The role of the pleiotropic genes of the major histocompatibility complex in evolution: The example of the three-spined stickleback

Plants and animals share their environments with a rich fauna of parasites and potentially, there is no species without at least one pathogen. Selection by parasites is recognized as one of the principal evolutionary processes and has probably even led to the evolution of sexual reproduction- a major paradox in evolutionary biology. To counter this constant threat, vertebrates have developed different immune defenses. In particularly the major histocompatibility complex (MHC) plays an important role in the vertebrate immune system. MHC molecules present self and non-self peptides to T-cells and when a foreign peptide is bound it triggers an immune reaction. Since a given MHC molecule can only present a limited number of different peptides, an immune response can be initiated against a limited amount of pathogens per MHC allele. Probably due to this limitation and the immense diversity of parasites, MHC genes present the highest degree of polymorphism reported in jawed vertebrates. This extreme intra- and inter-individual polymorphism is thought to be under balancing selection. The mechanisms by which MHC polymorphism could be maintained are fiercely debated and several theories have been proposed: heterozygote advantage, negative frequency dependent selection and/or habitat heterogeneity. In addition to its puzzling polymorphism, females have been shown to exert MHCdependent mate choice. Such preference may be better understood in the light of commonly cited indirect benefits of mate choice including the search for good genes, inbreeding avoidance and acquisition of resistance genes for the progeny. In the work detailed in this thesis, I (and my co-authors) used the three-spined stickleback as a model organism to investigate the role of the MHC in evolution. In an initial study (Chapter I), we explored whether MHC genes were correlated to important individual fitness traits. While we found a higher nest quality for males harboring an optimal MHC class IIB diversity, male breeding coloration was most intense at a maximal MHC class I diversity. Two MHC class I alleles were identified to be associated with a higher intensity of red coloration. In this study, we showed that individual sexual displays are MHC-dependent and thus could influence female mating decision. It had previously been demonstrated that fish with an intermediate MHC diversity were better at resisting parasites. Our results from chapter I showed that the MHC affects sexual characters as well. However, the ultimate measure of evolutionary fitness is the reproductive success. We combined this knowledge in a field experiment to test the relation between MHC and reproductive success- the ultimate factor of evolution (Chapter II). Using laboratory-bred families segregating for MHC genes, we found that reproductive success was higher at intermediate MHC diversity. In order to test for the generality of this phenomenon, we stocked the same field enclosures with wild-caught fish (Chapter III). We confirmed that with reference to their own MHC profile, female sticklebacks preferred to mate with males sharing an intermediate MHC diversity. Moreover, males with a specific MHC haplotype were better atresisting a common parasite (Gyrodactylus sp.) and were more likely to be selected by females. Interestingly, this translated directly into Darwinian fitness since fish harboring this specific MHC haplotype had a higher reproductive output. We concluded that females based their mating decision on a specific MHC haplotype conferring resistance against a common parasite. With this strategy, females achieve a two fold immunogenetic advantage: resistance against a wide range of parasites while keeping the T-Cell repertoire depletion to a minimum level and resistance against a currently common and costly parasite. The latter can be described as “good genes”. Such an interaction between host and parasite driving assortative mating is a pre-requisite to negative-frequency-dependent selection- a potential mechanism to explain MHC polymorphism. Our studied not only identified MHC as good genes but also suggested that females might adopt reproductive strategies that minimize disruption of locally adapted MHC genes against parasites. Such results motivated a reflection on the role of host MHC and parasite interaction in the evolution of population divergence. In chapter IV, we proposed that MHC genes could play a pleiotropic role in parasite resistance and female mating strategy. Individual MHC diversity is restricted compared to the entire allele pool due to a low optimal number of MHC loci. In contrasting niches, parasites communities differ and thus the pools of MHC alleles also differ between niches since not all alleles can be exhibited. Mate choice for the best adapted MHC genotype will favor local adaptations and will increase the pace of population differentiation and speciation. We therefore propose to consider MHC genes as potential “magic traits”. To investigate the pleiotropic role of MHC genes, we sampled stickleback populations from a lake and a connected river (chapter V), which previous studies have suggested represent ecomorphs. The two fish populations harbored different parasite communities. Most probably as a result of parasite pressure, we found substantial differences in MHC allele pools between the two ecomorphs. If our predictions are correct, we would expect females to exhibit assortative mating for males harboring locally adapted MHC alleles. We performed mate choice experiment based on the genetic component of odor cues, and found that female preference was directed towards males originating from the same habitat. Again, this suggests a pleiotropic role of MHC genes through both parasite resistance and mate choice and reinforces the idea of “magic traits”. Despite a range of observational evidence, no direct experimental study to date has revealed the connection of locally adapted MHC genes and local parasites in a natural habitat. We decided to challenge this question (Chapter VI)and for this test, we crossed three-spined sticklebacks from the same river and the same lake population as those used in the previous chapter. The populations displayed almost totally distinct MHC allele pools. Intercrossing F1 hybrids, we obtained a second fish generation segregating into pure river, pure lake and two hybrid MHC genotypes, while at the same time the genetic background was randomized. We used a double common garden experiment where fish were exposed simultaneously and reciprocally into the lake and into the river using replicated cages. We found that MHC genotypes were not equal in resisting parasites, which resulted in differences in parasite load between genotypes in turn. The total parasite load harbored by the fish appeared to result from an adaptive interaction between the innate and the acquired immune system of the hosts, which were driven by different parasite strategies. Generalist parasites were more successful in river fish lines, which naturally encounter low infection pressure from these parasites. However, when confronted with a stickleback specific parasite, which is commonly present in river habitats, the fish bearing lake MHC genotypes were more infected than fish carrying river MHC genotypes. In addition, we found support for an additive and synergistic role of genetic background and innate immunity in parasite resistance. Thus, this first test supports the adaptive connection between MHC and parasites and reinforces our previous suggestion of the pleiotropic role of MHC genes. Eventually, in the last chapter (VII), I report the initial results of another experimental test. We aimed at testing the potential role of negative frequency dependent selection in maintaining MHC polymorphism. Our design not only included selection by parasites but also included female mate choice. We created 6-replicated populations, each formed by 6 different laboratory-bred families in equal proportions and sex ratio. Three experimental populations were exposed to Camallanus lacustris, whilst the remaining three were confronted with Anguillicoloides crassus. These parasites were selected to be present in the original population were the founders of the lab-bred populations were caught. Letting the system evolve, we assumed that females would mate with partners that would possess a resistant allele against the experimental parasite, and thus increase the frequency of this allele in the following generation. Eggs were collected and raised in the laboratory, and the new stickleback generation was also exposed to both parasites. We hypothesize that fish from parents that faced one parasite, e.g. Camallanus lacustris, will be less heavily infected by this parasite than those derived from parents that encountered Anguillicoloides crassus and vice versa. As this experiment is running over several stickleback generations, the final dissection of the F2 fish has not yet taken place. Hence, I report the different steps that have been performed to date.Summary 6 Introduction 13 Host-parasite co-evolution 13 Parasites 13 Immune genes 14 Parasites and the evolution of sexual reproduction 20 Speciation: Another consequence of host-parasite co-evolution? 21 Species concept 21 Speciation mechanisms 22 The ecological concept of speciation 24 The Three-spined stickleback as model organism 27 Thesis outline 30 Chapter I: Individual MHC class I and MHC class II B diversities are associated with male and female reproductive traits in the three-spined stickleback 34 Chapter II: Life time reproductive success is maximised with optimal MHC diversity 49 Chapter III: MHC-based mate choice combines good genes and maintenance of MHC polymorphism 64 Chapter IV: Speciation accelerated and stabilized by pleiotropic MHC immunogenes 83 Chapter V: Ecological pleiotropy: pre- and post- zygotic reproductive isolation through MHC mediated selection 93 Chapter VI: MHC local adaptation: a field experiment 112 Chapter VII: Parasites maintaining MHC polymorphism: an experimental test for negative-frequency dependent selection 131 Conclusion 143 Acknowledgments 145 References 147 Glossary 172 Curriculum vitae 174 Appendix 17