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.

Amplification is the Primary Mode of Gene-by-Sex Interaction in Complex Human Traits

“…Sex differences in polygenic effects on traits are clearly of high evolutionary [22][23][24][25][26] and translational [27][28][29][30][31] importance. Yet with the exception of testosterone levels [32][33][34][35] , the genetic basis of sexual dimorphism is not well-understood 31 .…”

Systematic review reveals sexually antagonistic knockouts in model organisms

“…Only one of these studies, the VGLL polymorphism in Atlantic salmon affecting body size and age at maturity, was previously known to us (Barson et al, 2015). This was also the only study to describe the effect in terms of sexual antagonism, reflecting the problem of discipline‐specific terminology (Harper et al, 2021). This is far fewer variants than expected given 51 variants with sex‐opposite or sexually antagonistic effects were previously found in a systematic review of humans (Harper et al, 2021).…”

“…This was also the only study to describe the effect in terms of sexual antagonism, reflecting the problem of discipline‐specific terminology (Harper et al, 2021). This is far fewer variants than expected given 51 variants with sex‐opposite or sexually antagonistic effects were previously found in a systematic review of humans (Harper et al, 2021). Most examples come from the mouse model, which has been intensively studied in terms of its genetics for several decades.…”

“…To place the fitness impacts of the SA variants identified in this study in a broader context, we began by calculating their effect size ratios (Harper et al, 2021; Morrow & Connallon, 2013), where the positive effect of a variant in one sex is divided by the negative effect in the opposite sex. An effect size ratio of −1 for a particular variant indicates that the magnitude of the positive and negative effects are perfectly balanced, net positive variants will have effect size ratios below −1, and net negative variants will have effect size ratios above −1.…”

“…In the fruit‐fly Drosophila melanogaster , a recent genome‐wide association study of SA variation identified a number of candidate loci that warrant independent validation (Ruzicka et al, 2019). More recently, a systematic review of the biomedical literature found around 50 specific examples of genes in humans with sex‐opposite or SA effects on traits or disease risk (Harper et al, 2021). As predicted (Morrow & Connallon, 2013), allele frequencies at these loci were higher when the net effects were positive.…”