ausblenden:
Schlagwörter:
-
Zusammenfassung:
Infectious diseases have been among the top drivers of morbidity and mortality within human populations. Consequently, many human genes that orchestrate immune responses against pathogens exhibit signatures of natural selection. The outcome of this pathogen-mediated selection depends on the function of the gene’s product. Some genes, such as those targeting highly conserved structures of pathogens, exhibit very low levels of variation as they evolve
under strong purifying selection. On the other hand, balancing selection might result in
increased variation and maintenance of diversity. In jawed vertebrates, the extreme
polymorphism of the major histocompatibility complex (MHC) genes is hypothesized to have
evolved through this evolutionary process. Products of MHC genes, which are also known as human leukocyte antigen (HLA) in humans, bind and present short peptides of pathogens to the immune cells and initiate adaptive immune responses. Therefore, diverse pathogens presumably select for increased MHC diversity. Specific mechanisms of this selection process have been formulated, yet their relative roles in the maintenance of diversity continue to be
debated extensively. The aim of this thesis is to investigate the role of pathogen-mediated selection as well as other evolutionary forces in shaping the HLA diversity in humans. To this end, we employed several computational approaches, combined with the population genetics analysis of ancient and
modern human populations. Our first step was to focus on the commonly held assumption that high pathogen diversity selects for different HLA alleles. We analyzed the peptide diversity of 36 human pathogens and showed that peptides that are shared between pathogens constitute only a minority of the total peptidome pool. In other words, the peptide pool of each pathogen
is mostly unique to itself, exhibiting an immense diversity that our immune system faces. In order to explore the effect if this diversity on HLA genes, we used computational peptide-binding prediction algorithms and identified peptide sets that are bound by 321 common HLA class-I alleles. Our results revealed that some HLA alleles bind a larger fraction of peptides
from a few pathogens, suggesting a specialization towards these particular pathogens. Furthermore, this specialization was negatively correlated with the size of the peptide pool that the HLA allele binds. These results support a scenario that HLA alleles with small peptide repertoires can be maintained in populations by virtue of specialization against a few deadly
pathogens that emerge throughout human evolutionary history. On the other hand, generalist HLA alleles with large peptide repertoires continue protecting individuals against common pathogens. We next took a population genetics approach and analyzed two novel HLA datasets. The first dataset consists of 12 populations from sub-Saharan Africa. African populations are highly
underrepresented in genomics research, hampering both the successful medical applications in many countries and the detailed characterization of humans’ evolutionary history. We used both
genome-wide SNP data and targeted HLA sequencing data to investigate HLA diversity
patterns not only within African populations but also in a wider context by including population samples from Europe and East Asia. Our results highlight the excessive diversity across HLA genes within African populations. We observed that very old mutations constitute a significant
part of the diversity within MHC, suggesting that balancing selection favors ancient variation. In line with this observation, non-African populations appear to have maintained similarly high levels of HLA diversity despite the bottlenecks associated with the colonization of Eurasia.
Furthermore, contrary to what would be expected for a genomic region evolving under
divergent local selective pressures, population differentiation was not higher within MHC than it is in neutral regions. In fact, population differentiation at HLA genes largely reflects differentiation patterns of neutral markers indicating population history as the main determinant of differentiation. Targeted HLA sequencing analysis also supports the dominant role of balancing selection and population history in HLA evolution. The functional diversity within
HLA genes (i.e. functionally distinct HLA molecules) is maintained across different
populations while the differentiation patterns based on HLA alleles follow the continental
separations. Overall, these results suggest that demography and balancing selection on ancient variation play a major role in recent MHC evolution in humans while the role of adaptation to local pathogens appears much smaller. This conclusion is further supported by our analysis of the second novel HLA dataset. The dataset includes 129 ancient individuals from central Europe that lived between 5000 BCE to 2800 BCE. This period, namely the Neolithic, is associated
with drastic social and cultural changes. Domestication of animals and plants brought about a sedentary lifestyle, shifts in diet towards increased consumption of carbohydrate and milk products and, inadvertently, many novel pathogens. Consequently, it is commonly
hypothesized that as a result of these changes, many genes, specifically the immune genes would exhibit signatures of selection. Following that hypothesis, we used a novel pipeline to generate HLA genotypes of ancient individuals and analyzed the HLA diversity of early and late Neolithic farmers from Europe together with several modern European, East Asian and African populations. We found that although early Europeanfarmers migrated recently from
Anatolia, they exhibit high HLA diversity, potentially maintained by balancing selection.
Interestingly, major shifts in frequency of some HLA alleles were observed both between early and late farmers and between late farmers and modern Europeans. In fact, the differentiation between ancient and modern Europeans is almost comparable to the differentiation between modern Europeans and Asians. Although this observation is compatible with adaptive evolution in response to changes in pathogen landscape, our analysis based on peptide-binding predictions reveals that despite differences in allele frequencies, HLA allele pools of ancient and modern
populations are functionally similar. Indeed, focusing on the measles virus (MeV), which
emerged as a human pathogen after the Neolithic, we found that modern and ancient
populations bind similar numbers of MeV peptides. These results indicate that the major driver of the changes in allele frequencies was not the adaptation to novel or local pathogens. We suggest that the major differences between early and late Neolithic farmers were brought about
by the admixture with local hunter-gatherers while the differences between late farmers and modern Europeans are the result of admixture with steppe pastoralists. In line with this hypothesis, most common HLA alleles in modern central Europeans follow a north to south cline similar to the steppe ancestry. Therefore, it is likely that these alleles were introduced by
steppe pastoralists. We conclude that balancing selection and population history appears as the main drivers of recent HLA evolution while the effect of directional selection by pathogens is minimal.
