Systematic review reveals multiple sexually antagonistic polymorphisms affecting human disease and complex traits

Abstract

Sex differences in human disease risk and incidence are widely documented but their origins are poorly understood. An evolutionary model for differential disease risk between the sexes posits that alleles that are a risk factor (deleterious) in one sex may actually be protective (beneficial) in the opposite sex, and as such these sexually antagonistic (SA) alleles may be maintained in the population at frequencies higher than expected under purifying selection against unconditionally deleterious alleles. Sexual antagonism at genetic loci is thought to arise due to fundamental differences in how Darwinian fitness is maximized in the two sexes. Evidence for the occurrence of SA alleles comes from a number of non-human vertebrate and invertebrate models, which indicate they are common, although only very few specific SA loci have been identified. While there are many examples of loci with sex-limited effects in humans, there are no known examples of specific genetic variants that experience opposing patterns of selection between the sexes in humans. We hypothesized that this may be due to a mismatch in terminology between scientific disciplines rather than a genuine lack of examples. We therefore undertook a two-stage review of the literature for evidence of genetic polymorphisms in human disease and complex traits that exert their effects in a SA pattern, extracting effect sizes and allele frequencies where possible. In the first stage, we used search terms from evolutionary biology that aimed to identify examples of genes specifically referred to as having SA effects or were implicated in intralocus sexual conflict. In the second, we used terminology that was designed to capture as many examples as possible from the biomedical literature (e.g. sex-dependent, sex-opposite, men/women, boys/girls), which may have been missed in the first stage due to differences in terminology between the disciplines. The two searches gave very different results. The first stage returned 34 papers in total, none of which included evidence from humans of genetic loci with opposite effects on the two sexes. The second stage returned 881 papers, which after screening resulted in 70 papers, 32 were deemed eligible, and described 51 different genetic loci or haplotypes with sex-opposite effects on complex traits or disease. Of these, 22 loci increase disease risk or severity in one sex but are protective in the other. These loci may therefore be considered SA loci if the effects on disease risk translate into marginal effects on fitness through fecundity selection or mortality, although in no case were they referred to in terms of sexual antagonism. As predicted, alleles with net positive effects across the sexes occurred at a higher frequency than those with net negative effects. Our review reveals that significant long-term barriers to communication are occurring between scientific fields as a result of discipline-specific terminology. Less specific search terms revealed evidence that SA variants are maintained at relatively high frequencies in humans, apparently because while they contribute to the propensity for one sex to develop a range of diseases and disorders, they protect the other.