Molecular Evolution of Freshwater Snails with Contrasting Mating Systems

Molecular Evolution of Freshwater Snails with Contrasting Mating Systems Concetta Burgarella,*,1 Philippe Gayral,y,1 Marion Ballenghien,1 Aurelien Bernard,1 Patrice David,2 Philippe Jarne,2 Ana Correa,3 Sylvie Hurtrez-Boussès,3 Juan Escobar,z,1 Nicolas Galtier,1 and Sylvain Glemin1 1 Institut des Sciences de l’Evolution, UMR, CNRS 5554, Universite Montpellier II, Montpellier, France CEFE/CNRS Campus du CNRS 1919, Montpellier, France 3 MIVEGEC (Maladies Infectieuses et Vecteurs: Ecologie, Genetique, Evolution, Contr^ole), UMR (UM1-UM2-CNRS 5290-IRD224), IRD, Montpellier, France y Present address: Institut de Recherche sur la Biologie de l’Insecte, UMR 7261, CNRS, Universite François-Rabelais, Tours, France z Present address: Vidarium—Nutrition, Health and Wellness Research Center, Grupo Nutresa, Medellın, Colombia *Corresponding author: E-mail: . Associate editor: Stephen Wright 2 Because mating systems affect population genetics and ecology, they are expected to impact the molecular evolution of species. Self-fertilizing species experience reduced effective population size, recombination rates, and heterozygosity, which in turn should decrease the efficacy of natural selection, both adaptive and purifying, and the strength of meiotic drive processes such as GC-biased gene conversion. The empirical evidence is only partly congruent with these predictions, depending on the analyzed species, some, but not all, of the expected effects have been observed. One possible reason is that self-fertilization is an evolutionary dead-end, so that most current selfers recently evolved self-fertilization, and their genome has not yet been strongly impacted by selfing. Here, we investigate the molecular evolution of two groups of freshwater snails in which mating systems have likely been stable for several millions of years. Analyzing coding sequence polymorphism, divergence, and expression levels, we report a strongly reduced genetic diversity, decreased efficacy of purifying selection, slower rate of adaptive evolution, and weakened codon usage bias/GC-biased gene conversion in the selfer Galba compared with the outcrosser Physa, in full agreement with theoretical expectations. Our results demonstrate that self-fertilization, when effective in the long run, is a major driver of population genomic and molecular evolutionary processes. Despite the genomic effects of selfing, Galba truncatula seems to escape the demographic consequences of the genetic load. We suggest that the particular ecology of the species may buffer the negative consequences of selfing, shedding new light on the dead-end hypothesis. Key words: mating systems, molecular evolution, freshwater snails, selfing, selection, base composition. Life-history and ecological traits are expected to influence genome organization and evolution through their effects on population genetic processes (Lynch 2007). In particular, mating systems affect fundamental population genetic parameters, such as effective population size, recombination rates, and the efficacy of natural selection. Therefore, they potentially influence patterns of polymorphism, rates of molecular evolution, and genomic base composition (Jarne 1995; Charlesworth and Wright 2001; Wright et al. 2008; Glemin and Galtier 2012). Selfing is a widely distributed reproductive mode, fairly common in various families of plants and animals (Vogler and Kalisz 2001; Jarne and Auld 2006). Compared with outcrossing, selfing is expected to reduce the effective population size, Ne, because of nonindependent gamete sampling and prevalence of genome-wide homozygosity (Pollak 1987; Nordborg 2000). All other things being equal, Ne = N/(1 + FIS), where FIS is the Wright’s inbreeding coefficient, and N the effective size under panmixia, so that Ne is halved in case of complete self-fertilization (when FIS = 1). Ne is expected to be reduced further 1) by hitchhiking effects, such as background selection (Charlesworth et al. 1993; Kamran-Disfani and Agrawal 2014), because selfing also reduces effective recombination rates (Nordborg 2000); and 2) by bottleneck effects, which are thought to be more frequent in selfers because a single or a few individuals can create a new population (Schoen and Brown 1991; Jarne 1995). At the species levels, stronger population genetic structure (higher FST) is expected in selfers than in outcrossers, because of both higher local genetic drift and reduced gene flow through the male function (for organisms that disperse during the male haploid phase). High FST can increase global Ne (Whitlock and Barton 1997); however, when extinction–recolonization dynamics occur, selfing also reduces Ne at the species level (Ingvarsson 2002). This should be reinforced by the local reduction of Ne through hitchhiking effects. As levels of neutral genetic diversity are proportional to Ne, where  is the mutation rate, lower polymorphism levels are expected in selfers, both at the population and the species level (table 1). Along the same lines, the efficacy ß The Author 2015. Published by Oxford University Press on behalf of the Society for Molecular Biology and Evolution. All rights reserved. For permissions, please e-mail: Mol. Biol. Evol. 32(9):2403–2416 doi:10.1093/molbev/msv121 Advance Access publication May 14, 2015 2403 Article Introduction Abstract MBE Burgarella et al. . doi:10.1093/molbev/msv121 Table 1. Effects of Mating System on Genomic Features According to Predictions and Observed in This Study. Predicted Polymorphism Populations structure (FIS, FST) Selection efficacy (purifying selection, adaptive selection, selection on codon usage) Sexual conflicts (rapid evolution of male gametes) gBGC Observed Outcrossing +  + Selfing  +  Yes Yes Yes +  No +  Yes of selection, which depends on Nes (where s is the selection coefficient; Kimura 1983), should be lower in selfing species (table 1). If a large fraction of mutations are weakly deleterious, as supported empirically (Eyre-Walker and Keightley 2007), we thus expect higher nonsynonymous versus synonymous polymorphism (N/S) and divergence (dN/dS) ratios in selfing than in outcrossing lineages, reflecting the accumulation and fixation of deleterious mutations (Glemin 2007). For moderate to high selfing rates, purifying selection is also predicted to be reduced in subdivided populations (Roze and Rousset 2004). However, strongly deleterious and partly recessive mutations should be more easily purged in selfing species, somewhat leveling the above-mentioned effect (Glemin 2003). On the other hand, fixation of advantageous alleles should be reduced in selfers (at least when they are not too recessive), lowering the rate of adaptive substitutions (Glemin 2007). Similarly, bias in codon usage is expected to be weaker in selfers due to reduced efficacy of selection on translational speed/accuracy. Finally, mating systems might also affect the genome base composition through (...truncated)