The molecular genetic basis of herbivory between butterflies and their host plants

Abstract

Interactions between herbivorous insects and their host plants are a central component of terrestrial food webs and a critical topic in agriculture, where a substantial fraction of potential crop yield is lost annually to pests. Important insights into plant-insect interactions have come from research on specific plant defences and insect detoxification mechanisms. Yet, much remains unknown about the molecular mechanisms that mediate plant-insect interactions. Here we use multiple genome-wide approaches to map the molecular basis of herbivory from both plant and insect perspectives, focusing on butterflies and their larval host plants. Parallel genome-wide association studies in the cabbage white butterfly, Pieris rapae, and its host plant, Arabidopsis thaliana, pinpointed a small number of butterfly and plant genes that influenced herbivory. These genes, along with much of the genome, were regulated in a dynamic way over the time course of the feeding interaction. Comparative analyses, including diverse butterfly/plant systems, showed a variety of genome-wide responses to herbivory, as well as a core set of highly conserved genes in butterflies as well as their host plants. These results greatly expand our understanding of the genomic causes and evolutionary consequences of ecological interactions across two of nature's most diverse taxa, butterflies and flowering plants.

Combining Experimental Evolution and Genomics to Understand How Seed Beetles Adapt to a Marginal Host Plant

“…Nonetheless, we detected consistent differences in expression of cytochrome P450s between mung bean and lentil treatments. Cytochrome P450s are known to play a role in detoxification of plant secondary metabolites and insecticides [53,101,118,119], and this finding is consistent with a general pattern of increased plasticity of detoxification genes in herbivorous insects [76,120]. Because mortality is high for the M line when reared in lentil, we do not know if adaptation to lentil involved the evolution of increased plasticity of cytochrome P450s or if this differential expression was present in the ancestral M line and thus is mostly incidental to lentil adaptation.…”

“…We used these contrasts to identify the genes which show significant difference in gene expression for each pair of comparison. We then asked whether each of three specific classes of genes were over-represented among differentially expressed genes assuming binomial sampling: cytochrome P450s (known to be involved in detoxification of plant secondary chemicals [53,101]) and two classes of putative digestive enzymes, proteases and carboxylases. We classified genes as likely digestive proteases and carboxylases following the annotated genome provided by [102] (proteases = serine protease, trypsin, chymotrypsin, cathepsin, aspartic proteinase, lysosomal aspartic protease, cysteine protease or proteinase [88 genes]; carboxylases = amylase, cellulase, glucosidase or maltase [29 genes]).…”

“…Thus, population-genomic analyses and trait mapping approaches should be viewed as complementary, and combining the two methods is especially likely to reveal the genetic and phenotypic bases of adaptation (e.g., [29,[47][48][49][50][51]). Lastly, an even more comprehensive analysis would be to combine gene-expression data and genetic manipulations (e.g., RNA interference or genome editing) with population-genomic or trait-mapping data [3,52,53]. Gene expression data and genetic manipulations add a more mechanistic understanding to the link between genotype and phenotype [54][55][56], and in some cases can identify which genotype-phenotype associations are causal (e.g., [57]).…”

“…By integrating these three sets of genomic data, we are able to quantify parallelism in terms of the genetic basis of performance traits and allele-frequency change during adaptation (we test for parallelism in each of these patterns rather than using parallelism as a test for the process of selection), and obtain a more complete understanding of the genes and traits that allow C. maculatus to adapt to and persist on a marginal host plant. In terms of the latter, we specifically test for roles of digestive enzymes and detoxification genes as these types of genes have been implicated in host-plant adaptation in other systems [53,75,76].…”

Robust Cytonuclear Coordination of Transcription in Nascent Arabidopsis thaliana Autopolyploids

“…Quantifying transcription: Read counts per gene were determined from the BAM files using HTSeq [26] with the "intersection non-empty" setting using the Araport11 annotation file (Athaliana_447_Araport11.gene_exons.gff3) downloaded from Phytozome (https://phytozome.jgi. doe.gov/pz/portal.html).…”