MHC Opposites Attract? (or not)

I've mentioned MHC before; they're interesting from an evolutionary perspective because they're one of the most polymorphic loci in humans, likely due to their role in disease resistance and the "diversified portfolio" strategy which seems optimal over the long term (kind of like sex). The logic is pretty simple, pathogens are good at adapting because there are so many who replicate so fast so they figure out ways to get around common defensive strategies before a few host generations have come to pass. This increases the fitness of the rare alleles in the hosts, until they become not-so-rare because they rise in frequency, at which point the cycle starts all over again. Negative frequency dependent selection or feedback loop.

But obviously this story isn't just of interest as a model for long term evolutionary dynamics and how they are shaped by co-evolutionary forces. There are plenty of animal models which suggest that organisms attempt to mate with individuals who are different from them so that the offspring will have a diversified suite of alleles across the MHC loci. Many have wondered if this applies to humans. The data so far has been equivocal, confusing and contradictory. A short paper in PLoS Genetics answers little, but clarifies some, Is Mate Choice in Humans MHC-Dependent?:

...Several studies have reported a tendency for humans to prefer MHC-dissimilar mates, a sexual selection that would favour the production of MHC-heterozygous offspring, who would be more resistant to pathogens, but these results are unsupported by other studies. Here, we report analyses of genome-wide genotype data...and HLA types in African and European American couples to test whether humans tend to choose MHC-dissimilar mates. In order to distinguish MHC-specific effects from genome-wide effects, the pattern of similarity in the MHC region is compared to the pattern in the rest of the genome. African spouses show no significant pattern of similarity/dissimilarity across the MHC region...whereas across the genome, they are more similar than random pairs of individuals...We discuss several explanations for these observations, including demographic effects. On the other hand, the sampled European American couples are significantly more MHC-dissimilar than random pairs of individuals...and this pattern of dissimilarity is extreme when compared to the rest of the genome, both globally...and when broken into windows having the same length and recombination rate as the MHC (only nine genomic regions exhibit a higher level of genetic dissimilarity between spouses than does the MHC). This study thus supports the hypothesis that the MHC influences mate choice in some human populations.

As I said, the paper is a short read, and the Discussion does a good job of fleshing out various inferences. I've edited the chart to the left (the red text is mine), you can see immediately via inspection that the frequency distribution of Utah white couples and their relatedness differs from the Yoruba. The Utah white sample exhibited a tendency for the MHC loci in particular to be sharply deviated from random expectation in terms of relatedness of the total genome content. The MHC differed more than they should have. Disassortative mating along the dimensions of the MHC. For the Yoruba there wasn't really any statistically significant trend when it came to the MHC, but the couples were more closely related in terms of total genome content than one would have expected. Deviation from a random mating equilibrium, as if there was population substructure.

It is critical to note the nature and size of these samples, 30 Utah white and Yoruba couples each. I'm not worried about the size as much as the representativeness of either for some general insight into human nature. I do not know the recent demographic history of the Yoruba, though its seems plausible in a society where patrilineages are important that kinship ties would bias marriage patterns toward some non-trivial consanguinity. On the other hand, in the United States cousin marriage rates are very low, and mobility is relatively high compared to most of the world. But the Utah white sample is also peculiar in terms of its history; I am to understand these families were chosen in part because of their large sizes for medical studies. Because of their Mormon religion many of them have ancestors who were in polygynous relationships and were part of a massive demographic expansion on the American frontier. Therefore the authors compare them to Hutterites, another rather homogeneous population which has high fertility. They suggest that perhaps a maximal amount of MHC diversity is not the aim, but an optimal amount. The whites of Utah are more genetically homogeneous than the Yoruba, so there are more marginal returns on diversifying the MHC loci. But even this is certis paribus, reality is messier than this, and so there might be more than one optimal mix of alleles, and in many circumstances that optimal mix may simply be irrelevant where there are far more powerful parameters shaping the distribution of the relatedness curve. A small, hopeful piece of the puzzle nonetheless....