Gene duplication might not resolve intralocus sexual conflict

Abstract

Intralocus sexual conflict occurs when the fitness optima at a locus differ between the sexes [1xSexually antagonistic genes: experimental evidence. Rice, W.R. Science. 1992; 256: 1436–1439Crossref | PubMedSee all References[1]. This form of sexual antagonism has now been demonstrated in many taxa, from insects to birds to mammals [2xIntralocus sexual conflict. Bonduriansky, R. and Chenoweth, S.F. Trends Ecol. Evol. 2009; 24: 280–288Abstract | Full Text | Full Text PDF | PubMed | Scopus (249)See all References[2], and is thought to underlie the widespread sexual dimorphism seen in nature [3xSexual dimorphism, sexual selection and adaptation in polygenic characters. Lande, R. Evolution. 1980; 34: 292–305CrossrefSee all References[3]. Despite this, debate about the evolutionary importance of intralocus conflict continues, in part because it has been thought to be transient [4xIntralocus sexual conflict diminishes the benefits of sexual selection. Pischedda, A. and Chippindale, A.K. PLoS Biol. 2006; 4: E356Crossref | PubMed | Scopus (91)See all References[4]. This transience is a consequence of the many purported mechanisms that are thought to extinguish the sexually antagonistic selection generating intralocus conflict.One of these mechanisms was recently discussed in TREE by Gallach and Betran [5xIntralocus sexual conflict resolved through gene duplication. Gallach, M. and Betran, E. Trends Ecol. Evol. 2011; 26: 222–228Abstract | Full Text | Full Text PDF | PubMed | Scopus (41)See all References[5]. They provided a simple graphical model showing how gene duplication could lead to the resolution of intralocus conflict. It was suggested that resolution occurred when the duplicate gene, which initially had sexually antagonistic effects (increasing male and decreasing female fitness) evolved sex (tissue)-specific expression, which allowed males to achieve their fitness optima without interference from selection on the duplicated gene in females. Thus, gene duplication followed by sex-limited gene expression was seen to resolve intralocus conflict. However, although plausible at face value, this scenario ignores two issues, one inevitable and one potential, and both complicate conflict resolution.As has been discussed previously in TREE, even after gene duplication and the evolution of sex-limited gene expression, intralocus conflict still occurs, albeit in a weakened form [2xIntralocus sexual conflict. Bonduriansky, R. and Chenoweth, S.F. Trends Ecol. Evol. 2009; 24: 280–288Abstract | Full Text | Full Text PDF | PubMed | Scopus (249)See all References[2]. This is because deleterious mutations enter the gene pool and are sheltered from selection in the non-expressing sex [2xIntralocus sexual conflict. Bonduriansky, R. and Chenoweth, S.F. Trends Ecol. Evol. 2009; 24: 280–288Abstract | Full Text | Full Text PDF | PubMed | Scopus (249)See all References[2]. This deleterious input, either from mutation or gene flow, will displace the sex that is expressing the duplicate gene from their fitness optima. This insight highlights the fundamental basis of intralocus conflict: sexual differences in the fitness surface [2xIntralocus sexual conflict. Bonduriansky, R. and Chenoweth, S.F. Trends Ecol. Evol. 2009; 24: 280–288Abstract | Full Text | Full Text PDF | PubMed | Scopus (249)See all References[2]. For the situation being discussed, this surface is flat in the non-expressing sex (because they do not express the duplicated gene) and directional in the expressing sex as expression increases their fitness. Thus, gene duplication and sex-limited expression of the duplicated gene does not resolve sexual conflict as envisaged [5xIntralocus sexual conflict resolved through gene duplication. Gallach, M. and Betran, E. Trends Ecol. Evol. 2011; 26: 222–228Abstract | Full Text | Full Text PDF | PubMed | Scopus (41)See all References[5] because mutational input is inevitable even when there is no gene flow.The second issue, which also has the potential to complicate the basic scenario pictured by Gallach and Betran, was highlighted in a recent experimental evolution study [6xIntralocus sexual conflict unresolved by sex-limited trait expression. Harano, T. et al. Curr. Biol. 2010; 20: 2036–2039Abstract | Full Text | Full Text PDF | PubMed | Scopus (48)See all References[6] and relates to the genetic variance–covariance matrix. The study showed that even when sex-limitation occurs, genetic architecture can constrain intralocus conflict resolution. This is because characters within a sex are rarely (never?) uncorrelated with other characters and this intrasexual covariance, coupled with the appropriate intersexual covariance, can mean that selection on one sex still generates correlated responses in the other (Figure 1Figure 1). This of course has the potential to stop either from reaching their sex-specific fitness optima, and can occur even though the intersexual genetic covariance for the trait (or allele) in question is zero. I am not aware of any systematic studies of the covariance relationships for duplicated genes, so it is currently unclear how selection on traits encoded by them reverberates through the phenotype. Nonetheless, there is evidence that gene duplicates are extensively correlated with other genes and gene networks throughout the genome [7xUncovering genetic regulatory network divergence between duplicate genes using yeast eQTL landscape. Zhou, Y.Y. et al. J. Exp. Zool. B. 2009; 312: 722–733Crossref | Scopus (7)See all References[7]. As a result, the potential impacts of selection on duplicated genes via covariance with other genes cannot be ignored and, in addition to the evolution of sex-limited gene expression, genetic covariances would also need to evolve appropriately before this potential issue is negated.Figure 1A pictorial example of the potential impact of genetic architecture on intralocus sexual conflict. Arrows represent genetic correlations (owing to either pleiotropy or linkage). Trait 1 (or locus 1) is expressed in males (blue box in row labelled ‘Trait 1’), but not in females (black box covered by red cross in the same row), and is genetically correlated (black arrow) with Trait 2 (locus 2) in males (blue box in Trait 2 row). Trait 2 is also expressed in females (black box in Trait 2 row), and there is intersexual genetic covariance for this trait (the red arrow joining the blue and black boxes in the Trait 2 row). Selection on Trait 1 in males can generate a response in Trait 2 in females via these correlations and could, in principle, dislodge females from their Trait 2 optima even when there is no direct selection on Trait 2 in males (i.e. direct selection on Trait 2 is not sexually antagonistic). Therefore, even though there is sex-limited trait expression, there can be intralocus sexual conflict and this has a sex-limited trait at its source (here, Trait 1).View Large Image | Download PowerPoint SlideBased on all this, the suggestion that intralocus conflict could be resolved via gene duplication might not be 100% accurate. In the best-case scenario, and as Bonduriansky and Chenoweth previously reported in TREE, the conflict might be weakened, but this is not the same as being resolved.