Furthermore, the levels of apoLp-III, apoLp-II/I and apoLpR transcripts did not significantly differ between bees fed on royal jelly or beebread. Together, these results suggest that diet differentially regulates gene activity. The expression of apoLpR, which was up-regulated in bees fed syrup, reinforces this idea. The vasa gene is a germline marker in the ovaries ( Dearden,

2006). It is also expressed in the fat body of honey bee queens but not in queenright workers. This observation has led to the hypothesis that vasa may play a role in queen fertility ( Tanaka and Hartfelder, 2009). In the current study, we detected vasa expression in the fat body of queenless worker bees. Interestingly, vasa expression was up-regulated in the queenless bees fed beebread, which tended to have Metformin activated ovaries. This finding supports a possible role for this gene in fecundity. selleck If so, through the inhibition of vasa expression the infection may also have affected bee fecundity. Therefore, S. marcescens infection was costly to the honey bee, resulting in harmful effects on transcription, hemolymph protein storage and ovary activation. We had three main reasons to choose S. marcescens for the infections: (1) It is potentially pathogenic for insects

( Steinhaus, 1959) and was associated to septicaemia in adult honey bees ( Wille and Pinter, 1961). The isolation of S. marcescens from diseased honey bee larvae, followed by the reproduction of the disease experimentally, evidenced the pathogenicity of this microorganism ( El-Sanousi et al., 1987), (2) as we demonstrated ( Lourenço et al., 2009) S. marcescens was efficient in activating the honey bee immune system, (3) furthermore, and more importantly, S. marcescens is not lethal when the infection occurs orally, via food (see Steinhaus, 1959). Although S. marcescens is highly pathogenic when inoculated into the insect hemocoel, it is only mildly pathogenic when ingested ( Bulla et al., 1975). This feature is very important, considering that the accumulation of proteins in hemolymph, as

well as the ovary activation (in orphaned bees), occurs HSP90 gradually as the bees age. Thus, we used in our experiments a non-lethal bacterium, able to activate the immune system but allowing the survival, so that the infection costs in terms of transcription and storage of hemolymph proteins, and ovary activation, could be conveniently assessed. The infection did not appear to demand a significant cost from apolipophorins (apoLp-III, apoLp-II/I) and the apolipophorin receptor (apoLpR) transcriptions. In addition to its role in lipid transport, ApoLp-III has been shown to play a role in inducing antimicrobial proteins and phagocytosis by hemocytes (Wiesner et al., 1997 and Kim et al., 2004). It is known that ApoLp-III binds to bacterial surface components in Galleria melonella, thus playing an important role in the immune response ( Halwani et al., 2000).