Genomics of Adaptation in Herbivorous Insects

Abstract

Parasitism is one of the most successful life history strategies to have evolved: parasites make up a majority of all named species, and nearly every multicellular organism must contend with a diverse suite of parasites. Characterizing differences in patterns of genome evolution between parasites and free-living species could reveal general evolutionary and physiological strategies underlying the evolution of parasitism, with implications for goals ranging from understanding biodiversity to developing anti-parasite treatments. Dramatic differences in gene content between parasitic and non-parasitic lineages, as well as the absence of parasitic lineages from many higher taxonomic groups (for example, two-thirds of insect orders lack parasites of plants), might indicate that parasitism is difficult to evolve and requires many genome-wide changes. However, inferences of the genomic changes that enable free-living organisms to evolve into parasites have been hindered by the fact that typical patterns of genome evolution are not well characterized in the closest non-parasitic relatives of many well-studied parasitic lineages. A major question therefore remains unanswered: are dramatic genetic innovations (such as the birth of many new genes or dramatic re-organization of genome content) necessary for the evolution of parasitism, or is the modification of pre-existing genes and traits sufficient? The primary aim of this dissertation was to characterize the genetic basis and evolution of traits involved in overcoming major barriers to parasitism. I helped lead the development of a novel genomic model system for host-parasite interactions: leaf-mining (i.e., endo-parasitic) flies in the genus Scaptomyza that feed on mustard plants including the genetic model plant, Arabidopsis thaliana, and that are close relatives of the genetic model insect, Drosophila melanogaster. I then used approaches from evolutionary and quantitative genomics, functional genetics, and biochemistry to investigate genomic changes coupled with the evolution of parasitism of plants in this lineage. The availability of functional genetic and comparative genomic resources for many non-parasitic, microbe-feeding Drosophila is a key feature of this model system, which facilitated the identification of derived gene functions and patterns of genome evolution coupled with the transition to parasitism of plants. I found that Scaptomyza overcame some of the major barriers to parasitism through the modification of pre-existing genes and pathways rather than dramatic genetic innovations. Specifically, leaf-mining Scaptomyza evolved the capacity to efficiently detoxify the major defensive chemicals in their host plants through rapid molecular and functional evolution of paralogous genes encoding glutathione S-transferases, which are enzymes involved in detoxification of a broad range of electrophilic compounds across eukaryotes. The few other gene families evolving more rapidly in Scaptomyza than in its non-parasitic relatives also interact with dietary compounds, supporting the prediction that novel exposure to plant-derived toxins and nutrients is a major selective pressure for herbivorous insects. Lastly, I found evidence that a conserved pathway for peripheral sensory organ development is involved in the development of the serrated ovipositor of S. flava, a derived trophic organ associated with feeding and reproduction on plants. These findings challenge the paradigm that dramatic genetic innovations are required to overcome barriers to the evolution of parasitism. In addition to driving divergence among species, host-parasite interactions are also predicted to shape genetic variation within populations. In the second half of this dissertation, I investigated how genetic variation mediates the outcomes of plant-herbivore interactions. Through a genome-wide association study, I discovered that two classes of genes in the Arabidopsis genome control susceptibility to herbivory by S. flava: those affecting production of defensive chemicals and those affecting plant size. These findings, which bridge two competing theories regarding the genetic determinants of herbivory, resulted from an experimental approach that quantified both herbivore foraging behavior and feeding rates. Further, the effect of plant chemical defenses on feeding rate varied among populations of mustard-feeding Scaptomyza, which opens future avenues for dissecting genetic changes in Scaptomyza that determine how plant chemicals shape feeding behavior and performance. I conclude with a conceptual review that outlines key areas of future research required for understanding how interactions between plants and herbivores shape patterns of genetic variation.