Pathogen resistance and genetic variation at MHC loci

Abstract.— Balancing selection in the form of heterozygote advantage, frequency-dependent selection, or selection that varies in time and/or space, has been proposed to explain the high variation at major histocompatibility complex (MHC) genes. Here the effect of variation of the presence and absence of pathogens over time on genetic variation at multiallelic loci is examined. In the basic model, resistance to each pathogen is conferred by a given allele, and this allele is assumed to be dominant. Given that s is the selective disadvantage for homozygotes (and heterozygotes) without the resistance allele and the proportion of generations, which a pathogen is present, is e , fitnesses for homozygotes become (1 — s )^(n-1)e and the fitnesses for heterozygotes become (1 — s )^(n-2)e, where n is the number of alleles. In this situation, the conditions for a stable, multiallelic polymorphism are met even though there is no intrinsic heterozygote advantage. The distribution of allele frequencies and consequently heterozygosity are a function of the autocorrelation of the presence of the pathogen in subsequent generations. When there is a positive autocorrelation over generations, the observed heterozygosity is reduced. In addition, the effects of lower levels of selection and dominance and the influence of genetic drift were examined. These effects were compared to the observed heterozygosity for two MHC genes in several South American Indian samples. Overall, resistance conferred by specific alleles to temporally variable pathogens may contribute to the observed polymorphism at MHC genes and other similar host defense loci.     

Pathogen burden,co‐infection and major histocompatibility complex variability in the European badger (Meles meles)

Pathogen‐mediated selection is thought to maintain the extreme diversity in the major histocompatibility complex (MHC) genes, operating through the heterozygote advantage, rare‐allele advantage and fluctuating selection mechanisms. Heterozygote advantage (i.e. recognizing and binding a wider range of antigens than homozygotes) is expected to be more detectable when multiple pathogens are considered simultaneously. Here, we test whether MHC diversity in a wild population of European badgers (Meles meles) is driven by pathogen‐mediated selection. We examined individual prevalence (infected or not), infection intensity and co‐infection of 13 pathogens from a range of taxa and examined their relationships with MHC class I and class II variability. This population has a variable, but relatively low, number of MHC alleles and is infected by a variety of naturally occurring pathogens, making it very suitable for the investigation of MHC–pathogen relationships. We found associations between pathogen infections and specific MHC haplotypes and alleles. Co‐infection status was not correlated with MHC heterozygosity, but there was evidence of heterozygote advantage against individual pathogen infections. This suggests that rare‐allele advantages and/or fluctuating selection, and heterozygote advantage are probably the selective forces shaping MHC diversity in this species. We show stronger evidence for MHC associations with infection intensity than for prevalence and conclude that examining both pathogen prevalence and infection intensity is important. Moreover, examination of a large number and diversity of pathogens, and both MHC class I and II genes (which have different functions), provide an improved understanding of the mechanisms driving MHC diversity.     

Microsatellite loci in the house fly,Musca domestica L. (Diptera: Muscidae)

Genomic libraries from house flies enriched for (CA)₁₅ and (CAG)₁₀ repeats were constructed by using biotinylated probes. Twenty‐five loci were isolated and evaluated for polymorphisms in wild flies representing two geographically diverse populations. Fourteen of 19 dinucleotide loci, and one of six trinucleotide loci were polymorphic. One hundred and twenty‐seven alleles were detected, 39 of which were private. Average number of alleles per polymorphic locus was 8.4 ± 2.5 and average heterozygosity was 72 ± 4%. F_ST by the private allele method was 0.73. Three of 15 loci showed significant heterozygote deficiencies, attributed to null alleles. Five of 15 loci were amplified in the face fly, Musca autumnalis.     

Gene duplication and divergence produce divergent MHC genotypes without disassortative mating

Genes of the major histocompatibility complex (MHC) exhibit heterozygote advantage in immune defence, which in turn can select for MHC‐disassortative mate choice. However, many species lack this expected pattern of MHC‐disassortative mating. A possible explanation lies in evolutionary processes following gene duplication: if two duplicated MHC genes become functionally diverged from each other, offspring will inherit diverse multilocus genotypes even under random mating. We used locus‐specific primers for high‐throughput sequencing of two expressed MHC Class II B genes in Leach's storm‐petrels, Oceanodroma leucorhoa, and found that exon 2 alleles fall into two gene‐specific monophyletic clades. We tested for disassortative vs. random mating at these two functionally diverged Class II B genes, using multiple metrics and different subsets of exon 2 sequence data. With good statistical power, we consistently found random assortment of mates at MHC. Despite random mating, birds had MHC genotypes with functionally diverged alleles, averaging 13 amino acid differences in pairwise comparisons of exon 2 alleles within individuals. To test whether this high MHC diversity in individuals is driven by evolutionary divergence of the two duplicated genes, we built a phylogenetic permutation model. The model showed that genotypic diversity was strongly impacted by sequence divergence between the most common allele of each gene, with a smaller additional impact of monophyly of the two genes. Divergence of allele sequences between genes may have reduced the benefits of actively seeking MHC‐dissimilar mates, in which case the evolutionary history of duplicated genes is shaping the adaptive landscape of sexual selection.     

The ABCs of MHC

The major histocompatibility complex (MHC) contains the most diverse genes known in vertebrates. These genes encode cell‐surface molecules that play a central role in controlling immunological activity and, as a consequence, in tissue rejection, autoimmunity, and immune responses to infectious diseases. In vertebrates, there are many different MHC genes, most with many alleles. This is true for all primates studied thus far. Multiple loci and alleles allow for an increased peptide‐binding repertoire; their variety has a profound impact on an organism's ability to battle constantly evolving pathogens. The argument that infectious disease is a driving force for MHC variability is supported by observations that most of the allelic variation centers on the amino acid residues that directly interact with foreign peptides. However, while MHC diversity could be maintained through heterozygote advantage, frequency‐dependent selection, or both, the direct evidence that natural selection enhances diversity is limited. Indeed, it is not wholly clear whether selection operates only with respect to disease resistance or if behavioral and biological mechanisms also contribute to the extreme variation that has been observed for many species. Furthermore, reproductive behavior and biology may also help to maintain genetic variability at MHC loci.     

Balancing selection and heterozygote advantage in major histocompatibility complex loci of the bottlenecked Finnish wolf population

Maintaining effective immune response is an essential factor in the survival of small populations. One of the most important immune gene regions is the highly polymorphic major histocompatibility complex (MHC). We investigated how a population bottleneck and recovery have influenced the diversity and selection in three MHC class II loci, DLA‐DRB1, DLA‐DQA1 and DLA‐DQB1, in the Finnish wolf population. We studied the larger Russian Karelian wolf population for comparison and used 17 microsatellite markers as reference loci. The Finnish and Karelian wolf populations did not differ substantially in their MHC diversities ( = 0.047, P = 0.377), but differed in neutral microsatellite diversities ( = 0.148, P = 0.008). MHC allele frequency distributions in the Finnish population were more even than expected under neutrality, implying balancing selection. In addition, an excess of nonsynonymous compared to synonymous polymorphisms indicated historical balancing selection. We also studied association between helminth (Trichinella spp. and Echinococcus canadensis) prevalence and MHC diversity at allele and SNP level. MHC‐heterozygous wolves were less often infected by Trichinella spp. and carriers of specific MHC alleles, SNP haplotypes and SNP alleles had less helminth infections. The associated SNP haplotypes and alleles were shared by different MHC alleles, which emphasizes the necessity of single‐nucleotide‐level association studies also in MHC. Here, we show that strong balancing selection has had similar effect on MHC diversities in the Finnish and Russian Karelian wolf populations despite significant genetic differentiation at neutral markers and small population size in the Finnish population.     

Selection on MHC-linked microsatellite loci in sheep populations

Microsatellites within the major histocompatibility complex (MHC) region have received increasing attention as proxy measures of the level of polymorphism at the Mhc genes themselves. We assessed the diversity of microsatellite loci within or in close proximity of the Mhc genes in several breeds of domestic sheep (Ovis aries) and the wild Mouflon (Ovis orientalis musimon). This was compared to variation at other microsatellite loci scattered throughout the sheep genome. Significantly higher number of alleles were observed at the MHC microsatellites. The sheep breeds studied fell into high- and low-diversity group. This grouping is not related to the agricultural use of the breeds, whether for milk, meat or wool. It is, however, correlated with the geographic origins of the breeds. Southern breeds are genetically more diverse than northern breeds. The observed heterozygosity was in most cases lower than Hardy-Weinberg expectations. The potential impact of selective breeding by man on this is discussed. Neutrality tests indicated that for most of the breeds, the distribution of alleles at the MHC-linked microsatellites are more even than would be expected if the genes were neutral and sampled from populations under drift-mutation equilibrium. Hitchhiking due to tight linkage with alleles at the MHC loci that are under balancing selection is proposed as a possible explanation for this pattern.     

Are there Ubiquitous Parasite-driven Major Histocompatibility Complex Selection Mechanisms in Gray Mouse Lemurs?

A major goal of evolutionary biology is to understand how selection drives local adaptation. For example, the major histocompatibility complex (MHC) plays an important role in the immune system, and high levels of MHC variation are thought to be a form of adaptation in natural populations. Individual MHC composition may influence parasite resistance via advantages associated with 1) heterozygosity, because heterozygotes recognize a broader range of different antigens than homozygotes (heterozygote advantage); 2) highly variable amino acid sequences in MHC alleles, allowing individuals to bind a broader spectrum of parasite-derived peptides (divergent-alleles advantage, a mechanistic variant of the heterozygote advantage model); or 3) specific MHC alleles (rare allele advantage or frequency dependent selection). We investigated relationships between gastrointestinal nematode burden and both adaptive immune gene variability (MHC class II DRB) and neutral microsatellites in free-living gray mouse lemurs (Microcebus murinus) native to a dry deciduous forest population in western Madagascar to test these hypotheses. The individual MHC composition was related to parasite infestation. Specific MHC alleles were involved in parasite resistance and the presence of common alleles negatively influenced infestation intensity. We found no support for the heterozygote advantage hypothesis, but we did find support for the divergent-MHC allele advantage hypothesis: Individuals with very divergent MHC alleles carried fewer and less intense nematode infestations than individuals with more similar alleles in the more variable dry deciduous forest population. These results indicate that intestinal parasites are important selection pressures under natural conditions and suggest that different selection mechanisms are not mutually exclusive. In contrast, we detected no association between neutral overall individual genetic diversity (measured via 17 microsatellites) and parasite load. Finally, we investigated the ubiquity of parasite-driven selection mechanisms by comparing our results with a previous study of a mouse lemur population from the climatically different littoral forest in southeastern Madagascar, ca. 500 km away. This revealed that different specific MHC alleles were involved in parasite resistance in the 2 habitats, showing that gene-parasite associations are not consistent between populations.     

Major histocompatibility complex class II DAB alleles associated with intestinal parasite load in the vulnerable Chinese egret (Egretta eulophotes)

The maintenance of major histocompatibility complex (MHC) polymorphism has been hypothesized to result from many mechanisms such as rare‐allele advantage, heterozygote advantage, and allele counting. In the study reported herein, 224 vulnerable Chinese egrets (Egretta eulophotes) were used to examine these hypotheses as empirical results derived from bird studies are rare. Parasite survey showed that 147 (65.63%) individuals were infected with 1–3 helminths, and 82.31% of these infected individuals carried Ascaridia sp. Using asymmetric polymerase chain reaction technique, 10 DAB1, twelve DAB2, and three DAB3 exon 2 alleles were identified at each single locus. A significant association of the rare allele Egeu‐DAB2*05 (allele frequency: 0.022) with helminth resistance was found for all helminths, as well as for the most abundant morphotype Ascaridia sp. in the separate analyses. Egeu‐DAB2*05 occurred frequently in uninfected individuals, and individuals carrying Egeu‐DAB2*05 had significantly lower helminth morphotypes per individual (HMI) (the number of HMI) and the fecal egg count values. Further, the parasite infection measurements were consistently lower in individuals with an intermediate number of different alleles in the duplicated DAB loci. Significantly, heterozygosity within each DAB locus was not correlated with any parasite infection measurements. These results indicate that the diversity in MHC Egeu‐DAB gene is associated with intestinal parasite load and maintained by pathogen‐driven selection that probably operate through both the rare‐allele advantage and the allele counting strategy, and suggest that Egeu‐DAB2*05 might be a valuable indicator of better resistance to helminth diseases in the vulnerable Chinese egret.     

The genetics of the sixth and seventh components of complement in the dog: Polymorphism,linkage, locus duplication,and silent alleles

The complement components C6 and C7 exhibit genetic polymorphism in the domestic dog. In the case of C6, there is a single locus with a null allele and two structural alleles; in the case of C7, there are two linked loci, each with three structural alleles. There is a null allele or locus deletion at one of these loci. In all cases, inheritance is autosomal and codominant. The C7 loci are closely linked to each other and to C6. This complex is not close to the dog major histocompatibility complex (MHC) locus.     

‘True’ null allele detection in microsatellite loci: a comparison of methods,assessment of difficulties and survey of possible improvements

Null alleles are alleles that for various reasons fail to amplify in a PCR assay. The presence of null alleles in microsatellite data is known to bias the genetic parameter estimates. Thus, efficient detection of null alleles is crucial, but the methods available for indirect null allele detection return inconsistent results. Here, our aim was to compare different methods for null allele detection, to explain their respective performance and to provide improvements. We applied several approaches to identify the ‘true’ null alleles based on the predictions made by five different methods, used either individually or in combination. First, we introduced simulated ‘true’ null alleles into 240 population data sets and applied the methods to measure their success in detecting the simulated null alleles. The single best‐performing method was ML‐NullFreq_frequency. Furthermore, we applied different noise reduction approaches to improve the results. For instance, by combining the results of several methods, we obtained more reliable results than using a single one. Rule‐based classification was applied to identify population properties linked to the false discovery rate. Rules obtained from the classifier described which population genetic estimates and loci characteristics were linked to the success of each method. We have shown that by simulating ‘true’ null alleles into a population data set, we may define a null allele frequency threshold, related to a desired true or false discovery rate. Moreover, using such simulated data sets, the expected null allele homozygote frequency may be estimated independently of the equilibrium state of the population.     

Family‐assisted inference of the genetic architecture of major histocompatibility complex variation

With their direct link to individual fitness, genes of the major histocompatibility complex (MHC) are a popular system to study the evolution of adaptive genetic diversity. However, owing to the highly dynamic evolution of the MHC region, the isolation, characterization and genotyping of MHC genes remain a major challenge. While high‐throughput sequencing technologies now provide unprecedented resolution of the high allelic diversity observed at the MHC, in many species, it remains unclear (i) how alleles are distributed among MHC loci, (ii) whether MHC loci are linked or segregate independently and (iii) how much copy number variation (CNV) can be observed for MHC genes in natural populations. Here, we show that the study of allele segregation patterns within families can provide significant insights in this context. We sequenced two MHC class I (MHC‐I) loci in 1267 European barn owls (Tyto alba), including 590 offspring from 130 families using Illumina MiSeq technology. Coupled with a high per‐individual sequencing coverage (~3000×), the study of allele segregation patterns within families provided information on three aspects of the architecture of MHC‐I variation in barn owls: (i) extensive sharing of alleles among loci, (ii) strong linkage of MHC‐I loci indicating tandem architecture and (iii) the presence of CNV in the barn owl MHC‐I. We conclude that the additional information that can be gained from high‐coverage amplicon sequencing by investigating allele segregation patterns in families not only helps improving the accuracy of MHC genotyping, but also contributes towards enhanced analyses in the context of MHC evolutionary ecology.     

Locus specificity of polymorphic alleles and evolution by a birth-and-death process in mammalian MHC genes

We have conducted an extensive phylogenetic analysis of polymorphic alleles from human and mouse major histocompatibility complex (MHC) class I and class II genes. The phylogenetic tree obtained for 212 complete human class I allele sequences (HLA-A, -B, and -C) has shown that all alleles from the same locus form a single cluster, which is highly supported by bootstrap values, except for one HLA-B allele (HLA-B*7301). Mouse MHC class I loci did not show locus-specific clusters of polymorphic alleles. This was considered to be because of either interlocus genetic exchange or the confusing designation of loci in different haplotypes at the present time. The locus specificity of polymorphic alleles was also observed in human and mouse MHC class II loci. It was therefore concluded that interlocus recombination or gene conversion is not very important for generating MHC diversity, with a possible exception of mouse class I loci. According to the phylogenetic trees of complete coding sequences, we classified human MHC class I (HLA-A, -B, and -C) and class II (DRB1) alleles into three to five major allelic lineages (groups), which were monophyletic with high bootstrap values. Most of these allelic groups remained unchanged even in phylogenetic trees based on individual exons, though this does not exclude the possibility of intralocus recombination involving short DNA segments. These results, together with the previous observation that MHC loci are subject to frequent duplication and deletion, as well as to balancing selection, indicate that MHC evolution in mammals is in agreement with the birth-and-death model of evolution, rather than with the model of concerted evolution.     

Sixteen microsatellite loci from Halesia tetraptera (Styracaceae)

Sixteen microsatellite loci isolated from Halesia tetraptera are described. All 16 are polymorphic, with four to 13 alleles across 15–19 members of a single half‐sib family. Heterozygosity ranged from 0.067 to 1. One locus departs significantly from Hardy–Weinberg equilibrium in our test family. The test population shows significant heterozygote deficiency at this and two other loci. Thirteen loci exhibit significant linkage disequilibrium. These loci will be utilized in paternity analyses of geographically diverse half‐sib families.     

Genotyping strategy matters when analyzing hypervariable major histocompatibility complex‐Experience from a passerine bird

Genotyping of classical major histocompatibility complex (MHC) genes is challenging when they are hypervariable and occur in multiple copies. In this study, we used several different approaches to genotype the moderately variable MHC class I exon 3 (MHCIe3) and the highly polymorphic MHC class II exon 2 (MHCIIβe2) in the bluethroat (Luscinia svecica). Two family groups (eight individuals) were sequenced in replicates at both markers using Ion Torrent technology with both a single‐ and a dual‐indexed primer structure. Additionally, MHCIIβe2 was sequenced on Illumina MiSeq. Allele calling was conducted by modifications of the pipeline developed by Sommer et al. (BMC Genomics, 14, 2013, 542) and the software AmpliSAS. While the different genotyping strategies gave largely consistent results for MHCIe3, with a maximum of eight alleles per individual, MHCIIβe2 was remarkably complex with a maximum of 56 MHCIIβe2 alleles called for one individual. Each genotyping strategy detected on average 50%–82% of all MHCIIβe2 alleles per individual, but dropouts were largely allele‐specific and consistent within families for each strategy. The discrepancies among approaches indicate PCR biases caused by the platform‐specific primer tails. Further, AmpliSAS called fewer alleles than the modified Sommer pipeline. Our results demonstrate that allelic dropout is a significant problem when genotyping the hypervariable MHCIIβe2. As these genotyping errors are largely nonrandom and method‐specific, we caution against comparing genotypes across different genotyping strategies. Nevertheless, we conclude that high‐throughput approaches provide a major advance in the challenging task of genotyping hypervariable MHC loci, even though they may not reveal the complete allelic repertoire.     

The role of major histocompatibility complex diversity in vigour of fish males (Abramis brama L.) and parasite selection

The major histocompatibility complex (MHC) presents a group of genes with highly polymorphic loci involved in specific immune responses. The factors maintaining extensive MHC polymorphism have been questioned, considering three possible hypotheses of parasite‐mediated selection driving an extensive MHC diversity (i.e. heterozygote advantage, rare‐allele advantage, and favouring optimal MHC diversity). The patterns of MHC diversity of class IIB genes were investigated following two noncontradicting hypotheses, parasite‐driven selection and MHC‐based mating preferences, using males of common bream collected in the spawning period. Two allelic groups DAB1 and DAB3 were recognized from the phylogenetic analyses. Individuals expressed one or two alleles of the same or different allelic groups. Several individuals shared identical alleles; however, the presence of parasite species was not associated with the occurrence of a particular allele. The presence of different allelic groups (only DAB1, only DAB3, or both DAB1 and DAB3) in individuals was not associated with parasite presence or diversity. The expression of two DAB1 alleles was associated with higher endoparasite abundance. Moreover, nucleotide diversity in individuals expressing a single type of alleles (DAB1 or DAB3) increased with the abundance of ectoparasitic Dactylogyrus spp. (Monogenea) and Ergasilus sp. (Crustacea). This suggests that the expression of two alleles of a single allelic type is related to high metazoan parasite infection whereas no significant influence of parasitism on the combined allelic form (the presence of both DAB1 and DAB3 alleles) was found. Moreover, the expression of two alleles of a single allelic type was related to decreased immunocompetence measured by spleen size. The condition factor was higher in fish expressing the combined allelic type. Thus, the presence of alleles of different lineages in individuals appears to be advantageous for individual male fitness. The expression of a single allelic type was related to higher sexual ornamentation, which could support the role of MHC in the hypothesis of the sexual selection of ‘good genes’. © 2007 The Linnean Society of London, Biological Journal of the Linnean Society, 2007, 90 , 525–538.     

Between‐year variation of MHC allele frequencies in great reed warblers: selection or drift?

The major histocompatibility complex (MHC) genes are extremely polymorphic and this variation is assumed to be maintained by balancing selection. Cyclic interactions between pathogens and their hosts could generate such selection, and specific MHC alleles or heterozygosity at certain MHC loci have been shown to confer resistance against particular pathogens. Here we compare the temporal variation in allele frequencies of 23 MHC class I alleles with that of 23 neutral microsatellite markers in adult great reed warblers (a passerine bird) in nine successive cohorts. Overall, the MHC alleles showed a significantly higher variation in allele frequencies between cohorts than the microsatellite alleles, using a multi-variate genetic analysis (amova). The frequency of two specific MHC alleles, A3e (P = 0.046) and B4b (P = 0.0018), varied more between cohorts than expected from random, whereas none of the microsatellite alleles showed fluctuations exceeding the expectation from stochastic variation. These results imply that the variation in MHC allele frequencies between cohorts is not a result of demographic events, but rather an effect of selection favouring different MHC alleles in different years.     

Experimental evidence for major histocompatibility complex-allele-specific resistance to a bacterial infection

The extreme polymorphism found at some major histocompatibility complex (MHC) loci is believed to be maintained by balancing selection caused by infectious pathogens. Experimental support for this is inconclusive. We have studied the interaction between certain MHC alleles and the bacterium Aeromonas salmonicida, which causes the severe disease furunculosis, in Atlantic salmon (Salmo salar L.). We designed full-sibling broods consisting of combinations of homozygote and heterozygote genotypes with respect to resistance or susceptibility alleles. The juveniles were experimentally infected with A. salmonicida and their individual survival was monitored. By comparing full siblings carrying different MHC genotypes the effects on survival due to other segregating genes were minimized. We show that a pathogen has the potential to cause very intense selection pressure on particular MHC alleles; the relative fitness difference between individuals carrying different MHC alleles was as high as 0.5. A co-dominant pattern of disease resistance/susceptibility was found, indicative of qualitative difference in the immune response between individuals carrying the high- and low-resistance alleles. Rather unexpectedly, survival was not higher among heterozygous individuals as compared with homozygous ones.     

Detection of linkage between marker loci and loci affecting quantitative traits in crosses between segregating populations

Summary By making use of pedigree information and information on marker-genotypes of the parent and F-1 individuals crossed to form an F-2 population, it is possible to carry out a linkage analysis between marker loci and loci affecting quantitative traits in a cross between segregating parent populations that are at fixation for alternative alleles at the QTL, but share the same alleles at the marker loci. For two-allele systems, depending on marker allele frequencies in the parent populations, 2–4 times as many F-2 offspring will have to be raised and scored for markers and quantitative traits in order to provide power equivalent to that obtained in a cross between fully inbred lines. Major savings in number of F-2 offspring raised can be achieved by scoring each parent pair for a large number of markers in each chromosomal region and scoring F-1 and F-2 offspring only for those markers for which the parents were homozygous for alternative alleles. For multiple allele systems, particularly when dealing with hypervariable loci, only 10%–20% additional F-2 offspring will have to be raised and scored to provide power equivalent to that obtained in a cross between inbred lines. When a resource population contains novel favorable alleles at quantitative trait loci that are not present (or rare) in a commercial population, analyses of this sort will enable the loci of interest to be identified, mapped and manipulated effectively in breeding programs.Contribution no. 2124-E, 1987 series from The Agricultural Research Organization, The Volcani Center, Bet Dagan, Israel     

Selective pressures on MHC class II genes in the guppy (Poecilia reticulata) as inferred by hierarchical analysis of population structure

Genes of the major histocompatibility complex, which are the most polymorphic of all vertebrate genes, are a pre‐eminent system for the study of selective pressures that arise from host–pathogen interactions. Balancing selection capable of maintaining high polymorphism should lead to the homogenization of MHC allele frequencies among populations, but there is some evidence to suggest that diversifying selection also operates on the MHC. However, the pattern of population structure observed at MHC loci is likely to depend on the spatial and/or temporal scale examined. Here, we investigated selection acting on MHC genes at different geographic scales using Venezuelan guppy populations inhabiting four regions. We found a significant correlation between MHC and microsatellite allelic richness across populations, which suggests the role of genetic drift in shaping MHC diversity. However, compared to microsatellites, more MHC variation was explained by differences between populations within larger geographic regions and less by the differences between the regions. Furthermore, among proximate populations, variation in MHC allele frequencies was significantly higher compared to microsatellites, indicating that selection acting on MHC may increase population structure at small spatial scales. However, in populations that have significantly diverged at neutral markers, the population‐genetic signature of diversifying selection may be eradicated in the long term by that of balancing selection, which acts to preserve rare alleles and thus maintain a common pool of MHC alleles.