Development and application of genomic approaches for resolving complex traits among insect pests

Abstract

Adaptation to sudden and dramatic environmental change can occur if the genetic variation within a population has the capacity to respond to a selective pressure. One salient example is the insect-plant arms race, whereby plants counter herbivory through evolving chemical deterrents and insects respond by developing strategies to disarm them. Brassicaceae plants produce toxic glucosinolate metabolites to deter insect predation, yet the diamondback moth, Plutella xylostella L., coevolved a counteradaptation to circumvent this defense and became a brassica specialist and world-wide pest. In Kenya (c. 1999), a P. xylostella population underwent a surprising host plant range expansion and was found infesting sugar-snap pea crops (Pisum sativum; Fabaceae), raising concerns this surprising adaptative phenotype could undergo selection elsewhere. This thesis primarily focuses on the evolution and diversification of the Plutella genus and the genetic basis of a host plant range expansion in P. xylostella. Advances in genome sequencing technologies are improving opportunities to identify the genetic basis of adaptive traits, but also produce an increased volume of data leading to difficulties during data handling and processing. Here I developed two bioinformatic R package, ngsReports (Chapter 2) and geaR (Chapter 3) to overcome analytical bottlenecks encountered during this research, which facilitate quality assessment and analysis of high throughput genome sequence datasets. ngsReports interprets high-throughput Next Generation Sequencing (NGS) metrics, largely generated from Illumina platforms, enabling aggregation and visualization of quality control logs to rapidly identify sub-standard sequence data. geaR is a computationally inexpensive approach to perform established population genetic analyses using the Genomic Data Structure (GDS) format. Both packages were required for quality control and analysis of insect genomic datasets (Chapters 4-8). Plutella xylostella recently evolved resistance to diamide insecticides in Australia, although its cryptic ally, P. australiana, is highly susceptible and causes very little pest pressure on agriculture. Hybridization occurs between these species in laboratory crosses, raising concern that insecticide resistance alleles could be transferred interspecifically. To test this hypothesis, I examine whole genomes of P. xylostella and P. australiana populations collected across two consecutive years and developed sensitive methods for identifying gene flow between species (Chapter 4). Subsequent crosses between diamide resistant P. xylostella and susceptible P. australiana generated hybrid pupa which were used to produce a haploid chromosome-level (n=31) genome assembly of P. xylostella through trio binning. This process identified a point mutation within the insecticide’s target, the Ryanodine Receptor, causing field evolved diamide resistance in Australia for the first time (Chapter 6). The P. xylostella genome also established a resource for investigating the genetic basis of larval host plant range expansion (Chapter 7). Sequencing genetic crosses between Kenyan pea-adapted and wild-type P. xylostella strains revealed host adaptation is polygenic and polymorphic within the pea-strain despite ~17 years of laboratory captivity and sustained selection. Differential expression of larval midgut transcriptomes revealed an array of detoxification associated genes that responded to host plant diet. Similar approaches using head capsule tissue identified differential expression of olfaction and gustation pathway genes, and indicated decreased expression of gustatory and olfactory related genes were associated with pea adaptation. Computational approaches used to assess geneflow between Plutella species were re-applied to identify the genetic basis of the white pupa phenotype in Bactrocera dorsalis (oriental fruit fly) (Chapter 8). The white pupa genetic sexing marker has been used for decades to separate and discard females in rearing factories that supply males for pest control, yet the genetic basis remained unknown. The B. dorsalis white pupa phenotype was introgressed into the genetic background of B. tryoni (Queensland Fruit Fly) and genome-wide analysis with geaR identified the causal locus. After scanning for mutations across the white pupa locus, a single frame shift mutation was identified in a Major Facilitatory Superfamily gene. This provided a strong candidate for the white pupa gene and subsequent CRISPR/Cas9 mediated knock out in B. tryoni recapitulated the phenotype, confirming its role in puparium pigmentation. This thesis elucidates the genetic basis of agriculturally important phenotypes in major crop pest species and provides knowledge to advance integrated pest management strategies and crop protection. Bioinformatic packages developed provide a useful tool for researchers to carry out quality control of NGS datasets and evolutionary analysis of variant data when sample sizes are large. Furthermore, whole genome sequencing, RNA sequencing and a chromosome-level assembly of P. xylostella provide a powerful resource for future investigation of adaptive phenotypes in this agricultural pest.