We have examined HLA-A and HLA-B in a large sample of families of South Amerindians for non-random mating and for Mendelian segregation in their progeny. The proportion of couples sampled for mating averaged over 50% of the total matings with progeny for 10 of the 11 tribes with 10 or more known matings. In nearly all cases, HLA-sharing proportions were very close to those expected from random mating, suggesting strong negative-assortative mating for MHC is not present in these South Amerindians. In progeny from these matings and others in 23 tribes, there was a large deficiency of homozygotes (excess of heterozygotes) compared with Mendelian expectations. In particular, there was a deficiency in all types of matings except when the female was a homozygote and the male shared one allele with her, the reciprocal mating showed no such deficiency. This non-reciprocity suggests the importance of some type of maternal-fetal interaction. A model to describe these results showed very surprising properties for two alleles: no change in allelic frequency like neutrality, an excess of heterozygotes like heterozygote advantage, and a minimum fitness at intermediate allelic frequency like heterozygote disadvantage. For three or more alleles, the model has the qualities of a traditional balancing selection model.
ASJC Scopus subject areas