Abstract

Drosophila melanogaster is a model organism to study innate immunity in invertebrates. Temperate and tropical D. melanogaster populations, being exposed to different environments, potentially face distinct parasites and parasite pressure. Indeed, there is experimental evidence suggesting that tropical D. melanogaster populations survive longer than temperate ones to infection by the fungal parasite Beauveria bassiana. In the present work we test the generality of this conclusion and investigate if host populations differ in their molecular response to infection. We first exposed to B. bassiana infection two tropical (from Africa and South-East Asia) and two temperate (from Europe and North America) D. melanogaster out-crossed populations. We consistently found a significant effect of B. bassiana on Drosophila mortality, but we were not able to identify a significant difference in survival to infection among populations. These results indicate that tropical populations may not always survive better than temperate ones, and suggest that other environmental factors, such as humidity or local species richness may be more accurate predictors of immune competence. Subsequently, we recorded transcriptional response to B. bassiana in all D. melanogaster populations, both by microarray and RNA sequencing. To our knowledge this is the first time that transcriptional response to fungal infection has been determined in multiple D. melanogaster out-crossed populations. We found few or no genes significantly induced 8 hours after infection. On the other hand, we identified between 200 and 1,300 genes induced 24 hours after infection depending on the population. This means that transcriptional response to B. bassiana begins between 8 and 24 hours after infection. We reveal here that host populations respond differently at the molecular level, as shown by the large variation in the number of induced genes. We report that gene ontology categories related to translation, biosynthesis and reproduction are enriched in genes down-regulated upon infection, suggesting a metabolic cost of mounting the defence response. Next, we wanted to assess the selective pressures acting on induced candidate genes. We compared the genes induced in all populations to the ones induced specifically in each population and computed population genetic statistics for a subset of genes in each category. We noticed higher conservation at non-synonymous sites for commonly induced genes compared to population specific ones. This hints that common genes are under stronger selective constraints. Another topic we addressed in the present work is the effect of endosymbionts and trans-generational immune priming on D. melanogaster survival to B. bassiana. We tested for a protective effect of the endosymbiont Wolbachia pipientis in two D. melanogaster inbred lines. We did not find an effect of Wolbachia on survival to infection in two independent experimental replicates. In absence of infection, flies bearing Wolbachia had a lower fitness than cured ones. Therefore W. pipientis appears to have a negative effect on Drosophila general vigour, but no effect on mortality upon infection. Finally, we tested if flies whose parents were exposed to B. bassiana were less susceptible when infected by the same parasite. This would imply a transfer of immune memory from parents to offspring, which is called trans-generational immune priming. However, no evidence of immune transfer for two D. melanogaster out-crossed populations could be found. Yet, as trans-generational immune priming depends on host and parasite genotype, more experiments are needed to determine its generality in the D. melanogaster – B. bassiana system.