Endoparasitoid wasps are essential regulators of insect pests in ecosystems as well as important biological control auxiliaries. Traits important for parasitism success, such as the injection of venom proteins at oviposition, have thus been mainly studied. However, identification of the key genes involved among the large number of genes identified was still prevented by the lack of functional approaches. Here, we report the development of RNA interference (RNAi) in Leptopilina boulardi, a figitid endoparasitoid that performs its entire development inside the Drosophila host. Having set up conditions for in vitro development of parasitoid late larval stages or pupae, we first targeted the cinnabar gene by microinjecting double-stranded RNA (dsRNA), leading to its silencing and production of red-eyed individuals. We then demonstrated that expression of the gene encoding LbGAP, a virulence factor found in a high amount in L. boulardi venom, could be specifically and almost completely silenced. Finally, a time-course analysis revealed that LbGAP silencing lasted during the entire lifetime of L. boulardi. This is the first report of the efficient silencing of venom protein-encoding genes in parasitoid wasps. Overall, RNAi opens the way for a large-scale functional analysis of parasitoid venom factors as well as other traits involved in parasitism success and more largely in the biology of these ecologically important organisms.

The Drosophila endoparasitoid wasps Leptopilina boulardi and L. heterotoma (Hymenoptera: Cynipidae) are pro-ovigenic species, i.e., females contain their lifetime number of mature eggs at emergence. They are therefore able to immediately parasitize many hosts when present. In response to parasitoid oviposition, the larval host D. melanogaster can mount an immune response, encapsulation, that can destroy the parasitoid eggs. This response is counteracted by the venom the wasp injects during oviposition. Here, we estimated the amount of venom injected into a D. melanogaster host larva using immunodetection of venom proteins and we attempted to correlate this amount with the number of eggs a female can lay on successive days. The venom reservoir of L. boulardi contains enough venom for at least 100 ovipositions while that of L. heterotoma contains venom for about 16 ovipositions. While a female L. boulardi may have enough venom for three days of parasitism when 20 or 40 larval hosts were presented each day, L. heterotoma certainly needs to synthesize new venom to parasitize the number of hosts offered. Interestingly, parasitism stopped (L. boulardi), egg protection (L. heterotoma) and egg hatching decreased (both species) after three days of parasitism. Thus, although venom does not appear to be a limiting factor for parasitism, our data suggest that it may have less effectiveness on the egg protection and on egg/host development after high repetitive egg laying.