Probing at mechanisms of coevolution: investigating the genetic and molecular basis of host resistance and parasite infectivity in the Daphnia magna–Pasteuria ramosa system

Why is nature so diverse? If evolution by natural selection is about survival of the “fittest”, why, after billions of years, do we still see differences within the same species? The Red Queen Hypothesis claims that part of the answer to this question is: parasites. The Red Queen is familiarly known from the Alice in Wonderland stories. In Through the Looking Glass, the Red Queen says to Alice, “it takes all the running you can do, to keep in the same place”. But what does this have to do with parasites? Under the Red Queen Hypothesis, coevolution between hosts and parasites is driven by negative frequency-dependent selection, which essentially means that being common is bad. This type of selection then leads to time-lagged cycles in the frequencies of alleles underlying host resistance and parasite infectivity. Because being common is bad, it is unlikely that any particular host or parasite allele will go to fixation, and, thus, diversity is maintained within populations.
Despite being introduced nearly fifty years ago, and despite being implicated in several host–parasite systems, the Red Queen coevolution has proven challenging to demonstrate in natural populations. Part of the difficulty lies in the fact that the Red Queen hypothesis specifically relates to alleles; namely interacting alleles in the host and parasite, and these alleles can be difficult to identify. One helpful aspect in this regard is being able to define a distinct interaction phenotype, by which host genotypes can clearly be designated as resistant or susceptible to a particular parasite, and parasite genotypes can be designated as infectious or noninfectious to a particular host.
The Daphnia magna–Pasteuria ramosa system has become a model for studying host–parasite coevolution, in part, because such a strong interaction phenotype has been defined. Parasite spores have been shown to infect after successfully attaching to one of two sites on the host cuticle: the lining of the esophagus (=foregut) or the lining of the hindgut. The ability to attach depends on strong genotype-by-genotype interactions, and attachment phenotypes can be readily scored for hundreds of host–parasite genotype combinations.
In this thesis, I investigate the molecular and genetic basis of these attachment phenotypes, with the overarching goal to identify alleles underlying host resistance and parasite infectivity in order to characterize rules of interaction that drive coevolution and shape the diversity in this system. I start on the host side of the story, using QTL mapping to identify a new resistance locus that shows Mendelian segregation. The genomic region containing this locus includes several candidate genes, including glycosyltransferases, which have previously been implicated as candidates from other Pasteuria resistance loci. I next employ a proteomics approach, using molted D. magna cuticles, to identify host molecules that interact with parasite spores during attachment. I do not find differences between resistant and susceptible hosts, but the described cuticle proteome will serve as a tool for future studies.
Switching to the parasite side of the story, I start by surveying attachment diversity in a panel of P. ramosa isolates from several geographic origins. I describe several additional attachment sites that lead to infection, and I suggest attachment at each site may be mediated by different underlying genetics. I further show that isolates cluster by attachment specificity, and that genotype is a better predictor of attachment phenotype compared to geographic origin. I next analyze the attachment phenotypes as interaction networks, showing that site-specific attachment matrices tend to be nested, though foregut attachment and overall infectivity matrices are weakly modular. Finally, I return to proteomics to compare P. ramosa spores with varying infection specificities for differences in protein content or abundance. I show that the collagen-like protein PCL7 is by far the strongest candidate explaining the tested attachment phenotype, and that several additional PCLs and other genes may play a role as well. These results validate findings from a recent GWAS and confirm that proteomics can be used to identify candidate genes underlying parasite infectivity. Overall, my thesis advances our understanding of the vast diversity in this system, and it provides tools and genetic candidates that may be used for functional tests in future studies to further probe at the mechanisms of host resistance and parasite infectivity, and to ultimately test the Red Queen Hypothesis.