIA by AMD3100 acts via modulation of either of these, we measured specific anti-CII antibodies and DTH reactivity against CII. These tests were performed on day 14 after implantation of minipumps.
Total anti-CII IgG was determined in sera of the mice that were sacrificed for histological analysis. Titers of these antibodies in AMD3100-treated mice were not different from those in PBS-treated mice (Fig. 2a). The remainder of the sera were pooled and analysed for IgG2a, IgG2b and IgG1 isotypes against CII. IgG2a was below detection limit in both groups. We found no difference in the IgG2b and IgG1 concentrations between the two groups (Fig. 2b). Thus, the absence of clinical and histological symptoms of arthritis in the AMD3100-treated mice appeared not to be associated with any decreased antibody response against CII, nor with a switch between isotypes.
Cellular-immune responsiveness to CII was tested in mice that were immunized with CII/CFA and that were implanted on day 7 with osmotic minipumps containing AMD3100 or PBS. DTH testing was done on day 17 after immunization (i.e., day 10 of the treatment) by injecting 10 μg of CII in the right, and vehicle (PBS) in the left ear. Bars in Fig. 2c represent the percentages of swelling of the CII-challenged ears, normalized to the swelling of the PBS-challenged ears. No inhibition of the DTH response to CII was observed in the AMD3100-treated group indicating that AMD3100 did not interfere with the cellular immune response to CII.
AMD3100 blocks CXCL12-elicited cell migration in vivo and chemotaxis in vitro
To see whether AMD3100 inhibits CIA by blocking CXCL12-mediated tissue infiltration, we immunized a set of 16 mice with CII/CFA. At the time of disease onset (day 27) they were divided into two subgroups matched by average incidence and clinical score. One group was implanted with osmotic minipumps delivering AMD3100. In the control group, pumps were filled with PBS. An air pouch assay was done on day 14 after minipump implantation (day 41). In both the AMD3100-treated and the PBS-treated group, four of the mice received an injection into the air pouch with CXCL12 (2 μg in 1 ml of 0.9% (w/v) NaCl), and four other mice received an injection with 0.9% NaCl. Two hours after this challenge, cells were washed out of the air pouch using 2 ml PBS containing 2% (v/v) FCS.
Cell counts are shown in Fig. 3a. Mice implanted with PBS-delivering osmotic minipumps and challenged with 0.9% NaCl during the air pouch assay were considered as negative controls. In this group, an average of 1.5 ± 0.2 × 106 cells per mouse was obtained from the air pouch. Mice that carried PBS-delivering osmotic pumps and were injected with CXCL12 into the air pouch were considered as positive controls. In this group, we harvested an average of 3.4 ± 0.3 × 106 cells per mouse. This indicates specific infiltration of cells into the air pouch, in response to the chemokine CXCL12. Challenging mice with CXCL12 while they were treated with AMD3100 reduced the number of harvested cells to that of the negative control. The number of cells in the air pouch of AMD3100-treated mice after challenge with 0.9% NaCl was similar to that in the negative controls, indicating that the AMD3100-treatment did not, as such, affect the number of cells in the air pouch. Furthermore, flow cytometric analysis of the spleen and the lymph nodes did not reveal effects of AMD3100 on the number or proportions of CD4+, CD8+, CD19+ and CD11b+ cells. Together these data led us to conclude that treatment with AMD3100 is able to block CXCL12-elicited infiltration in vivo so as to prevent infiltration into inflamed tissues.
In vitro chemotactic assays performed on splenocytes of immunized mice allowed us to investigate the dose-dependent inhibition of CXCL12-elicited chemotaxis by AMD3100 (Fig. 3b). The percentage of cells that migrated in response to CXCL12 gradually decreased when the cells were pre-incubated with increasing concentrations of AMD3100. The dose-dependent inhibition by AMD3100 was confirmed in three additional experiments (pooled data are represented in Fig. 3c as the mean percentage of inhibition of CXCL12-elicited chemotaxis).
Flow cytometric analysis was performed after chemotaxis and revealed that CD4+, CD8+, CD19+ and CD11b+ cells were all attracted to CXCL12 with a chemotactic index of 2.5, 2.7, 6.9 and 3.4, respectively.
Expression of CXCL12 and presence of CXCR4+ cells in the arthritic joint
To collect further evidence for the hypothesis that AMD3100 protects mice from arthritis by blocking CXCL12-mediated leukocyte mobilization, we ascertained that CXCL12-elicited migration of immunocompetent cells to inflamed sites does take place during CIA development. Numbers of CXCL12 mRNA copies were found to be elevated in synovial cells of the inflamed joint, as evident from quantitative reverse transcription (RT)-PCR (Fig. 4a). Among cells harvested by synovial lavage from the arthritic joint, an average of 15% stained double positive for CXCR4 and CD11b, as investigated by flow cytometry (Fig. 4b). These data were confirmed in an additional experiment. Taken together, these findings are indicative of CXCL12-elicited recruitment of CXCR4+CD11b+ leukocytes to the joints as a mechanism contributing to CIA pathogenesis.
Influence of AMD3100 on cytokine production
We also considered the possibility that, in the course of CIA pathogenesis, CXCL12 might stimulate or enhance production of certain cytokines and that this might be a pathway by which AMD3100 could exert its protective action. To test this possibility, we looked at possible differences in the cytokine profiles of PBS- and AMD3100-treated mice. Eight mice were immunized with CII/CFA and treated on day 25 with AMD3100 (four mice) or PBS (four mice), using the osmotic minipumps. On day 35 post immunization (day 10 of the treatment), mice were bled and serum levels of IL-1β, IL-2, IL-4, IL-6, IL-10, IL-12, TNF-α and IFN-γ were determined by SearchLight proteome array. Only IL-6, IL-10, IL-12 and IFN-γ were detectable in the sera of mice. AMD3100 failed to change the levels of IL-10, IL-12 and IFN-γ, although blood levels of IL-6 were decreased in AMD3100-treated mice, a finding that was confirmed in additional experiments (data from these experiments are shown in Fig. 5a). Decreased systemic production of IL-6 in AMD3100-treated mice may be an indirect effect of inhibition of CXCL12-mediated cell traffic, as this might reduce formation of inflammatory tissue in joints and possibly other sites in the CII/CFA-immunized mice. Alternatively, inhibited IL-6 production might signify that CXCL12, aside from its chemotactic activity, directly activates certain CII/CFA-exposed leukocytes to produce this cytokine. To help distinguish between these two possibilities, we tested the ability of CXCL12 to induce the production of IL-6 in splenocyte cultures. Splenocytes of CII/CFA-immunized mice were cultured in the absence or presence of CXCL12 (0.5 μg/ml), with or without AMD3100 (25 μg/ml) (Fig. 5b). IL-6 was detectable in the supernatants of unstimulated cultures. Stimulation with CXCL12 or CXCL12 + AMD3100 did not alter the IL-6 production in the cultures, suggesting that the decreased IL-6 blood concentrations in the AMD3100-treated arthritic mice reflected an indirect, rather than a direct, CXCL12 action on IL-6 production.
CXCL12 facilitates osteoclast differentiation and activation
Osteoclast precursor cells and in vitro differentiated mature osteoclasts have been found to express CXCR4 [21-23], a finding that was confirmed in our laboratory (data not shown). To test the hypothesis that CXCL12 might facilitate osteoclast differentiation, splenocyte suspensions were cultured in the presence of M-CSF and RANKL, in the presence or absence of CXCL12 and/or AMD3100. After 6 days, osteoclasts were identified by staining for TRAP, a marker enzyme for osteoclasts. In cultures stimulated with M-CSF and RANKL, osteoclast differentiation could be observed (Fig. 6a–c). Addition of CXCL12 at a concentration of 0.1 μg/ml did not influence the number of osteoclasts (Fig. 6a). At 0.5 μg/ml, however, significantly higher numbers of osteoclasts were observed (Fig. 6b,d). Interestingly, when both AMD3100 and CXCL12 (in either concentration) were added, less differentiated osteoclasts appeared than in control cultures receiving only M-CSF and RANKL (Fig. 6a,b,e). Reduced differentiation of osteoclasts was not associated with increased mortality of splenocytes in these cultures (data not shown).
We also tested the effect of CXCL12 on the osteoclasts' ability to dissolve bone mineral. Splenocytes were cultured on quartz substrates, coated with a calcium phosphate film. Cells were stimulated with M-CSF + RANKL. After 6 days, multinucleated giant cells could be seen by microscopical examination, but resorption of the calcium phosphate film was not yet visible. On that day, the supernatant fluid was replaced with medium containing M-CSF alone, M-CSF + CXCL12 (0.5 μg/ml), M-CSF + CXCL12 + AMD3100 or M-CSF + AMD3100. Two days later, osteoclast activity was quantified as the ability to resorb the calcium phosphate film; 1 resorption pit is the area resorbed by 1 osteoclast, and the area of the pit correlates with osteoclast activity. The mean area of the resorption pits in the different conditions was calculated using a bioquant image analysis system (the data are presented in Fig. 7a and representative pictures of the resorbed areas on the calcium phosphate film are shown in Fig. 7b–d). It can be seen that CXCL12 significantly increased osteoclast activity, as evident from an increase in the resorbed area. When AMD3100 was added to the cultures with or without CXCL12, the osteoclast activity decreased significantly to a level beneath that of cultures with M-CSF alone (Fig. 7a). When the number of pits were counted and grouped according to their size, it appeared that CXCL12 also increased the number of pits, irrespective of their size, although resorption pits with a large area (>5,000 μm2) were most affected by CXCL12. In contrast, such large pits were barely detectable in cultures that had been treated with AMD3100 (Table 2).
Because AMD3100 decreased osteoclast differentiation (Fig. 6) and activation (Fig. 7) to a level beneath that of cultures where no exogenous CXCL12 was added, we verified whether splenocytes spontaneously produced CXCL12. To this end, splenocytes were cultured for 2 days without stimulation and CXCL12 concentrations in the supernatant were determined using the SearchLight proteome array. The mean level of CXCL12 in the supernatant of three independent splenocyte cultures was 85 ± 26 pg/ml.
Taken together, these data reveal a positive effect of CXCL12 on osteoclast differentiation and activity. Moreover, the inhibition of osteoclast differentiation and activation by the CXCR4 antagonist suggests an important role for endogenous CXCL12 in both processes.
Discussion
We had already established that CXCL12 plays an important role in the pathogenesis of murine CIA in the highly sensitive IFN-γR KO mouse [13]. In particular, treatment with the specific CXCL12 inhibitor AMD3100 had been shown to afford protection against CIA. Examination of the underlying mechanism led to the conclusion that AMD3100 interfered with CXCL12-mediated immigration of leukocytes in the joints, but also reduced the systemic DTH against CII that, in IFN-γR KO mice, is typically more pronounced than in wild-type mice.
Here, we demonstrate that AMD3100 reduced the incidence and progression of CIA in IFN-γR-competent mice, in a similar manner to that previously demonstrated in IFN-γR KO mice [13]. Thus, irrespective of whether the IFN-γ system is defective or intact, CXCL12 is a key cytokine in the pathogenesis of murine CIA.
Quantitative RT-PCR revealed an increased presence of CXCL12 mRNA in the inflamed synovium in comparison with normal synovium, and 15% of the cells that could be harvested from inflamed joints were found to be CD11b+CXCR4+ double positive. Splenocytes from mice subjected to the CIA immunisation schedule were found to display in vitro chemotactic responsiveness to CXCL12, an activity that was blocked by adding AMD3100. The in vivo relevance of these effects in mice immunized to develop CIA was examined with a subcutaneous air pouch system. CXCL12 injected into air pouches elicited immigration of leukocytes, an effect that was similarly blocked by AMD3100. These observations make it seem likely that in wild-type mice, as in IFN-γR-deficient ones, effects on leukocyte traffic constitute an important mechanism by which CXCL12 favours the pathogenesis of CIA.
In the case of IFN-γR KO mice, protection by AMD3100 treatment was associated with reduced cellular immune responsiveness to CII, as evident from reduced DTH in footpad swelling tests [13]. In wild-type mice, in contrast, AMD3100 did not affect DTH reactivity against CII. Basal DTH to CII following CIA induction was less pronounced in wild-type than in IFN-γR KO mice, however, and this in itself might account for it not being further reduced by AMD3100. Of note, another CXCL12-CXCR4 inhibitor, 4F-benzoyl-TN14003, has been shown to inhibit DTH to sheep red blood cells in normal Balb/c mice [24]. The difference in mouse strain and the use of a different antigen may account for the discrepancy between our findings. Failure of AMD3100 to affect anti-CII DTH reactivity in our wild-type mice suggests that, under the circumstances, cellular immunity to CII is perhaps not a key element by which CXCL12 influences the pathogenesis of CIA.
Formation of antibodies to CII was similarly not affected by AMD3100 treatment. Thus, although CXCL12 is known to act as a B cell growth factor [25], its possible action on humoral immunity to CII cannot be considered as a mechanism by which it acts as a disease-promoting factor in wild-type DBA/1 mice.
We also considered the possibility that CXCL12 favours CIA development by somehow affecting systemic cytokine production. In wild-type DBA/1 mice immunized to develop CIA, we found production of circulating IL-6, and levels of this cytokine, were reduced in mice treated with AMD3100. IL-6 is a crucial cytokine for CIA development because treatment with antibodies against IL-6 inhibits disease development [15]. In RA patients, serum IL-6 concentrations correlate with disease activity and decrease after effective treatment with disease modifying antirheumatic drugs. We wondered whether CXCL12 could induce IL-6 production and if this could be inhibited by AMD3100. If so, this would be an additional mechanism for AMD3100 to inhibit CIA development. We found, however, that IL-6 production was not increased in CXCL12-stimulated splenocyte cultures compared to unstimulated ones. This suggests that the lower levels of IL-6 in the serum of AMD3100-treated mice probably reflects the effectiveness of the CIA treatment, occurring for example by inhibition of leukocyte infiltration in the joint.
Furthermore, we investigated the possibility that CXCL12 can induce production of RANKL and thereby stimulate differentiation and/or activation of osteoclasts. If real, these activities might constitute part of the role of CXCL12 in CIA pathogenesis and might in part explain the protective effect of AMD3100. In fact, we found CXCL12 to be unable to induce RANKL or RANK expression and osteoclast differentiation in plain splenocyte cultures. The chemokine did potentiate induction of osteoclasts in cultures exposed to RANKL plus M-CSF, however, although it should be noted that the chemokine dose required to see this effect was in large excess of levels normally seen in CXCL12 production systems. Addition of AMD3100 to the system annihilated the CXCL12 effect but, intriguingly, reduced osteoclast induction to a level lower than that seen in cultures not exposed to CXCL12. A possible explanation may be that cultures exposed to M-CSF plus RANKL release endogenous CXCL12 at a level such that osteoclast induction is near to maximal, requiring supra high doses of exogenous CXCL12 for further augmentation.
Grassi et al. [26] reported that CXCL12 can enhance bone resorbing activity of osteoclasts. We similarly found that stimulation of osteoclasts with CXCL12 augmented their calcium phosphate-resorbing capacity. Moreover, addition of AMD3100 to osteoclast cultures reduced their resorbing potential. This inhibitory effect took place even if no exogenous CXCL12 had been added, showing that the osteoclast-activating activity of CXCL12 operates at concentrations within the endogenous physiological range.
Conclusion
As evident from our observations, we demonstrate that CXCL12 plays a crucial role in the CIA pathogenesis of fully IFN-γR-competent mice, as it was proven to be before in IFN-γR KO mice. The underlying mechanisms are diverse, however, and their relative impact may differ depending on whether the IFN-γ system is defective or intact. In both cases, effects on leukocyte migration to the inflamed joints seem to play an important role. An enhancing effect of CXCL12 on cellular immunity may play an additional important role in IFN-γR KO mice, which are considerably more sensitive to the disease; this mechanism seems to be of less importance in wild-type mice. Furthermore, we were able to document a potentiating effect of CXCL12 on osteoclast differentiation and activation, both of which were counteracted by AMD3100. These observations hold further promise for potential treatment of RA patients with CXCR4 antagonists.
Abbreviations
BSA = bovine serum albumin; CFA = complete Freund's adjuvant; CIA = collagen-induced arthritis; CII = collagen type II; DTH = delayed type hypersensitivity; ELISA = enzyme-linked immunosorbent assay; FCS = fetal calf serum; IFN = interferon; IFN-γR KO = IFN-γ receptor knock-out; IL = interleukin; M-CSF = macrophage colony-stimulating factor; PBS = phosphate-buffered saline; PCR = polymerase chain reaction; RA = rheumatoid arthritis; RANK = receptor activator of NF-κB; RANKL = receptor activator of NF-κB ligand; RT-PCR = reverse transcription polymerase chain reaction; TRAP = tartrate-resistant acid phosphatase.
Competing interests
The authors declare that they have no competing interests
Authors' contributions
CIA induction and the disease evaluation were done by BDK and YM. KV implanted the minipumps. PM performed the histological evaluations. Quantification of humoral and cellular response was done by BDK. In vitro and in vivo chemotactic assays were performed by SH and BDK, respectively. PCR and flow cytometry were done by HK. LG and BDK did the in vitro experiments for cytokine detection, osteoclast differentiation and the pit-forming assays. BDK, PM, SH and DS designed the study. BDK, PM and AB prepared the manuscript. All authors participated in the interpretation of the data.
Acknowledgements
We thank Tania Mitera and Chris Dillen for excellent assistance and helpful discussions. Studies in the authors' laboratories are funded by the Concerted Research Actions (GOA) Initiative of the Regional Government of Flanders, the Interuniversity Attraction Pole Program (IUAP) of the Belgian Federal Government, as well as grants from the National Fund for Scientific Research of Flanders (FWO). PM and SH are postdoctoral research fellows from the FWO Vlaanderen, and HK holds a fellowship from the FWO Vlaanderen.
Figures and Tables
Figure 1 Inhibition of collagen-induced arthritis in DBA/1 mice by treatment with AMD3100. Mice were immunized on day 0 with collagen type II in complete Freund's adjuvant and were observed for symptoms of arthritis. On day 27, when the first symptoms of arthritis appeared, the mice were divided into two groups in a way that a similar incidence and a similar average clinical score was reached in both groups. On this day, mice of one group were implanted with osmotic minipumps, delivering AMD3100 for two weeks at a constant rate of 600 μg/day. Mice of the other group were implanted with pumps containing PBS. The (a) cumulative incidence and (b) mean arthritic score ± standard error of the mean (SEM) for AMD3100-treated and control-treated mice are shown. Average group scores of arthritis were significantly different from day 30 onwards (p ≤ 0.05 on day 30; p ≤ 0.01 from day 31 till the end of the experiment, Mann-Whitney U test). (c) Evaluation of disease progression in mice with established arthritis at initiation of treatment with AMD3100. Circles represent percentage increase in scores of arthritis for individual mice at the end of the treatment. Data show the results of three individual experiments (explained in more detail in the legend of Table 1). (d-f) Histological analysis of the joints. On the last day of treatment, five mice out of both groups with a mean score representing the average group score, were selected and sacrificed. Paraffin sections of the fore and hind limbs were haematoxylin stained and histological examination was performed. Representative pictures are shown. (d) Joint of an AMD3100-treated mouse without clinical symptoms showing normal histological appearance. (e) Joints of arthritic AMD3100-treated mice show a weak infiltration of mono- and polymorphonuclear cells and hyperplasia of the synovium. (f) Joint section of a PBS-treated mouse, showing moderate to severe infiltration of leukocytes, hyperplasia and bone destruction.
Figure 2 AMD3100 in wild-type mice does not interfere with the humoral or the cellular response. At the end of the two week treatment (day 41), blood was collected from five mice out of each group. (a) Sera of individual mice were analyzed for total anti-CII IgG, using an absolute standard. Bars represent averages plus standard error of the mean of five mice. (b) Equal quantities of the sera were pooled for detection of anti-CII IgG2b and IgG1, using a standard in arbitrary U/ml; standard = 1,000 U/ml. (c) Delayed type hypersensitivity reactivity against CII. Ten mice were immunized with CII/complete Freund's adjuvant and implanted with osmotic pumps containing AMD3100 or PBS on day 7. On day 17 after immunization, five mice in each group were challenged with 10 μg of CII in the right ear and vehicle in the left. Delayed type hypersensitivity responses were measured as the percentage of ear swelling (i.e. 100 × the difference between the increase of thickness of the right and the left ear, divided by the thickness of the ear before challenge) at the indicated times. Bars represent averages ± standard error of the mean for five mice.
Figure 3 AMD3100 blocks CXCL12-elicited chemotaxis in vivo and in vitro. (a) Sixteen mice were immunized with collagen type II (CII) in complete Freund's adjuvant on day 0 and treated with AMD3100 or PBS in a similar way as described in the legend of Fig. 1. In vivo treatment is indicated along the X-axis. On the last day of treatment, a chemotactic assay was performed as described in Materials and methods. On that day, mice were injected with 2 μg of CXCL12 in 1 ml 0.9% NaCl (+) or 0.9% NaCl only (-) in a subcutaneous air pouch. Two hours after chemokine challenge, cells were washed out of the air pouch with 2 ml of PBS/FCS 2% and counted. Counts of the individual mice are shown (circles) and average ± standard error of the mean are indicated for each group (diamonds). (b,c) Dose-dependent inhibition by AMD3100 of CXCL12-elicited chemotaxis on total splenocytes. On day 21 post immunization with CII in complete Freund's adjuvant, spleens of three mice were pooled and a splenocyte suspension was prepared. Cell samples were pre-incubated for 10 minutes with AMD3100 at the indicated concentrations. Then, 5-μm filter inserts were loaded with 106 cells and transferred to a 24-well plate containing 100 ng/ml human CXCL12 in 600μl of buffer per well. After 3.5 h of incubation, the membrane inserts were removed and the cells in the wells were collected and counted by flow cytometry. The numbers of migrated cells of one representative experiment are shown in (b). (c) The experiment was confirmed by three additional experiments and the data of the experiments were pooled and represented as the percentage inhibition ± standard error of the mean of CXCL12-elicited chemotaxis by the indicated concentrations of AMD3100.
Figure 4 Presence of CXCL12 RNA and CXCR4+ cells in the arthritic joint. (a) Synovia of three collagen type II/complete Freund's adjuvant-immunized collagen-induced arthritic (CIA) mice and three naive mice were isolated on day 35 after immunization and total RNA was purified. Reverse-transcription was performed and cDNA was subjected to quantitative PCR and normalized to the amount of 18S RNA. (b) Joints of three other collagen type II/complete Freund's adjuvant-immunized mice were washed at day 35 with PBS/FCS 2%. Cells that were harvested from the joint were stained for the presence of CD11b using fluorescein isothiocyanate (FITC)-labeled antibodies, and for CXCR4 using phycoerythrin (PE)-labeled antibodies. (c) Control staining for CXCR4 using a PE-labeled rat IgG2b isotype control antibody. One representative experiment out of two is shown.
Figure 5 IL-6 levels in serum and in CXCL12-stimulated splenocyte cultures. (a) Eight mice were immunized with collagen type II/complete Freund's adjuvant (CIA) and implanted with osmotic minipumps delivering AMD3100 (four mice) at a constant rate of 600 μg/day or containing PBS (four mice). Blood was collected at day 10 of the treatment. Sera were pooled in each group and analysed for the presence of IL-6 (using a SearchLight Proteome array). Bars represent the average ± standard error of the mean of two independent experiments. (b) Splenocytes of three collagen type II/complete Freund's adjuvant-immunized mice were pooled and cultured in the absence of CXCL12, in the presence of CXCL12 (0.5 μg/ml) or in the presence of CXCL12 and AMD3100 (25 μg/ml). Supernatant was analysed after 48 h for the presence of IL-6. Bars represent the average ± standard error of the mean of three independent experiments.
Figure 6 CXCL12 stimulates and AMD3100 inhibits osteoclast differentiation. Splenocytes of three collagen type II/complete Freund's adjuvant-immunized mice were isolated and pooled. (a,b) Splenocytes were cultured for 6 days in the cups of a chamberslide, in the presence of the indicated stimuli (macrophage colony-stimulating factor (M-CSF), 20 ng/ml; receptor activator of NF-κB ligand (RANKL), 100 ng/ml; CXCL12, 0.1 μg/ml in (a), 0.5 μg/ml in (b); AMD3100, 25 μg/ml). After stimulation, cells were fixed and stained for the presence of tartrate-resistant acid phosphatase (TRAP). TRAP+ multinucleated (three or more nuclei) cells were counted within each cup. Bars represent averages ± standard error of the mean for four cultures. The asterisk represents p < 0.05 compared with the hatched bar (Mann-Whitney U-test). Representative pictures for TRAP-stained cultures stimulated with (c) M-CSF and RANKL alone and with added (d) CXCL12 (0.5 μg/ml) or (e) CXCL12 and AMD3100.
Figure 7 CXCL12 increases and AMD3100 inhibits osteoclast activity. Splenocytes of three collagen type II/complete Freund's adjuvant-immunized mice were isolated and pooled. Cell suspensions were cultured for 6 days on a quartz substrate coated with a calcium phosphate film in the presence of macrophage colony-stimulating factor (M-CSF, 20 ng/ml) and receptor activator of NF-κB ligand (RANKL, 100 ng/ml). At day 6 media were removed, cultures were provided with fresh media and stimulated as indicated (M-CSF, 20 ng/ml; CXCL12, 0.5 ng/ml; AMD3100, 25 μg/ml). Cells were removed from the quartz substrate after 2 days and resorption pits were visualized by light microscopy. The resorbed area was measured by a bioquant image analysis system. Bars represent the mean area resorbed by 1 osteoclast (average ± standard error of the mean), measured as the area of 1 resorption pit. The asterisk represents p < 0.001 compared with the M-CSF condition (Mann-Whitney U-test). Representative pictures of resorption pits are shown for the condition stimulated with (b) M-CSF, (c) M-CSF + CXCL12 and (d) M-CSF + CXCL12 + AMD3100. The data in this figure are representative for two independent experiments.
Table 1 Inhibition of the incidence and mean score of CIA by treatment with AMD3100
Experiment number Treatmenta Cumulative incidence (%) Score of arthritis (mean ± SEM)
Start of treatmentb End of treatmentc All miced Arthritic mice onlye
1 AMD3100 2/10 (20%) 3/10 (30%)f 1.4 ± 0.9f 4.7 ± 2.4
Control 2/12 (17%) 9/12 (75%) 5.2 ± 0.9 6.2 ± 0.6
2 AMD3100 1/7 (14%) 2/7 (29%) 0.9 ± 0.6f 3.0 ± 1.0
Control 2/7 (29%) 5/7 (71%) 2.4 ± 0.8 3.4 ± 0.7
3 AMD3100 2/8 (25%) 2/8 (25%)g 0.6 ± 0.5g 2.5 ± 1.5
Control 1/8 (13%) 7/8 (88%) 3.9 ± 1.4 4.3 ± 1.5
The table shows the results of three individual experiments. Male mice were immunized with collagen type II/complete Freund's adjuvant on day 0. aAt the day first symptoms appeared (day 27 in experiment 1 and 2, day 24 in experiment 3), mice were divided into two groups and were implanted with osmotic minipumps delivering AMD3100 at a constant rate of 600 μg/day or PBS in the control groups. Distribution of the mice between the two groups was done in a way that an equal incidence and a similar clinical score was reached in both groups. bArthritic incidence in both groups at the start of the treatment is shown. cAt the end of the treatment, there was a significant inhibition of the incidence in the AMD3100-treated group compared to the control in experiments 1 and 3 (fp < 0.05 and gp < 0.01, respectively; binomial proportion test). dAt the end of the treatment, the mean arthritic scores calculated for all mice were significantly different between the AMD3100-treated and control groups for all the three experiments (fp < 0.05 for experiments 1 and 2; gp < 0.01 for experiment 3; Mann-Whitney U-test). eAt the end of the treatment, the mean arthritic scores calculated for mice with symptoms of arthritis were not significantly different between the AMD3100-treated and control groups. SEM, standard error of the mean.
Table 2 CXCL12 increases and AMD3100 inhibits osteoclast activity
Number of pits with indicated resorption areaa
In vitro stimulationb 50–100 μm2 100–500 μm2 500–1,000 μm2 1,000–5,000 μm2 >5,000 μm2
Control 481 1,002 166 153 22
CXCL12 848 1,854 382 346 90
CXCL12 + AMD3100 598 1,143 193 164 5
AMD3100 580 1,008 171 140 6
aThe table shows the number of osteoclast resorption pits for five different surface intervals. bSplenocytes of collagen type II/complete Freund's adjuvant-immunized mice were cultured as described in the legend of Fig. 8, and the resorbed area was measured.
==== Refs
Bleul CC Farzan M Choe H Parolin C Clark-Lewis I Sodroski J Springer TA The lymphocyte chemoattractant SDF-1 is a ligand for LESTR/fusin and blocks HIV-1 entry Nature 1996 382 829 833 8752280 10.1038/382829a0
Oberlin E Amara A Bachelerie F Bessia C Virelizier JL Arenzana-Seisdedos F Schwartz O Heard JM Clark-Lewis I Legler DF The CXC chemokine SDF-1 is the ligand for LESTR/fusin and prevents infection by T-cell-line-adapted HIV-1 Nature 1996 382 833 835 8752281 10.1038/382833a0
Rossi D Zlotnik A The biology of chemokines and their receptors Annu Rev Immunol 2000 18 217 242 10837058 10.1146/annurev.immunol.18.1.217
Feng Y Broder CC Kennedy PE Berger EA HIV-1 entry cofactor: functional cDNA cloning of a seven-transmembrane, G protein-coupled receptor Science 1996 272 872 877 8629022
Kanbe K Takagishi K Chen Q Stimulation of matrix metalloprotease 3 release from human chondrocytes by the interaction of stromal cell-derived factor 1 and CXC chemokine receptor 4 Arthritis Rheum 2002 46 130 137 11817585 10.1002/1529-0131(200201)46:1<130::AID-ART10020>3.0.CO;2-D
Buckley CD Amft N Bradfield PF Pilling D Ross E Arenzana-Seisdedos F Amara A Curnow SJ Lord JM Scheel-Toellner D Salmon M Persistent induction of the chemokine receptor CXCR4 by TGF-beta 1 on synovial T cells contributes to their accumulation within the rheumatoid synovium J Immunol 2000 165 3423 3429 10975862
Nanki T Hayashida K El-Gabalawy HS Suson S Shi K Girschick HJ Yavuz S Lipsky PE Stromal cell-derived factor-1-CXC chemokine receptor 4 interactions play a central role in CD4+ T cell accumulation in rheumatoid arthritis synovium J Immunol 2000 165 6590 6598 11086103
Pablos JL Santiago B Galindo M Torres C Brehmer MT Blanco FJ Garcia-Lazaro FJ Synoviocyte-derived CXCL12 is displayed on endothelium and induces angiogenesis in rheumatoid arthritis J Immunol 2003 170 2147 2152 12574387
Bradfield PF Amft N Vernon-Wilson E Exley AE Parsonage G Rainger GE Nash GB Thomas AM Simmons DL Salmon M Buckley CD Rheumatoid fibroblast-like synoviocytes overexpress the chemokine stromal cell-derived factor 1 (CXCL12), which supports distinct patterns and rates of CD4+ and CD8+ T cell migration within synovial tissue Arthritis Rheum 2003 48 2472 2482 13130466 10.1002/art.11219
Blades MC Manzo A Ingegnoli F Taylor PR Panayi GS Irjala H Jalkanen S Haskard DO Perretti M Pitzalis C Stromal cell-derived factor 1 (CXCL12) induces human cell migration into human lymph nodes transplanted into SCID mice J Immunol 2002 168 4308 4317 11970972
Hatse S Princen K Bridger G De Clercq E Schols D Chemokine receptor inhibition by AMD3100 is strictly confined to CXCR4 FEBS Lett 2002 527 255 262 12220670 10.1016/S0014-5793(02)03143-5
Schols D Struyf S Van Damme J Este JA Henson G De Clercq E Inhibition of T-tropic HIV strains by selective antagonization of the chemokine receptor CXCR4 J Exp Med 1997 186 1383 1388 9334378 10.1084/jem.186.8.1383
Matthys P Hatse S Vermeire K Wuyts A Bridger G Henson GW De Clercq E Billiau A Schols D AMD3100 a potent and specific antagonist of the stromal cell-derived factor-1 chemokine receptor CXCR4, inhibits autoimmune joint inflammation in IFN-gamma receptor-deficient mice J Immunol 2001 167 4686 4692 11591799
De Klerck B Carpentier I Lories RJ Habraken Y Piette J Carmeliet G Beyaert R Billiau A Matthys P Enhanced osteoclast development in collagen-induced arthritis in interferon-gamma receptor knock-out mice as related to increased splenic CD11b+ myelopoiesis Arthritis Res Ther 2004 6 R220 R231 15142268 10.1186/ar1167
Matthys P Vermeire K Mitera T Heremans H Huang S Schols D De Wolf-Peeters C Billiau A Enhanced autoimmune arthritis in IFN-gamma receptor-deficient mice is conditioned by mycobacteria in Freund's adjuvant and by increased expansion of Mac-1+ myeloid cells J Immunol 1999 163 3503 3510 10477624
Matthys P Vermeire K Billiau A Mac-1(+) myelopoiesis induced by CFA: a clue to the paradoxical effects of IFN-gamma in autoimmune disease models Trends Immunol 2001 22 367 371 11429320 10.1016/S1471-4906(01)01937-8
Vermeire K Heremans H Vandeputte M Huang S Billiau A Matthys P Accelerated collagen-induced arthritis in IFN-gamma receptor-deficient mice J Immunol 1997 158 5507 5513 9164974
Campbell JJ Haraldsen G Pan J Rottman J Qin S Ponath P Andrew DP Warnke R Ruffing N Kassam N The chemokine receptor CCR4 in vascular recognition by cutaneous but not intestinal memory T cells Nature 1999 400 776 780 10466728 10.1038/23495
Levesque JP Hendy J Takamatsu Y Simmons PJ Bendall LJ Disruption of the CXCR4/CXCL12 chemotactic interaction during hematopoietic stem cell mobilization induced by GCSF or cyclophosphamide J Clin Invest 2003 111 187 196 12531874 10.1172/JCI200315994
Kahn J Byk T Jansson-Sjostrand L Petit I Shivtiel S Nagler A Hardan I Deutsch V Gazit Z Gazit D Overexpression of CXCR4 on human CD34+ progenitors increases their proliferation, migration, and NOD/SCID repopulation Blood 2004 103 2942 2949 15070669 10.1182/blood-2003-07-2607
Grassi F Piacentini A Cristino S Toneguzzi S Cavallo C Facchini A Lisignoli G Human osteoclasts express different CXC chemokines depending on cell culture substrate: molecular and immunocytochemical evidence of high levels of CXCL10 and CXCL12 Histochem Cell Biol 2003 120 391 400 14600836 10.1007/s00418-003-0587-3
Wright LM Maloney W Yu X Kindle L Collin-Osdoby P Osdoby P Stromal cell-derived factor-1 binding to its chemokine receptor CXCR4 on precursor cells promotes the chemotactic recruitment, development and survival of human osteoclasts Bone 2005 36 840 853 15794931 10.1016/j.bone.2005.01.021
Yu X Huang Y Collin-Osdoby P Osdoby P Stromal cell-derived factor-1 (SDF-1) recruits osteoclast precursors by inducing chemotaxis, matrix metalloproteinase-9 (MMP-9) activity, and collagen transmigration J Bone Miner Res 2003 18 1404 1418 12929930
Tamamura H Fujisawa M Hiramatsu K Mizumoto M Nakashima H Yamamoto N Otaka A Fujii N Identification of a CXCR4 antagonist, a T140 analog, as an anti-rheumatoid arthritis agent FEBS Lett 2004 569 99 104 15225616 10.1016/j.febslet.2004.05.056
Nagasawa T Kikutani H Kishimoto T Molecular cloning and structure of a pre-B-cell growth-stimulating factor Proc Natl Acad Sci USA 1994 91 2305 2309 8134392
Grassi F Cristino S Toneguzzi S Piacentini A Facchini A Lisignoli G CXCL12 chemokine up-regulates bone resorption and MMP-9 release by human osteoclasts: CXCL12 levels are increased in synovial and bone tissue of rheumatoid arthritis patients J Cell Physiol 2004 199 244 251 15040007 10.1002/jcp.10445
|
16277673
|
PMC1297565
|
CC BY
|
2021-01-04 16:02:39
|
no
|
Arthritis Res Ther. 2005 Aug 25; 7(6):R1208-R1220
|
utf-8
|
Arthritis Res Ther
| 2,005 |
10.1186/ar1806
|
oa_comm
|
==== Front
Arthritis Res TherArthritis Research & Therapy1478-63541478-6362BioMed Central London ar18111627767610.1186/ar1811Research ArticleElimination of rheumatoid synovium in situ using a Fas ligand 'gene scalpel' Zhang Haidi [email protected] Guangping [email protected] Gilda [email protected] H Ralph [email protected] Division of Pharmaceutics and Industry Pharmacy, School of Pharmacy, Long Island University, Brooklyn, NY, USA2 Department of Medicine, School of Medicine, University of Pennsylvania, Philadelphia, PA, USA3 Veterans Affairs Medical Center in Philadelphia, Philadelphia, PA, USA2005 8 9 2005 7 6 R1235 R1243 14 9 2004 1 11 2004 29 7 2005 3 8 2005 Copyright © 2005 Zhang et al.; licensee BioMed Central Ltd.This is an Open Access article distributed under the terms of the Creative Commons Attribution License (), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
Surgical synovectomy to remove the inflammatory synovium can temporarily ameliorate rheumatoid inflammation and delay the progress of joint destruction. An efficient medically induced programmed cell death (apoptosis) in the rheumatoid synovium might play a role similar to synovectomy but without surgical tissue damage. Gene transfer of Fas ligand (FasL) has increased the frequency of apoptotic cells in mouse and rabbit arthritic synovium. In this study, we investigated whether repeated FasL gene transfer could remove human inflammatory synovial tissue in situ and function as a molecular synovectomy. Briefly, specimens of human synovium from joint replacement surgeries and synovectomies of rheumatoid arthritis (RA) patients were grafted subcutaneously into male C.B-17 severe combined immunodeficiency (SCID) mice. Injections of a recombinant FasL adenovirus (Ad-FasL) into the grafted synovial tissue at the dosage of 1011 particles per mouse were performed every two weeks. Three days after the fifth virus injection, the mice were euthanized by CO2 inhalation and the human synovial tissues were collected, weighed and further examined. Compared to the control adenovirus-LacZ (Ad-LacZ) and phosphate buffered saline (PBS) injected RA synovium, the Ad-FasL injected RA synovium was dramatically reduced in size and weight (P < 0.005). The number of both synoviocytes & mononuclear cells was significantly reduced. Interestingly, an approximate 15-fold increased frequency of apoptotic cells was observed in RA synovium three days after Ad-FasL injection, compared with control tissues. In summary, our in vivo investigation of gene transfer to human synovium in SCID mice suggests that repeated intra-articular gene transfer of an apoptosis inducer, such as FasL, may function as a 'gene scalpel' for molecular synovectomy to arrest inflammatory synovium at an early stage of RA.
==== Body
Introduction
Rheumatoid arthritis (RA) is a potentially very disabling disease that is characterized by chronic synovitis, a hyperplasic synovial membrane, and finally cartilage and bone destruction. Overgrowth of fibroblast-like synoviocytes as well as their secretion of an impressive array of cytokines/chemokines, adhesion molecules and proteases play important roles in the pathogenesis of rheumatoid joint destruction [1-3]. Surgical synovectomy to remove the inflammatory synovium can temporarily ameliorate rheumatoid inflammation and delay the progress of joint destruction [4]. An efficient medically induced programmed cell death (apoptosis) in the inflammatory synovium [5-7] might play a role similar to surgical synovectomy.
Hyperplasia of the rheumatoid synovium may result from the imbalance between cell proliferation and apoptosis. Mutations in tumor suppressor genes such as p53, and elevated expression of proto-oncogenes and apoptosis inhibitors, such as c-myc, c-fos, c-ras, c-jun, and bcl-2 in RA synoviocytes, may lead to inadequate apoptosis and tumor-like proliferation of rheumatoid synoviocytes [8-11]. Thus, the induction of apoptosis by gene transfer of an apoptosis inducer or growth modulator in a sense or anti-sense orientation may function as a 'gene scalpel' for decreasing or eliminating the hyperplasia of the synovial membrane.
Fas, a membrane-bound death molecule, is highly expressed in RA synovial lining and sub-lining cells compared to osteoarthritis and normal synovial tissues [11]. Between 30% and 90% of cells in RA synovium are positive for Fas antigen, but Fas ligand (FasL) mRNA has been detected only in mononuclear cells in RA synovium [10]. The upregulation of Fas-mediated apoptosis in inflammatory synovium has been established by intra-articular administration of adenovirus mediated FasL gene [5,12], anti-Fas monoclonal antibodies [13-15], and FasL transfectants [16]. Current reports indicate that a medically induced upregulation of the Fas-mediated apoptosis pathway in inflammatory synoviocytes may provide a novel therapeutic strategy for RA treatment [5,12-18].
Compared to other current available gene delivery vehicles, adenoviral vector has high transduction efficiency into human and rabbit fibroblast-like synoviocytes in vitro and into the synovium of mice, rabbits, guinea pigs and rhesus monkeys in vivo (5,12,19,7). Thus, recombinant adenovirus could be an ideal vector for identifying dose dependent effects of certain transgenes expressed in arthritic joints, although it would not be an ideal vehicle for molecular synovectomy in clinical administration because of its strong immunogenicity.
Induction of apoptosis in RA synoviocytes by gene transfer may be an efficient approach for the treatment of synovitis because the inflammatory cytokines/chemokines and proteases/adhesion molecules in RA synovium will be limited once the producing cells have died. Determination of the effects of FasL gene transfer on human inflammatory synovium in vivo is an important step in progressing from mouse/rabbit gene therapies to human gene therapy. A novel experimental system incorporating the grafting of RA synovial tissue into severe combined immunodeficiency (SCID) mice [20-22] has provided an in vivo model for the evaluation of our hypothesis; the repeated gene transfer of an apoptosis inducer such as FasL, mediated by an efficient vehicle, might function as a gene scalpel for the removal of inflammatory synovium in situ.
Materials and methods
Rheumatoid arthritis synovial tissues
Synovium and cartilage were obtained from RA patients who were diagnosed according to the 1987 revised criteria of the American College of Rheumatology [23] and who underwent joint replacement surgeries and synovectomies. The fresh RA synovium and cartilage used for grafting into SCID mice were obtained from the same RA patient at the time of surgery, in accordance with the Institutional Review Board approved study protocol at University of Pennsylvania. The human tissues were kept on ice during the procedures. RA synovium from the knee, hip, or shoulder joints of six patients were examined for the effects of FasL gene transfer in vivo in SCID mice.
Animal model
Male C.B-17 SCID mice (Taconic, Germantown, NY, USA) aged 6 to 7 weeks were housed in the Research Animal Facility at the Veterans Affairs Medical Center in Philadelphia. SCID mice were kept at 72–75°F and handled under specific pathogen-free conditions. During surgical procedures, the mice were anesthetized intra-peritoneally with ketamine 2 mg and xylazine 0.4 mg in 0.2 ml PBS per mouse. RA synovial tissues were cut into 2 × 3 × 3 mm3 pieces. Synovium alone or synovium mixed with cartilage were grafted into SCID mice at 200 mg tissue per mouse subcutaneously on their backs approximately 30 minutes after the synovial tissues were removed from patients. The entire procedure was performed under sterile conditions.
Adenovirus preparation
The FasL adenovirus (Ad-FasL) and the LacZ adenovirus (Ad-LacZ) were generated as described previously [5,24] based on human type 5 adenovirus vector with the deletion of E1a, E1b, and a portion of the E3 region. The cDNA of FasL or LacZ was inserted between the cytomegalovirus enhancer/promoter and simian virus 40 late gene polyadenylation signal, respectively, in the 5' inverted terminal repeat regions of adenovirus type 5 at the place of the E1 deletion. Stock viruses were amplified in 293 cells with a variety of virus titer and serum concentrations. After 36 to 60 h of virus infection, cells were harvested and lysed by freeze and thaw cycles and purified through a two-round of cesium chloride gradient centrifugation. The cesium chloride was removed by a BioRad desalting column (Bio-Rad Laboratories, Hercules, CA, USA). The viruses were diluted with 10% glycerol in PBS (pH 7.4) to 2–4 × 1012 particles per ml, aliquoted and frozen at -80°C until use. The transducibility of Ad-LacZ and Ad-FasL was evaluated by detection of β-galactosidase transgene expression and apoptosis induction in human synovial fibroblasts in vitro.
Tissue X-gal staining
To detect the adenovirus vector mediated dose dependent transgene expression in vivo, RA synovium alone or with cartilage in vivo in the SCID mice were injected with Ad-LacZ at 109, 1010 or 1011 particles per mouse into the middle of the grafted tissues. Three days after virus injection, the synovium alone or with cartilage was collected and fixed in 4% paraformaldehyde on ice for 2 h, then washed three times with PBS and stained in a solution of 5 mM K4Fe(CN)6, 5 mM K3 Fe(CN)6, 2 mM MgCl2 in PBS containing 0.5 mg/ml of X-gal stain (5-bromo-4-chloro-3-indolyl-b-D-galactopyranoside; Sigma Chemical Co., St Louis, MO, USA) overnight at 37°C. Samples were then washed three times with PBS, photographed and embedded in paraffin blocks.
Treatment of the rheumatoid arthritis-SCID model
To identify the clinical potential of FasL gene transfer, RA synovium from six patients was grafted into SCID mice at 200 mg tissue per mouse subcutaneously. Two weeks after engraftment, 1011 particles per mouse of Ad-FasL or Ad-LacZ in 0.1 ml PBS, or 0.1 ml PBS only, were injected into the engrafted tissue location in the SCID mice. The injections were repeated every two weeks for a total of five times. Mice were euthanized by CO2 inhalation three days after the fifth virus injection. The human synovial tissues were collected, weighed, and used for further examinations. Ten samples from each treatment group were examined. The weight of each sample was measured individually for each group as shown in Table 1. Comparative statistical analysis of the effects of Ad-LacZ and Ad-FasL treatments on the weight of RA synovium was performed using the rank sum test. Some grafted tissues were collected three days after the first virus treatment for the detection of apoptosis and Fas and FasL protein expression in synovial cells.
Tissue hematoxylin and eosin staining
RA synovium and cartilage removed from the SCID mice were fixed in 10% phosphate-buffered formalin. After paraffin embedding, tissue sections (6 μm) were stained with hematoxylin and eosin for morphological evaluation.
Reverse transcription-PCR
Total RNA was isolated from the grafted RA synovium using the Ultraspectrum RNA System (Biotech Laboratories, Houston, TX, USA). The specific first-strand cDNA synthesis and amplification were performed in the one tube for two-step reactions with the Promega Access Systems (Promega Corp., Madison, WI, USA). The specific primers used for detection of the transgenic FasL mRNA expression correspond to the coding region sequence of FasL cDNA (forward primer: 5'-TCAGCTCTTCCACCTGCAG-3') and poly(A) region sequence of the viral vector (reverse primer: 5'-CACTGCATTCTAGTTGTGG-3') to generate a 688 base-pair (bp) DNA fragment. The pair of primers corresponding to β-actin cDNA (forward primer: 5'-GAAATCGTGCGTGACATTAAG-3' ; reverse primer: 5'-CTAGAAGCATTTGCGGTGGACG-3') was used to generate a 505 bp cDNA product as an internal control to evaluate the quality of RNA samples. Briefly, the first-strand cDNA was synthesized for 45 minutes at 48°C, and then amplified through 35 cycles at 94°C for 1 minute, 55°C for 1 minute and 72°C for 2 minutes. The final concentration of reagents was 1.5 mM MgSO4, 0.2 mM each dNTP, 1 μM each primer, 0.1u/ μl AMV reverse transcriptase and 0.1u/ μl Tfl DNA polymerase obtained from Promega Corp. Subsequently, aliquots of the PCR products were fractionated by electrophoresis on a 1% agarose gel and visualized by ethidium bromide staining. The PCR products of transgenic FasL cDNA were then purified from the gel using the QIAquick Gel Extraction Kit (QIAGEN, Valencia, CA, USA) and were confirmed by sequencing analysis using BigDye Terminator Ready Reaction Kits (PE Biosystems, Foster City, CA, USA).
Immunohistochemistry
An immunohistochemistry staining system from Santa Cruz (Santa Cruz Biotechnology, Inc., Santa Cruz, CA, USA) was used for detection of Fas and FasL protein in RA synovium. Grafted RA synovium samples from SCID mice were snap-frozen. The cryosections (6 μm) were fixed in cold acetone and treated with 3% hydrogen peroxide, blocked with 5% non-fat milk and 10% goat or horse serum, and incubated with the affinity-purified polyclonal rabbit anti-human Fas antibody or goat anti-FasL of mouse, rat and human antibody at room temperature for 1 h. Purified rabbit immunoglobulin G (IgG) or goat IgG was used as a control. After three washes with PBS, biotinylated goat anti-rabbit or horse anti-goat IgG was then added to the tissue. Positive staining was viewed subsequently using an avidin-peroxidase and diaminobenzidine color detection system combined with a hematoxylin counter-stain.
Terminal deoxyribonucleotidyl transferase-mediated dUTP nick end labeling
An ApopTag fluorescein in situ apoptosis detection kit (Intergen Company, Purchase, NY, USA) was used to detect apoptotic cells in RA synovium. The formalin-fixed and paraffin-embedded tissue sections (6 μm) were permeabilized with proteinase K (20 μg/ml) and then the 3'-OH ends of fragmented DNA were labeled with terminal deoxyribonucleotidyl transferase. The incorporated digitonigen-dUTP was detected by incubation with fluorescein-conjugated anti-digitonigen antibody at room temperature for 60 minutes, and positive reactions were revealed under fluorescence microscopy using appropriate excitation and emission filters after propidium iodide counter-staining and mounting (Vector Laboratories Inc., Burlingame, CA, USA). The apoptotic cells and total synovial cells were counted blindly to the specific treatments at 10 high power microscopic fields randomly selected for each section. Three sections at multiple layers were examined for each sample. The percentage of apoptotic cells was recorded and represented as mean and standard deviation for each group.
Results
Adenovirus vector mediated dose dependent transgene expression in human synovial tissues in vivo in SCID mice
We tested Ad-LacZ transgene expression at the dosage of 109, 1010, or 1011 particles per injection to determine the suitable experimental dosage of Ad-FasL in the in vivo RA-SCID model. The implanted human synovium only or with cartilage was harvested three days after virus injection and examined by X-gal staining (Fig. 1). The results show that the adenovirus vector mediates high dose dependent transgene expression in human synovium (Fig. 1a–d) but not in cartilage (Fig. 1e,f) in vivo three days after virus injection. The adenovirus vector infected synoviocytes, lymphocytes, macrophages, granulocytes and endothelial cells in human synovium, as indicated by LacZ transgene expression in these cell types (Fig. 1g–k). The most efficient dosage of adenovirus vector mediated gene transfer into human synovium in vivo was 1011 particles of virus per injection (Fig. 1a). This virus dosage was selected for further in vivo experiments with Ad-FasL.
Time dependent FasL transgene expression in engrafted human synovium in vivo in SCID mice
To investigate the adenovirus vector mediated, time dependent FasL transgene expression in engrafted human synovium in vivo, 1011 virus particles of Ad-FasL per dose were injected into the middle of the implanted human synovium in SCID mice. The same dosage of Ad-LacZ was injected into the synovium in vivo as a control. Tissue samples were harvested on days 2, 5, 7, 10 and 20. Transgenic FasL mRNA expression in tissue samples was detected by RT-PCR (Fig. 2). No transgenic FasL mRNA was detected in Ad-LacZ injected tissues samples (Fig. 2, lanes 7–11). The transgenic FasL mRNA was highly expressed on days 2 to 10 and obviously decreased on day 20 after virus injection (Fig. 2, lanes 2–6), which suggests that repeated administration of FasL gene transfer would be necessary to maintain a high level of FasL transgene expression in vivo. This experiment was repeated and the same gene expression profile was observed. Therefore, we decided to perform repeated Ad-FasL treatment every two weeks in our in vivo experiments.
FasL gene transfer increases the expression of FasL protein and the frequency of apoptotic cells in RA synovium
Fas protein is highly expressed in RA synovial lining cells and almost all kinds of inflammatory cells infiltrating the synovium, but FasL protein is poorly expressed in RA synovium [11]. We determined Fas and FasL protein expression in engrafted RA synovium by immunohistochemistry using polyclonal anti-Fas and anti-FasL antibodies. Our experimental results (Fig. 3a,b) are similar to those published by another group [11]. Three days after Ad-FasL gene transfer into RA synovium in vivo, augmented FasL protein expression was observed (Fig. 3c), and an increased frequency of apoptotic cells in RA synovium was detected by terminal deoxyribonucleotidyl transferase-mediated dUTP nick end labeling (TUNEL) staining (Fig. 3f). DNA fragmentation was determined in the location of the nucleus using a Leica TCS SPII laser scanning confocal microscope (Leica Microsystems, Heidelberg, Germany) (Fig. 3g–i). Few apoptotic cells were detected in RA synovium in the SCID mouse in vivo two weeks after tissue grafting. Three days after Ad-FasL gene transfer, the frequency of apoptotic cells increased 15 to 20-fold (Fig. 4). The frequency of apoptotic cells in PBS treated RA synovium underwent almost no change, and the frequency of apoptotic cells in Ad-LacZ injected RA synovium increased only slightly (Fig. 4).
Elimination of inflammatory RA synovium in vivo
Upregulation of Fas/FasL interaction in arthritic joints by virus vector mediated FasL gene transfer or anti-Fas antibody treatment intra-articularly can induce apoptosis in inflammatory synoviocytes, as has been identified by many groups [5,12-18]. Here the effect of continuous activation of Fas/FasL interaction through a repeated FasL gene transfer was explored to determine if this treatment could result in the death of inflammatory synoviocytes in RA synovium and elimination of RA tissue in vivo. The in vivo experiments were performed as explained in the Materials and methods section. Compared to the control Ad-LacZ treated RA synovium, the Ad-FasL injected RA-synovium was dramatically reduced in size (Fig. 5b) and weight (Table 1).
A significance test for the effects of Ad-LacZ and Ad-FasL treatments on the weight of RA synovium was performed using the rank sum test (n1 = n2 = 10, T1 = 154, T2 = 56, P < 0.005) (Table 1). Along with the approximate 15-fold increase in the frequency of apoptotic cells observed in the RA synovium three days after after Ad-FasL injection (Fig. 4) the number of both synoviocytes and mononuclear cells was greatly decreased in RA synovium after two months of treatment with repeated FasL gene transfers (Fig. 5d) compared with Ad-LacZ treated RA synovium (Fig. 5c). The RA synovium partially recovered, exhibiting features of normal human synovium after the two-month treatment with Ad-FasL (Fig. 5d).
Discussion
Molecular surgery to remove the pathological cells and tissues in situ by gene transfer could be an important part of molecular medicine. Molecular synovectomy using gene scalpels simplifies the current problems of gene delivery, gene target and gene expression regulation involved in human gene therapy. Because the induction of apoptosis in inflammatory synoviocytes is the main goal for intra-articular gene transfer, a transient, localized, immune tolerant and dose dependent gene transfer may be achieved by direct intra-articular injection of apoptosis inducers carried by a suitable vector, controlled to infect only synovial cells but not chondrocytes in cartilage. In this study, we have established that the inflammatory synovium from RA patients can be effectively reduced in vivo in a RA-SCID mouse model by repeated, locally administered adenovirus vector mediated FasL gene transfer, strongly supporting the gene scalpel concept.
C.B-17 SCID mice have severe combined immunodeficiency, resulting from a mutation on chromosome 16 responsible for deficient activity of an enzyme involved in DNA repair [25]. We chose to use C.B-17 SCID mice in this study because this strain has successfully hosted human synovium and has provided an ideal in vivo experimental environment to test how human inflammatory synovium responds to novel therapeutic reagents [21-23]. Unlike C.B-17 SCID beige mice, C.B-17 SCID mice have no deficiency of macrophages and natural killer cells, which is important for our experimental design because macrophage phagocytosis of apoptotic cells [26-28] is considered as one of the important mechanisms involved in the molecular synovectomy procedure. In theory, inflammatory synoviocytes in implanted RA synovium in SCID mice may consist of antigen specific human lymphocytes, proliferated human fibroblast-like synoviocytes, and infiltrated mouse macrophages and so on. An artificial feature of this model, however, is the absence of circulating human blood components when studying the properties of rheumatoid synovium [21]. In our experiments, the inflammatory histological features of rheumatoid synovium are preserved in implanted RA synovium in SCID mice. This combination of RA synovium and SCID mice allows for repeated FasL gene transfer mediated by adenovirus vector into RA synovium in vivo because the C.B-17 SCID mice lack an immune response to the adenovirus vector.
The elimination of synoviocytes from RA synovium after multiple applications of FasL gene transfer results in a significant reduction in RA synovium size and weight in vivo in the RA-SCID mice (Fig. 5; Table 1) along with dramatic synovial cell death detected by TUNEL staining (Figs 3g–I and 4). Recombinant adenovirus is a highly efficient vector for gene transfer into the synovium [5,7,12,19]. Adenovirus vector can infect a wide variety of cell types in human synovium including synoviocytes, lymphocytes, macrophages, granulocytes and endothelial cells, as determined by X-gal staining (Fig. 1g–k). In recent studies, endothelial cells were reported to constitutively express FasL and release soluble FasL [29-31]. These cells are not sensitive to Fas-mediated apoptosis and may play a role in the negative regulation of inflammation [30]. The mediation of cell death for lymphocytes and granulocytes strongly involves the Fas/FasL apoptosis pathway [32-35].
Primary cultured macrophages can survive 1 to 2 weeks after infection with 104 particles per cell of Ad-FasL (data not shown). Fas-associated death domain-like interleukin 1beta-converting enzyme (FLICE)-inhibitory protein (Flip), a negative regulator of Fas-induced apoptosis, is upregulated when monocytes differentiate into macrophages, which may confer the resistance macrophages have to Fas-mediated apoptosis [36,37]. The survival of macrophages after FasL gene transfer may be necessary for the phagocytosis of apoptotic cells during 'molecular synovectomy' using the 'FasL gene scalpel'. On the other hand, the survival of macrophages carrying viral vector antigens may activate the naive lymphocytes and initiate an anti-virus immune response if the adenovirus vector was used for clinical administration. Thus, the development of an immune tolerant gene delivery vehicle is an important task for the clinical use of FasL gene scalpel.
Compared with Ad-LacZ gene transfer, Ad-FasL gene transfer induces a significantly increased frequency of apoptotic cells in RA synovium (Figs 3e,f and 4), even though adenovirus vector itself shows a slight effect on enhancing apoptosis in synoviocytes (Fig. 4) and certain other kinds of cells [38,39]. From the point of view that repeated FasL gene transfer can remove RA synovium in vivo (Fig. 5b; Table 1), the FasL gene may function as a sharp scalpel for molecular synovectomy. It has been reported that Fas-mediated apoptosis is associated with activation of the Fas-associated death domain protein/Caspase-8/Caspase-3/poly(ADP-ribose)polymerase pathway and the c-Jun amino-terminal kinase/activator protein-1 pathway [40,41]. The former is thought to be the major signaling pathway required for Fas-mediated apoptosis in RA synoviocytes; the latter appears to be involved in the pathogeneses of inflammation by inducing pro-inflammatory cytokine/chemokine production [40,42]. Thus, the clinical potential of a FasL gene scalpel will depend on the clarification of the side effects of FasL intra-articular gene transfer. The combined administration of the intra-articular gene transfer of FasL with other genes and/or certain anti-inflammatory drugs might be an ideal complementary therapeutic approach for the local treatment of RA and other arthropathies.
A subset of chondrocytes located in the superficial zone of cartilage expresses the Fas antigen, and activation of the Fas receptor triggers apoptosis in these cells [43,44]. An adenovirus mediated short-term (24 h) FasL gene transfer in vivo in the rabbit knee only transduced synovium, but it did not affect the viability of chondrocytes in cartilage [12]. Our in vivo experiments with Ad-LacZ gene transfer on human synovium and cartilage in SCID mice presents similar results (Fig. 1e,f), but after a long-term (4 weeks) in vitro culture the chondrocytes in the superficial zone of cartilage were infected by Ad-LacZ 24 h after virus infection (unpublished data). This phenomenon suggests that the loss of cartilage matrix may have occurred during long-term in vitro culture. The cartilage matrix may be important for protection of chondrocytes from infection by adenovirus vector mediated gene transfer. Thus, the most suitable period for the treatment of rheumatoid arthritis with gene transfer would be in the early stage of the disease when the cartilage matrix is undamaged. The effects of long-term gene transfer of an apoptosis inducer such as FasL on the chondrocytes in cartilage in vivo require further investigation.
Currently, potential strategies for the treatment of RA using intra-articular gene transfer are focused on reducing the production and activation of inflammatory cytokines/chemokines and proteases/adhesion molecules, and inducing apoptosis of inflammatory synoviocytes. The latter may be more efficient in suppressing the inflammation by removing the producing and reactive cells responding to pro-inflammatory molecules. This potential therapeutic approach using 'gene scalpels' may function as a molecular synovectomy, not only for the treatment of RA, but also for the treatment of other arthropathies.
Conclusion
Our study elucidates firstly the in vivo therapeutic potential of molecular synovectomy using 'gene scalpels' for the treatment of RA locally through repeated intra-articular gene transfer of an apoptosis inducer. This approach may more efficiently arrest intra-articular inflammation and suppress the progress of RA by removing the cells producing and/or responding to pro-inflammatory molecules in arthritic joints.
Abbreviations
Ad-FasL = recombinant FasL adenovirus; Ad-LacZ = LacZ adenovirus; bp = base pair; FasL = Fas ligand; PBS = phosphate buffered saline; RA = rheumatoid arthritis; RT-PCR = reverse transcription-polymerase chain reaction; SCID = severe combined immunodeficiency; TUNEL = terminal deoxyribonucleotidyl transferase-mediated dUTP nick end labeling; X-gal = 5-bromo-4-chloro-3-indolyl β-d-galactopyranoside.
Competing interests
The author(s) declare that they have no competing interests.
Authors' contributions
GG provided the adenoviruses for study. GC contributed to the experiments and generated the histology data. RS coordinated the collection of the synovial tissues from patients in the clinic, carried out the pathological diagnosis of the human samples and provided advisory support for the study. HZ conceived of the study, participated in its design and coordination, carry out the in vivo study and molecular examinations, and drafted the manuscript. All authors read and approved the final manuscript.
Acknowledgements
This work was supported by grants 2T32AR07442-12 and R03 AR47983 from the National Institutes of Health/National Institute of Arthritis and Musculoskeletal and Skin Diseases (NIAMS), a research award from the University of Pennsylvania Research Foundation, and an investigator award from the Arthritis Foundation. We thank Dr Robert A Eisenberg for the advisory support with the NIH training grant; Dr James M Wilson for critical support with the collaboration; Drs William V Williams and Shigekazu Nagata for cells and reagents; Drs Philip Cohen and William V Williams for helpful discussion; Ms Marie Sieck for photographs; the Animal Facility at Veterans Affairs Research Center in Philadelphia and the Department of Orthopaedic Surgery at University of Pennsylvania Health System for assistance and cooperation with animal care and human sample collection.
Figures and Tables
Figure 1 Adenovirus vector mediated LacZ transgene expression in synovial tissues. Synovium alone or synovium with cartilage from joint replacement surgery and synovectomy was implanted subcutaneously on the lower backs of C.B-17 severe combined immunodeficiency (SCID) mice. LacZ adenovirus (Ad-LacZ) in 0.1 ml PBS was injected at a dosage of 1011, 1010, or 109 particles into the middle of the grafted tissues. An equal amount of PBS was injected as a control. Three days after virus injection, synovial tissues were collected and LacZ transgene expression in the synovial tissues was detected by X-gal staining. (a) The dose of 1011 virus particles of Ad-LacZ per injection resulted in high transgene expression in human synovium; (b) the dose of 1010 virus particles of Ad-LacZ per injection resulted in lower transgene expression; (c) the dose of 109 virus particles of Ad-LacZ per injection failed to induce LacZ transgene expression, similar to (d) PBS injected control tissues. (e) The expression of LacZ transgene only was apparent in synovium but not in cartilage three days after injection of 1011 virus particles (the red arrow indicates synovium; the white arrow indicates cartilage). (f) There was no evidence of chondrocyte infection by Ad-LacZ in cartilage, as demonstrated by X-gal/hematoxylin and eosin staining (original magnification 200×). A variety of cell types including (g) synoviocytes, (h) lymphocytes, (i) macrophages, (j) granulocytes and (k) endothelial cells could be infected by adenovirus vector, indicated by LacZ transgene expression detected with X-gal staining (g-k) (original magnification 400×).
Figure 2 Time dependent transgenic Fas ligand (FasL) mRNA expression in human synovium in vivo. The 1011 virus particles of recombinant FasL adenovirus (Ad-FasL) and LacZ adenovirus (Ad-LacZ) were injected into the grafted human synovium in severe combined immunodeficiency (SCID) mice in vivo. The injected tissue samples were harvested at days 2, 5, 7, 10, and 20 after virus injection. Total tissue RNA was isolated and the first-strand cDNA of transgenic FasL was detected by RT-PCR as described in Materials and methods. (a) Lane 1, PBS injected tissue; lanes 2–6, Ad-FasL infected tissues harvested on day 2, 5, 7, 10, and 20, respectively, after virus injection; lanes 7–11, the Ad-LacZ infected tissues harvested at the same time points. Ethidium bromide staining of fractionated DNA at 688 base pairs (bp) represents transgenic FasL mRNA expression; the DNA fragments at 505 bp represent β-actin expression as the internal control for the quality of mRNA samples. (b) Transgenic FasL mRNA in vivo was highly expressed on days 2–10 (lanes 2–5), and obviously decreased on day 20 (lane 6), indicated by the ratio of luminosity between transgenic FasL cDNA and the endogenous β-actin cDNA.
Figure 3 Detection of apoptotic cells and Fas/Fas ligand (FasL) protein expression in grafted rheumatoid arthritis (RA) synovium. The RA synovium in severe combined immunodeficiency (SCID) mice was harvested approximately two weeks after engraftment without or with three day treatment with 1011 particles of recombinant FasL adenovirus (Ad-FasL), LacZ adenovirus (Ad-LacZ), or PBS before it was collected. (a-c) Fas and FasL protein expression in RA synovium in vivo were examined using an immunohistochemistry staining system. (a) Fas protein is highly expressed on synoviocytes in RA synovium (original magnification 400×). (b) FasL protein is lacking in RA synoviocytes treated with Ad-LacZ (original magnification 400×); (c) but a significant increase in FasL protein expression in RA synoviocytes appears three days after Ad-FasL injection (original magnification 400×). (d-i) Apoptotic cells in RA synovium were examined using the ApopTag fluorescein in situ apoptosis detection kit (Intergen). The red propidium iodide counter-staining indicates the nucleus of all synovial cells in RA synovium. The green fluorescein stains the fragmented DNA in the nucleus. (d) Apoptotic cells in PBS treated RA synovium are almost undetectable (original magnification 200×). (e) A slightly increased apoptotic cell population was seen in Ad-LacZ treated RA synovium (original magnification 200×), (f) while around a 15-fold increased frequency of apoptotic cells in RA synovium was observed in Ad-FasL treated RA synovium (original magnification 200×; arrows point to TUNEL-positive cells). (g-i) The location of the fragmented DNA was observed using a laser scanning confocal microscope. (g) The nucleus of synovial cells in RA synovium was stained with propidium iodide (red; bar = 15 μm). (h) Fragmented DNA was stained with fluorescein (green; bar = 15 μm). (i) The localization of the fragmented DNA was determined by the overlap of the fluorescein staining and propidium iodide counter-staining (bar = 15 μm), which characterizes cell death.
Figure 4 Efficient increase of apoptosis incidence in rheumatoid arthritis (RA) synoviocytes by administration of recombinant FasL adenovirus (Ad-FasL). The percentage of apoptotic cells in grafted RA synovium in severe combined immunodeficiency (SCID) mice is compared to that in the RA synovium three days after treatments with Ad-FasL, LacZ adenovirus (Ad-LacZ) and PBS. Total synoviocytes and apoptotic cells were counted in 10 high power fluorescent microscopic fields randomly selected from each section, and three sections at multiple layers for each sample were examined. The percentage of apoptotic cells was recorded and is represented as the mean and standard deviation for each group. Open bars represent the percentage of apoptotic cells in RA synovium two weeks after engraftment without treatment; striped bars represent the percentage of apoptotic cells in RA synovium three days after Ad-FasL, Ad-LacZ and PBS injections.
Figure 5 Reduction of the size and cell density of rheumatoid arthritis (RA) synovium in vivo by repeated Fas ligand (FasL) gene transfer. (a) RA synovium, 200 mg tissue per mouse, was grafted into the back of each severe combined immunodeficiency (SCID) mouse subcutaneously. Two weeks after engraftment, 1011 particles of recombinant FasL adenovirus (Ad-FasL) or Ad-LacZ adenovirus (Ad-LacZ) in 0.1 ml PBS were injected into the engrafted RA synovium in the SCID mice. Injections were repeated every two weeks for a total of five times. Three days after the fifth injection, the mice were euthanized and the RA synovium samples were collected, weighed, and examined with hematoxylin and eosin (H&E) staining. (b) A comparative view of the size of the grafted RA synovium after two months of repeated treatment with control Ad-LacZ (tissue on the left side) and Ad-FasL (tissue on the right side). (c) H&E staining shows that both synoviocytes and mononuclear cells are greatly decreased in Ad-FasL treated RA synovium, in which a partial recovery of the normal features of synovium was observed (original magnification 200×); (d) but the Ad-LacZ treated RA synovium maintains the inflammatory infiltrates (original magnification 200×)
Table 1 Effect of repeated Ad-FasL or Ad-LacZ treatment on the weight of RA synovium
Ad-LacZ treated Ad-FasL treated
Weight (mg) Rank no. Weight (mg) Rank no.
162.00 16 17.00 4
122.00 13 72.00 11
171.00 17 29.00 6
159.00 15 62.00 9
203.00 20 11.00 3
184.00 19 46.00 8
98.00 12 34.00 7
177.00 18 10.00 2
156.00 14 0.00 1
68.00 10 25.00 5
n1 = 10 T1 = 154 n2 = 10 T2 = 56
P < 0.005 (Rank sum testa)
The repeated gene transfer of recombinant FasL adenovirus (Ad-FasL) and LacZ adenovirus (Ad-LacZ) into rheumatoid arthritis (RA) synovium in severe combined immunodeficiency (SCID) mice are the same as described in Fig. 5. Ten samples per treatment group were examined. After two months of treatment with viruses, the RA synovium samples were collected. The weight of each sample was measured individually for each group. aThe rank sum test was used to determine the P-value (P < 0.005).
==== Refs
Yocum DE Lafyatis R Remmers EF Schumacher HR Wilder RL Hyperplastic synoviocytes from rats with streptococcal cell wall-induced arthritis exhibit a transformed phenotype that is thymic-dependent and retinoid inhibitable Am J Pathol 1988 132 38 48 3260752
Harris EDJr Rheumatoid arthritis: Pathophysiology and implications for therapy N Engl J Med 1990 322 1277 1289 2271017
Firestein GS Invasive fibroblast-like synoviocytes in rheumatoid arthritis Arthritis Rheum 1996 39 1781 1790 8912499
Cozzolino F Gigliotti S Giuzio E Angrisani C Surgical synovectomy in the treatment of rheumatoid arthritis: the results obtained in a controlled study Chirurgia Degli Organi di Movimento 1991 76 341 346 1800046
Zhang H Yang Y Horton J Samoilova EB Judge TA Turka LA Wilson JM Chen Y Amelioration of collagen-induced arthritis by CD 95 (Apo-1/Fas)-Ligand gene transfer J Clin Invest 1997 100 1 7 9202049
Nishioka K Hasunuma T Kato T Sumida T Kobata T Apoptosis in rheumatoid arthritis: a novel pathway in the regulation of synovial tissue Arthritis Rheum 1998 41 1 9 9433863 10.1002/1529-0131(199801)41:1<1::AID-ART1>3.0.CO;2-V
Goossens PH Schouten GJ 't Hart BA Bout A Brok HP Kluin PM Breedveld FC Valerio D Huizinga TW Feasibility of adenovirus-mediated nonsurgical synovectomy in collagen-induced arthritis-affected rhesus monkeys Hum Gene Ther 1999 10 1139 1149 10340546 10.1089/10430349950018139
Migita K Honda S Yamasaki S Hirai Y Fukuda T Aoyagi T Kita M Ida H Tsukada T Kawakami A Regulation of rheumatoid synovial cell growth by ceramide Biochem Biophys Res Commun 2000 269 70 75 10694479 10.1006/bbrc.2000.2239
Michael VV Alisa KE Cell cycle implications in the pathogenesis of rheumatoid arthritis Frontiers Biosci 2000 5 D594 601
Asahara H Hasunuma T Kobata T Inoue H Muller-Ladner U Gay S Sumida T Nishioka K In situ expression of protooncogenes and Fas/Fas ligand in rheumatoid arthritis synovium J Rheum 1997 24 430 435 9058644
Zhang X Jiang M Zheng D Rheumatoid arthritis synoviocyte hyperplasia and expression of fas and bcl-2 genes Chung-Hua i Hsueh Tsa Chih 1998 78 175 178 10923523
Yao Q Glorioso JC Evans CH Robbins PD Kovesdi I Oligino TJ Ghivizzani SC Adenoviral mediated delivery of FAS ligand to arthritic joints causes extensive apoptosis in the synovial lining J Gene Med 2000 2 210 219 10894267 10.1002/(SICI)1521-2254(200005/06)2:3<210::AID-JGM101>3.0.CO;2-P
Iwakura Y Tosu M Youshida E Takiguchi M Sato K Kitajiam I Nishioka K Yamamoto K Takeda T Hatanaka M Induction of inflammatory arthropathy resembling rheumatoid arthritis in mice transgenic for HTLV-I Science 1991 253 1026 1028 1887217
Fujisawa K Asahara H Okamoto K Aono H Hasunuma T Kobata T Iwakura Y Yonehara S Sumida T Nishioka K Therapeutic effect of the anti-Fas antibody on arthritis in HTLV-I tax transgenic mice J Clin Invest 1996 98 271 278 8755634
Ogasawara J Watanabe-Fukunaga R Adachi M Matsuzawa A Kasugai T Kitamura Y Itoh N Suda T Nagata S Lethal effect of the anti-Fas antibody in mice Nature 1993 364 806 809 7689176 10.1038/364806a0
Okamoto K Asahara H Kobayashi T Matsuno H Hasunma T Kobata T Sumida T Nishioka K Induction of apoptosis in the rheumatoid synovium by Fas Ligand gene transfer Gene Therapy 1998 5 331 338 9614552 10.1038/sj.gt.3300597
Zhang H Gao G Williams W Wilson J Schumacher HR Induction of apoptosis in synovial fibroblasts from rheumatoid arthritis patients by Fas ligand gene transfer [abstract] Arthritis Rheum 1998 41 S142 10.1002/1529-0131(199810)41:10<1867::AID-ART20>3.0.CO;2-W
Wakisaka S Suzuki N Takeba Y Shimoyama H Nagafuchi H Takeno M Saito N Yokoe T Kaneko T Asai T Sakane T Modulation by proinflammatory cytokines of Fas/Fas ligand-mediated apoptotic cell death of synovial cells in patients with rheumatoid arthritis (RA) Clin Exp Immunol 1998 114 119 128 9764613 10.1046/j.1365-2249.1998.00701.x
Ikeda T Kubo T Arai Y Nakanishi T Kobayashi K Takahashi K Imanishi J Takigawa M Hirasawa Y Adenovirus mediated delivery to the joints of guinea pigs J Rheumatol 1998 25 1666 1673 9733444
SaKai K Matsuno H Morita I Nezuka T Tsuji H Shirai T Yonehara S Hasunuma T Nishioka K Potential withdrawal of rheumatoid synovium by the induction of apoptosis using a novel in vivo model of rheumatoid arthritis Arthritis Rheum 1998 41 1251 1257 9663483 10.1002/1529-0131(199807)41:7<1251::AID-ART16>3.0.CO;2-B
Geiler T Kriegsmann J Keyszer GM Gay RE Gay S A new model for rheumatoid arthritis generated by engraftment of rheumatoid synovial tissue and normal human cartilage into SCID mice Arthritis Rheum 1994 37 664 71
Jorgensen C Demoly P Noel D Mathieu M Piechaczyc M Gougat C Bousquet J Sany J Gene transfer to human rheumatoid synovial tissue engrafted in SCID mice J Rheumatol 1997 24 2076 2079 9375863
Arnett FC Edworthy SM Bloch DA McShane DJ Fries JF Cooper NS Healey LA Kaplan SR Liang MH Luthra HS The American Rheumatism Association 1987 revised criteria for the classification of rheumatoid arthritis Arthritis Rheum 1988 31 315 24 3358796
Gao GP Yang YP Wilson JM Biology of adenoviral vectors deleted of E1 and E4 for liver-directed gene therapy J Virol 1996 70 8934 8943 8971023
Jorgensen C Bologna C Sany J Autoimmunity. Insights provided by the SCID mouse model Revue Du Rhum 1995 62 519 524
Dini L Ruzittu MT Falasca L Recognition and phagocytosis of apoptotic cells Scanning Microsc 1996 10 239 251 9813609
Fadok VA Savill JS Haslett C Bratton DL Doherty DE Campbell PA Henson PM Different populations of macrophages use either the votronectin receptor or the phosphatidylserine receptor to recognize and remove apoptotic cells J Immunol 1992 149 4029 4035 1281199
Rossi AG McCutcheon JC Roy N Chilvers ER Haslett C Dransfield I Regulation of macrophage phagocytosis of apoptotic cells by camp J Immunol 1998 160 3562 3568 9531319
Mogi M Fukuo K Yang J Suhara T Ogihara T Hypoxia stimulates release of the soluble form of Fas ligand that inhibits endothelial cell apoptosis Lab Invest 2001 81 177 184 11232639
Walsh K Sata M Negative regulation of inflammation by Fas ligand expression on the vascular endothelium Trends Cardiovasc Med 1999 9 34 41 10189965 10.1016/S1050-1738(99)00006-7
Sata M Suhara T Walsh K Vascular endothelial cells and smooth muscle cells differ in expression of Fas and Fas ligand and in sensitivity to Fas ligand-induced cell death: implications for vascular disease and therapy Thromb Vasc Biol 2000 20 309 316
Lundy SK Boros DL Fas ligand-expressing B-1a lymphocytes mediate CD4(+)-T-cell apoptosis during schistosomal infection: induction by interleukin 4 (IL-4) and IL-10 Infect Immun 2002 70 812 819 11796615 10.1128/IAI.70.2.812-819.2002
Liles WC Kiener PA Ledbetter JA Aruffo A Klebanoff SJ Differential expression of Fas (CD95) and Fas ligand on normal human phagocytes: implications for the regulation of apoptosis in neutrophils J Exp Med 1996 184 429 440 8760796 10.1084/jem.184.2.429
Hanna N Vasquez P Pham P Heck DE Laskin JD Laskin DL Weinberger B Mechanisms underlying reduced apoptosis in neonatal neutrophils Pediatr Res 2005 57 56 62 15557111 10.1203/01.PDR.0000147568.14392.F0
O'Flaherty E Wong WK Pettit SJ Seymour K Ali S Kirby JA Regulation of T-cell apoptosis: a mixed lymphocyte reaction model Immunology 2000 100 289 299 10929050 10.1046/j.1365-2567.2000.00048.x
Perlman H Pagliari LJ Georganas C Mano T Walsh K Pope RM FLICE-inhibitory protein expression during macrophage differentiation confers resistance to fas-mediated apoptosis J Exp Med 1999 190 1679 1688 10587358 10.1084/jem.190.11.1679
Perlman H Pagliari LJ Liu H Koch AE Haines GK Pope RM 3rd Rheumatoid arthritis synovial macrophages express the Fas-associated death domain-like interleukin-1beta-converting enzyme-inhibitory protein and are refractory to Fas-mediated apoptosis Arthritis Rheum 1999 44 21 30 10.1002/1529-0131(200101)44:1<21::AID-ANR4>3.0.CO;2-8
Teramoto S Matsuse T Matsui H Ohga E Ishii T Nagase T Ouchi Y Recombinant E1-deleted adenoviral vectors induce apoptosis in a rat airway epithelial mucous goblet cell line Japan J Pharmacol 1999 81 73 78 10580373 10.1254/jjp.81.73
Teramoto S Matsuse T Matsui H Ohga E Ishii T Ouchi Y Recombinant E1-deleted adenovirus vector induces apoptosis in two lung cancer cell lines Eur Respir J 1999 13 1125 1132 10414415 10.1034/j.1399-3003.1999.13e31.x
Kobayashi T Okamoto K Kobata T Hasumuna T Nishioka K Apomodulation as a novel therapeutic concept for the regulation of apoptosis in rheumatoid synoviocytes Curr Opinin Rheumatol 1999 11 188 193 10.1097/00002281-199905000-00007
Okamoto K Kobayashi T Kobata T Hasumuna T Kato T Sumida T Nishioka K Fas-associated death domain protein is a Fas-mediated apoptosis modulator in synoviocytes Rheumatol 2000 39 471 480 10.1093/rheumatology/39.5.471
Naumann M Rudel T Wieland B Bartsch C Meyer TF Coordinate activation of activator protein 1 and inflammatory cytokines in response to Neisseria gonorrhoeae epithelial cell contact involves stress response kinases J Exp Med 1998 188 1277 1286 9763607 10.1084/jem.188.7.1277
Lotz M Hashimoto S Kuhn K Mechanisims of chondrocyte apoptosis Osteoarthritis Cartilage 1999 7 389 391 10419776 10.1053/joca.1998.0220
Hashimoto S Setareh M Ochs RL Lotz M Fas/Fas ligand expression and induction of apoptosis in chondrocytes Arthritis Rheum 1997 40 1749 1755 9336406
|
16277676
|
PMC1297566
|
CC BY
|
2021-01-04 16:02:40
|
no
|
Arthritis Res Ther. 2005 Sep 8; 7(6):R1235-R1243
|
utf-8
|
Arthritis Res Ther
| 2,005 |
10.1186/ar1811
|
oa_comm
|
==== Front
Arthritis Res TherArthritis Research & Therapy1478-63541478-6362BioMed Central London ar18121627767210.1186/ar1812Research ArticleRheumatoid arthritis seropositive for the rheumatoid factor is linked to the protein tyrosine phosphatase nonreceptor 22-620W allele Dieudé Philippe [email protected] Sophie [email protected] Laëtitia [email protected] Elisabeth [email protected] Elodie [email protected] Céline [email protected] Sandra [email protected] Thomas [email protected] Bernard [email protected]élis François [email protected] European Consortium on Rheumatoid Arthritis Families1 GenHotel-EA3886, Evry-Genopole, Evry, France2 Unité de Génétique Clinique, Fédération de Rhumatologie, Centre Viggo-Petersen, Hôpital Lariboisière, Assistance Publique des Hôpitaux de Paris, Paris, France3 Laboratoire Statistique et Génome, Evry-Genopole, Evry, France4 Consultation de Génétique Adulte, Centre Hospitalier Sud-Francilien, Evry-Corbeil, France2005 25 8 2005 7 6 R1200 R1207 14 5 2005 21 6 2005 15 7 2005 4 8 2005 Copyright © 2005 Dieudé et al.; licensee BioMed Central Ltd.This is an Open Access article distributed under the terms of the Creative Commons Attribution License (), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
The protein tyrosine phosphatase nonreceptor type 22 (PTPN22) gene encodes for lymphoid tyrosine phosphatase LYP, involved in the negative regulation of early T-cell activation. An association has recently been reported between the PTPN22-620W functional allele and rheumatoid factor-positive (RF+) rheumatoid arthritis (RA), among other autoimmune diseases. Expected linkage proof for consistency cannot be definitely produced by an affected sib-pair (ASP) analysis. Our aim was therefore to search for linkage evidence with the transmission disequilibrium test.
DNA from the French Caucasian population was available for two samples of 100 families with one RA patient and both parents, and for 88 RA index cases from RA ASP families. Genotyping was carried out by PCR-restriction fragment length polymorphism. The analysis was performed using the transmission disequilibrium test, genotype relative risk and ASP-based analysis.
The transmission disequilibrium test of the PTPN22-620W allele revealed linkage and association for RF+ RA (61% of transmission, P = 0.037). The genotype relative risk showed the risk allele in 34% of RF+ RA patients and in 24% of controls derived from nontransmitted parental chromosomes (P = 0.047, odds ratio = 1.69, 95% confidence interval = 1.03–2.78). The ASP investigation showed no enriched risk allele in RA multiplex families, resulting in a lack of power of ASP analysis, explaining the published negative results.
This study is the first to show linkage of PTPN22 to RF+ RA, consistent with PTPN22 as a new RA gene.
==== Body
Introduction
Rheumatoid arthritis (RA), the most common autoimmune disease, is thought to be a complex disease in which a combination of risk alleles from different susceptibility genes predisposes to the development of the disease, following exposure to as yet unknown environmental factors. Several genome scans have suggested multiple RA loci [1-8], and recent case-control association studies have suggested new RA genes [9,10]. However, only HLA-DRB1 alleles have been both linked to and associated with RA, fulfilling the criteria for a fully demonstrated genetic factor [11].
A genetic association involving a functional polymorphism of the protein tyrosine phosphatase nonreceptor type 22 (PTPN22) gene was reported to be associated with rheumatoid factor-positive (RF+) RA, with type 1 diabetes, with systemic lupus erythematosus and with autoimmune thyroid disease [12-19]. The PTPN22 gene encodes for the intracellular tyrosine phosphatase LYP, which acts as a negative regulator of early T-cell activation through binding to the Csk protein [20,21].
The PTPN22 single nucleotide polymorphism (SNP) (1858C/T) (rs 2476601) occurs as a result of an amino acid substitution of arginine for tryptophan at position 620 (R620W), in the P1 proline-rich domain. This domain is involved in the binding to the SH3 domain of Csk. Functional analysis showed that it affects the binding of LYP to Csk, leading to a lack of downregulation of T-cell activation, which is consistent with an increased susceptibility to autoimmunity for the 620W allele [12,14].
The PTPN22-1858T allele has been reported to be associated with RF+ RA in several case-control studies [12,18,19,22]. The first study, performed in a white North American population, reported an association between the PTPN22-1858T allele and RF+ sporadic RA (P = 6.6 × 10-4). This association was replicated in a different sample with multiplex RA cases (P = 5.6 × 10-8), the association being restricted to RF+ RA patients [12]. The second study, also performed in a white North American population, compared the frequency of the PTPN22 risk allele between the Study of New Onset Rheumatoid Arthritis cohort and the control sample of the previous study [12], observing the association between the 1858T allele and early RF+ RA. The study also suggested a stronger association for the homozygous genotype 1858T/T [22]. Three recent case-control studies performed in UK, Spanish and North-American Caucasian populations also found an association between the PTPN22-1858T allele and RA [18,19,23]. In contrast, the Spanish study observed no dose effect of the suspected allele [18]. The UK study found an increased frequency of the suspected PTPN22 allele in the RF+ RA cases and suggested a stronger association for the homozygous genotype [19]. The US study confirmed this association was restricted to RF+ RA and also showed a significantly higher risk for the homozygous genotype [23].
These findings provide strong evidence for the involvement of PTPN22 in RF+ RA susceptibility [24]. The linkage proof is so far lacking, however, as the linkage analysis of the North American Rheumatoid Arthritis Consortium RA-affected sib-pairs resource for this PTPN22 SNP was inconclusive [12]. The transmission disequilibrium test (TDT), simultaneously investigating linkage and association, is predicted to be more powerful than the affected sib-pair (ASP) analysis in demonstrating linkage for a factor such as PTPN22 [25,26]. Three family-based association and linkage studies using TDT analysis were recently reported, providing linkage evidence of PTPN22 to type 1 diabetes [15,27,28]. The aim of the present study was to test this PTPN22 polymorphism for linkage to RF+ RA in the French Caucasian population, taking advantage of the TDT.
Patients and methods
Study design and study population
A TDT linkage study was conducted to investigate the PTPN22-1858C/T SNP in RA for one Caucasian population. RA patients and family members were recruited through a national media campaign in France, which was followed by the selection of individuals who fulfill the American College of Rheumatology (formerly the American Rheumatism Association) 1987 revised criteria for RA [29], according to the physician in charge of the patient. All clinical data were reviewed by rheumatologists from our team (SL, LM or P Fritz). All individuals provided informed consent and the ethics committee of the Hôpital Bicêtre approved the study.
Transmission disequilibrium test RA samples
Inclusion criteria for the two samples of the 100 French Caucasian families investigated here were the participation of one RA patient and both parents, as well as a French Caucasian origin of the family, defined by the four grandparents being French Caucasian. Families with an additional sibling with RA or RA patients who were younger than 18 years old were excluded. RA characteristics of index cases from TDT samples 1 and 2 are summarized in Table 1.
Affected sib-pair RA sample
The 88 index RA patients from the French Caucasian ASP families that had been analyzed for a refined genome scan were investigated in this study [1]. Inclusion criteria for the sample of 88 families had been the participation of at least two siblings with RA and of French Caucasian origin, with all four grandparents being of European Caucasian origin. Families with RA patients younger than 18 years old were excluded. Of these 88 families, 81 had two affected siblings, six families had three affected siblings and one family had four affected siblings. Characteristics of the 88 RA index cases investigated in this study are summarized in Table 1. All ASP families had been previously genotyped for two microsatellite markers flanking the PTPN22 locus (D1S418 and D1S252) located at approximately -1 and +3 Mb, respectively, on either side of the PTPN22 locus, with heterozygosities of 80% and 81%, respectively [1,30].
Molecular genotyping methods
Genomic DNA was purified from fresh peripheral blood leukocytes by standard methods [31]. HLA-DRB1 typing (Dynal Classic high resolution and Sequence Specific Primers DR low resolution) and subtyping (Dynal Classic high resolution, for HLA-DRB1*01, HLA-DRB1*04, HLA-DRB1*11 and HLA-DRB1*13) were carried out using the PCR sequence-specific primers method (Dynal Biotech, Lake Success, NY, USA).
Genotyping of the PTPN22-1858C/T SNP was performed by PCR-restriction fragment length polymorphism. The sense and antisense primers were, respectively, 5'-GATAATGTTGCTTCAACGGAATTT-3' and 5'-CCATCCCACACTTTATTTTATACT-3'. The PTPN22-1858C/T transition at codon 620 eliminates a restriction site for RsAI in the 1858T allele. TDT RA sample 1 and sample 2 genotypes were checked with the PCR-restriction fragment length polymorphism using the XcmI enzyme, for which the 1858T allele creates a restriction site. Each genotype was interpreted independently by two of the investigators (EG and PD).
Rheumatoid factor status
The RF+ status was provided by the presence of at least one positive RF+ result during the disease course, as determined by latex fixation, by Waaler Rose assay or by laser nephelometry. The RF test was performed at least once for all TDT and ASP RA patients. The anti-cyclic citrullinated peptide status of RA patients was not available.
Hardy–Weinberg equilibrium check
The Hardy–Weinberg equilibrium of the PTPN22-1858C/T polymorphism was investigated using a chi-square test with one degree of freedom.
Analysis
We planned a linkage test of the PTPN22-1858T allele RA hypothesis, restricted to RF+ RA patients. This hypothesis was first tested using the TDT RA sample 1. In case linkage was observed, or at least suggested, a replication test was planned with the TDT RA sample 2 and a global analysis for all TDT RA families. We also investigated the PTPN22 putative genotype in the index ASP RA sample, taking advantage of the linkage data available at the PTPN22 locus, as previously described [32].
Test for linkage and association in the TDT RA samples
Linkage and association analysis were performed using the TDT [33] and the genotype relative risk (GRR) test [34]. The TDT compares the transmission of the SNP alleles from heterozygous parents to affected offspring, with Mendel's law expectation (50%), using a chi-square test with one degree of freedom. Similar to a case–control study, GRR compares the SNP genotypes distribution in RA cases and in 'controls' (controls are derived from nontransmitted parental chromosomes, for each family), using a chi-square test with the appropriate degree of freedom or the Fisher's exact test. P < 0.05 was considered significant.
Linkage-based test in the ASP RA sample [32]
Genetic factors are expected to be concentrated in families with multiples RA cases, such as ASP families. Within index RA cases, those sharing identical by descent chromosomes at the PTPN22 locus with their RA affected sib could be expected to concentrate further PTPN22 RA genetic factors. The putative PTPN22 genotypes were compared between the ASP RA index cases, the TDT RA cases and the controls from the TDT RA samples (controls are derived from nontransmitted parental chromosomes). For the linkage-based association test, the RF+ index cases that shared at least one allele identical by descent with their RF+ RA sib (IBD1 or IBD2) were used, taking advantage of the linkage data available at the PTPN22 locus [32].
Stratified linkage analysis based on the PTPN22-1858C/T genotypes
We conducted a linkage analysis using Allegro 1.1 software [35], taking into account the PTPN22-1858C/T genotypes, to select the subgroup of families with an index carrying the putative genotypes.
Power calculation
Assuming a PTPN22-1858T allele association similar to that of the North American population (14.8% allele frequency in RF+ RA cases and 8.7% in controls) [12], association analysis of our 100 TDT families (TDT RA sample 1) provides a 95% power to show a suggestion for association and a 53% power to reach statistical significance (P < 0.05). Our sample of 200 TDT families provides a 79% power for significance.
Results
Hardy–Weinberg equilibrium check
The PTPN22-1858C/T polymorphism was in Hardy-Weinberg equilibrium in the control samples investigated.
Test for linkage and association in the TDT RA samples
TDT RA sample 1
The PTPN22-1858T allele was more frequent in the RF+ RA cases than in the controls: 20% versus 11% (P = 0.022, odds ratio [OR] = 2.05, 95% confidence interval [CI] = 1.1–3.8). The allele frequency observed in rheumatoid factor-negative (RF-) RA patients was 16%, compared with 10.5% in the resulting controls (P = 0.52). Significant linkage to RF+ RA was observed with an excess of transmission of the 1858T allele from heterozygous parents to RA cases (66% versus 50%, n = 47, P = 0.029) (Table 2). The GRR analysis revealed a statistically significant increase in the frequency of genotypes carrying the PTPN22-1858T allele (1858C/T + 1858T/T) in RF+ RA cases (35%) compared with controls (21%) (Table 3).
TDT RA sample 2
We observed an excess of transmission of the PTPN22-1858T allele to RF+ RA patients that was not significant (56%, P = 0.45) (Table 2). The GRR analysis showed a nonsignificant increase of the genotypes carrying the 1858T allele in RF+ RA patients compared with controls (Table 3).
Combined analysis of TDT samples
No statistically significant difference was observed between samples 1 and 2, allowing pooling for combined analysis. PTPN22 linkage to RF+ RA was significant (T-allele transmission, 61%; n = 90, P = 0.037). By contrast, transmission to RF- RA followed Mendel's law exactly (50%) (Table 2). The PTPN22-1858T allele frequency was significantly increased in RF+ RA compared with controls (19% versus 13%, P = 0.029, OR = 1.62, 95% CI = 1.05–2.50). The GRR analysis showed a significant increase of PTPN22 genotypes carrying the PTPN22-1858T allele in RF+ RA patients compared with controls (34% versus 24%, P = 0.047, OR = 1.69, 95% CI = 1.03–2.78). In the RF- RA patients, genotype frequencies were identical to those of controls, in keeping with the 50% transmission (Table 3).
No correlation between the HLA-DRB1 shared epitope status (DRB1*0101, DRB1*0102, DRB1*0401, DRB1*0404, DRB1*0405, DRB1*0408, DRB1*1001) and the PTPN22 genotypes in RF+ RA index cases was observed (Table 4). Apart from the RF status, no specific clinical features (erosive disease, age at disease onset) were found to be associated with the PTPN22-1858T/C or PTPN22-1858T/T genotypes (data not shown).
Linkage-based test in the RA multiplex ASP sample
The frequency of the PTPN22-1858T allele was similar in the RF+ ASP RA cases compared with the RF+ TDT RA cases (18% versus 19%). The linkage-based subgroup (IBD1 or IBD2) of RF+ RA index cases with concordant RF+ RA sibs showed no increase in the frequency of the suspected allele, compared with RF+ TDT RA cases. The GRR analysis was consistent with those findings, the PTPN22-1858C/T or PTPN22-1858T/T genotype frequency being equal between RF+ RA ASP index cases and RF+ RA TDT cases (Table 5).
Stratified linkage analysis at the PTPN22 locus based on the 1858C/T genotypes
Previous linkage analysis of these ASP families had shown no linkage at the PTPN22 locus (P = 0.74) [1]. Stratified linkage analysis in RF+-concordant ASP, even after selection of the families with index cases carrying the PTPN22-1858C/T or PTPN22-1858T/T genotypes, still showed no linkage evidence at the PTPN22 locus (P = 0.69).
Discussion
We searched for the PTPN22-1858T allele linkage to RF+ RA using the TDT, which simultaneously tests linkage and association, avoiding the major drawback of inevitable imperfect matching between cases and controls. Here, we provide linkage evidence for RF+ RA to the PTPN22-1858T allele. We also observed association with the PTPN22-1858C/T or PTPN22-1858T/T genotypes and we report for the first time an estimation of the association in the French Caucasian population for RF+ RA (34% versus 24%, P = 0.047, OR = 1.69, 95% CI = 1.03–2.78). In ASP RF+ RA index cases, the 1858C/T or 1858T/T genotype has a similar frequency as the TDT RF+ RA index cases. The association appears to be independent from the HLA-DRB1 shared epitope.
Our findings therefore provide linkage evidence in support of PTPN22 as a new RF+ RA genetic factor, concurring with previously reported case–control studies [12,18,19,22,23]. We extend this observation to the French Caucasian population, in which the magnitude of the association is similar.
The linkage evidence provided by this study remains statistically modest. Further linkage studies are needed to definitively establish linkage of the PTPN22-1858T allele to RF+ RA. For the observed transmission disequilibrium of 61%, a TDT sample size of 232 families would be required to obtain, with 80% power, an independent replication of the linkage evidence reported here.
Genome scans are popular as they do not require any a priori hypothesis to detect disease loci. They are clearly unable to detect all disease loci, however, especially factors such as PTPN22. The increased power of the TDT over the ASP analysis for such factors was predicted long ago [26]. Our observation of the absence of a major increased frequency of the risk allele in multiplex ASP families when compared with sporadic cases, as reported by Begovich and colleagues [12], allows us to estimate the excess of allele sharing expected in the ASP analysis over the Mendel expectation of 50%. Using the estimation of the divergence from Mendel's law obtained from this study (61% transmission from heterozygous parents to RF+ RA patients, instead of 50%) and the genotype frequencies observed, the allele sharing expected is 52% for all families, 53% for the subgroup of RF+-concordant families and 56% for the small subgroup of RF+-concordant families with the 1858C/T or 1858T/T index case.
A huge sample size would therefore be required to demonstrate a significant excess of allele sharing over Mendel's law. In that regard, the PTPN22 situation is similar to that of the insulin gene in type 1 diabetes, for which the discrepancy between numerous association reports and the absence of linkage in ASP analysis was resolved using a TDT-like analysis [36]. This explains the complete absence of linkage evidence that we observed in our ASP analysis, in keeping with the absence of clear ASP linkage reported by Begovich and colleagues [12]. Further analysis using sophisticated software such as GIST might help clarify this point [37].
As indicated by Begovich and colleagues, the chromosome 1 linkage suggestion observed in the ASP analysis of the North American Rheumatoid Arthritis Consortium genome scan is not explained by the findings of the PTPN22 association [12]. New RA genes detected by such linkage suggestions, which could be expected to be stronger RA factors, remain to be discovered. Hence the major interest in genome scan persists, despite the lack of power for some RA genes, such as PTPN22.
Interestingly, transmission of the 1858T allele to RF- RA cases precisely followed Mendel's law, with genotype frequencies identical to controls, strengthening the evidence that the PTPN22-620W role is restricted to RF+ RA [12,19,22,23]. PTPN22 is probably the first example of a fully confirmed RA gene involved specifically in a precise aspect of RA clinical heterogeneity (RF+ RA). The absence of correlation between PTPN22 and HLA-DRB1 genotypes suggests that both RA genes could be involved in distinct gene combinations predisposing to RA, providing the first example of a clear genetic heterogeneity in RA.
Because the association is relatively modest, no genetic testing would be clinically indicated. Instead, the clinical relevance of the finding is likely to come through the better understanding of RA pathophysiology. It may lead to new therapeutic targets, aiming at the cause of RA, possibly shared by other autoimmune diseases.
Interestingly, all autoimmune diseases reported to be associated with the PTPN22-1858T allele are characterized by the production of autoantibodies [12-14,16,22], suggesting that the 620W variant of LYP could be implicated not only in T-cell activity regulation, but also in B-cell autoreactivity [24]. It will consequently be of major interest to test further for association of the PTPN22-1858T allele in RA families with a clustering of multiple autoimmune diseases to measure precisely this association with each disease [38-41].
Conclusion
Our findings provide linkage evidence for the involvement of the PTPN22-1858T allele in RF+ RA genetic susceptibility, in the French Caucasian population, independent of HLA-DRB1. This is in keeping with the proposal of PTPN22 as a new RA susceptibility gene.
Abbreviations
ASP = affected sib-pair; GRR = genotype relative risk; PCR = polymerase chain reaction; RA = rheumatoid arthritis; RF = rheumatoid factor; SNP = single nucleotide polymorphism; TDT = transmission disequilibrium test.
Competing interests
The author(s) declare that they have no competing interests.
Authors' contributions
PD and SG carried out the molecular genetic studies. LM, EP-T, EG, CP, SL and FC performed acquisition of the data. PD, SG, LM, CP, BP and FC analyzed and interpreted the data. LM, SL and TB made a substantial contribution to the acquisition of clinical data and the follow-up of the patients. All authors read and approved the final manuscript.
Acknowledgements
The authors thank the patients, their families, their physicians and Dr P Fritz (Centre Viggo-Petersen, Hôpital Lariboisière, Paris, France) for their participation. For funding, the authors thank Association Française des Polyarthritiques, Association Rhumatisme et Travail, Association Polyarctique and Groupe Taitbout, Association de Recherche pour la Polyarthrite, Société Française de Rhumatologie, Genopole, Conseil Régional Ile de France, Faculté de Médecine Lariboisière Saint-Louis and Ministère de la Recherche et de l'Enseignement Supérieur. Institutional support from Shering-Plough, Amgen, Pfizer and Wyeth was gratefully received. The authors thank Dr JF Prudhomme, Dr C Bouchier and Professor J Weissenbach at Genethon and Dr C De Toma, MF Legrand and Professor G Thomas at Fondation Jean Dausset-CEPH for technical support. They also thank M Dieudé and C Robinson for critical reading of the manuscript (Hôpital Notre-Dame, Montreal, Québec, Canada).
Figures and Tables
Table 1 Characteristics of rheumatoid arthritis (RA) index cases from the investigated samples
TDT RA sample 1 (n = 100) TDT RA sample 2 (n = 100) ASP RA sample (n = 88)
Female (%) 87 90 84
Mean age (± standard deviation) at disease onset (years) 32 (± 10) 31 (± 6) 40 (± 14)
Mean (± standard deviation) disease duration (years) 18 (± 7) 16 (± 8) 23 (± 10)
RA patients with bone erosions (%) 90 79 80
RA patients seropositive for rheumatoid factor (%) 81 76 84
RA patients carrying at least one HLA-DRB1 shared epitope allelea (%) 78 80 77
TDT, transmission disequilibrium test; ASP, affected sib-pair
aDRB1*0101, DRB1*0102, DRB1*0401, DRB1*0404, DRB1*0405, DRB1*0408, DRB1*1001.
Table 2 Linkage analysis of the PTPN22-1858T allele to rheumatoid factor-seropositive (RF+) rheumatoid arthritis (RA) using the transmission disequilibrium test (TDT)
Families Transmission [% (n)] P
TDT RA sample 1
TDT RA index cases (n = 100) 66 (55) 0.022
TDT RA index cases RF+ (n = 81) 66 (47) 0.029
TDT RA index cases RF- (n = 19) 63 (8) 0.48
TDT RA sample 2
TDT RA index cases (n = 100) 53 (57) 0.69
TDT RA index cases RF+ (n = 76) 56 (43) 0.45
TDT RA index cases RF- (n = 24) 43 (14) 0.59
All TDT RA families
All TDT RA index cases (n = 200) 59 (112) 0.059
All TDT RA index cases RF+ (n = 157) 61 (90) 0.037
All TDT RA index cases RF- (n = 43) 50 (22) 1
RF-, seronegative for rheumatoid factor. Transmission, percentage of heterozygous 1858C/T parents transmitting the 1858T allele. The plan was to test the hypothesis for RF+ RA; analysis for all RA and RF- RA were provided for the discussion.
Table 3 Association of PTPN22 genotypes carrying the 1858T allele and rheumatoid factor-seropositive (RF+) rheumatoid arthritis (RA)
PTPN22 genotypes [% (n)] Pa Odds ratio (95% confidence interval)
C/C C/T T/T C/T or T/T
TDT RA sample 1
All TDT RA index cases (n = 100) 65 (65) 31 (31) 4 (4) 35 (35)
Controlsb (n = 100) 79 (79) 20 (20) 1 (1) 21 (21) 0.029 2.05 (1.07–3.81)
TDT RA index cases RF+ (n = 81) 64 (52) 31 (25) 5 (4) 36 (29)
Controlsb (n = 81) 79 (64) 20 (16) 1 (1) 21 (17) 0.038 2.1 (1.04–4.24)
TDT RA index cases RF- (n = 19) 68 (13) 32 (6) 0 32 (6)
Controlsb (n = 19) 79 (15) 21 (4) 0 21 (4) 0.71
TDT RA sample 2
All TDT RA index cases (n = 100) 69 (69) 27 (27) 4 (4) 31 (31) 0.76
Controlsb (n = 100) 71 (71) 26 (26) 3 (3) 29 (29)
TDT RA index cases RF+ (n = 76) 68 (52) 28 (21) 4 (3) 32 (24)
Controlsb (n = 76) 74 (56) 24 (18) 2 (2) 26 (20) 0.47
TDT RA index cases RF- (n = 24) 71 (17) 25 (6) 4 (1) 29 (7)
Controlsb (n = 24) 63 (15) 33 (8) 4 (1) 37 (9) 0.76
All TDT RA families
All TDT RA index cases (n = 200) 67 (134) 29 (58) 4 (8) 33 (66)
Controlsb (n = 200) 75 (150) 23 (46) 2 (4) 25 (50) 0.078
TDT RA index cases RF+ (n = 157) 66 (104) 29 (46) 5 (7) 34 (53)
Controlsb (n = 157) 76 (120) 22 (34) 2 (3) 24 (37) 0.047 1.69 (1.03–2.78)
TDT RA index cases RF- (n = 43) 70 (30) 28 (12) 2 (1) 30 (13)
Controlsb (n = 43) 70 (30) 28 (12) 2 (1) 30 (13) 1
RF-, seronegative for rheumatoid factor.
aFollowing data previously reported in RA and because of the infrequency of the PTPN22-1858T/T genotype, it was combined with the 1858C/T genotype for the analysis.
bControls derived from nontransmitted parental chromosomes.
Table 4 PTPN22-1858 C/T genotypes distribution according to the HLA-DRB1 shared epitope (SE)
PTPN22 C/T or T/T PTPN22 C/C P
TDT RA sample 1 0.88
HLA-DRB1*SE/SE 10 16
HLA-DRB1*SE/X 18 34
HLA-DRB1*X/X 7 15
TDT RA sample 2 0.16
HLA-DRB1*SE/SE 13 17
HLA-DRB1*SE/X 14 35
HLA-DRB1*X/X 4 17
All TDT RA families 0.25
HLA-DRB1*SE/SE 23 33
HLA-DRB1*SE/X 32 69
HLA-DRB1*X/X 11 32
HLA-DRB1*SE/SE, two shared epitopes; HLA-DRB1*SE/X, one shared epitope; HLA-DRB1*X/X, zero shared epitope.
Table 5 PTPN22-1858C/T genotypes frequencies in the affected sib-pair (ASP) rheumatoid arthritis (RA) sample
PTPN22-1858C/T genotype frequencies [% (n)] Pa
C/C C/T T/T C/T or T/T
ASP RA index cases RF+ (n = 74) versus controlsb (n = 200) 66 (49) 31 (23) 3 (2) 34 (25) 0.15
ASP RA index cases RF+ and IBD1 or IBD2 at the PTPN22 locus with their RF+-concordant RA sib (n = 42) versus controlsb (n = 200) 71 (30) 26 (11) 3 (1) 29 (12) 0.63
All ASP RA index cases (n = 88) versus controlsb (n = 200) 69 (61) 29 (25) 2 (2) 31 (27) 0.32
RF+, seropositive for rheumatoid factor; IBD1 or IBD2, index sharing 1 or 2 identical by descent allele with the RA sib.
aFollowing data previously reported in RA and because of the infrequency of the PTPN22-1858T/T genotype, it was combined with the 1858C/T genotype for the analysis.
bControls derived from all non transmitted parental chromosomes.
==== Refs
Osorio YFJ Bukulmez H Petit-Teixeira E Michou L Pierlot C Cailleau-Moindrault S Lemaire I Lasbleiz S Alibert O Quillet P Dense genome-wide linkage analysis of rheumatoid arthritis, including covariates Arthritis Rheum 2004 50 2757 2765 15457443 10.1002/art.20458
Shiozawa S Hayashi S Tsukamoto Y Goko H Kawasaki H Wada T Shimizu K Yasuda N Kamatani N Takasugi K Identification of the gene loci that predispose to rheumatoid arthritis Int Immunol 1998 10 1891 1895 9885910 10.1093/intimm/10.12.1891
Cornelis F Faure S Martinez M Prud'homme JF Fritz P Dib C Alves H Barrera P de Vries N Balsa A New susceptibility locus for rheumatoid arthritis suggested by a genome-wide linkage study Proc Natl Acad Sci USA 1998 95 10746 10750 9724775 10.1073/pnas.95.18.10746
Jawaheer D Seldin MF Amos CI Chen WV Shigeta R Etzel C Damle A Xiao X Chen D Lum RF Screening the genome for rheumatoid arthritis susceptibility genes: a replication study and combined analysis of 512 multicase families Arthritis Rheum 2003 48 906 916 12687532 10.1002/art.10989
John S Shephard N Liu G Zeggini E Cao M Chen W Vasavda N Mills T Barton A Hinks A Whole-genome scan, in a complex disease, using 11,245 single-nucleotide polymorphisms: comparison with microsatellites Am J Hum Genet 2004 75 54 64 15154113 10.1086/422195
MacKay K Eyre S Myerscough A Milicic A Barton A Laval S Barrett J Lee D White S John S Whole-genome linkage analysis of rheumatoid arthritis susceptibility loci in 252 affected sibling pairs in the United Kingdom Arthritis Rheum 2002 46 632 639 11920398 10.1002/art.10147
Jawaheer D Seldin MF Amos CI Chen WV Shigeta R Monteiro J Kern M Criswell LA Albani S Nelson JL A genomewide screen in multiplex rheumatoid arthritis families suggests genetic overlap with other autoimmune diseases Am J Hum Genet 2001 68 927 936 11254450 10.1086/319518
Eyre S Barton A Shephard N Hinks A Brintnell W MacKay K Silman A Ollier W Wordsworth P John S Worthington J Investigation of susceptibility loci identified in the UK rheumatoid arthritis whole-genome scan in a further series of 217 UK affected sibling pairs Arthritis Rheum 2004 50 729 735 15022312 10.1002/art.20039
Suzuki A Yamada R Chang X Tokuhiro S Sawada T Suzuki M Nagasaki M Nakayama-Hamada M Kawaida R Ono M Functional haplotypes of PADI4, encoding citrullinating enzyme peptidylarginine deiminase 4, are associated with rheumatoid arthritis Nat Genet 2003 34 395 402 12833157 10.1038/ng1206
Tokuhiro S Yamada R Chang X Suzuki A Kochi Y Sawada T Suzuki M Nagasaki M Ohtsuki M Ono M An intronic SNP in a RUNX1 binding site of SLC22A4, encoding an organic cation transporter, is associated with rheumatoid arthritis Nat Genet 2003 35 341 348 14608356 10.1038/ng1267
Seldin MF Amos CI Ward R Gregersen PK The genetics revolution and the assault on rheumatoid arthritis Arthritis Rheum 1999 42 1071 1079 10366098 10.1002/1529-0131(199906)42:6<1071::AID-ANR1>3.0.CO;2-8
Begovich AB Carlton VE Honigberg LA Schrodi SJ Chokkalingam AP Alexander HC Ardlie KG Huang Q Smith AM Spoerke JM A missense single-nucleotide polymorphism in a gene encoding a protein tyrosine phosphatase (PTPN22) is associated with rheumatoid arthritis Am J Hum Genet 2004 75 330 337 15208781 10.1086/422827
Kyogoku C Langefeld CD Ortmann WA Lee A Selby S Carlton VE Chang M Ramos P Baechler EC Batliwalla FM Genetic association of the R620W polymorphism of protein tyrosine phosphatase PTPN22 with human SLE Am J Hum Genet 2004 75 504 507 15273934 10.1086/423790
Bottini N Musumeci L Alonso A Rahmouni S Nika K Rostamkhani M MacMurray J Meloni GF Lucarelli P Pellecchia M A functional variant of lymphoid tyrosine phosphatase is associated with type I diabetes Nat Genet 2004 36 337 338 15004560 10.1038/ng1323
Smyth D Cooper JD Collins JE Heward JM Franklyn JA Howson JM Vella A Nutland S Rance HE Maier L Replication of an association between the lymphoid tyrosine phosphatase locus (LYP/PTPN22) with type 1 diabetes, and evidence for its role as a general autoimmunity locus Diabetes 2004 53 3020 3023 15504986
Velaga MR Wilson V Jennings CE Owen CJ Herington S Donaldson PT Ball SG James RA Quinton R Perros P Pearce SH The codon 620 tryptophan allele of the lymphoid tyrosine phosphatase (LYP) gene is a major determinant of Graves' disease J Clin Endocrinol Metab 2004 89 5862 5865 15531553 10.1210/jc.2004-1108
Onengut-Gumuscu S Ewens KG Spielman RS Concannon P A functional polymorphism (1858C/T) in the PTPN22 gene is linked and associated with type I diabetes in multiplex families Genes Immun 2004 5 678 680 15526003 10.1038/sj.gene.6364138
Orozco G Sanchez E Gonzalez-Gay MA Lopez-Nevot MA Torres B Caliz R Ortego-Centeno N Jimenez-Alonso J Pascual-Salcedo D Balsa A Association of a functional single-nucleotide polymorphism of PTPN22, encoding lymphoid protein phosphatase, with rheumatoid arthritis and systemic lupus erythematosus Arthritis Rheum 2005 52 219 224 15641066 10.1002/art.20771
Steer S Lad B Grumley JA Kingsley GH Fisher SA Association of R602W in a protein tyrosine phosphatase gene with a high risk of rheumatoid arthritis in a British population: evidence for an early onset/disease severity effect Arthritis Rheum 2005 52 358 360 15641088 10.1002/art.20737
Cloutier JF Veillette A Association of inhibitory tyrosine protein kinase p50csk with protein tyrosine phosphatase PEP in T cells and other hemopoietic cells EMBO J 1996 15 4909 4918 8890164
Cloutier JF Veillette A Cooperative inhibition of T-cell antigen receptor signaling by a complex between a kinase and a phosphatase J Exp Med 1999 189 111 121 9874568 10.1084/jem.189.1.111
Gregersen PK Lee A Begovich A Massarotti E Weisman M Kent J Bombardier C A functional polymorphism of PTPN22 associates with seropositive but not seronegative rheumatoid arthritis in a cohort of patients with new onset RA (SONORA) [abstract] Arthritis Rheum 2004 S118
Lee A Li W Liew A Bombardier C Weisman M Massarotti E Kent J Wolfe F Begovich A Gregersen PK The PTPN22 R620W polymorphism associates with RF positive rheumatoid arthritis in a dose-dependent manner but not with HLA-SE status Genes Immun 2005 6 129 133 15674368 10.1038/sj.gene.6364159
Siminovitch KA PTPN22 and autoimmune disease Nat Genet 2004 36 1248 1249 15565104 10.1038/ng1204-1248
Spielman RS Ewens WJ The TDT and other family-based tests for linkage disequilibrium and association Am J Hum Genet 1996 59 983 989 8900224
Risch N Merikangas K The future of genetic studies of complex human diseases Science 1996 273 1516 1517 8801636
Qu H Tessier MC Hudson TJ Polychronakos C Confirmation of the association of the R620W polymorphism in the protein tyrosine phosphatase PTPN22 with type 1 diabetes in a family based study J Med Genet 2005 42 266 270 15744042 10.1136/jmg.2004.026971
Zhernakova A Eerligh P Wijmenga C Barrera P Roep BO Koeleman BP Differential association of the PTPN22 coding variant with autoimmune diseases in a Dutch population Genes Immun 2005 10.1038/sj.gene.6364220
Arnett FC Edworthy SM Bloch DA McShane DJ Fries JF Cooper NS Healey LA Kaplan SR Liang MH Luthra HS The American Rheumatism Association 1987 revised criteria for the classification of rheumatoid arthritis Arthritis Rheum 1988 31 315 324 3358796
Dib C Faure S Fizames C Samson D Drouot N Vignal A Millasseau P Marc S Hazan J Seboun E A comprehensive genetic map of the human genome based on 5,264 microsatellites Nature 1996 380 152 154 8600387 10.1038/380152a0
Sambrook P Fritsch E Maniatis T Cold Spring Harbor Molecular Cloning A Laboratory Manual 1989 1 2 New York: Cold Spring Harbor Laboratory Press
Dieude P Petit E Cailleau-Moindrault S Osorio J Pierlot C Martinez M Faure S Alibert O Lasbleiz S De Toma C Association between tumor necrosis factor receptor II and familial, but not sporadic, rheumatoid arthritis: evidence for genetic heterogeneity Arthritis Rheum 2002 46 2039 2044 12209506 10.1002/art.10101
Spielman RS McGinnis RE Ewens WJ Transmission test for linkage disequilibrium: the insulin gene region and insulin-dependent diabetes mellitus (IDDM) Am J Hum Genet 1993 52 506 516 8447318
Lathrop GM Estimating genotype relative risks Tissue Antigens 1983 22 160 166 6623452
Gudbjartsson DF Jonasson K Frigge ML Kong A Allegro, a new computer program for multipoint linkage analysis Nat Genet 2000 25 12 13 10802644 10.1038/75514
Julier C Hyer RN Davies J Merlin F Soularue P Briant L Cathelineau G Deschamps I Rotter JI Froguel P Insulin-IGF2 region on chromosome 11p encodes a gene implicated in HLA-DR4-dependent diabetes susceptibility Nature 1991 354 155 159 1944595 10.1038/354155a0
Li C Scott LJ Boehnke M Assessing whether an allele can account in part for a linkage signal: the Genotype-IBD Sharing Test (GIST) Am J Hum Genet 2004 74 418 431 14872409 10.1086/381712
Ittah M Gottenberg JE Proust A Hachulla E Puechal X Loiseau P Mariette X Miceli-Richard C No evidence for association between 1858 C/T single-nucleotide polymorphism of PTPN22 gene and primary Sjogren's syndrome Genes Immun 2005 6 457 458 15933742 10.1038/sj.gene.6364229
Criswell LA Pfeiffer KA Lum RF Gonzales B Novitzke J Kern M Moser KL Begovich AB Carlton VE Li W Analysis of families in the Multiple Autoimmune Disease Genetics Consortium (MADGC) collection: the PTPN22 620W allele associates with multiple autoimmune phenotypes Am J Hum Genet 2005 76 561 571 15719322 10.1086/429096
Hinks A Barton A John S Bruce I Hawkins C Griffiths CE Donn R Thomson W Silman A Worthington J Association between the PTPN22 gene and rheumatoid arthritis and juvenile idiopathic arthritis in a UK population: further support that PTPN22 is an autoimmunity gene Arthritis Rheum 2005 52 1694 1699 15934099 10.1002/art.21049
Skorka A Bednarczuk T Bar-Andziak E Nauman J Ploski R Lymphoid tyrosine phosphatase (PTPN22/LYP) variant and Graves' disease in a Polish population: association and gene dose-dependent correlation with age of onset Clin Endocrinol (Oxf) 2005 62 679 682 15943829 10.1111/j.1365-2265.2005.02279.x
|
16277672
|
PMC1297567
|
CC BY
|
2021-01-04 16:02:39
|
no
|
Arthritis Res Ther. 2005 Aug 25; 7(6):R1200-R1207
|
utf-8
|
Arthritis Res Ther
| 2,005 |
10.1186/ar1812
|
oa_comm
|
==== Front
Arthritis Res TherArthritis Research & Therapy1478-63541478-6362BioMed Central London ar18131627767410.1186/ar1813Research ArticleAbsence of autoantibodies against correctly folded recombinant fibrillin-1 protein in systemic sclerosis patients Brinckmann Jürgen [email protected] Nico 2El-Hallous Ehab 3Krieg Thomas 2Sakai Lynn Y 4Krengel Sven 1Reinhardt Dieter P 51 Department of Dermatology, University of Lübeck, Lübeck, Germany2 Department of Dermatology, University of Cologne, Cologne, Germany3 Department of Medical Molecular Biology, University of Lübeck, Lübeck, Germany4 Department of Biochemistry and Molecular Biology and Shriners Hospital for Children, Oregon Health and Science University, Portland, OR, USA5 Department of Anatomy and Cell Biology and Faculty of Dentistry, McGill University, Montreal, Canada2005 6 9 2005 7 6 R1221 R1226 1 6 2005 23 6 2005 11 7 2005 8 8 2005 Copyright © 2005 Brinckmann et al.; licensee BioMed Central Ltd.This is an Open Access article distributed under the terms of the Creative Commons Attribution License (), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
Autoantibodies against short recombinant fragments of fibrillin-1 produced in bacterial expression systems have been found in tight-skin mouse, systemic sclerosis, mixed connective tissue disease, and primary pulmonary hypertension syndrome. In patients with scleroderma, the frequency of anti-fibrillin-1 antibodies was 42% in Caucasians. Until now it has been unclear whether this immune response has a primary function in disease pathogenesis or is a secondary phenomenon. In the present study we analyzed the frequency of autoantibodies against two overlapping recombinant polypeptides spanning the N-terminal and C-terminal halves of human fibrillin-1, which were produced in human embryonic kidney (HEK-293) cells. Correct three-dimensional structures of the recombinant fibrillin-1 polypeptides were shown by electron microscopy and immunoreactivity with antibodies. Screening of fibrillin-1 antibodies was performed in 41 sera from systemic sclerosis patients and in 44 healthy controls with a Caucasian background. Microtiter plates were coated with the recombinant polypeptides of fibrillin-1 and incubated with 1:100 diluted sera. Positive binding was defined as being more than 2 SD above the mean of the control group. ELISAs showed that none of the sera of patients with systemic sclerosis contained autoantibodies against the N-terminal or C-terminal recombinant fibrillin-1 polypeptide. The data show the absence of autoantibodies against recombinant fibrillin-1 protein in Caucasian systemic sclerosis patients. Because the correct three-dimensional folding of the recombinant proteins has been substantiated by several independent methods, we conclude that autoantibodies against correctly folded fibrillin are not a primary phenomenon in the pathogenesis of systemic sclerosis.
==== Body
Introduction
Systemic sclerosis (SSc) is a connective tissue disease characterized by an excess deposition of collagen in skin and/or internal organs leading to malfunction and organ failure. The extent and progression of the fibrotic process presumably caused by the imbalance between extracellular matrix synthesis and degradation largely determines the prognosis of the disease. One hallmark of the disease is the presence of circulating autoantibodies against non-organ-specific nuclear and nucleolar antigens, which can be detected in at least 95% of patients. They include anti-centromere, anti-topoisomerase I and anti-RNA polymerase antibodies and are associated with distinct disease subtypes [1].
Heterozygous tight-skin mice (Tsk/+) are characterized by a phenotype of skin thickening and visceral fibrosis due to an increased deposition of extracellular matrix proteins in skin and organs. Furthermore, Tsk/+ mice develop lung emphysema and cardiac hypertrophy and have therefore been adopted as a potential genetic model of human SSc, cardiac hypertrophy and hereditary emphysema [2]. In a similar manner to human SSc, Tsk/+ mice produce autoantibodies against SSc-specific antigens such as topoisomerase I and RNA polymerase [3].
A duplication in the mouse fibrillin-1 gene was described for the Tsk/+ mouse, which is associated with premature death in utero for homozygous Tsk/Tsk animals [4]. Fibrillin-1 is one of the major structural components of microfibrils, which are extracellular supramolecular aggregates found in many elastic and non-elastic tissues (reviewed in [5]). Microfibrils are thought to be important in the assembly and organization of the elastic fibers by mediating tropoelastin deposition [6]. Fibrillin-1 and other members of the fibrillin family are repetitively aligned within microfibrils and constitute their structural backbone [7,8]. Murai and colleagues found that Tsk/+ mice spontaneously produce autoantibodies against a small recombinant protein spanning the proline-rich region of human fibrillin-1 [9]. This recombinant fragment comprises about 2% of the total fibrillin-1 molecule. Recently, the presence of autoantibodies against the same recombinant fibrillin-1 fragment has also been shown for sera from patients with SSc, localized scleroderma, mixed connective tissue disease and primary pulmonary hypertension syndrome [10-12]. Frequencies of autoantibodies showed remarkable differences between the ethnic groups studied. Choctaw American Indians and Japanese patients with SSc exhibited the highest frequency, with 81% and 78% respectively, whereas Caucasians with SSc were positive to a smaller extent with 34% [10].
In the present study we analyzed the autoantibody titer in Caucasian SSc patients against two overlapping recombinant fragments spanning the entire human fibrillin-1. One fragment constitutes the amino-terminal half of fibrillin-1 (amino acid residues 19 to 1,527) and the other fragment its carboxy-terminal half (residues 1,487 to 2,725). Before the analysis of antibody titers by ELISA, the proper folding of both recombinant proteins was shown by electron microscopy after rotary shadowing and binding of monoclonal and polyclonal antibodies by dot-blotting with or without previous reduction of the recombinant proteins.
Materials and methods
Patients and tissue specimens
Sera from Caucasian patients with SSc (n = 41; 29 female, 12 male; mean age 58.2 ± 14.3 years) and from healthy Caucasian controls (n = 44; 31 female, 13 male; mean age 46.9 ± 19.8 years) were studied. Patients with SSc were diagnosed in accordance with the American College of Rheumatology preliminary criteria for the classification of SSc [13]. Limited systemic sclerosis was present in 25 patients, and diffuse systemic sclerosis in 16. The range of disease duration was between 6 months and 27 years. The antibody profile showed positive titers of anti-nuclear antibodies for all patients. Of these, 16 had SCL-70, 13 anti-centromere, 1 RNA polymerase and 11 undifferentiated antibodies. Antibody testing consisted of the determination of the fluorescence pattern and titer on HEP2 cells (Viramed, Germany) as well as subsequent testing by a commercial ELISA for U1-RNP, Sm, Ro-SSA, La-SSB, Scl-70 and centromere reactivity (Orgentec, Germany). All samples were obtained after obtaining written consent from the donors under protocols approved by the local ethical committee.
Expression and production of recombinant fibrillin-1 polypeptides
The expression plasmids to express the N-terminal half (pDNSP-rF16) and the C-terminal half (pcDNA-rF6H) of human fibrillin-1 have previously been described in detail [14]. On the basis of SDS-PAGE and electron microscopy after rotary shadowing (see below), the purity of the recombinant fragments was more than 90%. Stable clones with these expression plasmids were obtained with human embryonic kidney (HEK-293) cells as described in detail [15]. The expression of pDNSP-rF16 in eukaryotic cells produces a secreted polypeptide (rF16) with the sequence Ala-Pro-Leu-Ala-Ser19-Val1,527-(His)6. The expression of pcDNA-rF6H in eukaryotic cells produces a secreted polypeptide (rF6H) with the sequence Ala-Pro-Leu-Ala-Asp1,487-Lys2,725-(His)6. Production and purification of rF16 and rF6H were performed as in the procedures described elsewhere [16].
Electron microscopy after rotary shadowing
The purified proteins were adjusted to a concentration of 0.25 mg/ml and dialyzed against 100 mM NH4HCO3. The samples were diluted with 0.05% (v/v) acetic acid to a final protein concentration of 60 μg/ml and mixed with glycerol to a final concentration of 50% (v/v) glycerol. Then 80 μl of the samples was sprayed onto freshly cleaved mica from a distance of 25 cm and dried under high vacuum (about 9 nbar) for about 2 to 3 hours in an Edwards Auto 306 vacuum coater. Rotary shadowing was performed by platinum evaporation for 15 s at 50 mA and 2.5 kV at an angle of 5° and a distance of 12 cm. The samples were rotated at 120 r.p.m., followed by coating with coal for stabilization for 2 s at 100 mA and 2.5 kV at an angle of 90°. The replicas were floated onto a very clean surface of distilled water and then supported with 400-mesh copper grids. Replicas were examined at 100 kV in a transmission electron microscope (Zeiss TEM 109).
Cell culture
Human dermal fibroblasts were derived from explant cultures of dissected tissues obtained from surgical samples after informed consent had been obtained. The cells were cultured in DMEM supplemented with 10% fetal bovine serum and penicillin/streptomycin (Invitrogen). Cells (106) were plated in a 60 mm dish and grown for 72 hours. The cell layers were washed with phosphate-buffered saline and then incubated for 24 hours in 3 ml of DMEM without serum. The conditioned medium was harvested and treated with 1 mM phenylmethylsulfonyl fluoride.
Dot-blot assay
Either 2 μg of purified recombinant proteins rF16 and rF6H or 1 ml of conditioned medium were transferred to nitrocellulose membranes using a dot-blot apparatus (Bio-Rad) with or without previous reduction of the proteins with 0.05 M dithiothreitol. After staining with Ponceau S, non-specific binding sites on the nitrocellulose membrane were blocked for 1 hour with Tris-buffered saline (TBS) containing 5% (w/v) non-fat milk. Nitrocellulose membranes were probed with a polyclonal antiserum against rF6H (diluted 1:500 [17]) and with monoclonal antibodies directed against rF6H (mAb 69, about 4 μg/ml) and rF16 (mAb 201 and mAb 26, both about 4 μg/ml [18]) followed by peroxidase-conjugated anti-rabbit or anti-mouse secondary antibody (diluted 1:800; Bio-Rad). Bound antibodies were revealed in accordance with the manufacturer's instructions by using the horseradish peroxidase developer 4-chloronaphthol (Bio-Rad).
ELISA assay
Microtiter plates were coated with 100 μl of 10 μg/ml purified recombinant human fibrillin-1 fragments rF16 and rF6H or BSA overnight at 4°C. After being washed three times with TBS containing 0.05% Tween 20 (TBS/Tween), the plates were blocked for 1 hour with 200 μl of 5% non-fat milk powder in TBS at room temperature (20°C). After being washed with TBS, the plates were incubated for 2 hours with 100 μl of test sera diluted 1:100 with TBS containing 5% non-fat milk powder at room temperature. After being washed three times with TBS/Tween, the plates were incubated for 1.5 hours with 100 μl of the horseradish peroxidase-conjugated secondary antibody (diluted 1:800) at room temperature (goat anti-rabbit for positive control sera, and goat anti-human for human sera; Sigma, Germany). After three washings with TBS/Tween, color development was achieved with 100 μl of 1 mg/ml 5-aminosalicylic acid in 0.02 M phosphate buffer (pH 6.8) and 1.5 μl/ml H2O2. Color development was stopped after 1 hour by the addition of 100 μl of 2 M NaOH. Absorbance was measured at 492 nm with an ELISA reader (Anthos, Austria). All experiments were run in parallel triplicates; the ELISA test was performed twice. The background binding of serum antibodies to BSA-coated wells was subtracted from the binding of serum to the respective rF16-coated and rF6H-coated wells after subtraction of the respective background of rF16-coated, rF6H-coated and BSA-coated blanks. Positive binding was defined as more than 2 SD above the mean of the control sera. The coefficient of variation was 7.4% (n = 10).
Results
Ultrastructural analysis of recombinant fibrillin-1 polypeptides
To analyze the molecular shape of the recombinant polypeptides, they were revealed by electron microscopy after rotary shadowing (Fig. 1). These results showed thread-like extended molecules for the recombinant polypeptides rF16 and rF6H representing the N-terminal and C-terminal halves of human fibrillin-1. At the termini of rF16 and rF6H the molecules occasionally adopted a curved shape.
The analysis of molecular dimensions revealed that the length of rF16 (73.1 ± 5.7 nm, n = 75) and rF6H (64.2 ± 5.9 nm, n = 56) corresponded well to the lengths for very similar constructs described previously [16] as well to the respective parts in full-length fibrillin-1 [19]. The extended shape of the recombinant proteins is a very good indicator of correct folding, because the molecular shape is determined by numerous intramolecular disulfide bridges stabilizing this extended structure [20,21].
Immunoreactive analysis of recombinant fibrillin-1 polypeptides
To analyze the immunoreactive properties of native fibrillin-1 synthesized by human dermal fibroblasts and the recombinant polypeptides rF16 and rF6H, dot-blotting under reducing and non-reducing conditions was performed (Fig. 2) Native fibrillin-1 reacts with monoclonal antibodies mAb 26 or mAb 201 or with polyclonal antibody anti-rF6H only under non-reducing conditions (not under reducing conditions). These data show that the antibodies primarily recognize epitopes in the correctly folded fibrillin-1 molecule but not in the denatured fibrillin-1 molecule. When the recombinant fibrillin-1 polypeptides rF16 and rF6H were tested in this assay, they showed much more reactivity in the non-reduced conformation than in the reduced conformation, showing that the corresponding epitopes are present in the same correct conformation as in native fibrillin-1. These data substantiate that the recombinant polypeptides are correctly folded.
ELISA analysis of sera from patients and controls by using rF16 and rF6H
A cutoff value was established for each ELISA as a value of 2 SD above the mean of 44 control sera. For rF16 the cutoff ELISA score was 0.072 and for rF6H it was 0.1. The analysis of 41 sera from Caucasian patients with systemic sclerosis showed that none of the sera exceeded the cutoff value for the N-terminal half of fibrillin-1. Furthermore, the ELISA score of all sera tested for the presence of antibodies against the C-terminal half of fibrillin-1 was in the normal range of the controls (Fig. 3).
Discussion
Mutations in the gene encoding fibrillin-1 have been documented for Marfan syndrome and some related disorders in humans, and for Tsk in animals [22,4]. The Tsk mutation in the fibrillin-1 gene, a 30-kilobase gene duplication of exons 17 to 40 containing a long centrally located stretch of calcium-binding epidermal growth factor-like domains, is accompanied by premature death in utero in homozygous mice, whereas mice heterozygous for the duplication are viable and show the tight-skin phenotype. The mutation results in a larger protein (418 kDa, as compared with 350 kDa in normal animals) which after incorporation along with wild-type fibrillin-1 seems to render all microfibrils more susceptible to proteolysis [23]. In a similar manner to SSc in humans, Tsk mice develop autoimmunity with antibodies against topoisomerase I and RNA polymerase. Recently, autoantibodies against a small 30 kDa human recombinant fibrillin-1 polypeptide covering the proline-rich region (residues 395 to 446) have been detected in 41% of Tsk mice [9].
Autoantibodies against the same recombinant fibrillin-1 polypeptide were also found in humans affected by SSc or primary pulmonary hypertension syndrome [10,12]. Especially in SSc, the frequency of anti-fibrillin-1 antibodies and the recognized epitopes differ according to the ethnic background of patients, as shown in a subsequent study [24]. In that study, reactivity against recombinant polypeptides covering the N-terminal end (residues 15 to 193), the proline-rich region (residues 367 to 425), and a stretch of calcium-binding epidermal growth factor-like domains (residues 1,326 to 1,549) was tested. Taking the different epitopes tested in that study together, Choctaw Native Americans, Japanese patients and African Americans revealed the highest levels with 100% and 80%, respectively. In the same study, sera from Caucasian SSc patients showed the presence of anti-fibrillin-1 antibodies in 42% of patients. Whether the occurrence of these autoantibodies has a primary role in the pathogenesis of SSc or is a secondary phenomenon is open to discussion.
In our study of 41 Caucasian patients with SSc, none of the sera showed positive reactivity against the recombinant polypeptide spanning either the N-terminal half or the C-terminal half of fibrillin-1. Structural studies by rotary shadowing and evaluation of molecular lengths showed that the recombinant fibrillin-1 polypeptides used resemble native molecules. They adopt the correct dimensions and extended conformations similar to regions observed in whole molecules of native fibrillin-1 purified from cell culture medium [19]. Various monoclonal and polyclonal antibodies recognize native fibrillin-1 only in a non-reduced (correctly folded) conformation but not in the reduced (misfolded) conformation because numerous intramolecular disulfide bonds stabilize the native conformation of fibrillins [20,21]. Similar binding properties of these monoclonal and polyclonal antibodies to the recombinant polypeptides rF16 and rF6H strongly support the notion that these polypeptides are folded correctly. Our data clearly show that SSc in Caucasians is not characterized by the presence of autoantibodies against properly folded fibrillin-1. This observation indicates that the presence of autoantibodies against fibrillin-1 does not have a primary role in the pathogenesis of the disease.
The recombinant fibrillin-1 antigens used in other studies showing a positive binding of antibodies obtained from SSc patients were relatively small (59, 179 and 224 residues) and were produced in bacterial expression systems [10,24]. No structural or functional characterization for these recombinant polypeptides is available to determine whether they adopt native or misfolded conformations. It is possible that the anti-fibrillin-1 autoantibodies detected with such recombinant polypeptides recognize cryptic or misfolded antigenic epitopes for example, which may become available after proteolytic fragmentation of fibrillin-1 in SSc or may be antibodies against cross-reacting antigens. In this light, one can speculate that these autoantibodies are a secondary phenomenon in SSc. This interpretation is further substantiated by a metabolic analysis of fibrillin-1 synthesized by SSc fibroblasts in cell culture, which revealed decreased amounts of abnormal microfibrils [25]. Furthermore, in the same study in vitro, data indicated that the amount of fibrillin-1 in the extracellular matrix produced by SSc cells diminished faster than in the matrix of control cells, arguing for a higher susceptibility to proteolytic degradation.
Conclusion
Our data clearly show that sera from 41 Caucasian SSc patients contained no autoantibodies against properly folded recombinant human fibrillin-1. These data therefore provide evidence that autoimmunity against fibrillin-1 is a secondary phenomenon in the pathogenesis of SSc in Caucasians.
Abbreviations
BSA = bovine serum albumin; DMEM = Dulbecco's modified Eagle's medium; ELISA = enzyme-linked immunosorbent assay; kDa = kilodaltons; mAb = monoclonal antibody; SSc = systemic sclerosis; TBS = Tris-buffered saline; Tsk = tight-skin mouse.
Competing interests
The author(s) declare that they have no competing interests.
Authors' contributions
Expression and production of recombinant polypeptides were performed by EE and DPR. Electron microscopy was performed by EE and DPR. ELISA analysis was performed JB, DPR, SK and NH. Immunoreactive analysis was performed by JB, LYS and DPR. Study design and coordination were performed by JB, NH and DPR. Editing of the manuscript was performed by JB, NH, DPR, TK and LYS. All authors read and approved the final manuscript.
Acknowledgements
We are grateful to Martina Alexander for excellent technical assistance. The work was supported by grants by the Deutsche Forschungsgemeinschaft (SFB367-A1, Br 1146/3-3), the Bundesminsterium für Bildung und Forschung (BMBF, German Network for Systemic Scleroderma), the Köln Fortune Program and the Canadian Institutes of Health Research (MOP-68836).
Figures and Tables
Figure 1 Recombinant amino-terminal (rF16)(a) and carboxy-terminal (rF6H)(b) halves of human fibrillin-1 were analyzed by electron microscopy after rotary shadowing. Representative images and histrograms of the measured lenghs of the recombinant fragments are shown. Note that both fragments showed thread-like extended molecules. The measurements are plotted as number of measurements, in 5 nm windows. The average length of rF16 was 73.1 ± 5.7 nm (mean ± SD; n = 75) and the average length of rF6H was 64.2 ± 5.9 nm (mean ± SD; n = 56).
Figure 2 Immunoreactive analysis of fibrillin-1 antibodies against recombinant fibrillin-1 polypeptides and against native fibrillin-1. Purified recombinant amino-terminal (rF16) and carboxy-terminal (rF6H) halves of human fibrillin-1 (2 μg of each) or 1 ml of conditioned medium (containing less than about 0.2 μg of fibrillin-1) produced by human dermal fibroblasts were transferred to nitrocellulose membranes with (upper panel) or without (lower panel) previous reduction by dithiothreitol. Nitrocellulose membranes were probed with a polyclonal antibody against rF6H (anti-rF6H) or with monoclonal antibodies (mAbs) 26, 201, and 69 directed against rF16 and rF6H. The dot-blots show that the binding of all antibodies depends markedly on the presence of disulfide bonds, which are crucial for the proper folding of epitopes in both native fibrillin-1 and the recombinant fragments.
Figure 3 Analysis of immunoreactivity of sera from systemic sclerosis patients and healthy controls of Caucasian origin. ELISA assays with the recombinant amino-terminal (rF16 (a)) and carboxy-terminal (rF6H (b)) halves of human fibrillin-1 are shown. Microtiter plates were coated with purified rF16 and rF6H or bovine serum albumin. The plates were incubated with test sera diluted 1:100 for 2 hours at room temperature. After incubation with horseradish peroxidase-conjugated secondary antibody and color development, the absorbance was determined by an ELISA reader. Positive binding was defined as more than 2 SD above the mean (dashed line) of the control sera. If the blank value exceeded the sample value the absorbance was set to zero in the figure. None of the sera showed a positive reactivity to rF16 or rF6H.
==== Refs
Harris ML Rosen A Autoimmunity in scleroderma: the origin, pathogenetic role, and clinical significance of autoantibodies Curr Opin Rheumatol 2003 15 778 784 14569210 10.1097/00002281-200311000-00016
Jimenez SA Christner PJ Murine animal models of systemic sclerosis Curr Opin Rheumatol 2002 14 671 680 12410090 10.1097/00002281-200211000-00008
Bona C Rothfield N Autoantibodies in scleroderma and tightskin mice Curr Opin Immunol 1994 6 931 937 7710717 10.1016/0952-7915(94)90016-7
Siracusa LD McGrath R Ma Q Moskow JJ Manne J Christner PJ Buchberg AM Jimenez SA A tandem duplication within the fibrillin 1 gene is associated with the mouse tight skin mutation Genome Res 1996 6 300 313 8723723
Ramirez F Pereira L The fibrillins Int J Biochem Cell Biol 1999 31 255 259 10216958 10.1016/S1357-2725(98)00109-5
Mecham RP Broekelmann T Davis EC Gibson MA Brown-Augsburger P Elastic fibre assembly: macromolecular interactions Ciba Found Symp 1995 192 172 181 8575256
Lin G Tiedemann K Vollbrandt T Peters H Batge B Brinckmann J Reinhardt DP Homo- and heterotypic fibrillin-1 and -2 interactions constitute the basis for the assembly of microfibrils J Biol Chem 2002 277 50795 50804 12399449 10.1074/jbc.M210611200
Charbonneau NL Dzamba BJ Ono RN Keene DR Corson GM Reinhardt DP Sakai LY Fibrillins can co-assemble in fibrils, but fibrillin fibril composition displays cell-specific differences J Biol Chem 2003 278 2740 2749 12429739 10.1074/jbc.M209201200
Murai C Saito S Kasturi KN Bona CA Spontaneous occurrence of anti-fibrillin-1 autoantibodies in tight-skin mice Autoimmunity 1998 28 151 155 9867127
Tan FK Arnett FC Antohi S Saito S Mirarchi A Spiera H Sasaki T Shoichi O Takeuchi K Pandey JP Autoantibodies to the extracellular matrix microfibrillar protein, fibrillin-1, in patients with scleroderma and other connective tissue diseases J Immunol 1999 163 1066 1072 10395706
Arnett FC Tan FK Uziel Y Laxer RM Krafchik BR Antohi S Bona C Autoantibodies to the extracellular matrix microfibrillar protein, fibrillin 1, in patients with localized scleroderma Arthritis Rheum 1999 42 2656 2659 10616014 10.1002/1529-0131(199912)42:12<2656::AID-ANR22>3.0.CO;2-N
Morse JH Antohi S Kasturi K Saito S Fotino M Humbert M Simonneau G Basst RJ Bona CA Fine specificity of anti-fibrillin-1 autoantibodies in primary pulmonary hypertension syndrome Scand J Immunol 2000 51 607 611 10849372 10.1046/j.1365-3083.2000.00733.x
Subcommittee for scleroderma criteria of the American Rheumatism Association Diagnostic and Therapeutic Criteria Committee Preliminary criteria for the classification of systemic sclerosis (scleroderma) Arthritis Rheum 1980 23 581 590 7378088
Jensen SA Reinhardt DP Gibson MA Weiss AS Protein interaction studies of MAGP-1 with tropoelastin and fibrillin-1 J Biol Chem 2001 276 39661 39666 11481325 10.1074/jbc.M104533200
Fox JW Mayer U Nischt R Aumailley M Reinhardt D Wiedemann H Mann K Timpl R Krieg T Engel J Recombinant nidogen consists of three globular domains and mediates binding of laminin to collagen type IV EMBO J 1991 10 3137 3146 1717261
Reinhardt DP Keene DR Corson GM Poschl E Bachinger HP Gambee JE Sakai LY Fibrillin-1: organization in microfibrils and structural properties J Mol Biol 1996 258 104 116 8613981 10.1006/jmbi.1996.0237
Tiedemann K Batge B Muller PK Reinhardt DP Interactions of fibrillin-1 with heparin/heparan sulfate, implications for microfibrillar assembly J Biol Chem 2001 276 36035 36042 11461921 10.1074/jbc.M104985200
Sakai LY Keene DR Engvall E Fibrillin, a new 350-kD glycoprotein, is a component of extracellular microfibrils J Cell Biol 1986 103 2499 2509 3536967 10.1083/jcb.103.6.2499
Sakai LY Keene DR Glanville RW Bachinger HP Purification and partial characterization of fibrillin, a cysteine-rich structural component of connective tissue microfibrils J Biol Chem 1991 266 14763 14770 1860873
Downing AK Knott V Werner JM Cardy CM Campbell ID Handford PA Solution structure of a pair of calcium-binding epidermal growth factor-like domains: implications for the Marfan syndrome and other genetic disorders Cell 1996 85 597 605 8653794 10.1016/S0092-8674(00)81259-3
Yuan X Downing AK Knott V Handford PA Solution structure of the transforming growth factor beta-binding protein-like module, a domain associated with matrix fibrils EMBO J 1997 16 6659 6666 9362480 10.1093/emboj/16.22.6659
Collod-Beroud G Le Bourdelles S Ades L Ala-Kokko L Booms P Boxer M Child A Comeglio P De Paepe A Hyland JC Update of the UMD-FBN1 mutation database and creation of an FBN1 polymorphism database Hum Mutat 2003 22 199 208 12938084 10.1002/humu.10249
Gayraud B Keene DR Sakai LY Ramirez F New insights into the assembly of extracellular microfibrils from the analysis of the fibrillin 1 mutation in the tight skin mouse J Cell Biol 2000 150 667 680 10931876 10.1083/jcb.150.3.667
Tan FK Arnett FC Reveille JD Ahn C Antohi S Sasaki T Nishioka K Bona CA Autoantibodies to fibrillin 1 in systemic sclerosis: ethnic differences in antigen recognition and lack of correlation with specific clinical features or HLA alleles Arthritis Rheum 2000 43 2464 2471 11083269 10.1002/1529-0131(200011)43:11<2464::AID-ANR13>3.0.CO;2-F
Wallis DD Tan FK Kielty CM Kimball MD Arnett FC Milewicz DM Abnormalities in fibrillin 1-containing microfibrils in dermal fibroblast cultures from patients with systemic sclerosis (scleroderma) Arthritis Rheum 2001 44 1855 1864 11508439 10.1002/1529-0131(200108)44:8<1855::AID-ART324>3.0.CO;2-Q
|
16277674
|
PMC1297568
|
CC BY
|
2021-01-04 16:02:40
|
no
|
Arthritis Res Ther. 2005 Sep 6; 7(6):R1221-R1226
|
utf-8
|
Arthritis Res Ther
| 2,005 |
10.1186/ar1813
|
oa_comm
|
==== Front
Arthritis Res TherArthritis Research & Therapy1478-63541478-6362BioMed Central London ar18151627767710.1186/ar1815Research ArticleRaloxifene reduces urokinase-type plasminogen activator-dependent proliferation of synoviocytes from patients with rheumatoid arthritis Guiducci S [email protected] Rosso A 1Cinelli M 1Perfetto F 1Livi R 1Rossi A 5Gabrielli A 3Giacomelli R 4Iori N 5Fibbi G 2Del Rosso M 2Cerinic M Matucci [email protected] Department of Internal Medicine, Division of Rheumatology, University of Florence, Florence, Italy2 Department of Experimental Pathology and Oncology, University of Florence, Florence, Italy3 Institute of Clinical Medicine, Hematology and Clinical Immunology, University of Ancona, Didactic pole, Torrette di Ancona, Ancona, Italy4 Department of Internal Medicine and Public Health, University of L'Aquila, L'Aquila, Italy5 Medical Direction, Eli Lilly Italia S.p.a., Sesto Fiorentino (FI), Italy2005 8 9 2005 7 6 R1244 R1253 10 9 2004 13 10 2004 28 7 2005 10 8 2005 Copyright © 2005 Guiducci et al.; licensee BioMed Central Ltd.This is an Open Access article distributed under the terms of the Creative Commons Attribution License (), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
Extracellular fibrinolysis, controlled by the membrane-bound fibrinolytic system, is involved in cartilage damage and rheumatoid arthritis (RA) synovitis. Estrogen status and metabolism seem to be impaired in RA, and synoviocytes show receptors for estrogens. Our aims in this study were to evaluate in healthy and RA synoviocytes the effects of Raloxifene (RAL), a selective estrogen receptor modulator (SERM), on: proliferation; the components of the fibrinolytic system; and chemoinvasion. The effects of RAL were studied in vitro on synoviocytes from four RA patients and four controls. Proliferation was evaluated as cell number increase, and synoviocytes were treated with 0.5 μM and 1 μM RAL with and without urokinase-plasminogen activator (u-PA) and anti-u-PA/anti-u-PA receptor (u-PAR) antibodies. Fibrinolytic system components (u-PA, u-PAR and plasminogen activator inhibitor (PAI)-1) were assayed by ELISA with cells treated with 0.5 μM and 1 μM RAL for 48 h. u-PA activity was evaluated by zymography and a direct fibrinolytic assay. U-PAR/cell and its saturation were studied by radioiodination of u-PA and a u-PA binding assay. Chemoinvasion was measured using the Boyden chamber invasion assay. u-PA induced proliferation of RA synoviocytes was blocked by RAL (p < 0.05) and antagonized by antibodies alone. The inhibitory effect of RAL was not additive with u-PA/u-PAR antagonism. RA synoviocytes treated with RAL showed, compared to basal, higher levels of PAI-1 (10.75 ± 0.26 versus 5.5 ± 0.1 μg/106 cells, respectively; p < 0.01), lower levels of u-PA (1.04 ± 0.05 versus 3.1 ± 0.4 ng/106 cells, respectively; p < 0.001), and lower levels of u-PAR (11.28 ± 0.22 versus 23.6 ± 0.1 ng/106 cells, respectively; p < 0.001). RAL also significantly inhibited u-PA-induced migration. Similar effects were also shown, at least partially, in controls. RAL exerts anti-proliferative and anti-invasive effects on synoviocytes, mainly modulating u-PAR and, to a lesser extent, u-PA and PAI-1 levels, and inhibiting cell migration and proliferation.
==== Body
Introduction
It is well known that sex hormones are implicated in the immune response. Estrogens enhance humoral immunity, while androgens and progesterone are natural immune-suppressors [1]. In rheumatoid arthritis (RA), sex hormones fuel synovitis. Synovial macrophages, monocytes and lymphocytes [2] possess functional androgen and estrogen receptors and metabolize gonadal hormones [3]. In RA, an association of estrogen gene polymorphism with age at onset has been observed [4].
In both male and female RA patients, low levels of androgens and a low androgen/estrogen ratio have been reported [5]. This supports a possible pathogenic immunosuppressive role for decreased androgen levels. In RA, normal serum androgen and low estrogen levels, but high synovial fluid estrogen and lower androgen levels, indicate that peripheral sex hormone metabolism may be involved in the manifestations of the disease and seems to play an important role in the immune-inflammatory local response [6].
A recent study provides a link between estrogen receptors (ER-alpha) on fibroblast-like synoviocytes and regulation of extracellular matrix (ECM) functionality by the system of matrix metalloproteinases (MMPs)/tissue inhibitors of matrix metalloproteinases (TIMPs) [7]. The expression and activity of MMPs, as well as the levels of TIMPs, is stimulated by 17beta-estradiol, but inhibited by progesterone.
Excessive extracellular proteolysis characterizes neoplastic cell invasion [8], tumor- or inflammation-associated angiogenesis [9] and breakdown of the articular cartilage in osteoarthritis [10]. As well as MMPs, the cell-associated serine proteases of the plasminogen activator/plasminutes system are also involved in extracellular proteolysis required for cell invasion, possibly including cartilage and subcondral bone degradation in RA.
In the fibrinolytic system, the urokinase-type plasminogen activator (u-PA) interacts with its membrane receptor (u-PAR) and activates the single-chain proenzyme plasminogen to the two-chain broad-spectrum serine proteinase plasmin, which is able to degrade ECM both directly and indirectly through activation of secreted pro-MMPs.
Membrane-type MMP undergoes a plasmin-dependent activation that enables it to activate membrane receptor-bound progelatinase A, thus triggering a multienzyme cascade leading to ECM destruction and subsequent cell invasion [11]. Additionally, cell-associated proteases are required for the activity of pro-angiogenic factors, which sustain the synovial pannus growth [12].
Synovial cells express membrane u-PAR, and cultured RA synoviocytes display a higher production of plasminogen activator inhibitor (PAI)-1 than in osteoarthritis and normal synoviocytes [13], suggesting that the plasminogen activator/plasminutes system is involved in the inflammatory remodeling of connective tissues occurring in arthritic joints. Beside its function in the plasminogen activation process, the u-PA/u-PAR interaction also induces plasmin-independent events, such as chemotaxis and chemokinesis [14], the proliferation [15,16] and differentiation [17] of RA synoviocytes and the autocrine secretion of u-PA [17]. Recently, our group has shown that, in healthy synoviocytes, the u-PA/u-PAR interaction determines chemotaxis, chemoinvasion and proliferation in a dose-dependent fashion [18].
More recently, we have shown that RA synoviocytes over-express u-PAR and PAI-1, under-express u-PA, and are more prone than their normal counterpart to spontaneous and u-PA-challenged invasion and proliferation [19].
Raloxifene (RAL) is a selective estrogen receptor modulator (SERM) [20,21] with anti-estrogen activity in uterus and breast tissues and pro-estrogen activity in bone [22,23]. In post-menopausal women, RAL is widely used in the prevention and treatment of osteoporosis due to its anti-resorptive activity and established efficacy in reducing the risk of vertebral fracture [24].
Depending on the endocrine balance, synoviocyte activity can be reduced or enhanced, leading to amelioration or exacerbation of synovitis, respectively. These considerations provide a link between hormonal status and the mechanism of ECM destruction in RA and open new avenues for possible future therapeutic intervention. While the regulatory effect of estrogen is partly targeted to synoviocyte-associated MMPs and TIMPs [7], little is known about the effects of RAL on healthy and RA synoviocytes and their fibrinolytic pattern.
Therefore, our aim was to observe the effects of RAL on the fibrinolytic components (u-PA, u-PAR and PAI-1) of RA synoviocytes in order to modulate their levels and to reduce their fibrinolytic-dependent cellular proliferation and invasion.
Materials and methods
Patients
Sex and age matched patients with RA and healthy controls were used as a source of synoviocytes. Synovial tissue was obtained from four RA patients undergoing surgery for synoviectomy or joint replacement, and from four controls undergoing orthopaedic surgery for knee trauma. Informed consent was obtained from the patients enrolled in this study and ethics approval was given by the local ethics committee.
Synovial cell cultures
Synovia was removed from knee joints, cut and subjected to a mild proteolytic treatment (0.05% trypsin, 0.5 mM EDTA in phosphate buffer saline, for 10 minutes at 37°C under gentle shaking). Trypsin was neutralized with FCS (Celbio, Milano, Italy) and cells were plated in culture dishes with RPMI 1640 (Cambrex Bio Science, Milano, Italy) supplemented with 10% FCS, 2 mM glutamine (Cambrex Bio Science) and penicillin-streptomycin (Cambrex Bio Science). Cell monolayers were used within the seventh passage in culture. The cells were considered type B fibroblast-like synoviocytes if negative by staining with anti-CD69, anti-CD14, anti-CD11b and anti-CD11c (Santa Cruz Biotechnology, Santa Cruz, CA, USA), positive by staining for the enzyme uridine-diphospho glucose dehydrogenase, and if they had a spindle-shaped, fibroblast-like morphologic appearance.
RAL (Lilly, Sesto Fiorentino (FI), Italy) was dissolved in methanol 100 mmol and further diluted with culture medium. Samples were analyzed both in basal conditions and after treatment with RAL (48 h). In preliminary experiments, we tested the effects of different doses of RAL in order to identify which dose would be more effective on synoviocytes without a lethal effect. Thus, the concentrations of 0.5 μM and 1 μM RAL were chosen for further experiments and used in this study. These concentrations were approved by the manufacturer (Lilly) and seem to be comparable with normal therapeutic doses, even if it is impossible to know which dose really reaches the synovia.
A431 cell culture
A431 is a human epidermoid cancer cell line from a cervix squamous cell carcinoma. A431 cells were cultivated in DMEM with 4.5 g/L glucose (Cambrex Bio Science), supplemented with 10% FCS, 2 mM glutamine and penicillin-streptomycin. When cells were at confluence they were washed and incubated with DMEM 0.2% FCS for 48 h. Then the conditioned medium was centrifuged to remove particles and stocked at -80°C. It was used as a chemoattractant reagent in the chemoinvasion assay because of its known property to secrete chemoattractant soluble factors [25-27].
Proliferation assay
Cell growth was quantified in subconfluent cell monolayers. Synoviocytes were seeded in 24 multi-well plates (15,000 cells/well) with 10% FCS in RPMI 1640. After 48 h incubation, cells were washed three times with serum-free medium and incubated in 0.2% FCS medium for an additional 48 h. Cells were then incubated for 48 h in 10% FCS medium (positive control); 0.2% FCS medium (negative control); 0.2% FCS with u-PA (Serono, Roma, Italy) 500 ng/ml with/without u-PA or u-PAR antagonists. The agonists were anti-human u-PA monoclonal antibody 5B4 (mAb 5B4) and anti-u-PAR monoclonal antibody 3936 (mAb 3936) (American Diagnostica, Montreal, Canada), which were used at 1.5 μg/ml. Both mAb 5B4 and mAb 3936 sterically impede the u-PA/u-PAR interaction and were developed against the u-PA-binding site of u-PAR. In preliminary experiments, we tested the effects of different doses of u-PA to identify what dose would be more efficacious in inducing synoviocyte proliferation. Thus, the concentration of 500 ng/ml u-PA was chosen.
Each experimental point was performed in triplicate. At the end of incubation cells were counted. Samples were analyzed both in basal conditions and after treatment with RAL 0.5 μM and 1 μM (48 h).
Analysis of u-PA, u-PAR and PAI-1 levels
Samples were analyzed for u-PAR, u-PA and PAI-1 using commercially available ELISA kits (IMUBIND, American Diagnostica, Montreal, Canada) according to the manufacturer's instructions. Briefly, synoviocytes were seeded in six multi-well plates (25,000 cells/well) with 10% FCS in RPMI 1640. After 48 h of incubation, cells were washed three times with serum-free medium and incubated in 0.2% FCS medium for an additional 48 h. Cells were then treated with 0.5 μM or 1 μM RAL for 48 h. At the end of incubation, cells were detached, counted and lysed by a lysis buffer, as suggested by the manufacturer. The lysates were replaced in their original well and incubated for 1 h at 4°C to allow exhaustive extraction of undetached material. Cell extracts were centrifuged and stocked at -80°C until u-PAR analysis. Culture mediums were collected, centrifuged and stocked at -80°C until u-PA and PAI-1 determination. The results were correlated to the standard curve, within the range of linearity. Each sample was evaluated in triplicate and with two different dilutions. The u-PA assay reveals u-PA antigen, independent of its catalytic activity; the u-PAR and PAI-1 assays measure both occupied and unoccupied receptor, and both free and u-PA-coupled PAI-1, respectively.
The sensibility levels were: 10 pg of u-PA/ml of sample; 0.1 ng of u-PAR/ml of sample; 1 ng of PAI-1 /ml of sample.
Analysis of u-PA enzymatic activity
u-PA enzymatic activity was evaluated by zymography. Culture medium samples were collected as described above and concentrated by centrifugation at 8,000 rpm for 30 minutes in centricon tubes (Amicon Division, Beverly, MA, USA) with 30 kDa molecular weight cut-off pores. The samples were subjected to SDS-PAGE (10%) under non-reducing conditions and migrated proteins were transferred onto 0.45 μm pore-size nitrocellulose filters (Bio-Rad Laboratories, Richmond, California, USA) in a 0.04 M phosphate buffer (pH 6.5) and run for 2 h under a current of 0.4 A. The nitrocellulose filter was removed and placed on an indicating layer containing casein and plasminogen. After overnight incubation at 37°C, u-PA digestion of plasminogen showed clear bands of lysis in the cloudy casein background, corresponding to the position of plasminogen activators in the polyacrylamide gel. u-PA 0.1 U/ml was used as a positive control. Zymograms were then scanned. Samples were analyzed both in basal conditions and after treatment with 0.5 μM or 1 μM RAL (48 h).
Because zymography was performed after protein separation by SDS-PAGE, which uncoupled u-PA from PAI-1, we used a direct fibrinolytic assay of cell membrane-associated plasminogen activators to study the final fibrinolytic balance of synoviocytes. Briefly, synoviocyte monolayers were treated with acidic wash (50 mM glycine-HCl buffer, pH 3.0, 0.1 M NaCl) to uncouple both u-PA and u-PA/PAI-1 complexes from cell surface u-PAR; Hepes buffer 0.5 M NaCl, pH 7.5, was added after 3 minutes. Aliquots of the acidic wash were dotted on fibrin plates prepared according to a method previously described [17] and fibrinolytic activity was evaluated after 16 h of incubation at 37°C, measuring the diameter of the lysis areas. Each sample was evaluated in triplicate.
Radioiodination of u-PA and u-PA binding assay
u-PA was subjected to radioiodination with 125I-Na, using Iodogen (Pierce Eurochemie BV, Holland), following the instructions of the manufacturer. Iodinated protein was separated from non-incorporated radioactivity by gel filtration on Sephadex G-25 (Pharmacia, Milan, Italy). The specific activity obtained in the preparation used in this study was 17.8 μCi/μg. Saturation binding experiments were performed as previously described [28]. Binding experiments were performed either on untreated cell monolayers or following acidic wash (50 mM glycine-HCl buffer, pH 3.0, 0.1 M NaCl) to uncouple u-PA from cell surface u-PAR; Hepes buffer 0.5 M NaCl, pH 7.5, was added after 3 minutes.
Migration assays
The Boyden chamber procedure was used to evaluate cell migration [14]. The method is based on the passage of cells across porous filters against a concentration gradient of the migration effector. A 48-well micro-chemotaxis chamber (Neuroprobe, Gaithersburg, MD, USA) was used. The two wells were separated by a polyvinyl-pyrrolidine-free polycarbonate filter with an 8 μm pore size (Neuroprobe). To evaluate chemoinvasion, the filter was coated with Matrigel (50 μg/filter) (Becton Dickinson Labware, Two Oak Park, Bedford, MA, USA). Test solutions were dissolved in serum-free medium and placed in the lower wells. 50 μl of cell suspension (12,500 cells) were added to each upper well. u-PA 100 ng/ml and/or RAL (0.5 μM or 1 μM) were added to the upper and/or lower well. In the experiment with neutralizing antibodies, the anti-u-PA mAb 5B4 (1.5 μg/ml) was placed in the lower wells, while the anti-u-PAR mAb 3936 (1.5 μg/ml) was incubated with the cell suspension. The chamber was incubated at 37°C for 5 h after which the filter was removed and fixed with methanol. Non-migrating cells on the upper surface of the filter were removed with a cotton swab. Cells were stained with Diff-Quick (Mertz-Dade AG, Dade International, Milan, Italy) and counted using a light microscope (40×) in 10 random fields for each well. Mobilization was measured by the number of cells moving across the Matrigel and the filter pores and spread on the lower surface of the filter. Each experimental point was performed in triplicate. Mean values of migrated cells for each experimental point were calculated.
Statistics
A non-parametric Mann-Whitney test for independent samples was used to compare results from healthy and RA synoviocytes for the levels of u-PA, u-PAR and PAI-1. The results were expressed as mean ± standard deviation. The proliferation induced by u-PA and RAL was evaluated by two tailed t-test for independent samples and by analysis of variance (ANOVA with Bonferroni correction). Migration was measured as a percentage of the healthy basal response and was expressed as the mean ± standard error of the mean.
Results
u-PA-dependent proliferation of synoviocytes
u-PA induced a dose-dependent proliferation in both normal and RA synoviocytes, reaching a maximum at 500 ng/ml (p < 0.001 for any dose in both cell lines) (Fig. 1). Proportionally, u-PA increased cell proliferation in RA and healthy synoviocytes, without reaching a significant difference between the two cell lines at any u-PA dose.
These results show that u-PA has a pro-proliferative effect on both healthy and RA synoviocytes and identify 500 ng/ml u-PA as the minimal dose providing the maximal proliferative effect. This dose has been used in the following experiments on u-PA-dependent proliferation.
Effects of Raloxifene on synoviocyte proliferation
RAL inhibited cell proliferation in a dose-dependent manner in both normal (p < 0.001) and RA synoviocytes (p < 0.001), reaching a maximal effect at 1 μM (Fig. 2a). The RAL induced decrease in cell proliferation did not significantly differ between RA and healthy synoviocytes for any RAL dose. These results indicate that RAL exerts an anti-proliferative effect on both healthy and RA synoviocytes that is effective at doses as small as 0.5 μM.
Effects of Raloxifene on u-PA-dependent proliferation
In both normal and RA synovial cells, RAL significantly reduced u-PA- and 10% FCS-dependent proliferation (p < 0.001) (Fig. 2b). RA synoviocytes are as prone as normal synoviocytes to spontaneous (0.2% FCS) proliferation and to proliferation challenged both with 10% FCS and 500 ng/ml u-PA (Fig. 2b). In particular, serum-dependent proliferation (10% FCS) did not significantly differ from u-PA-dependent proliferation in both healthy and RA cell lines.
The proliferative effect elicited by 500 ng/ml of u-PA in 0.2% FCS was significantly reduced by the mAb antagonists of u-PA (mAb 5B4) and u-PAR (mAb 3936) in healthy (p < 0.001) and RA synoviocytes (p < 0.001) (Fig. 2b). This indicates that u-PA/u-PAR interaction is required for the pro-proliferative effect of u-PA on normal and RA synoviocytes.
RAL blocked both FCS-dependent and u-PA-dependent proliferation to a similar extent. RAL reduction of u-PA-dependent proliferation was similar to the inhibition caused by the antagonist antibodies (p < 0.001 in normal and RA synoviocytes) (Fig. 2b) and was not increased by co-treatment with them, indicating a common target for RAL and anti u-PA/u-PAR antibodies (data not shown). This suggests that inhibition of the expression of critical members of the cell-associated fibrinolytic system is a likely target of RAL treatment.
u-PA, u-PAR and PAI-1 levels in normal and RA synoviocytes
As our ELISA assay for u-PA measures u-PA antigen independent of its interaction with PAI-1, the data shown in Fig. 3a indicate that u-PA released into the culture medium by RA synoviocytes was significantly lower than u-PA produced by healthy synoviocytes (3.1 ± 0.4 versus 10.05 ± 0.04 ng/106 cells, respectivle; p < 0.05).
A zymographic assay of u-PA performed on aliquots of the culture medium confirmed the data obtained using antibodies. Culture mediums of RA synoviocytes display a lower u-PA activity than medium obtained from healthy synoviocytes (Fig. 4a).
The direct fibrinolytic assay, performed on fibrin plates as described, indicated the following: the culture medium of healthy synoviocytes showed a higher fibrinolytic activity than RA synoviocytes, thus confirming the data obtained with zymography, again indicating lower constitutive u-PA production in RA synoviocytes (Fig. 4b, left side); and the fibrinolytic activity exhibited by aliquots of the acidic wash was slightly higher in RA than in healthy synoviocytes (Fig. 4b, right side). This means that, regardless of the lower amount of secreted u-PA, a larger availability of u-PAR on the surface of RA synoviocytes enabled RA synovial cells to bind a higher amount of the ligand. These results have also been confirmed by radioligand binding experiments. Healthy synoviocytes exhibited 95 ± 15 × 103 u-PAR/cell, with a Kd ranging from 1.75 to 1.85 nM, and the receptor number was 140 ± 25 × 103/cell following acidic treatment prior to performing binding experiments. RA synoviocytes had 180 ± 30 × 103 uPAR/cell before and 230 ± 48 × 103 u-PAR/cell after acidic wash, while the Kd did not change. Higher levels of u-PAR are found in cell lysates of RA than of healthy synoviocytes (23.6 ± 0.10 versus 9.2 ± 0.05 ng/106 cells, respectively; p < 0.05), also determined using ELISA (Fig. 3b).
Moreover, RA synovial cells release into culture medium higher amounts of PAI-1 than healthy synoviocytes (5.5 ± 0.1 versus 2.9 ± 0.1 μg/106 cells, respectively; p < 0.05; Fig. 3c).
Effects of Raloxifene on u-PA, u-PAR and PAI-1 levels in normal and rheumatoid arthritis synoviocytes
RAL reduces u-PA levels in healthy and RA synoviocytes (p < 0.0001; Fig. 3a). It does this dose dependently: 1 μM RAL is more efficient than 0.5 μM in healthy (p < 0.01) and in RA synoviocytes (p < 0.05). RAL-dependent reduction of u-PA levels at 0.5 μM and 1 μM is significantly higher (p < 0.05) in healthy than in RA synoviocytes.
A zymographic assay performed on culture medium treated with 0.5 μM and 1 μM RAL confirmed the data obtained with the u-PA assay. In fact, RAL reduces the u-PA enzymatic activity in a dose dependent manner, in both healthy and RA synoviocytes (Fig. 4a).
The direct fibrinolytic assay on fibrin plates, performed with aliquots of culture medium, indicated that RAL-dependent reduction of cell-associated fibrinolytic activity is more pronounced in RA than in normal synoviocytes (Fig. 4b, left side). Fig. 4b (right side) shows a RAL-dependent reduction of u-PA in the acidic wash of both normal and RA synoviocytes.
RAL significantly reduces u-PAR levels in cell lysates from RA synoviocytes (p < 0.0001; Fig. 3b); 1 μM RAL is more efficient than 0.5 μM at reducing u-PAR levels in healthy and RA synoviocytes (p < 0.01). The effect of 0.5 μM and 1 μM RAL is significantly higher in RA than in healthy synoviocytes (p < 0.05). Radioligand binding experiments showed that, upon RAL treatment, the receptor number before and after acidic treatment was similar in healthy synoviocytes, whereas a dramatic decrease of both unoccupied u-PAR (85 ± 18 × 103 receptor/cell, before acidic wash) and total u-PAR (150 ± 25 × 103 receptor/cell after acidic wash) was observed in RA synoviocytes.
RAL significantly increases PAI-1 levels in healthy and RA synoviocytes (p < 0.0001; Fig. 3c). This is true for both 0.5 μM and 1 μM RAL (healthy synoviocytes, p < 0.01 and p < 0.001, respectively; RA synoviocytes, p < 0.001 and p < 0.01, respectively). RAL 1 μM is more efficient than RAL 0.5 μM at increasing PAI-1 levels in both healthy and RA synoviocytes (p < 0.001). RAL 0.5 μM and 1 μM have a significantly higher effect on increasing PAI-1 levels in RA than in healthy synoviocytes (p < 0.05).
u-PA-dependent synoviocyte chemoinvasion
u-PA-dependent chemoinvasion was dose-dependent (Fig. 5a), with a maximal effect at 100 ng/ml for both healthy and RA synoviocytes. Basal migration was more pronounced (more than 40%) in RA than in healthy synoviocytes. This difference is maintained at each concentration of u-PA in the lower well of the migration chamber. An increase in invasion observed after a 5 h incubation with 100 ng/ml u-PA was counteracted by the incubation of invasive cells with mAb antagonists of u-PA (mAb 5B4) and u-PAR (mAb 3936) (Fig. 5b). This indicates that u-PA/u-PAR interaction is required for the pro-invasive effect of u-PA on normal and RA synoviocytes.
Effects of Raloxifene on u-PA-dependent synoviocyte chemoinvasion
RAL inhibits cell chemoinvasion in a dose-dependent manner in both normal and RA synoviocytes, reaching a maximum at 1 μM (Fig. 5B); this decrease did not significantly differ between healthy and RA cells at any RAL dose.
RAL significantly reduced u-PA-induced chemoinvasion in healthy and RA synoviocytes; 1 μM RAL reduced chemoinvasion to basal values. When cells were incubated together with RAL in the upper chamber, it caused similar effects on chemoinvasion. In fact, RAL was able to inhibit migration induced by u-PA in a dose-dependent way with a maximal action at a dose of 1 μM and this decrease in mobility was proportionally similar in healthy and RA synoviocytes (data not shown).
Upon combining the results shown in Figs 3, 4 and 5, one is tempted to think that RAL works mainly through the reduction in the levels of u-PAR and, therefore, by blocking u-PA-dependent chemoinvasion. In fact, the RAL reduction of u-PA-dependent migration was similar to the inhibition caused by the antagonist mAbs (Fig. 5b) and is not increased by co-treatment with them, indicating that RAL and anti-u-PA/u-PAR antibodies have a common target (data not shown).
Discussion
This is the first study on the effects of RAL on RA synoviocytes and on their fibrinolytic pattern and function. The presence of functional estrogen receptors (ER-alpha) in synoviocytes has been identified [7,29]. Our data show that RAL acts as an inhibitor of the fibrinolytic system, exhibiting similar effects on both healthy and RA synoviocytes, with the activity on RA synoviocytes being more pronounced. Indeed, the drug inhibits cell proliferation and chemoinvasion in a dose-dependent manner, reduces u-PA and u-PAR levels, increases PAI-1 levels and blocks u-PA by acting on the specific interaction between u-PA and u-PAR.
Despite increasing evidence on the significance of sex hormones in RA, their etiopathological role and potential long term effect on RA progression still remain unclear. Also, the link between sex hormones and the fibrinolytic system is still to be elucidated in RA synovial cells.
In human breast cancer cells, estradiol inhibited the expression and secretion of u-PA, t-PA and PAI-1 proteins. Zymography confirmed the inhibitory effect of estradiol on u-PA activity [30]. The regulation of expression of genes encoding u-PA and PAI-1 by estradiol and different SERMs has been described in human breast cancer cells [31]. Two different SERMS (4-hydroxytamoxifen and RAL) had a concentration-dependent agonistic (estradiol-like) effect on the regulation of these genes. In contrast, a pure anti-estrogen alone had no effect but could block the action of estradiol and SERMs. This demonstrates that RAL has an estrogen-like effect on human breast cancer cells with respect to the regulation of u-PA and PAI-1 gene expression [31].
These findings prompted us to investigate the effects of RAL on the fibrinolytic system of RA synovial cells. In fibroblast-like synoviocytes, we found a link between the functionality of another proteolityc system (the u-PA/u-PAR-PAI-1 system) and a SERM such as RAL. In RA, the formation and invasiveness of synovial pannus, supported by angiogenesis, is linked to serine proteinases, mainly u-PA [11], produced in high quantity by synoviocytes [12,18,19].
The u-PA/u-PAR-PAI-1 system is an organizer of cell-ECM contacts and covers the full range of activities required to promote invasion and angiogenesis and to disrupt cell attachment sites. We have recently shown that RA synoviocytes display a fibrinolytic machinery (u-PA, u-PAR and PAI-1) addressed toward an invasive pattern [19]. RAL inhibits the proliferation of RA synoviocytes dose dependently and blocks the proliferation induced by u-PA, mainly by reducing u-PAR levels without affecting the Kd of the u-PA/u-PAR interaction (as shown in this study by ELISA and radioligand binding experiments) and, therefore, by reducing the specific interaction between u-PA and u-PAR following reduction of u-PAR.
Cellular migration, linked to cellular adhesiveness, is an important process for the invasion of articular cavity and extra-articular tissues, typical of RA. Our data show that RAL inhibits migration of RA synoviocytes, thus potentially contributing to the modulation of the growth of synovial pannus. u-PAR regulates pericellular proteolysis and cell surface adhesion receptors, fundamental events in the first steps of invasion and angiogenesis. RAL further reduces u-PA levels derived from RA synoviocytes, thus potentially reducing the invasive and angiogenetic potential of these cells.
The increased expression of PAI-1 in RA synoviocytes is in agreement with previous data showing higher production of PAI-1 in cultured RA synoviocytes [12,19]. In RA, elevated PAI-1 levels could act as an ECM-stabilizing molecule by blocking extracellular proteolysis, thus providing cells with a substrate favoring cell movement. At the same time, PAI-1 promotes cell detachment [32]. RAL increases PAI-1, thereby blocking the u-PA-dependent ECM degradation and reducing cell movement. The whole in vitro scenario is that RAL reduces u-PAR and u-PA levels and increases PAI-1 levels. This action may significantly contribute to reducing the pro-invasive and pro-angiogenic pattern of RA synoviocytes, although the clinical relevance of these findings require further study.
The over-expression of u-PAR in RA may depend on the need to activate the fibrinolytic pathway in order to degrade and invade ECM, as well as to promote interaction between u-PAR and vitronectin, which provides the adhesive grip necessary for cell locomotion, events required in all the invasive pathologies [33]. RAL seems to exert an upstream blocking of the membrane-bound fibrinolytic system, which also stops the proteolytic activities of MMPs.
Conclusion
RAL modulates in vitro the components and functionality of the membrane-bound fibrinolytic system, exhibiting a more intense activity in RA synoviocytes compared to their normal counterpart. RAL inhibits u-PA-induced proliferation and u-PA-induced cell migration dose dependently. The effect of RAL in the reduction of the formation of synovial pannus and the radiological progression of RA warrants further study.
Abbreviations
DMEM = Dulbecco's modified Eagle medium; ECM = extracellular matrix; ELISA = enzyme-linked immunosorbent assay; FCS = fetal calf serum; mAb = monoclonal antibody; MMP = matrix metalloproteinase; PAI = plasminogen activator inhibitor; RA = rheumatoid arthritis; RAL = Raloxifene; SERM = selective receptor estrogen modulator; TIMP = tissue inhibitor of matrix metalloproteinase; u-PA = urokinase-type plasminogen activator; u-PAR = urokinase-type plasminogen activator receptor.
Competing interests
The authors declare that they have no competing interests.
Authors' contributions
SG isolated synoviocytes, carried out the chemoinvasion assay and participated in drafting the manuscript. ADR participated in drafting the manuscript and performed the statistical analysis. MC carried out the ELISA assay, proliferation experiments and participated in drafting the manuscript. RL performed the fibrinolytic assay. FP, AR, AG, RG and NI participated in the design of the study. GF and MDR coordinated the study. MMC conceived the study.
Figures and Tables
Figure 1 Urokinase-type plasminogen activator u-PA/u-PA receptor-dependent proliferation in healthy (H) and rheumatoid arthritis (RA) synoviocytes. Synoviocytes were seeded in 24 multi-well plates (15,000 cells/well) with 10% FCS in RPMI 1640. After 48 h incubation, cells were washed three times with serum-free medium and incubated in 0.2% FCS medium for an additional 48 h. At this time, cells were incubated for 48 h in 0.2% FCS containing increasing concentrations of u-PA (50 to 1,000 ng/ml). Proliferation of H and RA synoviocytes is reported as a function of u-PA concentration. Each point represents the mean ± standard deviation of three experiments performed in triplicate on four normal and four RA synovial cell lines.
Figure 2 Effects of Raloxifene (RAL) on proliferation in healthy (H) and rheumatoid arthritis (RA) synoviocytes. (a) Proliferation of H and RA synoviocytes as a function of RAL concentration. Synoviocytes were seeded in 24 multi-well plates (15,000 cells/well) with 10% FCS in RPMI 1640. After 24 h incubation, cells were treated with increasing concentrations of RAL (0.5 to 2 μM). (b) Proliferation of H and RA synoviocytes under control conditions (0.2% FCS as a negative control, and 10% FCS as a positive control), in 0.2% FCS under the effect of 500 ng/ml urokinase-type plasminogen activator (u-PA) and in the presence of 0.5 and 1 μM RAL or the monoclonal antibodies (mAbs) 5B4 and 3936, which impair u-PA/u-PAR binding (see text for details). In both (a) and (b), each point represents the mean ± standard deviation of three experiments performed in triplicate on four normal and four RA synovial cell lines. § = p < 0.01 versus basal (0.2% FCS); # = p < 0.001 versus 10% FCS or u-PA, for both H and RA.
Figure 3 Fibrinolytic pattern of healthy (H) and rheumatoid arthritis (RA) synoviocytes treated with Raloxifene. Samples were analyzed both in basal conditions and after treatment with 0.5 and 1 μM Raloxifene (48 h). (a) Urokinase-type plasminogen activator (u-PA) antigen was quantified on aliquots of the culture medium by ELISA assay and reported as ng/106 synovial cells. (b) u-PA receptor (u-PAR) was quantified by ELISA assay on aliquots of cell lysates and reported as ng/106 synovial cells. (c) Plasminogen activator inhibitor (PAI)-1 was also measured in aliquots of the culture medium, and reported as μg/106 synovial cells. H refers to four different synovial cell cultures from healthy individuals, RA refers to synovial cell cultures from four different rheumatoid arthritis patients. In (a-c), each point represents the mean ± standard deviation of three experiments performed in triplicate on each synovial cell line. § = p < 0.01 versus basal (0.2% FCS).
Figure 4 Urokinase-type plasminogen activator (u-PA) activity in healthy (H) and rheumatoid arthritis (RA) synoviocytes treated with Raloxifene (RAL). Samples were analyzed both in basal conditions and after treatment with 0.5 and 1 μM RAL (48 h). (a) Zymographic assay of aliquots of the culture medium. u-PA digestion of plasminogen shows clear bands of lysis in the cloudy casein background of the indicating layer. Shown here are the effects on one H and one RA line; the other lines furnished similar results. (b) Direct fibrinolytic assay of cell membrane-associated plasminogen activators. Left side: lysis areas of culture medium under the indicated experimental conditions. Right side: lysis areas of aliquots of the acidic wash (Ac), indicating the activity eluted from the cell surface. The diameter (mm) of the lysis areas were measured as an index of fibrin digestion. H refers to four different synovial cell cultures from healthy individuals; RA refers to synovial cell cultures from four different RA patients Each point represents the mean ± standard deviation of three experiments performed in triplicate on each synovial cell line. § = p < 0.01 versus control (C).
Figure 5 Urokinase-type plasminogen activator (u-PA)/u-PA receptor (u-PAR)-dependent chemoinvasion in healthy (H) and rheumatoid arthritis (RA) synoviocytes. (a) Cell invasion of Matrigel-coated filters by H and RA synoviocytes as a function of u-PA concentration. Migration was stimulated by increasing concentrations of u-PA (5 to 250 ng/ml) in the lower well of the migration chamber. (b) Percent increase of Matrigel invasion in H and RA synoviocytes treated with Raloxifene (RAL) or neutralizing antibodies. Basal, invasion of H and RA synovial cells in the presence of 0.2% FCS in the lower well; C+, invasion stimulated by conditioned medium of A431 cell line, used as a sure chemotactic agent; RAL 0.5 and RAL 1, invasion challenged with 0.5 μM and 1 μM RAL in the lower well; u-PA, invasion challenged with 100 ng/ml of u-PA in the lower well; u-PA+RAL 0.5 and u-PA+RAL 1, invasion challenged with 100 ng/ml u-PA in the presence of 0.5 and 1 μM RAL in the lower well; u-PA+5B4 and u-PA+3936, invasion challenged with 100 ng/ml u-PA in the presence of 1.5 μg/ml of monoclonal antibodies 5B4 and 3936 in the lower and upper well, respectively. In (a) and (b), each point represents the mean ± standard deviation of three experiments performed in triplicate on four normal and four RA synovial cell lines. § = p < 0.01 versus basal (0.2% FCS); # = p < 0.001 versus u-PA, for both H and RA.
==== Refs
Cutolo M Villaggio B Craviotto C Pizzorni C Seriolo B Sulli A Sex hormones and rheumatoid arthritis Autoimmun Rev 2002 1 284 289 12848982 10.1016/S1568-9972(02)00064-2
Cutolo M Accardo S Villaggio B Clerico P Bagnasco M Coviello D Carruba G Lo Casto M Castagnetta L Presence of estrogen-binding sites on macrophage-like synoviocytes and CD8+, CD29+, CD45RO+ T lymphocytes in normal and rheumatoid synovium Arthritis Rheum 1993 36 1087 1097 8343185
Cutolo M Accardo S Villaggio B Barone A Sulli A Coviello DA Carabbio C Felli L Miceli D Farruggio R Androgen and estrogen receptors are present in primary cultures of human synovial macrophages J Clin Endocrinol Metab 1996 81 820 827 8636310 10.1210/jc.81.2.820
Ushiyama T Mori K Inoue K Huang J Nishioka J Hukuda S Association of oestrogen receptor gene polymorphisms with age at onset of rheumatoid arthritis Ann Rheum Dis 1999 58 7 10 10343533
Cutolo M Seriolo B Villaggio B Pizzorni C Craviotto C Sulli A Androgens and estrogens modulate the immune and inflammatory responses in rheumatoid arthritis Ann NY Acad Sci 2002 966 131 142 12114267
Cutolo M Sulli A Pizzorni C Craviotto C Straub RH Hypothalamic-pituitary-adrenocortical and gonadal functions in rheumatoid arthritis Ann NY Acad Sci 2003 992 107 117 12794051
Khalkhali-Ellis Z Seftor EA Nieva DR Handa RJ Price RH JrKirschmann DA Baragi VM Sharma RV Bhalla RC Moore TL Hendrix M Estrogen and progesterone regulation of human fibroblast-like synoviocyte function in vitro: implications in rheumatoid arthritis J Rheumatol 2000 27 1622 1631 10914842
Mignatti P Rifkin DB Biology and biochemistry of proteinase in tumor invasion Physiol Rev 1993 73 161 195 8419965
Bouck N Stellmach V Hsu SC How tumors become angiogenetic Adv Cancer Res 1996 69 135 174 8791681
Hollander AP Dieppe PA Atkins RM Elson CJ Hypothesis: cartilage catabolic cofactors in human arthritis J Rheumatol 1993 20 223 234 8474056
Vassalli JD Pepper MS Membrane proteases in focus Nature 1994 370 14 15 8015594 10.1038/370014a0
Smeets TJ Barg EC Kraan MC Smith MD Breedveld FC Tak PP Analysis of the cell infiltrate and expression of proinflammatory cytokines and matrix metalloproteinases in arthroscopic synovial biopsies: comparison with synovial samples from patients with end stage, destructive rheumatoid arthritis Ann Rheum Dis 2003 62 635 638 12810425 10.1136/ard.62.7.635
Matucci Cerinic M Generini S Partsch G Pignone A Dini G Konttinen YT Del Rosso M Synoviocytes from osteoarthritis and rheumatoid arthritis produce plasminogen activators and plasminogen activator inhibitor-1 and display u-PA receptor on their surface Life Sci 1998 63 441 453 9718068 10.1016/S0024-3205(98)00293-8
Fibbi G Ziche M Morbidelli L Magnelli L Del Rosso M Interaction of urokinase with specific receptors stimulates mobilization of bovine adrenal capillary endothelial cells Exp Cell Res 1988 179 385 395 2847935 10.1016/0014-4827(88)90277-7
Kirchheimer JC Wojta J Christ G Binder BR Proliferation of a human epidermal cell line stimulated by urokinase FASEB J 1987 1 125 128 3038646
Anichini E Fibbi G Pucci M Caldini R Chevanne M Del Rosso M Production of second messengers following chemotactic and mitogenic urokinase-receptor Exp Cell Res 1994 213 438 448 8050501 10.1006/excr.1994.1221
Fibbi G Magnelli L Pucci M Del Rosso M Interaction of urokinase with the receptor of human keratinocytes stimulates release of urokinase-like plasminogen activator Exp Cell Res 1990 187 33 38 2153567 10.1016/0014-4827(90)90112-N
Fibbi G Pucci M Serni U Matucci Cerinic M Del Rosso M Antisense targeting of the urokinase receptor blocks urokinase-dependent proliferation, chemoinvasion, and chemotaxis of human synovial cells and chondrocytes in vitro Proc Assoc Am Phys 1998 110 1 11 9460078
Guiducci S Del Rosso A Cinelli M Margheri F D'Alessio S Fibbi G Del Rosso M Rheumatoid synovial fibroblasts constitutively express the fibrinolytic pattern of invasive tumor-like cells Clin Exp Rheumatol 2005 23 364 372 15971425
Kauffman RF Bryant HU Selective estrogen receptor modulators Drug News Perspectives 1995 8 531 539
Bryant HU Glasebrook AL Yang NN Sato M A pharmacological review of raloxifene J Bone Mineral Metabolism 1996 14 1 9 10.1007/BF01771666
El-Hajj Fuleihan G Tissue-specific estrogens. The promise for the future New Eng J Med 1997 337 1686 1687 9385130 10.1056/NEJM199712043372309
Evans GL Turner RT Tissue-selective actions of estrogen analogs Bone 1995 17 4 Suppl 181S 190S 8579915
Walsh BW Kuller LH Wild RA Paul S Farmer M Lawrence JB Shah AS Anderson PW Effects of raloxifene on serum lipids and coagulation factors in healthy postmenopausal women JAMA 1998 279 1445 1451 9600478 10.1001/jama.279.18.1445
Stoppelli MP Tacchetti C Cubellis MV Corti A Hearing VJ Cassani G Appella E Blasi F Autocrine saturation of pro-urokinase receptors on human A431 cells Cell 1986 45 675 684 3011276 10.1016/0092-8674(86)90782-8
Tsuji T Kawada Y Kai-Murozono M Komatsu S Han SA Takeuchi K Mizushima H Miyazaki K Irimura T Regulation of melanoma cell migration and invasion by laminin-5 and alpha3beta1 integrin (VLA-3) Clin Exp Metastasis 2002 19 127 134 11964076 10.1023/A:1014573204062
Reiss M Stash EB Vellucci VF Zhou ZL Activation of the autocrine transforming growth factor alpha pathway in human squamous carcinoma cells Cancer Res 1991 51 6254 6262 1933886
Del Rosso M Pedersen N Fibbi G Pucci M Dini G Anichini E Blasi F Selective localization of receptors for urokinase amino-terminal fragment at substratum contact sites of an in vitro-established line of human epidermal cells Exp Cell Res 1992 203 427 434 1333982 10.1016/0014-4827(92)90017-3
Ushiyama T Inoue K Nishioka J Expression of estrogen receptor related protein (p29) and estradiol binding in human arthritic synovium J Rheumatol 1995 22 421 426 7540206
Levenson AS Kwaan HC Svoboda KM Weiss IM Sakurai S Jordan VC Oestradiol regulation of the components of the plasminogen-plasminutes system in MDA-MB-231 human breast cancer cells stably expressing the oestrogen receptor Br J Cancer 1998 78 88 95 9662256
Levenson AS Svoboda KM Kwaan HC Jordan VC Agonist activity of antiestrogen-receptor complexes to regulate urokinase plasminogen activator (uPA) and plasminogen activator inhibitor type 1 (PAI-1) endogenous gene expression in breast cancer cells Cancer Lett 1998 125 215 220 9566718 10.1016/S0304-3835(97)00516-8
Ronday HK Smits HH Quax PH van der Pluijm G Lowik CW Breedveld FC Verheijen JH Bone matrix degradation by the plasminogen activation system. Possible mechanism of bone destruction in arthritis Br J Rheumatol 1997 36 9 15 9117184 10.1093/rheumatology/36.1.9
Del Rosso M Fibbi G Pucci M D'Alessio S Del Rosso A Magnelli L Chiarugi V Multiple pathways of cell invasion are regulated by multiple families of serine proteases Clin Exp Metastasis 2002 19 193 207 12067200 10.1023/A:1015531321445
|
16277677
|
PMC1297569
|
CC BY
|
2021-01-04 16:02:40
|
no
|
Arthritis Res Ther. 2005 Sep 8; 7(6):R1244-R1253
|
utf-8
|
Arthritis Res Ther
| 2,005 |
10.1186/ar1815
|
oa_comm
|
==== Front
Arthritis Res TherArthritis Research & Therapy1478-63541478-6362BioMed Central London ar18161627767510.1186/ar1816Research ArticlePolymorphism in the tumour necrosis factor receptor II gene is associated with circulating levels of soluble tumour necrosis factor receptors in rheumatoid arthritis Glossop John R [email protected] Peter T [email protected] Nicola B [email protected] Derek L [email protected] Staffordshire Rheumatology Centre, University Hospital of North Staffordshire, Stoke-on-Trent, UK2005 7 9 2005 7 6 R1227 R1234 17 3 2005 25 4 2005 28 7 2005 10 8 2005 Copyright © 2005 Glossop et al.; licensee BioMed Central Ltd.This is an Open Access article distributed under the terms of the Creative Commons Attribution License (), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
Levels of soluble tumour necrosis factor receptors (sTNFRs) are elevated in the circulation of patients with rheumatoid arthritis (RA). Although these receptors can act as natural inhibitors of tumour necrosis factor-α, levels of sTNFRs in RA appear to be insufficient to prevent tumour necrosis factor-α induced inflammation. The factors that regulate circulating levels of sTNFRs are unclear, but polymorphisms in the tumour necrosis factor receptor genes may play a role. We investigated the relationship between polymorphisms in the tumour necrosis factor receptor I (TNF-RI) and II (TNF-RII) genes and levels of sTNFRs in two groups of Caucasian RA patients: one with early (disease duration ≤2 years; n = 103) and one with established disease (disease duration ≥5 years; n = 151). PCR restriction fragment length polymorphism analysis was used to genotype patients for the A36G polymorphism in the TNF-RI gene and the T676G polymorphism in TNF-RII. Levels of sTNFRs were measured using ELISA. We also isolated T cells from peripheral blood of 58 patients with established RA with known TNF-R genotypes, and release of sTNFRs into the culture medium was measured in cells incubated with or without phytohaemagglutinin. Serum levels of the two sTNFRs (sTNF-RI and sTNF-RII) were positively correlated in both populations, and the level of each sTNFR was significantly higher in the patients with established disease (P < 0.0001). Multiple regression analyses corrected for age, sex and disease duration revealed a significant trend toward decreasing sTNF-RI and sTNF-RII levels across the TNF-RII genotypes (TT > TG > GG) of patients with established disease (P for trend = 0.01 and P for trend = 0.03, respectively). A similar nonsignificant trend was seen for early disease. No relationship with the TNF-RI A36G polymorphism was observed. sTNFRs released by isolated T cells exhibited a similar trend toward decreasing levels according to TNF-RII genotype, although only the association with levels of sTNF-RII was significant. Strong correlations were found between levels of circulating sTNFRs and levels released by T cells in vitro. Our data indicate that the T676G polymorphism in TNF-RII is associated with levels of sTNFRs released from peripheral blood T cells, and with circulating levels of sTNFR in patients with RA.
==== Body
Introduction
Tumour necrosis factor (TNF)-α is a pleiotropic cytokine that is important in the pathogenesis of rheumatoid arthritis (RA), in which it plays a role in cartilage degradation, bone resorption, adhesion molecule expression, leucocyte infiltration, enzyme production and cytokine synthesis (see reviews by Brennan and coworkers [1] and Choy and Panayi [2]). The actions of TNF-α are mediated through binding to two distinct cell surface receptors, namely tumour necrosis factor receptor I (TNF-RI) and II (TNF-RII) [3,4]. Both are transmembrane glycoproteins with a three domain structure: a multiple cysteine-rich motif bearing an extracellular domain that facilitates ligand binding; a hydrophobic membrane spanning domain; and an intracellular domain that mediates signal transduction. The receptor molecules share significant homology in their extracellular domains but they have distinct intracellular domains [5]. Most significantly, TNF-RI, but not TNF-RII, possesses a death domain that can transduce the signal for cell death [6]. The two receptors appear to promote distinct TNF-α-induced cellular responses, although both are capable of inducing the nuclear factor-κB (NF-κB) and apoptotic pathways [7-10], providing some evidence of receptor function redundancy.
In addition to membrane bound forms, both TNF receptors can exist as soluble proteins. These are soluble variants of the extracellular domains [11-13] and are derived from the membrane bound form by the proteolytic actions of a disintegrin metalloproteinase called TNF-α converting enzyme (TACE) [14]. They retain their ligand binding capacity after cleavage [11,13] and can act as natural inhibitors of TNF-α by sequestering soluble TNF-α and preventing it from binding to membrane-bound TNF receptor. The levels of soluble TNF receptors (sTNFRs) are elevated in the serum and synovial fluid of RA patients [15-17], but these levels appear to be insufficient to prevent the chronic inflammation promoted by TNF-α [16]. Furthermore, the expression of membrane-bound TNF receptor is increased on a variety of cells in RA synovium [18,19], facilitating prolonged TNF-α signalling and the continuation of TNF-α regulated processes. The factors that regulate the levels of sTNFR are unclear, but polymorphisms within the TNF receptor genes may play a role.
The genes encoding TNF-RI and TNF-RII have been mapped to chromosomes 12p13 and 1p36, respectively [20]. Numerous polymorphisms are present in these genes [21-23] and some have been investigated for their association with RA [24-30]. An association has been reported between a single nucleotide polymorphism (SNP) in exon 6 (T676G) of the TNF-RII gene and susceptibility to familial but not sporadic RA [24,25]. Two studies in sporadic RA showed no association between the T676G polymorphism in the TNF-RII gene and RA severity [27,29], although one report has suggested an association with functional severity [28]. The A36G polymorphism in exon 1 of the TNF-RI gene has been associated with a protective role in familial RA [30].
In this study we report that polymorphism in the TNF-RII gene, but not the TNF-RI gene, is associated with circulating levels of TNF receptors in a population of Caucasian RA patients, and that this polymorphism is also associated with levels of sTNFRs released in vitro by isolated T cells from RA patients.
Materials and methods
Patients
Two groups of Caucasian RA patients were studied. The first group had early disease (duration ≤2 years; n = 103) and the second had established disease (duration ≥5 years; n = 151; Table 1). The patients were all of British origin and resident in North Staffordshire, England, and satisfied the 1987 American College of Rheumatology criteria for RA [31]. All patients were receiving anti-inflammatory and/or antirheumatic therapy, with the majority of patients with established disease (>90%) being treated with one or more disease-modifying antirheumatic drugs. Steroids and cytotoxic drugs such as azathioprine or cyclophosphamide were being received by a small minority of individuals (<5%). No patients were being treated with anti-TNF-α agents. Radiographic damage was measured by scoring radiographs of the hands and feet using the method proposed by Larsen and coworkers [32], and functional outcome was assessed using the Health Assessment Questionnaire (HAQ) [33]. In patients with early RA, HAQ measurements were taken at recruitment into the study and at 5 years of follow up.
Serum, separated from clotted peripheral blood (5 ml) from each patient, was stored at -70°C until required for measurement of sTNFRs. Synovial fluid was also collected from 45 patients who presented with knee effusions at the time of blood collection. Fluids were centrifuged and separated from the resulting cell pellet, before storage at -70°C. The study was approved by the North Staffordshire local research ethics committee.
Cell isolation and culture
T cells were isolated from fresh peripheral blood samples (4 ml) from RA patients with established disease (n = 58). Cell isolation was by negative selection using a modified density gradient centrifugation technique that utilizes novel tetrameric antibody complexes (RosetteSep; Stemcell Technologies Inc., Vancouver, Canada). Isolated T cells (2 × 105 cells/200 μl) were cultured in RPMI 1640 synthetic culture medium supplemented with 10% heat-inactivated foetal bovine serum, 100 units/ml penicillin, 100 μg/ml streptomycin and 10% autologous serum, in 96-well cell culture plates. Cultures were incubated, with or without phytohaemagglutinin (PHA; 10 μg/ml), at 37°C in a 5% carbon dioxide humidified air environment for 48 hours. Cell supernatants were then harvested and stored at -20°C until required for analysis of sTNFR levels.
Genomic DNA isolation
Peripheral blood samples (4 ml) collected in EDTA tubes were obtained from each patient and were stored at -20°C. After thawing at 37°C, the genomic DNA was isolated using a DNAce MegaBlood Kit procedure as directed by the manufacturer (Bioline, London, UK).
PCR primers
The following primer sequences were used to amplify a 183 base pair fragment containing the SNP at nucleotide 36 in exon 1 of the TNF-RI gene [21]: forward 5'-GAG CCC AAA TGG GGG AGT GAG AGG-3', and reverse 5'-ACC AGG CCC GGG CAG GAG AG-3'.
A 242 base pair fragment containing the SNP at nucleotide 676 in exon 6 of the TNF-RII gene was amplified with the following primer sequences [34]: forward 5'-ACT CTC CTA TCC TGC CTG CT-3'; and reverse 5'-TTC TGG AGT TGG CTG CGT GT-3'.
PCR amplification and single nucleotide polymorphism genotyping
The fragment of interest from each of the TNF receptor genes was amplified using an identical reaction mixture and conditions that were described previously [27]. All amplification reactions were performed in a Flexigene Thermal Cycler unit (Techne [Cambridge] Limited, Cambridge, UK) using a 96-well, full-skirt heating block. During amplification wells were capped with PCR cap strips. Following amplification the products were stored at 4°C until required for genotyping by restriction fragment length polymorphism analysis [27].
ELISA
Serum, synovial fluid and T cell supernatant levels of sTNF-RI and sTNF-RII were quantified using the respective Duoset ELISA Development Kit as directed by the manufacturer (R&D Systems Europe, Abingdon, UK). For determination of sTNF-RI levels, sera, synovial fluids and T-cell supernatants were diluted 1:10, 1:50 and 1:3, respectively. For soluble TNF-RII, sera, synovial fluids and T-cell supernatants were diluted 1:20, 1:80 and 1:4, respectively. All samples were run in duplicate with the appropriate standards on 96-well microplates.
Statistical analysis
The relationship between the two sTNFRs was assessed using Spearman's rank correlation, whereas differences in sTNFR levels between early and established RA were assessed using the Mann–Whitney U-Test. Multiple regression analysis was used to assess the relationship between each sTNFR and age (corrected for sex and disease duration), and between the TNF receptor genotypes and sTNFR levels (corrected for age, sex and disease duration). Where necessary the data were normalized by logarithmic transformation before analysis. All data were analyzed using the Number Cruncher Statistical Software package for Windows (NCSS 2000, NCSS Statistical Software, Kaysville, Utah, USA) P < 0.05 were considered statistically significant.
Results
sTNFR levels in rheumatoid arthritis
Both sTNFRs were detected in the sera of all patients studied. Consistent with the findings reported by Cope and coworkers [16], levels of sTNF-RII were approximately three times greater on average than those of sTNF-RI. A strong positive correlation was observed between the levels of the two sTNFRs in both patient populations (Rs > 0.45; P < 0.0001). The levels of each sTNFR were also found to increase significantly with age in both patient groups (P < 0.0001), and this was independent of disease duration. Similar associations between sTNFR levels and age were seen in male and female patients (data not shown). Also, the median level of each sTNFR was significantly higher in patients with established disease than in those with early disease (P < 0.0001); this association remained after correction for age.
TNF-RI A36G single nucleotide polymorphism and sTNFR serum levels
The A36G SNP genotype frequencies in each population and the respective mean levels of both sTNFRs are shown in Table 2. The observed allele frequencies for the A and G alleles were 64.1% and 35.9%, respectively, in the early RA population and 55.0% and 45.0% in the established RA population. The allele and genotype frequencies are broadly comparable to those reported elsewhere [24,27,30], although there is a suggestion from this study that the GG genotype is more frequent in patients with established disease. There were no significant differences in the serum levels of either sTNFR between the three genotypes in patients with early disease or with established disease (Table 2). This finding was also observed when the two populations were combined and the analysis repeated (data not shown).
TNF-RII T676G single nucleotide polymorphism and sTNFR serum levels
The genotype and sTNFR level data for the T676G polymorphism in patients with early and established disease are shown in Table 3. The T and G alleles had frequencies of 77.2% and 22.8%, respectively, in both the early and established disease populations, and these frequencies were similar to those previously reported [24-29]. In established RA, analysis by multiple regression with correction for age, sex and disease duration revealed a significant association between TNF-RII genotype and the levels of sTNF-RI (P for trend = 0.01) and sTNF-RII (P for trend = 0.03) in the order TT > TG > GG. An identical trend was seen for levels of sTNF-RI and sTNF-RII in patients with early disease, although these associations were not significant (P = 0.3 and P = 0.055, respectively). In addition, the levels of sTNF-RI and sTNF-RII were significantly associated with TNF-RII genotype (P for trend = 0.02 and P for trend = 0.01, respectively) when the two populations were combined and analyzed by multiple regression with correction for age, sex and disease duration.
TNF receptor polymorphisms and sTNFR synovial fluid levels
Synovial fluids collected at the same time as sera were available in 45 patients. Mean levels of sTNF-RI and sTNF-RII in the synovial fluids were significantly higher (7,736 and 18,120 pg/ml, respectively) than in the paired sera, but there was no direct correlation between levels in the synovial fluid and serum. No association was found between synovial fluid sTNFR levels and the A36G TNF-RI or 676G TNF-RII genotypes (data not shown).
TNF receptor polymorphisms and clinical outcome measures
We showed previously that polymorphisms in the TNF-RI and TNF-RII genes were not associated with radiographic or functional severity in a cross-sectional study of patients with RA [27]. Similar findings were later reported by van der Helm-van Mil and coworkers [29], although another study by Constantin and colleagues [28] suggested an association of the TNF-RII G allele with worse functional (HAQ) outcome in early RA patients followed up for 5 years.
In the present study we again found no association between TNF-RI or TNF-RII polymorphisms and cross-sectional measures of radiographic or functional severity in patients with early or established disease (data not shown). In a similar manner to that reported by Constantin and coworkers [28], we also investigated the association between the TNF-RII polymorphism and functional severity of the early RA patients examined at baseline and at 5 years follow up. There was no significant difference in HAQ scores between patients with and those without the G allele at baseline (1.41 versus 1.60; P = 0.1) or after 5 years of follow up (1.41 versus 1.50; P = 0.9). There was also no significant difference in HAQ score progression.
Analysis of other clinical parameters associated with disease severity (extra-articular disease/nodules, rheumatoid factor, surgery, mechanical joint score, etc.) revealed no differences between TNF-RII genotypes (data not shown). However, in a separate study on anaemia in RA, involving many of these patients, we reported an association between carriage of the TNF-RII T allele and anaemia of chronic disease [35].
TNF receptor polymorphisms and levels of sTNFR released by isolated T cells
We investigated whether there was any association between polymorphism in the TNF receptor genes and levels of sTNFRs released into the culture medium of unstimulated and stimulated T cells from RA patients. No association was found between the TNF-RI A36G polymorphism and levels of sTNFRs released (data not shown). However, a significant trend was found in levels of sTNF-RII released into culture medium by both unstimulated and stimulated T cells according to the TNF-RII genotype in the order TT > TG > GG (unstimulated and stimulated, respectively:P for trend = 0.049 and P for trend = 0.02; Table 4). Similar trends for release of sTNF-RI were seen in unstimulated and stimulated T cells, although these did not achieve statistical significance.
Relationship between circulating levels of sTNFR and in vitro release from T cells
We examined whether the levels of sTNFRs in the circulation of RA patients were reflected in the levels of sTNFRs released by peripheral blood T cells in vitro. Strong correlations were found between serum levels of both sTNFRs and levels of these receptors released from isolated T cells (Table 5). The circulating levels of each sTNFR were strongly correlated with levels released by both unstimulated and stimulated T cells.
Discussion
We investigated whether SNPs in the TNF receptor genes are associated with circulating levels of the naturally occurring soluble form of these receptor molecules in patients with early and established RA. We report evidence of an association between the T676G polymorphism in TNF-RII and serum levels of both sTNF-RI and sTNF-RII in patients with established disease, with a trend toward decreasing levels across the genotypes in the order TT > TG > GG. An identical trend was observed in patients with early disease, although the data failed to reach statistical significance. There was no evidence of any association between the TNF-RI A36G polymorphism and the levels of either sTNFR, in early or established RA.
No association was found between TNF receptor genotypes and synovial fluid levels of sTNFRs. This is probably not surprising because no correlation was found between synovial fluid and serum levels of sTNFRs, which is consistent with previous data [17]. We suggest that the high levels of sTNFRs seen in synovial fluids reflect the high degree of local inflammation, where genetic regulation of sTNFR levels by the TNF-RII gene is likely to have less impact than other factors in such an inflammatory environment. In contrast, the levels seen in the circulation are more likely to reflect genetic regulation of sTNFR levels, because any genetic influence is less likely to be overwhelmed by inflammatory factors.
Our finding of an association between the TNF-RII polymorphism and circulating levels of sTNFRs is reinforced by our in vitro studies, which show an identical trend in the release of sTNFRs, according to genotype, by isolated T cells from RA patients. We also demonstrated that the levels of sTNFR released by T cells in vitro are very closely correlated with the levels of circulating sTNFRs in these patients. Release of sTNFRs was greatest in T-cell cultures from patients carrying the TNF-RII TT genotype. The same trend was seen both in unstimulated and PHA stimulated cells, although the association with TNF-RII genotype was significant only for levels of sTNF-RII.
We also measured sTNFR release by isolated monocytes in vitro and found a similar relationship between sTNFR levels and TNF-RII genotype in cells stimulated with or without lipopolysaccharide. However, this did not reach significance, and the correlation between TNF receptor levels released by monocytes and serum levels was weaker than for T cells (unpublished observations). In multiple regression analyses of serum TNF receptor levels, which included levels released by T cells and monocytes as independent variables, we found that only levels released by T cells were associated with serum levels (unpublished observations).
The association of the TNF-RII T676G polymorphism with circulating sTNF-RII levels is consistent with a previous study [36] that demonstrated higher levels of sTNF-RII in healthy individuals carrying a T allele. However, the association with sTNF-RI levels was unexpected because the two TNF receptor genes are encoded on separate chromosomes. It is not clear how polymorphism within the TNF-RII gene might influence levels of sTNF-RI, although the strong correlation between the levels of these soluble receptors indicates that their production and/or release are closely linked.
The T676G polymorphism in exon 6 of the TNF-RII gene occurs within the fourth cysteine-rich domain of the extracellular domain, close to a point where the proteolytic cleavage site for TACE is thought to lie [37]. The polymorphism results in a nonconservative amino acid substitution in which arginine, with a highly basic side chain, replaces methionine, which has a nonpolar side chain (methionine → arginine, M196R). The location and nature of this polymorphism suggests the possibility that processing of membrane bound TNF-RII by TACE might be affected. However, functional analysis of this polymorphism in TNF-RII transfected HeLa cells revealed no effects on the release of soluble receptors from the cell surface, nor any effect on physical binding parameters [38]. In contrast, our findings suggest that this polymorphism may play a role in the regulation of soluble receptor release in T cells. However, the possibility that the association is with another polymorphism in linkage disequilibrium cannot be ruled out.
Recently, the TNF-RII 196R variant was shown to have a significantly lower ability to induce direct NF-κB signalling via TNF-RII and to enhance TNF-RI dependent TNF-α induced apoptosis [39]. The diminished ability of the 196R variant to induce NF-κB activation is paralleled by a diminished induction of NF-κB dependent target genes involved in antiapoptotic or proinflammatory functions. It is possible that in certain cell types or under particular experimental conditions that the reduced ability of the 196R variant to induce NF-κB dependent genes may lead to reduced release of sTNFRs (e.g. through lower production of proteins that are important in regulating the cleavage and release of these receptors).
The mean serum levels of sTNF-RII were approximately three times greater than for sTNF-RI in each patient population. The levels of sTNF-RII in culture medium from unstimulated T cells were also about three times greater than those of sTNF-RI, although this increased to approximately five times greater after stimulation of T cells with PHA. This can be explained by increased release of sTNF-RII, but not of sTNF-RI, after PHA stimulation. This is similar to the situation previously observed in lipopolysaccharide stimulated monocytes and alveolar macrophages, in which sTNF-RII but not sTNF-RI release was enhanced after stimulation [40,41]. These differences in soluble receptor release may have important consequences for TNF-α signalling (e.g. increased release of sTNF-RII may reduce the ability of cells to be activated by interactions with membrane bound TNF-α on surrounding cells) [42]. Localized differences in the concentrations of soluble receptors may also have significant effects on inhibition or promotion of TNF-α activity, depending on the tissue compartment and the level of TNF-α present.
Previous studies in healthy individuals reported an age associated significant increase in serum levels of both sTNFRs [43,44], whereas a study conducted in RA patients failed to identify any correlation between sTNFR levels and age [16]. In another study conducted in healthy individuals [45] the levels of sTNF-RII were lower in older (50–67 years) than in younger individuals (25–35 years). In both populations studied here, we found a highly significant association between increasing levels of both sTNFRs and age, which was independent of disease duration. An explanation for these conflicting data is not yet evident.
The clinical relevance of our findings is unclear at present, but it has been shown that familial susceptibility to RA is associated with the TNF-RII G allele and particularly the GG genotype [24,25], which was associated with the lowest sTNF-RII levels in our study. It is possible that lower levels of sTNFRs may contribute to the development of RA if a particular threshold of TNF-α activity is exceeded in genetically susceptible individuals. Genetic regulation of TNF receptor levels may also influence the long-term outcome of the disease and response to anti-TNF-α therapy. There is some evidence that individuals carrying the TNF-RII G allele exhibit poorer response to anti-TNF-α therapy [46]. Constantin and coworkers [28] also suggested that the G allele is associated with worse functional outcome, based on 5 years of follow up of early RA patients, although we did not find an association in a previous cross-sectional study [27] and could not confirm the findings of those investigators in the present study. Therefore, the association of the TNF-RII T676G polymorphism with functional severity is uncertain. However, we have provided recent evidence that the T allele (associated with higher sTNFR serum levels and increased release from T cells in the present study) may be associated with anaemia of chronic disease in RA [35]. Compared with nonanaemic patients, those with anaemia of chronic disease also have serum levels of sTNF-RI and sTNF-RII that are about 30% higher (P ≤ 0.007), which is consistent with the T allele association.
Although the differences in sTNFR serum levels between TNF-RII genotypes are not large, it is noteworthy that exactly the same trend was seen throughout for serum levels in early and established RA, and for unstimulated and stimulated T-cell cultures. Several studies have shown that circulating sTNFR levels and/or polymorphisms in the TNF-RII gene are associated with heart failure, hypertension, obesity and insulin resistance, and differences in serum levels of a similar magnitude to those found in this study were shown to be clinically relevant in these conditions [47-52]. Our findings may thus be of particular importance in RA, in which there is evidence of increased risk for cardiovascular disease and metabolic syndrome abnormalities [53].
Conclusion
Our results indicate that there is an association between the T676G SNP in the TNF-RII gene and levels of sTNFRs released by T cells of RA patients. This finding is reinforced by an association between this polymorphism and circulating levels of sTNFRs in established RA. Although various inflammatory factors may influence the release of TNF receptors, our data indicate that genetic regulation involving the TNF-RII gene may play some role in determining circulating levels in RA.
Abbreviations
ELISA = enzyme-linked immunosorbent assay; HAQ = health assessment questionnaire; NF-κB = nuclear factor-κB; PCR = polymerase chain reaction; PHA = phytohaemagglutinin; RA = rheumatoid arthritis; SNP = single nucleotide polymorphism; sTNFR = soluble tumour necrosis factor receptor; TACE = TNF-α converting enzyme; TNF = tumour necrosis factor; TNF-RI = tumour necrosis factor receptor I; TNF-RII = tumour necrosis factor receptor II.
Competing interests
The author(s) declare that they have no competing interests.
Authors' contributions
JRG carried out genotyping, cell isolation, cell culture, and ELISA work, and wrote the first draft of the paper. PTD gave advice on patient selection, study design and interpretation of data. NBN carried out some genotyping and ELISA work. DLM conceived and oversaw the study, carried out statistical analyses and interpretation of data, and finalized the manuscript. The final manuscript was read and approved by all authors.
Acknowledgements
This study was supported by the Haywood Rheumatism Research and Development Foundation.
Figures and Tables
Table 1 Characteristics of the two rheumatoid arthritis patient populations
Population Early disease Established disease
Number 103 151
Male/Female (n) 44/59 63/88
Age (years; mean ± SD) 55.0 ± 13.4 60.2 ± 11.1
Age at onset (years; mean ± SD) 54.3 ± 13.4 46.6 ± 12.1
Disease duration (years; mean ± SD) 0.7 ± 0.5 13.7 ± 6.3
Rheumatoid factor ever positive (%) 52/92 (56.5) 86/126 (68.3)
Nodule positive (%) 4/102 (3.9) 28/150 (18.7)
SD, standard deviation.
Table 2 TNF-RI A36G single nucleotide polymorphism genotype frequencies and sTNFR levels
Genotype n (%) sTNF-RI (pg/ml) sTNF-RII (pg/ml)
Early RA
AA 41 (39.8) 1,543 ± 597 4,435 ± 1,898
AG 50 (48.5) 1,426 ± 629 4,302 ± 1,672
GG 12 (11.7) 1,303 ± 447 4,566 ± 1,490
Established RA
AA 48 (31.8) 1,827 ± 758 5,740 ± 1,942
AG 70 (46.4) 1,688 ± 674 5,475 ± 2,020
GG 33 (21.8) 1,757 ± 559 5,857 ± 2,393
Shown are tumor necrosis factor receptor I (TNF-RI) A36G single nucleotide polymorphism genotype frequencies and soluble tumor necrosis factor receptor (sTNFR) levels in rheumatoid arthritis (RA) patients with early (n = 103) and established (n = 151) disease. sTNFR levels are expressed as the mean ± standard deviation. No significant differences in sTNFR levels were found between any of the genotypes in either population. sTNF-RII, soluble tumor necrosis factor receptor II.
Table 3 TNF-RII T676G single nucleotide polymorphism genotype frequencies and sTNFR levels
Genotype n (%) sTNF-RI (pg/ml) sTNF-RII (pg/ml)
Early RA
TT 63 (61.2) 1,503 ± 704 4,690 ± 1,961
TG 33 (32.0) 1,451 ± 370 3,961 ± 1,242
GG 7 (6.8) 1,094 ± 240 3,648 ± 697
Established RA
TT 91 (60.3) 1,816 ± 705 5,837 ± 2,219
TG 51 (33.7) 1,633 ± 642 5,375 ± 1,921
GG 9 (6.0) 1,700 ± 570 5,187 ± 1,066
Shown are tumour necrosis factor receptor II (TNF-RII) T676G single nucleotide polymorphism (SNP) genotype frequencies and soluble tumour necrosis factor receptor (sTNFR) levels in rheumatoid arthritis (RA) patients with early (n = 103) and established (n = 151) disease. sTNFR levels are expressed as the mean ± standard deviation. Multiple regression analyses of log transformed data corrected for age, sex and disease duration revealed a significant trend of decreasing soluble tumour necrosis factor receptor I (sTNF-RI) and sTNF-RII levels across the genotypes (order: TT > TG > GG) of patients with established disease (P for trend = 0.01 and P for trend = 0.03, respectively). A similar nonsignificant trend was seen for patients with early disease (P = 0.3 and P = 0.055, respectively).
Table 4 Association between TNF-RII T676G single nucleotide polymorphism genotype and sTNFR levels released by T cells
Genotype n (%) sTNF-RI (pg/ml) sTNF-RII (pg/ml)
Unstimulated T cells
TT 38 (65.5) 166.8 ± 57.8 582.2 ± 259.6
TG 15 (25.9) 144.1 ± 78.2 428.1 ± 222.3
GG 5 (8.6) 137.2 ± 68.9 398.8 ± 194.9
Stimulated T cells
TT 38 (65.5) 178.0 ± 57.9 998.3 ± 355.6
TG 15 (25.9) 146.5 ± 75.5 769.5 ± 292.8
GG 5 (8.6) 141.6 ± 75.7 724.4 ± 167.3
Shown is the association between tumour necrosis factor receptor II (TNF-RII) T676G single nucleotide polymorphism genotype and soluble tumour necrosis factor (sTNFR) levels released by T cells isolated from rheumatoid arthritis (RA) patients (n = 58). sTNFR levels are expressed as mean ± standard deviation. Levels of sTNFR released into culture medium of isolated T cells exhibited a similar trend of decreasing levels of both receptors according to TNF-RII genotype (order: TT > TG > GG), although only the associations with sTNF-RII were significant (unstimulated and stimulated cells, respectively: P for trend = 0.049 and P for trend = 0.02; multiple regression analysis corrected for age, sex and disease duration). sTNF-RI, soluble tumor necrosis factor receptor I.
Table 5 Correlation between serum levels of sTNFR and levels released by isolated T cells
Serum levels Unstimulated T cells Stimulated T cells
sTNF-RI sTNF-RII sTNF-RI sTNF-RII
sTNF-RI 0.883 0.818 0.858 0.692
sTNF-RII 0.865 0.923 0.837 0.763
Shown is the correlation between serum levels of soluble tumour necrosis factor receptor (sTNFR) and levels released by isolated T cells from rheumatoid arthritis (RA) patients (n = 58). Spearman correlation coefficients are shown. P < 0.0001 for all correlations. sTNF-RI, soluble tumor necrosis factor receptor I; sTNF-RII, soluble tumor necrosis factor receptor II.
==== Refs
Brennan FM Maini RN Feldmann M TNFα: a pivotal role in rheumatoid arthritis? Br J Rheumatol 1992 31 293 298 1581770
Choy EHS Panayi GS Cytokine pathways and joint inflammation in rheumatoid arthritis N Engl J Med 2001 344 907 916 11259725 10.1056/NEJM200103223441207
Hohmann HP Remy R Brockhaus M van Loon AP Two different cell types have different major receptors for human tumor necrosis factor (TNF alpha) J Biol Chem 1989 264 14927 14934 2549042
Brockhaus M Schoenfeld H Schlaeger E Hunziker W Lesslauer W Loetscher H Identification of two types of tumor necrosis factor receptors on human cell lines by monoclonal antibodies Proc Natl Acad Sci USA 1990 87 3127 3131 2158104
Dembic Z Loetscher H Gubler U Pan YC Lahm HW Gentz R Brockhaus M Lesslaur W Two human TNF receptors have similar extracellular, but distinct intracellular, domain sequences Cytokine 1990 2 231 137 1966549 10.1016/1043-4666(90)90022-L
Tartaglia LA Ayres TM Wong GH Goeddel DV A novel domain within the 55kd TNF receptor signals cell death Cell 1993 74 845 853 8397073 10.1016/0092-8674(93)90464-2
Tartaglia LA Weber RF Figari IS Reynolds C Palladino MA JrGoeddel DV The two different receptors for tumor necrosis factor mediate distinct cellular responses Proc Natl Acad Sci USA 1991 88 9292 9296 1718003
Hohmann HP Brockhaus M Baeuerle PA Remy R Kolbeck R van Loon AP Expression of the types A and B tumor necrosis factor (TNF) receptors is independently regulated, and both receptors mediate activation of the transcription factor NF-kappa B. TNF alpha is not needed for induction of a biological effect via TNF receptors J Biol Chem 1990 265 22409 417 2176217
Tartaglia LA Rothe M Hu YF Goeddel DV Tumor necrosis factor's cytotoxic activity is signaled by the p55 TNF receptor Cell 1993 73 213 216 8386591 10.1016/0092-8674(93)90222-C
Heller RA Song K Fan N Chang DJ The p70 tumor necrosis factor receptor mediates cytotoxicity Cell 1992 70 47 56 1320461 10.1016/0092-8674(92)90532-H
Engelmann H Novick D Wallach D Two tumor necrosis factor-binding proteins purified from human urine: evidence for immunological cross-reactivity with cell surface tumor necrosis factor receptors J Biol Chem 1990 265 1531 1536 2153136
Nophar Y Kemper O Brake busch C Engelmann H Zwang R Aderka D Holtmann H Wallach D Soluble forms of tumor necrosis factor receptors (TNF-Rs). The cDNA for the type I TNF-R, cloned using amino acid sequence data of its soluble form, encodes both the cell surface and a soluble form of the receptor EMBO J 1990 9 3269 3278 1698610
Kohno T Brewer MT Baker SL Schwartz PE King MW Hale KK Squires CH Thompson RC Venice JL A second tumor necrosis factor receptor gene product can shed a naturally occurring tumor necrosis factor inhibitor Proc Natl Acad Sci USA 1990 87 8331 8335 2172983
Reddy P Slack JL Davis R Cerretti DP Kozlosky CJ Blanton RA Shows D Peschon JJ Black RA Functional analysis of the domain structure of tumor necrosis factor-α converting enzyme J Biol Chem 2000 275 14608 14614 10799547 10.1074/jbc.275.19.14608
Barrera P Boerbooms ANTH Janssen EM Sauerwein RW Gallati H Mulder J de Boo T Demacker PN van de Putte LB van der Meer JW Circulating soluble tumor necrosis factor receptors, interleukin-2 receptors, tumor necrosis factor α, and interleukin-6 levels in rheumatoid arthritis: longitudinal evaluation during methotrexate and azathioprine therapy Arthritis Rheum 1993 36 1070 1079 8393676
Cope AP Aderka D Doherty M Engelmann H Gibbons D Jones AC Brennan FM Maini RM Wallach D Feldmann M Increased levels of soluble tumor necrosis factor receptors in the sera and synovial fluid of patients with rheumatic diseases Arthritis Rheum 1992 35 1160 1169 1329774
Steiner G Studnicka-Benke A Witzmann G Höfler E Smolen J Soluble receptors for tumor necrosis factor and interleukin-2 in serum and synovial fluid of patients with rheumatoid arthritis, reactive arthritis and osteoarthritis J Rheumatol 1995 22 406 412 7783054
Deleuran BW Chu C-Q Field M Brennan FM Mitchell T Feldmann M Maini RN Localization of tumor necrosis factor receptors in the synovial tissue and cartilage-pannus junction in patients with rheumatoid arthritis: Implications for local actions of tumor necrosis factor α Arthritis Rheum 1992 35 1170 1178 1329775
Brennan FM Gibbons DL Mitchell T Cope AP Maini RN Feldmann M Enhanced expression of TNF receptor mRNA and protein in mononuclear cells isolated from rheumatoid arthritis synovial joints Eur J Immunol 1992 22 1907 1912 1320571
Baker E Chen LZ Smith CA Callen DF Goodwin R Sutherland GR Chromosomal location of the human tumor necrosis factor receptor genes Cytogenet Cell Genet 1991 57 117 118 1655358
Pitts SA Olomolaiye OO Elson CJ Westacott CI Bidwell JL An MspA1 I polymorphism in exon 1 of the human TNF receptor type I (p55) gene Eur J Immunogenet 1998 25 269 270 9615917 10.1046/j.1365-2370.1998.00100.x
Pitts SA Olomolaiye OO Elson CJ Westacott CI Bidwell JL Identification of a rare BglII polymorphism in the promoter region of the human TNF receptor type I (p55) gene Eur J Immunogenet 1998 25 271 272 9615918 10.1046/j.1365-2370.1998.00111.x
Pantelidis P Lympany PA Foley PJ Fanning GC Welsh KI du Bois RM Polymorphic analysis of the high-affinity tumor necrosis factor receptor 2 Tissue Antigens 1999 54 585 591 10674973 10.1034/j.1399-0039.1999.540608.x
Barton A John S Ollier WER Silman A Worthington J Association between rheumatoid arthritis and polymorphism of tumor necrosis factor receptor II, but not tumor necrosis factor receptor I, in caucasians Arthritis Rheum 2001 44 61 65 11212177 10.1002/1529-0131(200101)44:1<61::AID-ANR9>3.0.CO;2-Q
Dieudé P Petit E Cailleau-Moindrault S Osorio J Pierlot C Martinez M Faure S Alibert O Lasbleiz S De Toma C Association between tumor necrosis factor receptor II and familial, but not sporadic, rheumatoid arthritis: Evidence for genetic heterogeneity Arthritis Rheum 2002 46 2039 2044 12209506 10.1002/art.10101
Bridges SL JrJenq G Moran M Kuffner T Whitworth WC McNicholl J Single-nucleotide polymorphisms in tumor necrosis factor receptor genes: Definition of novel haplotypes and racial/ethnic differences Arthritis Rheum 2002 46 2045 2050 12209507 10.1002/art.10463
Glossop JR Nixon NB Dawes PT Hassell AB Mattey DL No association of polymorphisms in the tumor necrosis factor receptor I and receptor II genes with disease severity in rheumatoid arthritis J Rheumatol 2003 30 1406 1409 12858434
Constantin A Dieude P Lauwers-Cances V Jamard B Mazieres B Cambon-Thomsen A Cornelis F Cantagrel A Tumor necrosis factor receptor II gene polymorphism and severity of rheumatoid arthritis Arthritis Rheum 2004 50 742 747 15022314 10.1002/art.20113
van der Helm-van Mil AH Dieude P Schonkeren JJ Cornelis F Huizinga TW No association between tumour necrosis factor receptor type 2 gene polymorphism and rheumatoid arthritis severity: a comparison of the extremes of phenotypes Rheumatology 2004 43 1232 1234 15252214 10.1093/rheumatology/keh314
Dieude P Osorio J Petit-Teixeira E Moreno S Garnier S Cailleau-Moindrault S Stalens C Lasbleiz S Bardin T Prum B European Consortium on Rheumatoid Arthritis Families A TNFR1 genotype with a protective role in familial rheumatoid arthritis Arthritis Rheum 2004 50 413 419 14872483 10.1002/art.20055
Arnett FC Edworthy SM Bloch DA McShane DJ Fries JF Cooper NS Healey LA Kaplan SR Liang MH Luthra HS The AmericanRheumatism Association 1987 revised criteria for the classification of rheumatoid arthritis Arthritis Rheum 1988 31 315 324 3358796
Larsen A Dale K Eck M Radiographic evaluation of rheumatoid arthritis and related conditions by standard reference films Acta Radiol (Diagn Stockh) 1977 18 481 91 920239
Fries JF Spitz P Kraines RG Holman HR Measurement of patient outcome in arthritis Arthritis Rheum 1980 23 137 45 7362664
Komata T Tsuchiya N Matsushita M Hagiwara K Tokunaga K Association of tumor necrosis factor receptor 2 (TNFR2) polymorphism with susceptibility to systemic lupus erythrematosus Tissue Antigens 1999 53 527 533 10395102 10.1034/j.1399-0039.1999.530602.x
Glossop JR Dawes PT Hassell AB Mattey DL Anemia in rheumatoid arthritis. Association with polymorphism in the tumor necrosis factor I and II genes J Rheumatol 2005
Stark GL Dickinson AM Jackson GH Taylor PR Proctor SJ Middleton PG Tumour necrosis factor receptor type II 196M/R genotype correlates with circulating soluble receptor levels in normal subjects and with graft-versus-host disease after sibling allogeneic bone marrow transplantation Transplantation 2003 76 1742 1749 14688526 10.1097/01.TP.0000092496.05951.D5
Herman C Chernajovsky Y Mutation of proline 211 reduces shedding of the human p75 TNF receptor J Immunol 1998 160 2478 2487 9498793
Morita C Horiuchi T Tsukamoto H Hatta N Kikuchi Y Arinobu Y Otsuka T Sawabe T Harashima S Nagasawa K Association of tumor necrosis factor receptor type II polymorphism 196R with systemic lupus erythematosus in the Japanese: Molecular and functional analysis Arthritis Rheum 2001 44 2819 2827 11762942 10.1002/1529-0131(200112)44:12<2819::AID-ART469>3.0.CO;2-2
Till A Rosenstiel P Krippner-Heidenreich A Mascheretti-Croucher S Croucher PJ Schafer H Scheurich P Seegert D Schreiber S The Met196Arg variation of human TNFR2 affects TNF-alpha-induced apoptosis by impaired NF-κB-signalling and target gene expression J Biol Chem 2005 280 5994 6004 15572357 10.1074/jbc.M411541200
Leeuwenberg JF Dentener MA Buurman WA Lipopolysaccharide LPS-mediated soluble TNF receptor release and TNF receptor expression by monocytes. Role of CD14, LPS binding protein, and bactericidal/permeability-increasing protein J Immunol 1994 152 5070 5076 7513728
Galve-de Rochemonteix B Nicod LP Dayer JM Tumor necrosis factor soluble receptor 75: the principal receptor form released by human alveolar macrophages and monocyes in the presence of interferon gamma Am J Respir Cell Mol Biol 1996 14 279 287 8845179
Grell M Douni E Wajant H Lohden M Clauss M Maxeiner B Georgopoulos S Lesslaur W Kollias G Pfizenmaier K The transmembrane form of tumour necrosis factor is the prime activating ligand of the 80 kDa tumour necrosis receptor Cell 1995 83 793 802 8521496 10.1016/0092-8674(95)90192-2
Gerli R Monti D Bistoni O Mazzone AM Peri G Cossarizza A Di Gioacchino M Cesarotti ME Doni A Mantovani A Chemokines, sTNF-Rs and sCD30 serum levels in healthy aged people and centenarians Mech Ageing Dev 2000 121 37 46 11164458 10.1016/S0047-6374(00)00195-0
Hasegawa Y Sawada M Ozaki N Inagaki T Suzumura A Increased soluble tumor necrosis factor receptor levels in the serum of elderly people Gerontology 2000 46 185 188 10859456 10.1159/000022157
Albright JW Albright JF Soluble receptors and other substances that regulate proinflammatory cytokines in young and aging humans J Gerontol A Biol Sci Med Sci 2000 55 B20 B25 10719759
Fabris M Tolusso B Di Poi E Assaloni R Sinigaglia L Ferraccioli G Tumor necrosis factor-α receptor II polymorphism in patients from southern Europe with mild-moderate and severe rheumatoid arthritis J Rheumatol 2002 29 1847 1850 12233877
Torre-Amione G Kapadia S Lee J Durand J-B Bies RD Young JB Mann DL Tumor necrosis factor-α and tumor necrosis factor receptors in the failing human heart Circulation 1996 93 704 711 8640999
Benjafield AV Wang XL Morris BJ Tumor necrosis factor receptor 2 gene (TNFRSF1B) in genetic basis of coronary artery disease J Mol Med 2001 79 109 115 11357933 10.1007/s001090000168
Glenn CL Wang WYS Benjafield AV Morris BJ Linkage and association of tumor necrosis factor receptor 2 locus with hypertension, hypercholesterolemia and plasma shed receptor Hum Mol Genet 2000 9 1943 1949 10942422 10.1093/hmg/9.13.1943
Dahlqvist SR Arlestig L Sikstrom C Linghult S Tumor necrosis receptor type II (exon 6) and interleukin-6 (-174) gene polymorphisms are not associated with family history but tumor necrosis factor receptor type II is associated with hypertension in patients with rheumatoid arthritis from northern Sweden Arthritis Rheum 2002 46 3096 3098 12428254 10.1002/art.10592
Fernandez-Real J-M Vendrell J Ricart W Broch M Gutierrez C Casamitjana R Oriola J Richart C Polymorphism of the tumor necrosis factor-α receptor 2 gene is associated with obesity, leptin levels and insulin resistance in young subjects and diet-treated type 2 diabetic patients Diabetes Care 2000 23 831 837 10841005
Fernandez-Real J-M Lainez B Vendrell J Rigla M Castro A Penrroja G Broch M Perez A Richart C Engel P Shedding of TNF-α receptors, blood pressure, and insulin sensitivity in type 2 diabetes mellitus Am J Physiol Endocrinol Metab 2002 282 E952 E959 11882518
Sattar N McCarey DW Capell H McInnes IB Explaining how 'high-grade' systemic inflammation accelerates vascular risk in rheumatoid arthritis Circulation 2003 108 2957 2963 14676136 10.1161/01.CIR.0000099844.31524.05
|
16277675
|
PMC1297570
|
CC BY
|
2021-01-04 16:02:40
|
no
|
Arthritis Res Ther. 2005 Sep 7; 7(6):R1227-R1234
|
utf-8
|
Arthritis Res Ther
| 2,005 |
10.1186/ar1816
|
oa_comm
|
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.