As evolutionary biologists, we are often curious about the genomic origins of our favorite adaptations. Although some innovations certainly arose de novo, many more originated through the process of whole-gene or within-gene duplication. Following whole-gene duplication events, at least one gene copy is thought to be under relaxed selective constraints, meaning that mutations can accumulate within the gene and potentially give rise to novel adaptive traits. In this thesis, I aim to identify how gene duplication events have helped caterpillars cope with toxic host plants. Building upon the already-extensive literature on plant-insect coevolution, I highlight the complexity underlying detoxification phenotypes.

The research presented in Chapters II and III focuses on a family of genes coding for insect nitrile-specifier proteins (NSPs). These NSP-like genes are the canonical detoxification genes in Pierinae butterflies, allowing for the detoxification of the glucosinolates defenses present in their host plants. Importantly, the NSP-like gene family was formed through gene duplication events, with two key genes (NSP and MA) originating from the same ancestral gene. In Chapter II, Crispr-Cas9 methods were used to create lines of Pieris brassicae that lacked functional copies of NSP and/or MA. Through feeding assays on natural host plants, we showed that either NSP or MA are necessary for larval survival on plants containing aliphatic or benzyl glucosinolates – but not indole glucosinolates. Further, NSP seemed to be specialized for aliphatic glucosinolate detoxification, suggesting that some degree of subfunctionalisation occurred following gene duplication.  Expanding on these findings, we focused on the regulatory consequences of NSP-like family gene knockouts in Chapter III, looking specifically at the transcriptomic response to three host plants with vastly different glucosinolate profiles. We ultimately discovered that the response to host plant change was strongest when all NSP-like genes were functional, suggesting that the NSP-like gene family is part of a larger regulatory response to host plant defenses. 

While the above chapters center around gene duplicates that have already been associated with adaptations to host chemistry, there are likely more gene families out there that have been important for caterpillars overcoming ever-escalating plant defenses. In Chapters I and IV, I sought to identify some of these families using comparative genomic analyses. In Chapter I, I used genomes from across the Lepidoptera to see if diet breadth could be correlated with gene family sizes. I found that two serine protease families were larger in specialists and that a family of glutathione-S-transferases was larger in generalists. Due to the scope of the study, I was unable to associate gene duplication events with any particular host plant toxins. This knowledge gap ultimately led to the development of the work in Chapter IV, which centered specifically on instances of gene duplication and death in the Pieridae that occurred following major chemical changes in hosts. In addition to confirming that NSP-like genes are lost upon shifts away from Brassicales-feeding, we found that a subset of sulfotransferases existed in higher copy number in species that feed on glucosinolates.

Overall, this thesis shows that gene duplication may be important for insect dietary transitions, and that gene duplicates can become specialized to dynamically respond to host plant chemical profiles. It also provides a starting point for future studies, as open questions remain about the role of general detoxification mechanisms during initial transitions on to plants with novel chemical defenses.