The HLA‐B landscape of Africa: Signatures of pathogen‐driven selection and molecular identification of candidate alleles to malaria protection

Abstract

Human leukocyte antigen (HLA) genes play a key role in the immune response to infectious diseases, some of which are highly prevalent in specific environments, like malaria in sub-Saharan Africa. Former case-control studies showed that one particular HLA-B allele, B*53, was associated with malaria protection in Gambia, but this hypothesis was not tested so far within a population genetics framework. In this study, our objective was to assess whether pathogen-driven selection associated with malaria contributed to shape the HLA-B genetic landscape of Africa. To that aim, we first typed the HLA-A and -B loci in 484 individuals from 11 populations living in different environments across the Sahel, and we analysed these data together with those available for 29 other populations using several approaches including linear modelling on various genetic, geographic and environmental parameters. In addition to relevant signatures of populations' demography and migrations history in the genetic differentiation patterns of both HLA-A and -B loci, we found that the frequencies of three HLA alleles, B*53, B*78 and A*74, were significantly associated with Plasmodium falciparum malaria prevalence, suggesting their increase through pathogen-driven selection in malaria-endemic environments. The two HLA-B alleles were further identified, by highthroughput sequencing, as B*53:01:01 (in putative linkage disequilibrium with one HLA-C allele, C*04:01:01:01) and B*78:01 in all but one individuals tested, making them appropriate candidates to malaria protection. These results highlight the role of environmental factors in the evolution of the HLA polymorphism and open key perspectives for functional studies focusing on HLA peptide-binding properties.

Deciphering the fine nucleotide diversity of full HLA class I and class II genes in a well‐documented population from sub‐Saharan Africa

“…Although these negative values are not significantly different from 0, they might represent a signal of demographic expansion (rather than purifying selection), as such a signal is expected in the Mandenka population based on both its known demographic history (see section 1) and other genetic studies (eg, Reference ). This would also support the hypothesis that the evolution of the HLA‐A polymorphism (at least at the level of its exon 2 region, as suggested by the present study) is closer to neutrality and more prone to reveal demographic signals . At the opposite of locus A exon 2 along the second axis of the PCA lie DPB1 exons 2 and 3.…”

“…The HLA allelic profile characterising the West African Mandenka population (the basic statistics of which are given in our companion Population Report ) brings us new insights into the mechanisms that drove the evolution of this polymorphism, mostly because this population lives in a region where infectious diseases, like malaria, are highly prevalent. Surprisingly, the Mandenkalu do not exhibit a particularly high frequency of B*53:01:01 (Allele Frequency [AF] = 0.06)—the most commonly recognised HLA class I allele protective to this disease—contrary to what is observed in most West‐African regions as a putative result of resistance to Plasmodium falciparum . However, the most frequent HLA‐B allele observed in the Mandenkalu, B*35:01:01 (AF = 0.16), has been reported as protective to malaria in Ghana and its predicted peptide‐binding profile is similar to that of B*53:01:01 , which also suggests a protective effect to this disease (this is also the case, although to a lesser extent, for the following 2 most common alleles found in this population, B*15:03:01 (AF = 0.08) and B*78:01:01 (AF = 0.08)).…”

“…In this context, the analysis of the Mandenka population is particularly motivating because it lives in a region where several infectious diseases are highly prevalent. In addition, although sub-Saharan African populations, which represent more than 2000 ethno-linguistic groups, 38 have been extensively studied for different genetic markers [39][40][41][42][43][44] including HLA, 29,[45][46][47][48] to our knowledge their HLA molecular variability has not been analysed so far at its greatest degree of detail, that is, the nucleotide level. In this study, we thus investigated the nucleotide polymorphism of 8 extended HLA genes in a sample of the Mandenka population by using full-gene NGS-MiSeq high-throughput sequencing.…”

Genetic history of the African Sahelian populations

“…While the DARC gene, which encodes the Duffy blood group antigen, is known to act as a receptor for the malaria pathogen Plasmodium vivax , explaining the positive selection of the Duffy FY*O allele in Africa, the HLA‐B gene has been suggested to be a target of positive selection favouring the B*53 allele in Plasmodium falciparum malaria endemic environments. This hypothesis, which was initially proposed on the basis of a case‐control study carried out in Gambia, was recently tested at the continental level by typing the HLA‐A and ‐B genes in a set of 11 Sahelian populations living in both malaria‐endemic and non‐endemic environments and by comparing their genetic profiles to those of 29 other populations from all over Africa …”

“…By applying both linear modelling and high‐throughput HLA sequencing, the same study suggested that the frequencies of two HLA‐B alleles, B*53:01:01 (likely associated to HLA‐C*04:01:01:01 ) and B*78:01 increased in malaria‐endemic regions (mostly western Africa) as a result of positive selection due to their putative protection against P. falciparum. These two alleles (and possibly others) may have become beneficial at the onset of malaria in the region, such a phenomenon of soft selective sweep (involving positive selection of several markers simultaneously, ie, new selective pressures on the standing genetic variation) representing a peculiar case of pathogen‐driven selection . Noteworthy is the higher occurrence of B*78 in the Fulani (the only HLA‐typed nomadic Sahelian population living in a malaria endemic region) compared to the other western African populations where B*53 is predominantly observed, which, together with high frequencies of other alleles like B*27 and B*37 (both >10%, see Figure ), makes this population diverge from others at the HLA‐B locus.…”