Gene Drive and Sexual Selection in House Mice

At first sight, biological organisms appear as harmonious entities, armed with features exquisitely fine-tuned to its survival and reproduction. This is no accident: Darwin’s theory of evolution by natural selection entails that genes spread through populations, exactly because they contribute to organismal fitness. However, biologists are uncovering ever more cases of genetic entities, so called drive genes, that are at odds with the notion of the organism as the (sole) fitness-maximising agent. By violating the Mendelian rules of inheritance, drive genes successfully spread through populations, often despite detrimental consequences for their carriers. The t haplotype in house mice (Mus musculus domesticus) is the paradigm example of a drive element. This cluster of genes, occupying about one third of mouse chromosome 17, is fatal to the fitness of its hosts: t/t homozygotes die from recessive lethal mutations during embryogenesis. On its own, this would predict immediate extinction of the t haplotype. Yet, t haplotypes systematically manipulate gametogenesis in their favour. As a result +/t heterozygous males transmit the t haplotype, instead of the usual 50%, to up to 90% of their progeny. This selfish disruption of the fair Mendelian ratios has allowed t haplotypes to spread through house mouse populations around the world in spite of the harm they incur to the individuals and populations harbouring them. The stable presence of a drive gene has profound evolutionary consequences. Once the ‘fair’ rules of meiosis have failed at maintaining the integrity and function of the organism, we expect selection on the organism to evolve other measures to suppress drive element’s selfish acts. For example, it has been suggested that females may avoid fertilisation by t haplotype carrying sperm, because such avoidance will protect her offspring form the t related fitness costs. Two mechanisms of sexual selection against drive genes have received particular attention in the literature. First, females could avoid t fertilisation by avoiding t-haplotype-carrying males prior to mating. Second, females could avoid t fertilisation by systematically mating with several males (termed polyandry). This second hypothesis is based on the premise that drive-carrying males are compromised in their sperm competitive ability. In this thesis, I have investigated the joint evolution of a gene drive and the female mating behaviour. Using a variety of methodological approaches, ranging from theoretical modelling to laboratory experiments to data analysis from a natural population, I have addressed the following two questions. (1) How does female mating behaviour affect the frequency dynamics of a drive gene? Understanding the evolutionary forces that determine the frequency of drive genes in natural populations is a long-standing focus of evolutionary biology. In the case of the t haplotype, naturally observed frequencies are typically much lower than expected based on drive and homozygote lethality alone (this discrepancy between the standard model and data has been termed ‘the t frequency paradox’). In this dissertation, I show that the inclusion of female mating behaviour to the standard model, polyandry in particular, greatly improves our t frequency predictions and can largely explain the low t frequencies observed in nature. In a mate choice experiment, we show that females are indeed able to avoid fertilisation by drive-carrying males. The absence of clear social preferences during the choice test suggested that this fertilisation bias is largely driven by polyandry and subsequent sperm competitive effects. The importance of sperm competitive effects was corroborated in a second study. Here, we provide direct evidence that t carrying males are heavily compromised in their sperm competitive ability. Accordingly, t-carrying males only sired 19% of offspring when competing fertilisation with a wild-type male. We further show that this disadvantage has direct implications on population t frequencies. We found that in a selection experiment, where mice were kept under strictly monogamous or polyandrous conditions over the course of 20 generations, t frequencies have significantly decreased in the polyandrous selection lines, while remaining constant and high in the monogamous lines. For the first time in any drive system, we provide evidence that such sperm competitive effects are directly relevant under natural conditions. In an intensively monitored house mouse population outside Zürich, we found that the reproductive success of t males was particularly strongly affected by sperm competition. Moreover, females are highly polyandrous: over 47% of litters born during the 4.5 year observation period were sired by more than one father. In line with the ‘polyandry hypothesis’, we observe a decline in population t frequencies during the investigation period. (2) How does the presence of drive gene affect the evolution of female mating behaviour? Understanding the evolutionary forces that drive the evolution of female mating behaviour (such as mate choice and polyandry) is a highly debated topic in evolutionary research. It has been hypothesised that mate choice and/or polyandry is beneficial to females because they will result in fertilisation by males of a high genetic quality (‘good genes’ or ‘good sperm’ hypotheses). Yet conventional genetic mechanisms are usually insufficient to maintain variation in male genetic quality, thereby rendering any form of choice obsolete (this problem is generally known as the ‘lek paradox’). Using a theoretical model, we show here that the presence of a drive gene can greatly facilitate the evolution of female choice, even in circumstances where such a choice is associated with direct fitness costs. First, costly drive-male avoidance is beneficial to females because it helps them avoid drive related fitness costs. Second, costly drive-male avoidance is evolutionarily stable, because gene drive maintains variation in male genetic quality at equilibrium. As a result, the lek paradox is largely avoided. Despite this compelling theoretical argument, we have found little evidence that the presence of the t haplotype has triggered the evolution of female drive avoidance in the circumstances considered here. While polyandry helped females avoid t related litter losses in the laboratory, we find no signs of selection on polyandry rates under natural conditions. Moreover, we find little evidence that polyandry is heritable. Thus, even in the case of drive-triggered selection on polyandry, it is unlikely that the trait would respond to selection