Gut bacteria differentially affect egg production in the anautogenous mosquito Aedes aegypti and facultatively autogenous mosquito Aedes atropalpus (Diptera: Culicidae)

Abstract

Background: Aedes aegypti and A. atropalpus are related mosquitoes that differ reproductively. Aedes aegypti must blood-feed to produce eggs (anautogenous) while A. atropalpus always produces a first clutch of eggs without blood-feeding (facultatively autogenous). We recently characterized the gut microbiota of A. aegypti and A. atropalpus that were reared identically in the laboratory. Here, we assessed the effects of specific members of the gut microbiota in A. aegypti and A. atropalpus on female fitness including egg production.

Methods: Gnotobiotic A. aegypti and A. atropalpus larvae were colonized by specific members of the gut microbiota. Survival, development time, size and egg production for each treatment was then compared to axenic and conventionally reared larvae.

Results: Most species of bacteria we tested supported normal development and egg production by A. aegypti but only one betaproteobacterium, a Comamonas, supported development and egg production by A. atropalpus to equivalent levels as conventionally reared females. Aedes atropalpus females colonized by Comamonas contained similar stores of glycogen and protein as conventionally reared females, whereas females colonized by Aquitalea did not. Small differences in bacterial loads were detected between gnotobiotic and conventionally reared A. aegypti and A. atropalpus, but this variation did not correlate with the beneficial effects of Comamonas in A. atropalpus.

Conclusions: Specific members of the gut microbiota more strongly affected survival, size and egg production by A. atropalpus than A. aegypti.

Keywords: Clutch size; Development; Microbiota; Oogenesis; Reproduction.

Fig. 1

Development of Aedes aegypti and A. atropalpus larvae that were axenic, inoculated with a single bacterial species, or conventionally reared (non-sterile). a Survival from egg hatching to adult emergence differed among treatments for A. aegypti (Fisher’s exact test: P = 0.0005) and A. atropalpus (Fisher’s exact test: P < 0.0001). An asterisk (*) above a given bar indicates the treatment significantly differed from the non-sterile control by post-hoc pair-wise comparisons with Bonferroni correction. b Development time from egg hatching to pupation differed among treatments for A. aegypti (ANOVA: F _(5,673) = 8.0, P < 0.0001) and A. atropalpus (ANOVA: F _(5,880) = 211.3, P < 0.0001). c Size as estimated by forewing length did not differ among treatments for A. aegypti (ANOVA: F _(5,183) = 1.2, P = 0.29) but did differ for A. atropalpus (ANOVA: F _(5,129) = 32.7, P < 0.0001). Asterisks above the bars in (b) or (c) indicate means that significantly differ from the non-sterile control as determined by Dunnett’s test (P < 0.01). A minimum of 5 replicate dishes and 100 larvae per treatment were assayed for survival and development times. A single forewing from a minimum of 20 randomly selected adult females per treatment was measured to estimate adult size. The bars in (b) and (c) present mean values with 95 % confidence intervals for each treatment

Fig. 2

Mature egg formation by Aedes aegypti and A. atropalpus adult females from larvae that were inoculated with a single bacterial species or conventionally reared (Non-sterile). a The proportion of females that produced one or more mature eggs did not differ among treatments for A. aegypti (Fisher’s exact test: P > 0.05) but did differ for A. atropalpus (Fisher’s exact test: P < 0.0001). An asterisk above a given bar (*) indicates the treatment significantly differed from the non-sterile control by post-hoc pair-wise comparisons with Bonferroni correction. b Total clutch sizes (sum of the number of eggs laid and the number of mature eggs in the ovaries) did not differ among treatments for A. aegypti (ANOVA: F _(5,183) = 2.3, P > 0.05) but did differ for A. atropalpus (ANOVA: F _(5,127) = 13.0, P < 0.0001). c Number of eggs laid by females in a given treatment that produced at least one mature egg did not not differ for A. aegypti (ANOVA: F _(5,183) = 2.2, P > 0.05) but did differ for A. atropalpus (ANOVA: F _(5,127) = 18.4, P < 0.0001). Bars in (b) and (c) present mean values with 95 % confidence intervals while asterisks (*) in (b) and (c) indicate treatments that significantly differ from the non-sterile control (Dunnett’s test; P < 0.01)

Fig. 3

Total lipid (a), glycogen (b) and protein c in Aedes atropalpus adult females from larvae that were inoculated with Aquitalea, Comamonas, or conventionally reared (non-sterile). Lipid (ANOVA: F _(3,35) = 96.0, P < 0.0001), glycogen (ANOVA: F _(3,35) = 6.2, P = 0.002), and protein (ANOVA: F _(3,35) = 6.2, P = 0.002) significantly differed among treatments. For each nutrient, an asterisk above a bar indicates means significantly differ from the non-sterile control (Dunnett’s test; P < 0.01). Two adult females with gut removed were analyzed per replicate with 10 replicates analyzed for each treatment and nutrient. Bars indicate mean values with 95 % confidence intervals

Fig. 4

Bacterial loads in fourth instar gnotobiotic larvae colonized by a single bacterium and conventionally reared (non-sterile) fourth instars as measured by plate counts. Each gnotobiotic treatment and the conventionally reared (non-sterile) control are indicated on the X-axis. A minimum of 4 individuals was assayed per treatment. Bars indicate mean bacteria per larva with 95 % confidence intervals. Bacterial loads overall significantly differed among treatments for Aedes aegypti (ANOVA: F _(5,18) = 13.9, P < 0.0001) and A. atropalpus (ANOVA: F _(5,34) = 12.4, P < 0.0001). Asterisks (*) indicate treatments that significantly differ from conventionally reared larvae as determined by a post hoc Dunnett’s test (P < 0.01)

Fig. 5

CFU counts in 24 h old fourth instars (Larvae), 6–12 h adult females, and 72 h adult females. Treatments and the number of individuals analyzed per life stage are the same as in Fig. 4. For A. aegypti, bacterial loads significantly differed among treatments in larvae (ANOVA: F _(5,18) = 13.9, P < 0.0001) and 6–12 h adults (ANOVA: F _(5,22) = 19.6, P < 0.0001), but did not differ in 72 h adults (ANOVA: F _(5,16) = 1.6, P = 0.21). Between stage comparisons indicated that bacterial loads significantly differed between larvae, 6–12 h adults and 72 h adults (ANOVA: F _(2,71) = 31.6, P < 0.0001; followed by a Tukey-Kramer HSD test). For A. atropalpus, bacterial loads significantly differed among treatments in larvae (ANOVA: F _(5,34) = 12.4, P < 0.0001), 6–12 h adults (ANOVA: F _(5,33) = 8.9, P < 0.0001) and 72 h adults (ANOVA: F _(5,36) = 6.8, P = 0.0002). Between stage comparisons indicated that larvae (*) had higher bacterial loads than 6–12 h or 72 h adults, which did not differ from one another (NS) (ANOVA: F _(1,118) = 86.6, P < 0.0005; followed by a Tukey-Kramer HSD test). For A. aegypti adults (6–12 h) and A. atropalpus adults (6–12 and 72 h), Dunnett’s tests indicated treatment differences were due to higher colony counts for gnotobiotic individuals colonized by Chryseobacterium (red data points)