Differential gene expression in a tripartite interaction: Drosophila, Spiroplasma and parasitic wasps

Abstract

Background Several facultative bacterial symbionts of insects protect their hosts against natural enemies. Spiroplasma poulsonii strain sMel (hereafter Spiroplasma), a male-killing heritable symbiont of Drosophila melanogaster, confers protection against some species of parasitic wasps. Several lines of evidence suggest that Spiroplasma-encoded ribosome inactivating proteins (RIPs) are involved in the protection mechanism, but the potential contribution of the fly-encoded functions (e.g., immune response), has not been deeply explored. Methods Here we used RNA-seq to evaluate the response of D. melanogaster to infection by Spiroplasma and parasitism by the Spiroplasma-susceptible wasp Leptopilina heterotoma, and the Spiroplasma-resistant wasp Ganaspis sp. In addition, we used quantitative (q)PCR to evaluate the transcript levels of the Spiroplasma-encoded Ribosomal inactivation protein (RIP) genes. Results In the absence of Spiroplasma infection, we found evidence of Drosophila immune activation by Ganaspis sp., but not by L. heterotoma, which in turn negatively influenced functions associated with male gonad development. As expected for a symbiont that kills males, we detected extensive downregulation in the Spiroplasma-infected treatments of genes known to have male-biased expression. We detected very few genes whose expression patterns appeared to be influenced by the Spiroplasma-L. heterotoma interaction, and these genes are not known to be associated with immune response. For most of these genes, parasitism by L. heterotoma (in the absence of Spiroplasma) caused an expression change that was at least partly reversed when both L. heterotoma and Spiroplasma were present. It is unclear whether such genes are involved in the Spiroplasma-mediated mechanism that leads to wasp death and/or fly rescue. Nonetheless, the expression pattern of some of these genes, which reportedly undergo expression shifts during the larva-to-pupa transition, is suggestive of an influence of Spiroplasma on the development time of L. heterotoma-parasitized flies. One of the five RIP genes (RIP2) was consistently highly expressed independently of wasp parasitism, in two substrains of sMel. Finally, the RNAseq data revealed evidence consistent with RIP-induced damage in the ribosomal (r)RNA of the Spiroplasma-susceptible, but not the Spiroplasma-resistant, wasp. Acknowledging the caveat that we lacked adequate power to detect the majority of DE genes with fold-changes lower than 3, we conclude that immune priming is unlikely to contribute to the Spiroplasma-mediated protection against wasps, and that the mechanism by which Ganaspis sp. resists/tolerates Spiroplasma does not involve inhibition of RIP transcription.