Parasitoids

These are parasites that grow to relatively large sizes within their host and almost inevitably kill it when they complete their development and emerge from the host. They include hymenopteran and dipteran insects, nematomorphs, mermithid nematodes, oenonid polychaetes andCordyceps fungi. Remarkably, the ability to alter host behaviour markedly around the time of parasite emergence, in ways that benefit the parasite, has evolved independently in multiple lineages of parasitoids (e.g. Maitland,Reference Maitland1994; Eberhard,Reference Eberhard2000; Thomaset al.Reference Thomas, Schmidt-Rhaesa, Martin, Manu, Durand and Renaud2002; Poulin and Latham,Reference Poulin and Latham2002; Grosmanet al.Reference Grosman, Janssen, de Brito, Cordeiro, Colares, Fonseca, Lima, Pallini and Sabelis2008). As their large size and high virulence would not allow the long-term persistence of species with widespread infections throughout a host population, only species with life history traits that lead to a sustainable exploitation of host resources have persisted over evolutionary time; therefore, parasitoids typically occur at low prevalence and low mean intensity of infection (close to a single parasite per infected host).

Parasitic castrators

These parasites induce the total or near-total suppression of host reproduction early in the infection, and use the resources that the host would have invested in its reproduction for their own reproduction. This is their main difference from parasitoids, which they resemble in many other aspects. Indeed, castrators share many traits with parasitoids (Kuris,Reference Kuris1974): they also attain relatively large body sizes, and occur at low prevalence and low intensities of infection in their host populations. Parasitic castrators include taxa such as ascothoracican and rhizocephalan barnacles, entoniscid isopods, strepsipteran insects, and the juvenile stages of some helminths, most notably the intramolluscan stages of trematodes.

Directly transmitted parasites

Parasites using this strategy require a single host individual per generation, in which they induce variable but intensity-dependent pathology. Directly transmitted parasites include copepods, cyamid amphipods, lice, mites, monogeneans, many nematodes, many fungi and many taxa of protists (Eimeria, Giardia); many bacteria and viruses would also fall within this category. The infection mode used by these various taxa to get inside the host or attach to the host's outside surfaces vary considerably, but these mechanistic differences do not detract from the more fundamental life history similarities uniting these parasites. Certain parasites present unusual features, but nevertheless fit within this strategy. For instance, the many single-celled parasites capable of within-host multiplication, and entomopathogenic nematodes which multiply within their host after the latter is killed by symbiotic bacteria carried by the nematodes, are still fundamentally directly transmitted parasites. At the population level, directly transmitted parasites can display a range of prevalence and mean intensity values, but are almost invariably characterized by an aggregated distribution in which most individual hosts harbour few or no parasites, whereas a few hosts harbour the majority of the parasite population (Shaw and Dobson,Reference Shaw and Dobson1995; Poulin,Reference Poulin2007,Reference Poulin2013). Direct transmission is probably the first strategy adopted by any taxon following its transition to a parasitic mode of life; later, during the course of evolution, extra hosts can be added to the life cycle and other fundamental traits may change, leading to the adoption of any of the other, derived strategies.

Trophically transmitted parasites

These parasites require two or more hosts of different species, in a particular order, to complete a single generation, and are acquired by their definitive host (the one in which they mature and reproduce; almost always a vertebrate) via predation on the preceding host. Trophically transmitted parasites include all trematodes (except schistosomes, in which trophic transmission has been secondarily lost) and cestodes, acanthocephalans, pentastomids, numerous nematodes and many taxa of protists (e.g.Toxoplasma, Sarcocystis). The distribution of trophically transmitted parasites among their hosts is characterized by aggregation (Poulin,Reference Poulin2013), just like that of directly transmitted parasites. They can be quite virulent to their intermediate hosts. However, the most notable effect on the intermediate host is the ability of many trophically transmitted parasites to manipulate the phenotype of the intermediate host in ways that increase its susceptibility to predation by the definitive host (Moore,Reference Moore2002; Poulin,Reference Poulin2010). Some of these manipulations result in spectacular alterations in the behaviour or appearance of the intermediate host that are much too specific and targeted to be the products of chance. This adaptive trait has evolved independently in numerous lineages of trophically transmitted parasites, indicating that it is part of the converging suite of core traits characterizing this parasitic strategy.

Vector-transmitted parasites

These parasites require two hosts of different species to complete one generation: a vertebrate host and a micropredator (see below) that serves as a vector transporting the new generation of parasites from its host of origin to a new host. Vector-transmitted parasites include filaroid nematodes and numerous taxa of protists (Plasmodium, Babesia, Leishmania, Trypanosoma, etc.); of course, numerous fungi, bacteria and viruses also are vector-transmitted. Parasites using this strategy are invariably small, both in absolute terms and relative to their host, and potentially highly virulent through replication within the host. In addition, manipulation of vector behaviour to increase transmission efficiency has evolved independently in several lineages of vector-transmitted parasites (Moore,Reference Moore1993), adding another typical feature to this strategy.

Micropredators

Parasites using this strategy use an indeterminate number of host individuals (of the same or different species) per generation. Each association with a host is brief, lasting from seconds to days depending on the species, and is followed by a period spent off the host moulting or laying eggs. Most micropredators feed on host blood, and include taxa such as leeches, branchiuran crustaceans, gnathiid isopods, blood-sucking dipteran insects (mosquitoes, sand flies, etc.), fleas, ticks, lampreys and vampire bats. Unless they are large relative to their host (e.g. a lamprey feeding on a medium-sized fish), their direct impact on host fitness is generally minimal; however, by vectoring virulent pathogens (see above), their indirect impact can be severe.

The categorization of parasites into the above six strategies requires only a consideration of how many hosts they require per generation, and of what they do to these hosts. Of course, on more detailed physiological or anatomical levels, evidence of convergence is also easy to find. For example, many directly transmitted and most trophically transmitted parasites inhabit the vertebrate gut. As a consequence, phylogenetically unrelated parasites that have adopted these strategies display similar adaptations to an anaerobic environment and similar defences against enzymatic attack. Similarly, numerous unrelated taxa have independently evolved multi-host life cycles by adding a predator (upward incorporation) or a prey (downward incorporation) of the original host to exploit both participants of a predator–prey interaction (Parkeret al.Reference Parker, Chubb, Ball and Roberts2003). Many such lineages that converged toward the trophically transmitted parasite strategy have since independently evolved other strikingly similar adaptations to their life cycle, including manipulation of the intermediate host (see above), asexual multiplication in the intermediate host (Moore and Brooks,Reference Moore and Brooks1987; Galaktionov and Dobrovolskij,Reference Galaktionov and Dobrovolskij2003), and facultative truncation of the life cycle (Poulin and Cribb,Reference Poulin and Cribb2002; Levsen and Jakobsen,Reference Levsen and Jakobsen2002; Andreassenet al.Reference Andreassen, Ito, Ito, Nakao and Nakaya2004; Lefebvre and Poulin,Reference Lefebvre and Poulin2005).

Perhaps the most convincing line of evidence for the universality of these six strategies as end points of evolution comes from the fact that they also capture the range of host exploitation strategies seen among parasites of plants. As hosts, plants are in many ways very different from animals. Their lower metabolic rate means that less energy can be extracted from them per gram of tissue or per unit time than from an animal. Plants have tough cell walls and thick external surfaces, with no easy access to their internal tissues comparable to an animal's mouth. Yet, with the exception of trophically transmitted parasites (because plants are autotrophic), all other parasite strategies are represented among organisms exploiting plants as their hosts (see Poulin,Reference Poulin2011). Among arthropods feeding on plants, there are micropredators that take small meals of sap from many plants during their lifetime (e.g. leafhoppers). These often carry vector-transmitted parasites from plant to plant, including pathogenic fungi, bacteria and viruses (Agrios,Reference Agrios1997). Other arthropods, like scale insects, have a strategy identical to that of directly transmitted parasites (Edwards and Wratten,Reference Edwards and Wratten1980). And yet others can be seen as parasitoids. Consider a brood of caterpillars hatching on a small shrub: as they increase in size and before they pupate (equivalent to emerging from the host), they can consume a large amount of leaf tissue and cause the photosynthetic death of the plant host. To these arthropod parasites, we could add fungi that behave as castrators (Roy,Reference Roy1993), and plant-parasitic nematodes which are essentially all directly transmitted parasites (Jasmeret al.Reference Jasmer, Goverse and Smant2003). Finally, parasitic angiosperms (flowering plants) that derive sustenance from other plants can also be categorized into the strategies defined above for animal parasites. Most, like mistletoes and dodder, have opted for a strategy superficially different from but fundamentally identical to that of directly transmitted parasites (Press and Graves,Reference Press and Graves1995; Poulin,Reference Poulin2011). Some, like strangler figs, are true parasitoids (Putz and Holbrook,Reference Putz and Holbrook1989).

Just as the evolution of parasitism appears inevitable, having occurred repeatedly and independently in all major lineages of unicellular and multicellular organisms, their subsequent convergence toward one of the above strategies also seems inescapable (see Poulin,Reference Poulin2011, for further evidence and discussion). These adaptive strategies consist of particular combinations of traits that have survived the selection process, while lineages adopting different trait combinations have gone extinct. In the next section, we look for comparable general or convergent trends in the evolution of parasite genomes.

Genomic reduction and compaction

Parasitism has evolved independently on multiple occasions within different taxonomic groups (Blaxteret al.Reference Blaxter, De Ley, Garey, Liu, Scheldeman, Vierstraete, Vanfleteren, Mackey, Dorris, Frisse, Vida and Thomas1998; Poulin,Reference Poulin2011; Blouin and Lane,Reference Blouin and Lane2012). Thus far, genome-mining has yet to uncover a single universal gene, or subset of genes, associated with parasitism (Rödelspergeret al.Reference Rödelsperger, Streit and Sommer2013). Rather, selective pressures associated with parasitism, such as metabolic and spatial economy and cell multiplication speed (Cavalier-Smith,Reference Cavalier-Smith2005), have placed similar evolutionary constraints on genomes of parasitic organisms leading to a convergence towards genome reduction or compaction. Although not universal (e.g. Kikuchiet al.Reference Kikuchi, Cotton, Dalzell, Hasegawa, Kanzaki, McVeigh, Takanashi, Tsai, Assefa, Cock, Dan Otto, Hunt, Reid, Sanchez-Flores, Tsuchihara, Yokoi, Larsson, Miwa, Maule, Sahashi, Jones and Berriman2011; Raffaele and Kamoun,Reference Raffaele and Kamoun2012), nuclear genome reduction in parasitic organisms (compared with their free-living relatives) has been observed in a wide range of taxa, including amoebozoans (Glöckner and Noegel,Reference Glöckner and Noegel2013), bacteria (Touchonet al.Reference Touchon, Barbier, Bernadet, Loux, Vacherie, Barbe, Rocha and Duchaud2011), fungi (Cushion,Reference Cushion2004; Heinzet al.Reference Heinz, Williams, Nakjang, Noël, Swan, Goldberg, Harris, Weinmaier, Markert, Becher, Bernhardt, Dagan, Hacker, Lucocq, Schweder, Rattei, Hall, Hirt and Embley2012), mites (Mounseyet al.Reference Mounsey, Willis, Burgess, Holt, McCarthy and Fischer2012), nematodes (Kikuchiet al.Reference Kikuchi, Cotton, Dalzell, Hasegawa, Kanzaki, McVeigh, Takanashi, Tsai, Assefa, Cock, Dan Otto, Hunt, Reid, Sanchez-Flores, Tsuchihara, Yokoi, Larsson, Miwa, Maule, Sahashi, Jones and Berriman2011; Rödelspergeret al.Reference Rödelsperger, Streit and Sommer2013) and parasitoid hymenopterans (Ardila-Garciaet al.Reference Ardila-Garcia, Umphrey and Gregory2010). The most extreme examples of genome reduction occur in intracellular parasites. For instance, the microsporidianEncephalitozoon intestinalis has the smallest known eukaryotic nuclear genome at 2·3 Mb (Corradiet al.Reference Corradi, Pombert, Farinelli, Didier and Keeling2010), which is much smaller than that of many free-living bacteria (Corradi and Slamovits,Reference Corradi and Slamovits2011). Although no complete nuclear genome for free-living relatives of the Platyhelminthes has been sequenced and annotated, the 106–147 Mb nuclear genomes of tapeworms (Olsonet al.Reference Olson, Zarowiecki, Kiss and Brehm2012; Tsaiet al.Reference Tsai, Zarowiecki, Holroyd, Garciarrubio, Sanchez-Flores, Brooks, Tracey, Bobes, Fragoso, Sciutto, Aslett, Beasley, Bennett, Cai, Camicia, Clark, Cucher, De Silva, Day, Deplazes, Estrada, Fernández, Holland, Hou, Hu, Huckvale, Hung, Kamenetzky, Keane, Kiss, Koziol, Lambert, Liu, Luo, Luo, Macchiaroli, Nichol, Paps, Parkinson, Pouchkina-Stantcheva, Riddiford, Rosenzvit, Salinas, Wasmuth, Zamanian, Zheng, Cai, Soberón, Olson, Laclette, Brehm and Berriman2013) are smaller than that of the free-living planarianSchmidtea (700 Mb) (see Olsonet al.Reference Olson, Zarowiecki, Kiss and Brehm2012) and those of the parasitic trematodesSchistosoma mansoni (363 Mb) andClonorchis sinensis (516 Mb) (Berrimanet al.Reference Berriman, Haas, LoVerde, Wilson, Dillon, Cerqueira, Mashiyama, Al-Lazikani, Andrade, Ashton, Aslett, Bartholomeu, Blandin, Caffrey, Coghlan, Coulson, Day, Delcher, DeMarco, Djikeng, Eyre, Gamble, Ghedin, Gu, Hertz-Fowler, Hirai, Hirai, Houston, Ivens, Johnston, Lacerda, Macedo, McVeigh, Ning, Oliveira, Overington, Parkhill, Pertea, Pierce, Protasio, Quail, Rajandream, Rogers, Sajid, Salzberg, Stanke, Tivey, White, Williams, Wortman, Wu, Zamanian, Zerlotini, Fraser-Liggett, Barrell and El-Sayed2009 and Wanget al.Reference Wang, Chen, Huang, Sun, Men, Liu, Luo, Guo, Lv, Deng, Zhou, Fan, Li, Huang, Hu, Liang, Hu, Xu and Yu2011, respectively). Furthermore, based on a comparison of C-values, or amount of DNA within a haploid nucleus, available in the Animal Genome Size Database, those of trematodes and cestodes are generally lower than those reported for free-living planarian flatworms (Gregory,Reference Gregory2013) and suggest nuclear genome reduction in parasitic flatworms consistent with most other taxa.

In addition, convergent trends of genomic reductions associated with parasitism are not limited to the nucleus. Organelle genome reduction is an evolutionary trend common in many parasitic organisms (Rocha and Danchin,Reference Rocha and Danchin2002; Keeling and Slamovits,Reference Keeling and Slamovits2005; Cafasso and Chinali,Reference Cafasso and Chinali2012). For instance, mitochondria are either absent or retained as mitochondria-derived organelles (e.g. mitosomes) in a variety of taxa including apicomplexans (Corradi and Slamovits,Reference Corradi and Slamovits2011; Hikosakaet al.Reference Hikosaka, Kita and Tanabe2013), microsporidians (Katinkaet al.Reference Katinka, Duprat, Cornillot, Méténier, Thomarat, Prensier, Barbe, Peyretaillade, Brottier, Wincker, Delbac, El Alaoui, Peyret, Saurin, Gouy, Weissenbach and Vivarès2001) and amoebozoans (Glöckner and Noegel,Reference Glöckner and Noegel2013) or reduced in size such as in ciliates (Burgeret al.Reference Burger, Zhu, Littlejohn, Greenwood, Schnare, Lang and Gray2000), red algae (Hancocket al.Reference Hancock, Goff and Lane2011), trematodes and cestodes (Leet al.Reference Le, Blair and McManus2002), and nematodes (Huet al.Reference Hu, Chilton and Gasser2003). Nevertheless, mitochondrial genome sizes in one family (Mermithidae) of nematodes parasitic in insects far exceed those reported for free-living nematodes, an exception to the general trend with no clear explanation (Lagiszet al.Reference Lagisz, Poulin and Nakagawa2013).

Should genomic reductions, whether nuclear or organellar, result in a loss of genetic information? Not necessarily. Genome compaction is generally achieved through a reduction in the number of mobile or repetitive genetic elements (e.g. Kissinger and de Barry,Reference Kissinger and DeBarry2011; Olsonet al.Reference Olson, Zarowiecki, Kiss and Brehm2012), protein shortening (e.g. Katinkaet al.Reference Katinka, Duprat, Cornillot, Méténier, Thomarat, Prensier, Barbe, Peyretaillade, Brottier, Wincker, Delbac, El Alaoui, Peyret, Saurin, Gouy, Weissenbach and Vivarès2001) or reduced intergenic spacers and introns (e.g. Keeling and Slamovits,Reference Keeling and Slamovits2005). Generally, in parasitic organisms, genome compaction has led to an increase in gene density compared with their free-living relatives (e.g. Keeling and Slamovits,Reference Keeling and Slamovits2005; Rödelspergeret al.Reference Rödelsperger, Streit and Sommer2013).

Functional loss

As a result of evolving alongside their respective hosts and relying on the latter to provide shelter and resources, parasitic organisms from a wide range of taxa have either lost or simplified/reduced several metabolic pathways (Mülleret al.Reference Müller, Mentel, van Hellemond, Henze, Woehle, Gould, Yu, van der Giezen, Tielens and Martin2012). For instance, many parasitic plant species lack chlorophyll and have lost their photosynthetic ability (de Pamphilis and Palmer,Reference de Pamphilis and Palmer1990; Revillet al.Reference Revill, Stanley and Hibberd2005), the apicomplexanCryptosporidium parvum has lost the ability to synthesize nucleotidesde novo (Striepenet al.Reference Striepen, Pruijssers, Huang, Li, Gubbels, Umejiego, Hedstrom and Kissinger2004) and cestodes are unable to synthesize cholesterolde novo (Olsonet al.Reference Olson, Zarowiecki, Kiss and Brehm2012). In other taxa, the loss of function is incomplete such as in some microsporidians, which have lost all but two of 21 genes involved in glycolysis (Keelinget al.Reference Keeling, Corradi, Morrison, Haag, Ebert, Weiss, Akiyoshi and Tzipori2010). Furthermore, several groups of parasitic organisms having evolved in anaerobic environments, such as amoebozoans and microsporidians, have undergone a secondary reduction of the mitochondria (Mülleret al.Reference Müller, Mentel, van Hellemond, Henze, Woehle, Gould, Yu, van der Giezen, Tielens and Martin2012), having lost nearly all mitochondrial function, including the ability to generate ATP (Heinzet al.Reference Heinz, Williams, Nakjang, Noël, Swan, Goldberg, Harris, Weinmaier, Markert, Becher, Bernhardt, Dagan, Hacker, Lucocq, Schweder, Rattei, Hall, Hirt and Embley2012; Heinz and Lithgow,Reference Heinz and Lithgow2013 and references therein). On the other hand, many other groups of parasitic organisms have conserved 36 of 37 genes typically found in animal mitochondria (only lacking the subunit 8 of the adenosine triphosphatase complex), including nematodes (Huet al.Reference Hu, Chilton and Gasser2003; Kanget al.Reference Kang, Sultana, Eom, Park, Soonthornpong, Nadler and Park2009; Liuet al.Reference Liu, Wang, Song, Li, Ai, Yu and Zhu2013), monogeneans (Kanget al.Reference Kang, Kim, Lee, Kim, Min and Park2012), cestodes (Jeonet al.Reference Jeon, Lee, Kim, Hwang and Eom2005,Reference Jeon, Kim and Eom2007; Kimet al.Reference Kim, Jeon, Kang, Sultana, Kim, Eom and Park2007; Parket al.Reference Park, Kim, Kang, Jeon, Kim, Littlewood and Eom2007) and trematodes (Leet al.Reference Le, Blair and McManus2001a,Reference Le, Humair, Blair, Agatsuma, Littlewood and McManusb,Reference Le, Blair and McManus2002), whereas this subunit has been lost in the mitochondria and transferred to the nucleus in ciliates (Burgeret al.Reference Burger, Zhu, Littlejohn, Greenwood, Schnare, Lang and Gray2000). Other losses appear unrelated to the parasitic mode of lifeper se, though they may be indirect consequences of other evolutionary changes associated with parasitism; for instance, cestodes have lost several homeobox gene families (Tsaiet al.Reference Tsai, Zarowiecki, Holroyd, Garciarrubio, Sanchez-Flores, Brooks, Tracey, Bobes, Fragoso, Sciutto, Aslett, Beasley, Bennett, Cai, Camicia, Clark, Cucher, De Silva, Day, Deplazes, Estrada, Fernández, Holland, Hou, Hu, Huckvale, Hung, Kamenetzky, Keane, Kiss, Koziol, Lambert, Liu, Luo, Luo, Macchiaroli, Nichol, Paps, Parkinson, Pouchkina-Stantcheva, Riddiford, Rosenzvit, Salinas, Wasmuth, Zamanian, Zheng, Cai, Soberón, Olson, Laclette, Brehm and Berriman2013), which may account for their simplified morphology. Overall, however, the generality associated with these functional losses is that selective pressures towards genome reduction have resulted independently in the loss of entire gene families or functions and in the reduction of metabolic pathways primarily associated with a free-living existence (Katinkaet al.Reference Katinka, Duprat, Cornillot, Méténier, Thomarat, Prensier, Barbe, Peyretaillade, Brottier, Wincker, Delbac, El Alaoui, Peyret, Saurin, Gouy, Weissenbach and Vivarès2001; Corradiet al.Reference Corradi, Pombert, Farinelli, Didier and Keeling2010).

Functional gains

Horizontal or lateral gene transfer (HGT) from prokaryotes to eukaryotes is identified as a process by which new genes of adaptive significance may have been acquired by parasitic organisms (Alsmarket al.Reference Alsmark, Foster, Sicheritz-Ponten, Nakjang, Embley and Hirt2013). Such transfers appear to have played a significant role in the evolution of virulence (Gardineret al.Reference Gardiner, McDonald, Covarelli, Solomon, Rusu, Marshall, Kazan, Chakraborty, McDonald and Manners2012). For instance, it is believed that ancestral bacterivorous nematodes acquired cell wall-degrading enzymes from soil bacteria via HGT, thus gaining the ability to parasitize plants (Keen and Roberts,Reference Keen and Roberts1998; Abadet al.Reference Abad, Gouzy, Aury, Castagnone-Sereno, Danchin, Deleury, Perfus-Barbeoch, Anthouard, Artiguenave, Block, Caillaud, Coutinho, Dasilva, De Luca, Deau, Esquibet, Flutre, Goldstone, Hamamouch, Hewezi, Jaillon, Jubin, Leonetti, Magliano, Maier, Markov, McVeigh, Pesole, Poulain, Robinson-Rechavi, Sallet, Ségurens, Steinbach, Tytgat, Ugarte, van Ghelder, Veronico, Baum, Blaxter, Bleve-Zacheo, Davis, Ewbank, Favery, Grenier, Henrissat, Jones, Laudet, Maule, Quesneville, Rosso, Schiex, Smant, Weissenbach and Wincker2008; Paganiniet al.Reference Paganini, Campan-Fournier, Da Rocha, Gouret, Pontarotti, Wajnberg, Abad and Danchin2012). For example, the plant parasiteBursaphelenchus xylophilus possesses glycoside hydrolase enzymes capable of breaking down the plant cell wall most likely acquired from bacteria or possibly fungi (Kikuchiet al.Reference Kikuchi, Cotton, Dalzell, Hasegawa, Kanzaki, McVeigh, Takanashi, Tsai, Assefa, Cock, Dan Otto, Hunt, Reid, Sanchez-Flores, Tsuchihara, Yokoi, Larsson, Miwa, Maule, Sahashi, Jones and Berriman2011). Pathogenicity in fungi is also hypothesized to have originated via HGT through the acquisition of secreted cell wall proteins (Butleret al.Reference Butler, Rasmussen, Lin, Santos, Sakthikumar, Munro, Rheinbay, Grabherr, Forche, Reedy, Agrafioti, Arnaud, Bates, Brown, Brunke, Costanzo, Fitzpatrick, de Groot, Harris, Hoyer, Hube, Klis, Kodira, Lennard, Logus, Martin, Neiman, Nikolaou, Quail, Quinn, Sanots, Schitzberger, Sherlock, Shah, Silverstein, Skrzypek, Soll, Staggs, Stansfield, Stumpf, Sudbery, Srikantha, Zeng, Berman, Berriman, Heitman, Gow, Lorenz, Birren, Kellis and Cuomo2009). Other examples of genes acquired via HGT include enzymes involved in the nucleotide salvage pathway inC. parvum (Striepenet al.Reference Striepen, Pruijssers, Huang, Li, Gubbels, Umejiego, Hedstrom and Kissinger2004) and ATP transporters in microsporidians (Pombertet al.Reference Pombert, Selman, Burki, Bardell, Farinelli, Solter, Whitman, Weiss, Corradi and Keeling2012). Although the evolution of many novel genes in parasitic organisms has arisen via HGT, only those under positive selection have undergone extensive gene duplication and successfully integrated into their new respective genomes (Blaxter,Reference Blaxter2007). Despite a common trend towards genome reduction, examples of selective expansion in the genomes of parasitic organisms are plentiful (e.g. Huanget al.Reference Huang, Mullapudi, Sicheritz-Ponten and Kissinger2004; Loftuset al.Reference Loftus, Anderson, Davies, Alsmark, Samuelson, Amedeo, Roncaglia, Berriman, Hirt, Mann, Nozaki, Suh, Pop, Duchene, Ackers, Tannich, Leippe, Hofer, Bruchhaus, Willhoeft, Bhattacharya, Chillingworth, Churcher, Hance, Harris, Harris, Jagels, Moule, Mungall, Ormond, Squares, Whitehead, Quail, Rabbinowitsch, Norbertczak, Price, Wang, Guillén, Gilchrist, Stroup, Battacharya, Lohia, Foster, Sicheritz-Ponten, Weber, Singh, Mukherjee, El-Sayed, Petri, Clark, Embley, Barrell, Fraser and Hall2005). For instance, approximately 40% of genes in theGiardia genome have undergone gene duplication, with the majority of these involved in evading the host's immune response (Sunet al.Reference Sun, Jiang, Flores and Wen2010) and selective expansion of secreted cell wall proteins following HGT has been observed in parasitic fungi (Butleret al.Reference Butler, Rasmussen, Lin, Santos, Sakthikumar, Munro, Rheinbay, Grabherr, Forche, Reedy, Agrafioti, Arnaud, Bates, Brown, Brunke, Costanzo, Fitzpatrick, de Groot, Harris, Hoyer, Hube, Klis, Kodira, Lennard, Logus, Martin, Neiman, Nikolaou, Quail, Quinn, Sanots, Schitzberger, Sherlock, Shah, Silverstein, Skrzypek, Soll, Staggs, Stansfield, Stumpf, Sudbery, Srikantha, Zeng, Berman, Berriman, Heitman, Gow, Lorenz, Birren, Kellis and Cuomo2009). However, some parasitic organisms have evolved ways of expanding certain gene families associated with evading their host's immune system while maintaining a highly compact genome. For instance, certain rickettsiales bacteria have pseudogenes that possess conserved 5′ and 3′ ends flanking a hypervariable region. When recombined into a functional expressed site, these pseudogenes generate new antigenic variants (Braytonet al.Reference Brayton, Knowles, McGuire and Palmer2001). Also, antigenic diversity inPlasmodium falciparum is achieved by maintaining large, multicopy, and hypervariable gene families such as variant antigen repertoire (var) genes (Suet al.Reference Su, Heatwole, Werthmeier, Guinet, Herrfeldt, Peterson, Ravetch and Wellems1995). Throughout the genome ofP. falciparum, approximately 60var genes encode proteins composed of conserved basic architecture with hypervariable amino acid sequence translating into a potentially unlimited repertoire of membrane proteins (see Dzikowskiet al.Reference Dzikowski, Templeton and Deitsch2006). By encoding a single membrane protein at any given time, as the host mounts an immune response against this particular variant, sub-populations encoding different membrane proteins are selected for, leading to a proliferation of the parasitic infection (see Dzikowskiet al.Reference Dzikowski, Templeton and Deitsch2006). Furthermore, recent evidence suggests that many molecules involved in epigenetic memory, regulation of gene expression, and immune evasion in apicomplexans have been acquired via HGT from eukaryotes (Kishoreet al.Reference Kishore, Stiller and Deitsch2013).

Lineages that have adopted parasitism as a mode of life are under strong selective pressures at the genomic level, with potentially important evolutionary consequences. Genome reductions/compactions may provide a significant adaptive potential, which has allowed the evolution of an intracellular lifestyle or the exploitation of intermediate hosts orders of magnitude smaller than their definitive hosts. The loss of genes associated with processes/functions redundant with those carried out by the hosts and strong selection towards genomes that are A-T rich (Moran,Reference Moran1995; Cavalier-Smith,Reference Cavalier-Smith2005) are all means to preserve energy, whereas the gain of new genes via HGT and their establishment into the genomes has led to the evolutionary specialization of parasites. Although many of the processes leading to parasitism may have arisen by chance, they have evolved through positive selection. Overall, from a genomic perspective, parasites are highly efficient at what they do and are nothing like the overly simplified version of their free-living relatives often depicted in textbooks.