Wolbachia-associated bacterial protection in the mosquito Aedes aegypti

Abstract

Background: Wolbachia infections confer protection for their insect hosts against a range of pathogens including bacteria, viruses, nematodes and the malaria parasite. A single mechanism that might explain this broad-based pathogen protection is immune priming, in which the presence of the symbiont upregulates the basal immune response, preparing the insect to defend against subsequent pathogen infection. A study that compared natural Wolbachia infections in Drosophila melanogaster with the mosquito vector Aedes aegypti artificially transinfected with the same strains has suggested that innate immune priming may only occur in recent host-Wolbachia associations. This same study also revealed that while immune priming may play a role in viral protection it cannot explain the entirety of the effect.

Methodology/findings: Here we assess whether the level of innate immune priming induced by different Wolbachia strains in A. aegypti is correlated with the degree of protection conferred against bacterial pathogens. We show that Wolbachia strains wMel and wMelPop, currently being tested for field release for dengue biocontrol, differ in their protective abilities. The wMelPop strain provides stronger, more broad-based protection than wMel, and this is likely explained by both the higher induction of immune gene expression and the strain-specific activation of particular genes. We also show that Wolbachia densities themselves decline during pathogen infection, likely as a result of the immune induction.

Conclusions/significance: This work shows a correlation between innate immune priming and bacterial protection phenotypes. The ability of the Toll pathway, melanisation and antimicrobial peptides to enhance viral protection or to provide the basis of malaria protection should be further explored in the context of this two-strain comparison. This work raises the questions of whether Wolbachia may improve the ability of wild mosquitoes to survive pathogen infection or alter the natural composition of gut flora, and thus have broader consequences for host fitness.

Figure 1. Survival curves of Wolbachia-infected (circle) and Wolbachia-uninfected (square) Drosophila infected with pathogenic bacteria (solid) or mock infected with PBS (open).

(A&E) E. carotovora, (B&F) B. cepacia, (C&G) S. typhimurium and (D&H) M. marinum. Error bars are SEM calculated from the three replicates. * P-value<0.05, ** P-value<0.01, *** P-value<0.001 denote differences in survival between Wolbachia infected and uninfected lines by Log-rank statistics (Table S1A).

Figure 2. Survival curves of Wolbachia-infected (circle) and Wolbachia-uninfected (square) mosquitoes infected with pathogenic bacteria (solid) or mock infected with PBS (open).

(A&E) E. carotovora, (B&F) B. cepacia, (C&G) S. typhimurium and (D&H) M. marinum. Error bars are SEM calculated from the three replicates. * P-value<0.05, ** P-value<0.01, *** P-value<0.001 denote differences in survival between Wolbachia infected and uninfected lines by Log-rank statistics (Table S1B). The relative risk ratio [EXP(B)] of Wolbachia uninfected to infected lines with 95% confidence intervals shown in parentheses is reported on graphs where significant.

Figure 3. Median (with interquartile range) fold change in pathogen density variable hours post infection (hpi) in mosquitoes.

Five pairs of individuals were used for E. carotovora (A), B. cepacia (B), S. typhimurium (C) and M. marinum (D). (Mann-Whitney U-test; * P-value<0.05, ** P-value<0.01).

Figure 4. Median (with interquartile range) relative Wolbachia density after infection in mosquitoes.

Five pairs of individuals were used for E. carotovora (A), B. cepacia (B), S. typhimurium (C) and M. marinum (D). (Mann-Whitney U-test; * P-value<0.05, ** P-value<0.01).