INTRODUCTION
Homo sapiens is certainly the most investigated species regarding its infectious diseases. More than 1,400 species are listed as human pathogens (Cleaveland et al. Reference Cleaveland, Laurenson and Taylor2001; Woolhouse and Gowtage-Sequeria, Reference Woolhouse and Gowtage-Sequeria2005) and at least 60 per cent of these are zonootic (Taylor et al. Reference Taylor, Latham and Woolhouse2001). Documenting and understanding ecological, historical and biogeographical associations between humans and parasites has been the subject of numerous studies (May, Reference May1958; Cockburn, Reference Cockburn and Cockburn1967; McNeil, Reference McNeil1976; Dobson and Carper, Reference Dobson and Carper1996; Guernier et al. Reference Guernier, Hochberg and Guégan2004; Wolfe et al. Reference Wolfe, Panosian Dunavan and Diamond2007), in which it has been emphasized that humans have gained their parasites either through descent (i.e. inherited from a common ancestor) or by acquiring from either wild or domesticated animal species in sympatry.
Co-phylogenetic studies such as between primates and Pneumocystis spp. (Hugot et al. Reference Hugot, Demanche, Barriel, Dei-Cas and Guillot2003), lice (Reed et al. Reference Reed, Light, Allen and Kirchman2007) or viruses (Switzer et al. Reference Switzer, Salemi, Shanmugam, Gao, Cong, Kuiken, Bhullar, Beer, Vallet, Gautier-Hion, Tooze, Villinger, Holmes and Heneine2005) have provided examples showing that parasites may have been inherited by descent from the common ancestors of Homo sapiens and/or close relatives. Moreover, Davies and Pedersen (Reference Davies and Pedersen2008) showed that infectious diseases are more often shared between pairs of primate species, including humans, that are phylogenetically related but also that live in the same geographical region. However, most parasite species infecting modern humans have come from domestic and wild animals through hunting or domestication (Cleaveland et al. Reference Cleaveland, Laurenson and Taylor2001; Weiss, Reference Weiss2001; Woolhouse and Gowtage-Sequeria, Reference Woolhouse and Gowtage-Sequeria2005; Perrin et al. Reference Perrin, Herbreteau, Hugot, Morand, Morand and Krasnov2010).
Perrin et al. (Reference Perrin, Herbreteau, Hugot, Morand, Morand and Krasnov2010) using the checklist of Ashford and Crew (Reference Ashford and Crewe1998) of 402 parasite species (helminths, arthropods and protozoans) in humans showed that carnivores are the most likely to share their parasite species with humans (124 parasite species – 31%), followed by ruminants and pigs (83 species – 21%), rodents (66 species – 17%) and horses and other equids (nine parasite species – 2%). Humans and canids cohabited long before dogs became domesticated, which occurred around 15,000 years ago (i.e. before other mammals). Ruminants and pigs are considered among the oldest domesticated groups after dogs, from about 6,000 to 8,000 years ago. The high number of parasites shared with rodents is explained by their being used for meat and by a long history of several species living as human commensals, feeding on food storage, waste and rubbish. Horses were domesticated between 6,000 years and 3,000 years later than dogs and cattle (Horwitz and Smith, Reference Horwitz and Smith2000). Interestingly, elephants are not known to have donated any parasites and rarely donate zoonoses, maybe because they lived with humans at low abundance (Wolfe et al. Reference Wolfe, Panosian Dunavan and Diamond2007).
These first observations strongly suggest that interpreting geographic origins and dispersion of shared human-animal parasites needs to take into account the domestication process. Archaeological studies suggest large-scale domestication of plants and animals between 10,000–7,000 cal years BP (Gupta, Reference Gupta2004). Indeed, the number of parasite species shared between domesticated animals and humans is positively related to time since domestication (Horwitz and Smith, Reference Horwitz and Smith2000), a pattern already observed by McNeil (Reference McNeil1976) (see Perrin et al. Reference Perrin, Herbreteau, Hugot, Morand, Morand and Krasnov2010).
The potential sources of plants and animals suitable for domestication seem not to be randomly distributed on continents, and it appears that the majority of domestic animals originated from the Middle East, Central, Southwest, and Southern Asia (Diamond, Reference Diamond1997; Gupta, Reference Gupta2004; Larson et al. Reference Larson, Dobney, Albarella, Fang, Matisoo-Smith, Robins, Lowden, Finlayson, Brand, Willerslev, Rowley- Conwy, Andersson and Cooper2005; Driscoll et al. Reference Driscoll, Menotti-Raymond, Roca, Hupe, Johnson, Geffen, Harley, Delibes, Pontier, Kitchener, Yamaguchi, O'Brien and Macdonald2007; Naderi et al. Reference Naderi, Rezaei, Pompanon, Blum, Negrini, Naghash, Balkız, Mashkour, Gaggiotti, Ajmone-Marsan, Kence, Vigne and Taberlet2008). However, eight of the 15 temperate diseases investigated in the Wolfe et al. (Reference Wolfe, Panosian Dunavan and Diamond2007) study reached humans from domestic animals (diphtheria, influenza A, measles, mumps, pertussis, rotavirus, smallpox and tuberculosis) and only three of the ten tropical diseases investigated originated from animal domestication, which questions the potential dispersion of zoonotic diseases.
Here, my aim is to summarize the recent advances in the phylogeographics and phylogenetics of human pathogens and to show how these studies contribute to our understanding of the building of human parasite communities. In particular, three major origins and distributions of human parasites have been tested using phylogeographic studies: (1) the “out of Africa” pattern where parasites followed the dispersal and expansion of modern humans in and out of Africa, (2) the “domestication” pattern where parasites were captured in the domestication centres and then dispersed more widely and (3) the “globalization” pattern, which reflects the distribution of parasites in relation to historical and more recent trade routes. The classification of these three mechanisms does not preclude the notion that they are not mutually exclusive as some examples reported here may show.
THE PHYLOGEOGRAPHY OF PARASITES
Even if important progress has been made in documenting the geographic distribution of parasites and pathogens, their origins and dispersion are less well known (Morand and Krasnov, Reference Morand and Krasnov2010). Phylogeographic approaches originate in historical biogeography (Katinas et al. Reference Katinas, Posadas and Crisci2003) in an attempt to take account of invasive processes and the mechanisms of geographic colonization (Avise, Reference Avise2000), with earlier studies related to post-glacial re-colonization in Europe. The development of phylogenetic studies may then permit exploration of these origins and dispersion of hosts and their parasites and microbes on fine geographic and temporal scales (Nieberding et al. Reference Nieberding, Morand, Libois and Michaux2004; Nieberding and Olivieri, Reference Nieberding and Olivieri2007).
The phylogeographic method
As evolutionary processes take place in a dynamic geographical context, patterns of genetic variation are strongly structured in space and time (Hewitt, Reference Hewitt2001). Phylogeography investigates the processes governing the geographical distributions of lineages within and among closely related species (Avise, Reference Avise2000). Phylogeographic studies of pathogens and parasites are more recent than phylogenetic studies (Holmes, Reference Holmes2004) and most of them argue that phylogeographic studies of parasites may help explain the phylogeography of their hosts (Nieberding et al. Reference Nieberding, Morand, Libois and Michaux2004; Wirth et al. Reference Wirth, Meyer and Achtman2005; Nieberding and Olivieri, Reference Nieberding and Olivieri2007). Phylogeographic studies may also help identify timing, rate and origins of parasite emergence as exemplified by Borrelia burgdorferi the causative agent of Lyme disease (Gatewood Hoen et al. Reference Gatewood Hoen, Margos, Bent, Diuk-Wasser, Barbour, Kurtenbach and Fish2009).
The main distinction between phylogeographic and phylogenetic analyses is the addition of spatial information so that the inferences from the former allow testing of hypotheses concerning the association of clades (branches) of phylogenetic trees (haplotypes) according to spatial distribution. There has been substantial improvement in statistical inference methods from nested clade analysis (see Templeton, Reference Templeton2010) to the more recent use of approximate Bayesian computation (ABC) (Knowles, Reference Knowles2009; Beaumont et al. Reference Beaumont, Nielsen, Robert, Hey, Gaggiotti, Knowles, Estoup, Panchal, Corander, Hickerson, Sisson, Fagundes, Chikhi, Beerli, Vitalis, Cornuet, Huelsenbeck, Foll, Yang, Rousset, Balding and Excoffier2010). A promising new investigative approach has been developed by Lemey et al. (Reference Lemey, Rambaut, Drummond and Suchard2009), who used Bayesian modelling of character evolution, here geographical locations, for the inference of ancestral states, in this case the geographical location of ancestral nodes and migration events.
Rather like co-phylogenetic studies, co-phylogeographic studies aim to compare phylogeographic trees obtained for both hosts and their parasites in a spatial context, while aiming to use similar statistical methods based on null hypotheses (Nieberding et al. Reference Nieberding, Durette-Desset, Vanderpoorten, Casanovas, Ribas, Deffontaine, Feliu, Morand, Libois and Michaux2008). Quantitative tests of mechanisms that may generate co-phylogeographic scenarios, however, are still in their infancy (but see Nieberding et al. Reference Nieberding, Jousselin, Desdevises, Morand and Krasnov2010).
“Out of Africa” pattern
There is a consensus that modern humans left Africa up to 150,000 years ago (from 60,000–150,000 years before present, ybp; Cann et al. Reference Cann, Toma, Cazes, Legrand, Morel, Piouffre, Bodmer, Bodmer, Bonne-Tamir, Cambon-Thomsen, Chen, Chu, Carcassi, Contu, Du, Excoffier, Ferrara, Friedlaender, Groot, Gurwitz, Jenkins, Herrera, Huang, Kidd, Kidd, Langaney, Lin, Mehdi, Parham, Piazza, Pistillo, Qian, Shu, Xu, Zhu, Weber, Greely, Feldman, Thomas, Dausset and Cavalli-Sforza2002). Humans dispersed out of Africa toward the Middle East ∼60,000–150,000 ybp and then independently to Europe and Asia (Cavalli-Sforza et al. Reference Criscione, Anderson, Sudimack, Peng, Jha, Williams-Blangero and Anderson1994), and probably in two major waves to Asia (Rasmussen et al. Reference Rasmussen, Guo, Wang, Lohmueller, Rasmussen, Albrechtsen, Skotte, Lindgreen, Metspalu, Jombart, Kivisild, Zhai, Eriksson, Manica, Orlando, De La Vega, Tridico, Metspalu, Nielsen, Ávila-Arcos, Moreno-Mayar, Muller, Dortch, Gilbert, Lund, Wesolowska, Karmin, Weinert, Wang, Li, Tai, Xiao, Hanihara, van Driem, Jha, Ricaut, de Knijff, Migliano, Gallego Romero, Kristiansen, Lambert, Brunak, Forster, Brinkmann, Nehlich, Bunce, Richards, Gupta, Bustamante, Krogh, Foley, Lahr, Balloux, Sicheritz-Pontén, Villems, Nielsen, Wang and Willerslev2011). Dispersals to the Americas occurred around 14,000 ybp by peoples of East Asian ancestry who crossed the Bering Straits in two major migrations (Schurr and Sherry, Reference Schurr and Sherry2004). Finally, it is known that the islands of the Western Pacific were populated by people who originated in Taiwan around 5,500 ybp (Gray et al. Reference Gray, Drummond and Greenhill2009).
Human microbes and parasites are used as markers of these human dipsersals (Dominguez-Bello and Blaser, Reference Dominguez-Bello and Blaser2011). A variety of human pathogens such as Haemophilus influenzae (Musser et al. Reference Musser, Kroll, Granoff, Richard Moxon, Brodeur, Campos, Dabernat, Frederiksen, Hamel, Hammond, Høiby, Jonsdottir, Kabeer, Kallings, Khan, Kilian, Knowles, Koornhof, Law, Li, Montgomery, Pattison, Piffaretti, Takala, Len Thong, Wall, Ward and Selander1990), human polyomavirus JCV (Agostini et al. Reference Agostini, Yanagihara, Davis, Ryschkewitsch and Stoner1997; Sugimito et al. Reference Sugimoto, Hasegawa, Kato, Zheng, Ebihara, Taguchi, Kitamura and Yogo2002; Zheng et al. Reference Zheng, Sugimoto, Hasegawa, Kobayashi, Kanayama, Rodas, Mejia, Nakamichi, Guo, Kitamura and Yogo2003), the human T cell lymphotropic virus I (HTLV-I) (Miura et al. Reference Miura, Fukunaga, Igarashi, Yamashita, Ido, Funahashi, Ishida, Washio, Ueda and Hashimoto1994), Mycobacterium tuberculosis (Kremer et al. Reference Kremer, van Soolingen, Frothingham, Haas, Hermans, Martín, Palittapongarnpim, Plikaytis, Riley, Yakrus, Musser and van Embden1999), Mycobacterium leprae (Monot et al. Reference Monot, Honore, Garnier, Araoz, Coppee, Lacroix, Sow, Spencer, Truman, Williams, Gelber, Virmond, Flageul, Cho, Ji, Paniz-Mondolfi, Convit, Young, Fine, Rasolofo, Brennan and Cole2005) the human pathogenic fungus Histoplasma capsulatum (Kasuga et al. Reference Kasuga, Taylor and White1999), Streptococcus mutans (Caufield et al. Reference Caufield, Saxena, Fitch and Li2007) and Helicobacter pylori (Ghose et al. Reference Ghose, Perez-Perez, Dominguez-Bello, Pride, Bravi and Blaser2002; Falush et al. Reference Falush, Wirth, Linz, Pritchard, Stephens, Kidd, Blaser, Graham, Vacher, Perez-Perez, Yamaoka, Mégraud, Otto, Reichard, Katzowitsch, Wang, Achtman and Suerbaum2003; Wirth et al. Reference Wirth, Wang, Linz, Novick, Lum, Blaser, Morelli, Falush and Achtman2004) all show geographic structures. These parasites are hypothesized to have accompanied humans during their ancient and recent dispersals, and investigation of their population structures may help us to understand human evolutionary history (Wirth et al. Reference Wirth, Meyer and Achtman2005). For this purpose bacteria and viruses have been the most intensively investigated organisms(Table 1).
Table 1. Phylogeographic studies of microbes and parasites of humans
MLST (Multilocus sequence typing) consists of sequencing several housekeeping gene fragments for a total concatenated length of 3–4 kb.
SNP: single nucleotide polymorphism.
* large sequence polymorphisms.
Helicobacter pylori is a bacterium that colonizes the stomachs of most humans and can cause chronic gastric pathology. Wirth et al. (Reference Wirth, Wang, Linz, Novick, Lum, Blaser, Morelli, Falush and Achtman2004) showed that Helicobacter pylori can distinguish between closely related Buddhist and Muslim populations in Ladakh (India). This bacterium is also divided into several populations with distinct geographical distributions on a more global scale (Falush et al. Reference Falush, Wirth, Linz, Pritchard, Stephens, Kidd, Blaser, Graham, Vacher, Perez-Perez, Yamaoka, Mégraud, Otto, Reichard, Katzowitsch, Wang, Achtman and Suerbaum2003). Molecular studies and phylogenetic reconstructions have revealed that these populations are in fact derived from ancestral populations that arose in Africa (Linz et al. Reference Linz, Balloux, Moodley, Manica, Liu, Roumagnac, Falush, Stamer, Prugnolle, van der Merwe, Yamaoka, Graham, Perez-Trallero, Wadstrom, Suerbaum and Achtman2007) followed by extensive diversification in Central and East Asia. Subsequent worldwide spread can be attributed to human dispersals such as the prehistoric colonization of Polynesia and the Americas, the Neolithic introduction of farming to Europe, the Bantu expansion within Africa and the slave trade.
Hepatitis G virus (HGV or GBV-C) is an RNA flavivirus that is widely distributed with geographically divergent isolates. Phylogenetic analyses that include sequences from chimpanzee isolates suggest an ancient African origin (Muerhoff et al. Reference Muerhoff, Leary, Sathar, Dawson and Desai2005). Loureiro et al. (Reference Loureiro, Alonso, Pacheco, Uzcategui, Villegas, León, De Saéz, Liprandi, López and Pujol2002) showed that HGV was introduced from Asia to America by early human dispersals.
JC virus, a member of Polyomaviridae family, is ubiquitous in human populations and primary infection occurs asymptomatically during childhood. JCV is transmitted mainly from parents to children during prolonged cohabitation, which has enabled its use for retracing human dispersals (Sugimoto et al. Reference Sugimoto, Hasegawa, Kato, Zheng, Ebihara, Taguchi, Kitamura and Yogo2002). Although there is no outgroup that can be used to root JCV trees, the phylogenetic analyses of JCV DNA sequences conducted by Sugimoto et al. (Reference Sugimoto, Hasegawa, Kato, Zheng, Ebihara, Taguchi, Kitamura and Yogo2002) suggest an African origin of the virus. Phylogenetic analyses have revealed geographic structures that are compatible with the dispersal and expansion of humans (Hammer et al. Reference Hammer, Karafet, Redd, Jarjanazi, Santachiara-Benerecetti, Soodyall and Zegura2001). JVC was also used to investigate the relationships between East Asian populations (Japanese and Koreans) and Native Americans (Amerinds and Na-Denes) (Zheng et al. Reference Zheng, Sugimoto, Hasegawa, Kobayashi, Kanayama, Rodas, Mejia, Nakamichi, Guo, Kitamura and Yogo2003). Differences between European and Africa/Asian JCV sub-populations seemed to indicate that human population structures alone cannot account for diversity patterns in the virus and that some other factors must have played a role in the genetic differentiation of the virus (Wooding, Reference Wooding2001). Indeed, phylogenies of human populations and JCV are not concordant, indicating extensive horizontal gene transmission in the virus (Shackelton et al. Reference Shackelton, Rambaut, Pybus and Holmes2006).
The Mycobacterium tuberculosis complex is clustered in six lineages associated with geographic locations and phylogeographic studies suggest that these associations reflect ancient human dispersals (Gagneux and Small, Reference Gagneux and Small2007). As emphasized by Achtman (Reference Achtman2008), however, most analyses have focused in Europe and North America on strains isolated from infected immigrants coming from developing countries. The M. tuberculosis complex is supposed to be part of a highly diversified protospecies that has infected hominids since their origins and was subject to an extreme genetic bottleneck following the emergence of the modern humans and their dispersal in and out of Africa (Gutierrez et al. Reference Gutierrez, Brisse, Brosch, Fabre and Omais2005). Indeed, the previously designated M. canetti, from Djibouti, shows a greater genetic diversity. Although the ancestral form of M. leprae, the agent of leprosy, is still not known, phylogeographic studies support an African origin followed by early dispersion of the modern human (Monot et al. Reference Monot, Honoré, Garnier, Zidane, Sherafi, Paniz-Mondolfi, Matsuoka, Taylor, Donoghue, Bouwman, Mays, Watson, Lockwood, Khamesipour, Dowlati, Jianping, Rea, Vera-Cabrera, Stefani, Banu, Macdonald, Sapkota, Spencer, Thomas, Harshman, Singh, Busso, Gattiker, Rougemont, Brennan and Cole2009).
Phylogenetic studies have shown that the occurrence of Taenia in humans predates the domestication of cattle and swine by Neolithic farmers, suggesting that their ancestors first became infected while consuming raw meat such as partially consumed and discarded prey items of carnivores and scavengers (Hoberg et al. Reference Hoberg, Alkire, de Queiroz and Jones2001). Taenia accompanied early human dispersion out of Africa, and swine and cattle are thought to have acquired infections with Taenia species during their early domestication (Hoberg et al. Reference Hoberg, Alkire, de Queiroz and Jones2001).
“Domestication” pattern
Dogs were domesticated 15,000 years ago and studies on the patterns of phylogeographic variation indicate an East Asian origin (Savolainen et al. Reference Savolainen, Zhang, Luo, Lundeberg and Leitner2002). The study of Bourhy et al. (Reference Bourhy, Reynes, Dunham, Dacheux, Larrous, Thi Que Huong, Xu, Yan, Miranda and Holmes2008) on the phylogeography of dog rabies virus (RABV) showed that the RABV from terrestrial mammals comprises six major distinct phylogeographic clades. RABV is hypothesized to have an ancestry that lies with domestic dogs from the south of India and evolutionary diversification, using coalescent-based methods, is estimated to have occurred within the last 1500 years. Moreover, Bourhy et al. (Reference Bourhy, Reynes, Dunham, Dacheux, Larrous, Thi Que Huong, Xu, Yan, Miranda and Holmes2008) hypothesized that the dog has served as the main vector for RABV transmission to other taxa of the Canidae such as foxes and raccoons as well as several species of Herpestidae and Mephitidae.
Cattle and swine have had extensive epidemiological interactions with humans. The results of several phylogenetic studies suggest that these animals were not only the sources of parasite and microbe infections for humans, but they were also recipients of parasites and microbes from humans acquired, in the opposite direction, as a result of domestication.
Another example is Mycobacterium bovis, which has been shown to arise from a M. tuberculosis strain. This evolutionary scenario has been confirmed by genomics, in particular by the fact that M. bovis has a smaller chromosome than M. tuberculosis (Smith et al. Reference Smith, Hewinson, Kremer, Brosch and Gordon2009). The insertion of M. bovis strains within West African strains of M. tuberculosis suggests an African origin of M. bovis. Unfortunately, investigations of both domesticated and wild Asian buffaloes are lacking.
Anthrax is caused by Bacillus anthracis, which is derived from a Bacillus cereus ancestor. Molecular clock estimates that the separation of B. anthracis from its closest outgroup occurred more than 17,000 years ago (Van Ert et al. Reference Van Ert, Easterday, Huynh, Okinaka, Hugh-Jones, Ravel, Zanecki, Pearson, Simonson, U'Ren, Kachur, Leadem-Dougherty, Rhoton, Zinser, Farlow, Coker, Smith, Wang, Kenefic, Fraser-Liggett, Wagner and Paul Keim2007); this associates the radiation of B. anthracis with the domestication of cattle and suggests human-mediated dispersal. However, the geographic origins of anthrax have not been established largely because of the lack of studies conducted in the domestication centres.
Recent works on the phylogeny and phylogeography of measles virus (MeV) suggest a recent emergence of measles around the 11th and 12th centuries from the closely related rinderpest virus (RPV) of ruminants. This more recent origin, if it is confirmed, challenges the hypothesis of an pathogen emergence associated with the early domestication of ruminants in the Fertile Crescent or in Asia.
The history of pig domestication involves multiple centres of domestication in Asia, where pigs were first domesticated in the Neolithic period from several lineages of wild boar, and in Europe, where pigs were domesticated from distinct and presumably genetically restricted ancestors (Larson et al. Reference Larson, Dobney, Albarella, Fang, Matisoo-Smith, Robins, Lowden, Finlayson, Brand, Willerslev, Rowley- Conwy, Andersson and Cooper2005, Reference Larson, Albarella, Dobney, Rowley-Conwy, Schibler, Tresset, Vigne, Edwards, Schlumbaum, Dinu, Balacsescu, Dolman, Tagliacozzo, Manaseryan, Miracle, Van Wijngaarden-Bakker, Masseti, Bradley and Cooper2007). China and Europe thus represent different domestications or breed formation processes (Megens et al. Reference Megens, Crooijmans, San Cristobal, Hui, Li and Groenen2008). Pigs and wild boar are hosts to several potential zoonotic parasites, but few of them have been studied in depth phylogenetically.
Ancestors of human Ascaris are hypothesized to be derived from nematodes hosted by wild boar at the very start of their domestication, but there is little evidence to support this hypothesis (Criscione et al. Reference Criscione, Anderson, Sudimack, Peng, Jha, Williams-Blangero and Anderson2007). Recently, Zhou et al. (Reference Zhou, Li, Yuan, Hu and Peng2011) investigated the phylogeography of Ascaris lumbricoides and A. suum. This study is not without sampling bias as it lacks inclusion of an external outgroup. Indeed, inclusion of an external outgroup would have permitted inference of the direction of host origin.
Trichinella spiralis is the second zoonotic roundworm that has been examined in a phylogeographic context. Rosenthal et al. (Reference Rosenthal, LaRosa, Zarlenga, Dunams, Chunyu, Mingyuan and Pozio2008) investigated the diversification of T. spiralis using Trichinella nativa and T. murrelli, both parasites of wildlife, as outgroups. Their study suggested that European lineages of T. spiralis originated several thousand years ago when pigs were first domesticated there and hypothesized that, more recently, Europeans have introduced T. spiralis to America via infected pigs and/or rats. The lower genetic diversity observed in European lineages of T. spiralis compared to Asian lineages is in accordance with the lower diversity of European wild boar in comparison with Asian wild boar.
“Globalization” pattern
Human societies have been engaged in trading activities for a very long time and they not only exchanged goods but also infectious diseases (Mc Neill, Reference McNeil1976; Diamond, Reference Diamond1997). Plague, syphilis, smallpox and leprosy are strongly associated with trade but also with wars and the resulting human displacements.
Yersinia pestis is the agent of two old pandemic plagues, Justinian's plague (541–767 AD) and the Black Death (1346–1800s). The third and last pandemic was related to shipping from Hong Kong, which carried infected rats and fleas to the entire globe in 1894 (Achtman, Reference Achtman2008). Three centres of origin of these pandemic plagues are depicted according to phylogeographic studies, Africa, Central and East Asia. The spread of all these plagues is most likely linked with commercial trading. Unfortunately, Y. pestis from endemic areas in Central and East Asia has not been analysed in relation to isolates from other parts of the world (Achtman, Reference Achtman2008).
The comparative genomic analysis of Mycobacterium leprae, the agent of leprosy, suggests an African origin and its actual distribution explained by dispersal patterns of early humans and by trade routes, such as the Silk Road, which have contributed to the spread leprosy (Monot et al. Reference Monot, Honore, Garnier, Araoz, Coppee, Lacroix, Sow, Spencer, Truman, Williams, Gelber, Virmond, Flageul, Cho, Ji, Paniz-Mondolfi, Convit, Young, Fine, Rasolofo, Brennan and Cole2005, Reference Monot, Honoré, Garnier, Zidane, Sherafi, Paniz-Mondolfi, Matsuoka, Taylor, Donoghue, Bouwman, Mays, Watson, Lockwood, Khamesipour, Dowlati, Jianping, Rea, Vera-Cabrera, Stefani, Banu, Macdonald, Sapkota, Spencer, Thomas, Harshman, Singh, Busso, Gattiker, Rougemont, Brennan and Cole2009). However, in contrast to the Black Death, leprosy is thought to have originated in either Europe or the Middle East and then spread to China via the Silk Road and thereafter to the Far East (Monot et al. Reference Monot, Honoré, Garnier, Zidane, Sherafi, Paniz-Mondolfi, Matsuoka, Taylor, Donoghue, Bouwman, Mays, Watson, Lockwood, Khamesipour, Dowlati, Jianping, Rea, Vera-Cabrera, Stefani, Banu, Macdonald, Sapkota, Spencer, Thomas, Harshman, Singh, Busso, Gattiker, Rougemont, Brennan and Cole2009).
Different varieties of treponemal diseases exist, with syphilis due to the spirochete Trepanoma pallidum sub-species pallidum. A recent phylogenetic investigation has given new insights on the diversification of T. pallidum, which arose in Africa and Southeast Asia (T. pallidum subspecies pertenue I and II) (non-venereal infection) and spread subsequently to the Middle East/Eastern Europe, in the form of endemic syphilis, and to the Americas, in the form of New World yaws (Harper et al. Reference Harper, Ocampo, Steiner, George and Silverman2008). These results support the Columbian theory of the origin of syphilis, as an American T. pallidum strain was re-introduced to the Old World giving rise to the progenitor of modern syphilis-causing strains.
A recent study on smallpox virus illustrates the problem of the data information. Li et al. (Reference Li, Carroll, Gardner, Walsh, Vitalis and Damon2007) show that two primary VARV clades may have diverged from an ancestral African rodent-borne variola-like virus either around 16,000 or around 68,000 ybp, depending on which historical records (East Asian or African) are used to calibrate the molecular clock. The numbers of events needing to fit the date with the two scenarios differ. A parsimony analysis has to be performed, which would likely suggest an Asian origin of smallpox with subsequent dispersion by trade routes to Africa, and from Africa to Americas by slave trade.
Limitations of phylogeographic studies
Five main limitations and potential associated pitfalls can be identified from this brief overview of phylogeographic studies on human pathogens and parasites. The first one is linked to the infectious or parasitic agent investigated in phylogenetic reconstruction. Although largely used in phylogeographic studies, viruses do not seem to be the best candidates for investigating the “out of Africa” pattern because of their high rates of mutations (Holmes, Reference Holmes2004). However, viruses may be of real interest for investigating recent historical events. Bacteria are also widely used, mostly to test early modern human dispersals (out of Africa) with great success (e.g. Helicobacter pylori). Parasites such as protozoans, helminths or arthropods rarely feature in phylogeographic studies of human parasites, although they appear to be excellent candidates for the investigation of co-speciation and co-phylogeographics (see Criscione et al. Reference Criscione, Anderson, Sudimack, Peng, Jha, Williams-Blangero and Anderson2007; Nieberding and Olivieri, Reference Nieberding and Olivieri2007).
The second limitation is related to the genetic markers, which will depend on the agent in question. While viruses and bacteria can be fully sequenced and comparative genomics help in investigating genomic evolution, these technological advances are used less for macroparasites. Most phylogeographic studies of macroparasites have been based on the sequencing of one (mitochondrial) or ideally several genes, both mitochondrial and nuclear (but see Criscione et al. Reference Criscione, Poulin and Blouin2005; Criscione, Reference Criscione2008). Microsatellites may have complementary significance in inferring more recent dynamics.
The third and major limitation lies in the samples themselves and the potential flaws of analyses due to sample bias. Phylogeographics need the establishment of a sampling protocol adapted to the hypothesis to be tested and the size of the geographic area investigated. However, most studies are not based on a specified sample protocol in relation to hypotheses under investigation and it appears that, in some studies, samples have been gathered more or less at random. Although crucial for examining dispersal (dispersion routes in the current context), the choice of the outgroup is also rarely explained and not even mentioned at all in some studies. This problem of sampling may lie in the fact that many studies simply apply a phylogenetic analysis to isolates of the organisms investigated, without any apparent awareness of the concepts and methods of phylogeographics (see Avise, Reference Avise2000; Nieberding and Olivieri, Reference Nieberding, Durette-Desset, Vanderpoorten, Casanovas, Ribas, Deffontaine, Feliu, Morand, Libois and Michaux2005). Moreover, as emphasized by Beheregaray (Reference Beheregaray2008) in an exhaustive review of phylogeographic studies, there is a great need to acquire more phylogeographic data outside of Occidental countries, which account for around 80% of the studies whereas Asia accounts for 15% and Africa just 7%. This is crucial for parasites and microbes of domestic animals in the domestication centres.
The fourth limitation lies in the methods used for genealogical or tree reconstruction. Most of the studies presented here have used standard phylogenetic reconstructions and few have used more recent advanced tools. Investigating recent evolutionary events takes account not only of population structure but also demography and potential genetic exchanges, as a Bayesian modelling approach may permit (Knowles, Reference Knowles2009).
The final limitation relates to the dating of diversification events. This is always difficult with parasites due to the lack of fossil records against which to calibrate molecular information. Calibration depends on life cycle data (generation times) and mutation rates, but assumes that these will be roughly constant through time and lineage. For example, in the case of T. spiralis, the microsatellite mutations during parasite generations was calculated to estimate that Western and Asian lineages of T. spiralis would have diverged 18,000 years ago (Rosenthal et al. Reference Rosenthal, LaRosa, Zarlenga, Dunams, Chunyu, Mingyuan and Pozio2008). Such calculations have been used for several microbes and have indicated time to the most recent common ancestor (the age) is 10,000–71,000 years for Salmonella Typhi (Roumagnac et al. Reference Roumagnac, Weill, Dolecek, Baker, Brisse, Chinh, Le, Acosta, Farrar, Dougan and Achtman2006), 13,000 years for Y. pestis (Achtman et al. Reference Achtman, Morelli, Zhu, Wirth, Diehl, Kusecek, Vogler, Wagner, Allender, Easterday, Chenal-Francisque, Worsham, Thomson, Parkhill, Lindler, Carniel and Keim2004), 15,000–20,000 years for the M. tuberculosis complex (Kapur et al. Reference Kapur, Whittam and Musser1994), and 40,000 years for E. coli O157:H7 (Zhang et al. Reference Zhang, Qi, Albert, Motiwala, Alland, Hyytia-Trees, Ribot, Fields, Whittam and Swaminathan2006). These calculations should be compared with other sources of information, both historical and archaeological.
CONCLUSION
The building of human parasite communities through “out of Africa”, domestication and globalization mechanisms has found its significance in evolutionary biogeography where processes are invasion, establishment and expansion (Lockwood et al. Reference Lockwood, Hoopes and Marchetti2007). Phylogeography provides the tools to explore these events and their linkages (Avise, Reference Avise2000). Although studies on human parasites and microbes are classified here through these three patterns, they are not mutually exclusive. A clear example is human Taenia species originating in Africa, which has followed human dispersal and later switched to domesticated animals. Moreover, recent trading activities may have obscured the oldest patterns, i.e. “out of Africa” and domestication centres. This brief overview of phylogeographic studies of human parasites and microbes raises two main observations, one on the small number of investigation on parasites and the second on the need to investigate parasites of domestic animals.
Parasites (helminths, ectoparasitic arthropods, fungi and even protozoans) have been poorly investigated compared to bacteria and viruses. This is quite intriguing as some species could be good candidates for testing the “out of Africa” pattern such as the fungi Pneumocystis spp. (Hugot et al. Reference Hugot, Demanche, Barriel, Dei-Cas and Guillot2003) or lice (Reed et al. Reference Reed, Light, Allen and Kirchman2007) compared to viruses that may evolve too fast for depicting old dispersion events (Holmes, Reference Holmes2004). The pinworm, Enterobius vermicularis, seems to be particularly relevant as it can be found at archaeological sites and its prevalence seems to be linked with changes in settlement habits and population densities (Hugot et al. Reference Hugot, Reinhard, Gardner and Morand1999).
There is a lack of information on the phylogeny and phylogeographics of parasites and microbes of domestic animals. The ‘domestication’ pattern should be completed by investigations of parasites and microbes that are specific to domesticated animals (and with relatives known in wild animals). For example, the recent molecular epidemiology of Mycoplasma capricolum, the causative agent of contagious caprine pleuropneumonia (CCPP) of domestic goats, showed a potential origin centre in the Middle East, which is presumably a domestication centre for the domestic goat. However, the existence of a distinct Asian cluster favours the hypothesis that this disease has been endemic in Asia for a long time (Manso-Silvan et al. Reference Manso-Silvan, Dupuy, Chu and Thiaucourt2011). Another example is described by Troell et al. (Reference Troell, Engström, Morrison, Mattsson and Höglund2006), who investigated the population genetic structure of the nematode Haemonchus contortus on a worldwide scale and showed the existence of geographic clusters. Phylogeographic investigations may depict the spreading routes of domesticated animals in Asia and in Africa. Phylogeography is a functional tool that may also help with investigations of old and recent emerging diseases (Keim and Wagner, Reference Keim and Wagner2009). As already emphasized by Wolfe et al. (Reference Wolfe, Panosian Dunavan and Diamond2007), most studies have been based on specimens collected opportunistically from domestic animals with no systematic surveys over the spectrum of domestic and wild animals for almost any particular parasite. Among the potentially interesting organisms, Wolfe et al. (Reference Wolfe, Panosian Dunavan and Diamond2007) cited the mite Sarcoptes scabei as the agent of sarcoptic mange, which affects wild mammals worldwide dramatically. Again, it is hypothesized that S. scabei originated in human populations and then spread to domesticated animals, which in turn, transmitted sarcoptic mange to a wide range of wild mammals (Fain, Reference Fain1978).
We are still far from understanding the patterns and processes behind the building of human parasite communities. As this review emphasizes, phylogeographics is an important tool for determining the origins, distributions and dispersal patterns of human parasites and microbes. A major research effort should be initiated for zoonotic parasites and microbes that are shared with domestic animals (including the commensal rodents). As the review also highlights, we have to move our attention and effort towards Asia where almost all of the animal domestication processes took place and from where the first global commercial trade emerged, and we have to leave our Eurocentrism out of phylogeography.
ACNOWLEDGEMENTS
I thank Professor Sarah Randolph for her invitation to contribute to this special issue. I thank three reviewers for their helpful comments. This work is supported by ATP-CIRAD “Emergence et risques sanitaires” and the PIR-CNRS-MIE “Homme-Pathogènes : une longue co-évolution”.
