Interactions between the microbiome and mating influence the female's transcriptional profile in Drosophila melanogaster

Abstract

Drosophila melanogaster females undergo a variety of post-mating changes that influence their activity, feeding behavior, metabolism, egg production and gene expression. These changes are induced either by mating itself or by sperm or seminal fluid proteins. In addition, studies have shown that axenic females-those lacking a microbiome-have altered fecundity compared to females with a microbiome, and that the microbiome of the female's mate can influence reproductive success. However, the extent to which post-mating changes in transcript abundance are affected by microbiome state is not well-characterized. Here we investigated fecundity and the post-mating transcript abundance profile of axenic or control females after mating with either axenic or control males. We observed interactions between the female's microbiome and her mating status: transcripts of genes involved in reproduction and genes with neuronal functions were differentially abundant depending on the females' microbiome status, but only in mated females. In addition, immunity genes showed varied responses to either the microbiome, mating, or a combination of those two factors. We further observed that the male's microbiome status influences the fecundity of both control and axenic females, while only influencing the transcriptional profile of axenic females. Our results indicate that the microbiome plays a vital role in the post-mating switch of the female's transcriptome.

Figure 1

Fecundity and transcript abundance differences between axenic and control females. (A) Fecundity of axenic and control females after a single mating to axenic or control males. Each point represents eggs laid by one female over the course of 54 h. The first letter refers to the female’s microbiome status, the second letter refers to the male’s microbiome status (A = axenic, C = Control). Sample sizes: n = 49 for AA, n = 47 for AC, n = 46 for CA and n = 50 for CC. Groups were compared using a generalized linear mixed model with a Poisson response distribution. ***$p<0.001$; **$p<0.01$ (B). Multidimensional scaling plot of replicates and samples used in the study. Microbiome status is represented by colour, and sample origin represented by shape. (C) Volcano plot showing the results of a differential expression analysis of control virgin females relative to axenic virgin females. Significant genes (FDR < 0.05, >two-fold) are shown in pink. (D) Volcano plot showing the results of a differential expression analysis of control mated females relative to axenic mated females, averaged across the two male microbiome states. Significant genes (FDR < 0.05, >two-fold) are shown in green (E) Overlap of genes that are influenced by the microbiome in virgin and mated females. Figures were produced using R.

Figure 2

Tissue enrichment scores for differentially abundant transcripts across 14 female tissues. (A) Heatmap for 124 “core” genes whose RNA levels are influenced by the microbiome in both mated and virgin females. (B) Heatmap for 247 genes whose RNA levels are influenced by the microbiome in mated females only. Enrichment scores were calculated using TPM values from FlyAtlas2. Both heatmaps were generated with the R package pheatmap (v. 1.0.12). (AP = anal plate; M = mated; SG = salivary gland; SP = spermathecae; TAG = Thoracico-abdominal ganglion; V = virgin). Figures were produced using R.

Figure 3

Genes whose mRNA levels are influenced by the microbiome in mated females have roles in metabolism and neuronal functions. (A) Gene Set Enrichment Analysis (GSEA) showing that transcripts involved in glycolitic metabolism are generally up-regulated in mated females with a microbiome relative to axenic mated females. The top panel of the GSEA plot shows the running enrichment score for a rank-orderd list of genes that are involved in hydrolase activity (genes are ranked based on the log2 fold change between mated control and mated axenic females, in decreasing order). Each vertical black line represents a gene that is involved in hydrolase activity. (B) Tissue enrichment Z-scores of genes encoding hydrolases, that are differentially expressed between mated axenic and mated control females. These genes show a strong expression bias in the female midgut (abbreviated tissue samples are the same as in Fig. 2). (C) Heatmap showing mean centered log2 TPM for 14 genes with sensory or neuronal functions, whose transcript abundance is altered in mated females depending on whether they have a microbiome or not. (The panel A figure was generated using the gseaplot2 function of the R package enrichplot (v. 1.4.0) and heatmaps in B and C were generated with the R package pheatmap (v. 1.0.12).(A = axenic, C = control). The first letter refers to the female’s microbiome status, the second letter refers to the male’s microbiome status. Female microbiome and mating status are indicated above the heatmap and in the key. Figures were produced using R.

Figure 4

Analysis of female transcriptome changes that are influenced by the male’s microbiome status. (A) Volcano plot showing the results of a differential expression analysis assessing the effect of male microbiome status on control females (top) and axenic females (bottom). (B) Heatmap of normalized, batch-adjusted abundance values (TPM; transcripts per million) for transcripts that are altered in axenic females after mating with an axenic or control male. The heatmap was generated with the R package pheatmap (v. 1.0.12). (C) Barplots of TPM values for a subset of genes that are influenced by the male’s microbiome in axenic mated females. Error bars represent standard error, and points represent quadruplicate TPM values. Figures were produced using R.