Mosquitoes rely on their gut microbiota for development

Abstract

Field studies indicate adult mosquitoes (Culicidae) host low diversity communities of bacteria that vary greatly among individuals and species. In contrast, it remains unclear how adult mosquitoes acquire their microbiome, what influences community structure, and whether the microbiome is important for survival. Here, we used pyrosequencing of 16S rRNA to characterize the bacterial communities of three mosquito species reared under identical conditions. Two of these species, Aedes aegypti and Anopheles gambiae, are anautogenous and must blood-feed to produce eggs, while one, Georgecraigius atropalpus, is autogenous and produces eggs without blood feeding. Each mosquito species contained a low diversity community comprised primarily of aerobic bacteria acquired from the aquatic habitat in which larvae developed. Our results suggested that the communities in Ae. aegypti and An. gambiae larvae share more similarities with one another than with G. atropalpus. Studies with Ae. aegypti also strongly suggested that adults transstadially acquired several members of the larval bacterial community, but only four genera of bacteria present in blood fed females were detected on eggs. Functional assays showed that axenic larvae of each species failed to develop beyond the first instar. Experiments with Ae. aegypti indicated several members of the microbial community and Escherichia coli successfully colonized axenic larvae and rescued development. Overall, our results provide new insights about the acquisition and structure of bacterial communities in mosquitoes. They also indicate that three mosquito species spanning the breadth of the Culicidae depend on their gut microbiome for development.

Keywords: bacteria; development; evolution; insects; microbial biology.

Figure 1

Life cycle of Ae. aegypti, An. gambiae and Ge. atropalpus and samples collected for 454 pyrotagging analysis. Each species oviposits eggs that hatch in aquatic habitats where the larval stage feeds and acquires bacteria that colonize the digestive tract. Larvae undergo metamorphosis after the fourth instar to form pupae that float on the surface of the aquatic habitat. Adults emerge from the pupal stage and persist in terrestrial habitats. Newly emerged adult mosquitoes often imbibe water from the aquatic habitat. Adults of each species also feed on sugar sources. Ge. atropalpus are autogenous and oviposit a first clutch of eggs without taking a blood meal. Adult female Ae. aegypti and An. gambiae are anautogenous and must blood-feed on a vertebrate host to lay eggs. Circles indicate life stages sampled for pyrosequencing for Ae. aegypti (blue), An. gambiae (yellow), and Ge. atropalpus (red), respectively. For conventionally reared mosquitoes, samples were collected from the water in which larvae developed, and from fourth instars of each species. For Ae. aegypti, conventionally reared pupae were surface sterilized to produce NBF adults and BF adults after blood feeding on a surface sterilized host. BF adults then laid STR eggs. Eggs laid by conventionally reared Ae. aegypti females were named ConR eggs (see Materials and methods for additional details).

Figure 2

Bacterial composition at the phylum and family levels in laboratory reared Ae. aegypti, An. gambiae and Ge. atropalpus. Water and larval samples (fourth instars) were analyzed for each species. Adult and egg samples were analyzed for only Ae. aegypti. Each bar graph presents the proportion of sequencing reads assigned to a given bacterial family. Only categories >2% are presented. NBF, non-blood fed adult female; BF, adult female 24 h after blood feeding on a host; ConR, eggs laid by conventionally reared Ae. aegypti females; STR, eggs laid by females that emerged from surface sterilized pupae.

Figure 3

Overlap between OTUs in (A) the water habitat larvae of each mosquito species was reared in, (B) the larval stage of each species, and (C) water and larvae for each species.

Figure 4

Survival of first instar Ae. aegypti, An. gambiae, and Ge. atropalpus to adulthood under three conditions: axenic larvae fed sterilized diet (Axenic); axenic larvae fed sterilized diet plus bacteria from the conventional laboratory habitat (With bacteria), and conventional, non-sterile larvae fed sterilized diet (Non-sterile). A minimum of 50 larvae per treatment was assayed. The proportion of larvae that developed into adults is presented. Columns indicate mean values with 95% confidence intervals. ANOVA analyses were conducted separately for each species with multiple comparisons performed by Tukey-Kramer HSD (NS, treatments not significantly different; ***, Axenic treatment significantly different from the With Bacteria and Non-sterile treatments, p< 0.0001).

Figure 5

Several bacterial isolates from Ae. aegypti and E. coli rescue development of axenic larvae. (A) Proportion of axenic first instars that developed into adults when fed: sterilized diet only (Axenic), heat-killed bacteria plus sterilized diet (Heat killed bacteria), sterilized diet in bacteria-conditioned water (Bacteria conditioned), or sterilized diet plus different bacterial isolates. Non-sterile larvae fed sterilized diet served as the positive control. A minimum of 40 larvae per treatment was assayed per treatment. Columns present mean values with 95% confidence intervals for each treatment. *** indicates a significant difference for a given treatment relative to the positive control as determined by a post-hoc Dunnett’s test (p<0.0001). (B) Development time of axenic larvae to adulthood when fed sterilized diet and different bacteria. Non-sterile larvae served as the positive control. Columns present mean values with 95% confidence intervals for each treatment. ANOVA detected no differences among treatments.

Figure 6

Most bacterial isolates colonized axenic Ae. aegypti. Axenic first instars were fed sterilized diet plus the indicated bacterial isolate. DNA was then isolated from an individual and used as template with universal or taxon specific primers. The agarose gel shows ethidium bromide stained PCR products. Lane 1, molecular mass marker; Lane 2, universal primers plus DNA from a non-sterile first instar; Lane 3, universal primers plus DNA from an axenic first instar; Lanes 4–6, Acinetobacter specific primers plus template from Acinetobacter (control), a fourth instar or an adult. The same treatments are then shown for Aeromonas (Lanes 7–9), Aquitalea (Lanes 10–12), Chryseobacterium (Lanes 13–15), Microbacterium (Lanes, 16, 17) Paenibacillus (Lanes 18–20), E. coli (Lanes 21–23). At least 10 individuals were examined for each treatment with all outcomes being identical to what is presented in the figure.