Inter- and intra-species variation in genome-wide gene expression of Drosophila in response to parasitoid wasp attack

Abstract

Background: Parasitoid resistance in Drosophila varies considerably, among and within species. An immune response, lamellocyte-mediated encapsulation, evolved in a subclade of Drosophila and was subsequently lost in at least one species within this subclade. While the mechanisms of resistance are fairly well documented in D. melanogaster, much less is known for closely related species. Here, we studied the inter- and intra-species variation in gene expression after parasitoid attack in Drosophila. We used RNA-seq after parasitization of four closely related Drosophila species of the melanogaster subgroup and replicated lines of D. melanogaster experimentally selected for increased resistance to gain insights into short- and long-term evolutionary changes.

Results: We found a core set of genes that are consistently up-regulated after parasitoid attack in the species and lines tested, regardless of their level of resistance. Another set of genes showed no up-regulation or expression in D. sechellia, the species unable to raise an immune response against parasitoids. This set consists largely of genes that are lineage-restricted to the melanogaster subgroup. Artificially selected lines did not show significant differences in gene expression with respect to non-selected lines in their responses to parasitoid attack, but several genes showed differential exon usage.

Conclusions: We showed substantial similarities, but also notable differences, in the transcriptional responses to parasitoid attack among four closely related Drosophila species. In contrast, within D. melanogaster, the responses were remarkably similar. We confirmed that in the short-term, selection does not act on a pre-activation of the immune response. Instead it may target alternative mechanisms such as differential exon usage. In the long-term, we found support for the hypothesis that the ability to immunologically resist parasitoid attack is contingent on new genes that are restricted to the melanogaster subgroup.

Keywords: Drosophila speciesm; Evolution immune response; Parasitoid wasp; RNAseq.

Fig. 1

Experimental design and workflow. Schematic representation of samples, experimental design and analysis workflow

Fig. 2

Summary of the number of significant DEG. Summary of the number of significant DEG in each contrast a for 5h and b for 50h after parasitoid attack. The number of DEG is shown in horizontal barplots (blue: species-specific normalization, yellow: all-species normalization). Vertical barplots show the overlap in differentially expressed genes in the categories indicated by the filled dots

Fig. 3

Expression of the significantly DEG after parasitoid attack in 4 Drosophila species at 5h. (Log2-transformed) counts per million (CPM) of parasitized (red dots) and control (blue triangles)

Fig. 4

Expression of the significantly DEG after parasitoid attack in 4 Drosophila species at 50h. (Log2-transformed) counts per million (CPM) of parasitized (red dots) and control (blue triangles)

Fig. 5

Heatmap of significantly DEG in D. melanogaster at 5h. Colors represent the CPM values for all lines, treatments and replicates. The first 12 columns present the gene expression in larvae that were parasitized (par) in lines that had been selected (S) for increased resistance and their respective unselected control lines (C), and the second 12 columns the expression in not-parasitized control (ctl) larvae

Fig. 6

Heatmap of significantly DEG in D. melanogaster at 50h. Colors represent the CPM values for all lines, treatments and replicates. The first 12 columns present the gene expression in larvae that were parasitized (par) in lines that had been selected (S) for increased resistance and their respective unselected control lines (C), and the second 12 columns the expression in not-parasitized control (ctl) larvae

Fig. 7

Heatmap of DEG from all analyses, species and time points. Genes were included only if they were differentially expressed in at least one species, line or timepoint, and were expressed in all species (49 genes were excluded that were not present or not expressed in some species). The data for D. yakuba at 50h is excluded due to technical problems. The colors of the heatmap indicate the (log2-transformed) median fold change of the three replicates. The fold change was calculated as the ratio of counts per million of parasitized to control samples. The distance matrix was calculated using Pearson correlation and clustered using “ward.D2”. Each gene was annotated with eight functional and three orthologous categories (right panel, where the presence is indicated by a black filled squares)