Investigating patterns of variability in functional protein-coding genes is fundamental to identifying the basis for population and species adaptation and ultimately, for predicting evolutionary potential in the face of environmental change. The Major Histocompatibility Complex (MHC), a family of immune genes, has been one of the most emphasized gene systems for studying selection and adaptation in vertebrates due to its significance in pathogen recognition and consequently, in eliciting host immune response. Pathogen evasion of host resistance is thought to be the primary mechanism preserving extreme levels of MHC polymorphism and shaping immunogenetic patterns across host populations and species. In this thesis, I examined the evolution of two equine MHC genes, DRA and DQA, over the history of the genus Equus and across free-ranging plains zebra (E. quagga) populations of southern Africa: Etosha National Park (ENP), Namibia and Kruger National Park (KNP), South Africa. Furthermore, I evaluated the relationships between the DRA locus and parasite intensity in E. quagga of ENP, to elucidate the mechanisms by which parasites have shaped diversity at the MHC.
In equids, the full extent of diversity and selection on the MHC in wild populations is unknown. Therefore, in this study, I molecularly characterized MHC diversity and selection across equid species to shed light on its mode of evolution in Equus and to identify specific sites under positive selection. Both the DRA and DQA exhibited a high degree of polymorphism and more intriguingly, greater allelic diversity was observed at the DRA than has previously been shown in any other vertebrate taxon. Global selection analyses of both loci indicated that the majority of codon sites are under purifying selection which may be explained by functional constraints on the protein. However, maximum likelihood based codon models of selection, allowing for heterogeneity in selection across codons, suggested that selective pressures varied across sites. Furthermore, at the DQA locus, all sites predicted to be under positive selection were antigen binding sites, implying that a few selected amino acid residues may play a significant role in equid immune function. Observations of trans-species polymorphisms and elevated genetic diversity were concordant with the hypothesis that balancing selection is acting on these genes.
Over the past half century, the role of neutral versus selective processes in shaping genetic diversity has been at the center of an ongoing dialogue among evolutionary biologists. To determine the relative influence of demography versus selection on the DRA and DQA loci, I contrasted diversity patterns of neutral and MHC data across the E. quagga populations of ENP and KNP. Neutrality tests, along with observations of elevated diversity and low differentiation across populations relative to nuclear intron data, provided further evidence for balancing selection at these loci among E. quagga populations. However, at the DRA locus, differentiation was comparable to results at microsatellite loci. Furthermore, zebra in ENP exhibited reduced levels of diversity relative to KNP due to a highly skewed allele frequency distribution that could not be explained by demography. These findings were indicative of spatially heterogeneous selection and suggested directional selection and local adaptation at the DRA locus.
There still remains a great deal of discussion over the mechanisms by which pathogens preserve immune gene diversity. The leading hypotheses that have been predominantly considered are: (i) heterozygote advantage (i.e. overdominant selection), (ii) rare allele advantage (i.e. frequency-dependent selection), and (iii) spatiotemporally fluctuating selection. An increasing number of studies have investigated MHC-parasite relationships to reconcile this debate, with conflicting results. To elucidate the mechanism driving the population-level patterns of diversity at the DRA locus, I examined relationships between this locus and both gastrointestinal (GI) and ectoparasite intensity in plains zebra of ENP. I discovered antagonistic pleiotropic effects of particular DRA alleles, with rare alleles predicting increased GI parasitism and common alleles associated with higher tick burdens. These results supported a frequency-dependent process and because maladaptive `susceptibility alleles' were found at reduced frequencies, suggested that GI parasites exert strong selective pressure at this locus. Furthermore, heterozygote advantage also played a role in decreasing GI parasite burden, but only when a common allele was paired with a more divergent allele, implying that frequency-dependent and overdominant selection are acting in synchrony. These results indicated that an immunogenetic tradeoff may modulate resistance/susceptibility to parasites in this system, such that with MHC-based resistance to GI parasitism, a fitness cost is incurred to the host in the form of increased ectoparasite susceptibility. It is also suggested that these selective mechanisms are not mutually exclusive.
In conclusion, these results provided species and population-level evidence for selection on the equid MHC, and highlighted the complexity in which selection operates in natural systems. In addition to heterogeneity in selective pressures at the molecular-level (across a gene region), selection likely varies spatiotemporally across populations due to fluctuations in pathogen regimes. Furthermore, pleiotropic effects of multiple pathogens can obscure our ability to understand adaptive processes. Given the level of complexity in which selection operates, I emphasize the necessity of incorporating multiple lines of evidence, using both neutral and adaptive data, to illuminate how selection operates. Finally, I also highlight the importance of considering the selective effects of multiple pathogens on host immunogenetics to better understand MHC function and adaptation.