Enterococci Mediate the Oviposition Preference of Drosophila melanogaster through Sucrose Catabolism

Abstract

Sucrose, one of the main products of photosynthesis in plants, functions as a universal biomarker for nutritional content and maturity of different fruits across diverse ecological niches. Drosophila melanogaster congregates to lay eggs in rotting fruits, yet the factors that influence these decisions remains uncovered. Here, we report that lactic acid bacteria Enterococci are critical modulators to attract Drosophila to lay eggs on decaying food. Drosophila-associated Enterococci predominantly catabolize sucrose for growing their population in fly food, and thus generate a unique ecological niche with depleted sucrose, but enriched bacteria. Female flies navigate these favorable oviposition sites by probing the sucrose cue with their gustatory sensory neurons. Acquirement of indigenous microbiota facilitated the development and systemic growth of Drosophila, thereby benefiting the survival and fitness of their offspring. Thus, our finding highlights the pivotal roles of commensal bacteria in influencing host behavior, opening the door to a better understanding of the ecological relationships between the microbial and metazoan worlds.

Figure 1

The innate oviposition behavior in response to fermented diet. (a) A diagram of the egg laying preference assay with the 2-choice cage. The surface of fly food was augmented with bacteria to generate a fermented diet for 48 h in the incubator, whereas the control was used with H₂O. Each food item was chopped into two halves, and each half was placed into the 2-choice cages. Mated females with yeast paste were transferred to the 2-choice cage and allowed to lay eggs for 16 h. The numbers of eggs were counted on each half, and the oviposition preference was calculated. (b) The quantification of egg laying preference for fermented fly food by wild-type Oregon R (OR) and Canston S (CS). H₂O: water, EF: Enterococcus faecium, AA: acetic acid; mocks are two halves of fly food with water or EF). The one-sample t-test was used to assess the mean deviance of each column from 0; ANOVA tests with LSD post hoc analysis were used to calculate significant differences between columns, n = 6–14. (c) The stimulation of egg laying with fermentation. Twenty females were transferred into each cage of the whole-forced cage with a control or fermented diet, respectively, and the average number of eggs was calculated. ANOVA tests with LSD post hoc analysis, n = 6–8. Mean ± SEM; Symbols: NS p > 0.05; *p < 0.05; **p < 0.01; ***p < 0.001.

Figure 2

Role of gustatory system and sucrose receptor neurons in oviposition choices. (a) The positional preference for fermentation. Wild-type (WT)  flies were averse to a fermented fly food, while Orco mutants deficient in odor were neutral. Females were presented the following 2-choice as depicted in Fig. 1a, and position was examined. The one-sample t-test was used to assess the mean deviance of each column from 0; ANOVA tests with LSD post hoc analysis were used to calculate significance differences between columns; n = 28–46. (b) The feeding preference for fermented food. Females were presented the following 2-choice food combinations, and one half food was supplemented with dye. After feeding, the amount of dye in fly gut contents was quantified. The one-sample t-test was used to assess the mean deviance of each column from 0; n = 8. (c) Screening of candidate sensory modalities for oviposition selection for fermentation. The indicated animals were allowed to choose using two-way food preference assays. For vision, WT flies in darkness and ninaB were used; for olfaction, antennaectomized females (surgically removing the primary olfactory organs) and Orco mutants (unable to respond to most olfactory stimuli) were used; for gustation, the forelegs that contain gustatory sensilla were surgically ablated. ANOVA tests with LSD post hoc analysis, n = 6–12. (d) The role of Gr5a; Gr64a neurons in the oviposition preference for fermentation. Gr33a ¹ mutants, IR76b ¹ mutants and ΔGr5a; ΔGr64a double mutants were used, and the oviposition index was evaluated. ANOVA tests with LSD post hoc analysis, n = 8–15. (e) Hyperpolarizing Gr5a and Gr64a neurons reduced the oviposition preference for fermentation. Animals carrying Gr64a-GAL4;Gr5a-GAL or UAS-Kir2.1 were used as a negative control. Mann-Whitney Test, n = 5–8. Mean ± SEM. Symbols: NS p > 0.05; *p < 0.05; **p < 0.01; ***p < 0.001.

Figure 3

Sucrose modulated the oviposition preference for fermentation. (a) Females avoided to deposit eggs on the casein-cornmeal-agar media titrated to different sucrose (+Sucrose%) compared to fly food without sucrose (−Sucrose). n = 12, one of two replicates. (b) Sucrose deprivation impaired the oviposition preference for fermentation. Fly food was control food (+sucrose) or deprived of sucrose (−sucrose). The 2-choice cage of each food was assembled with H₂O or bacteria. n = 8, one of three replicates. (c) Sucrose replenishment attenuated this oviposition preference in a dose-dependent manner. Sucrose was added to fermented fly food (EF + Sucrose), and the oviposition preference of EF + sucrose food was compared to fermented food (H₂O). ANOVA tests with LSD post hoc analysis were used to calculate significant differences between columns. n = 12, one of two replicates. (d) The α-glucosidase inhibitor, acarbose, diminished the oviposition preference for fermentation. Acarbose was added to fermented fly food (EF + acarbose), and the oviposition preference of EF + acarbose food was compared to fermented food (H₂O). n = 12, one of two replicates. (e) Bacterial cells were dispensable to trigger the oviposition preference for fermentation. Frozen bacterial cells were supplemented on the surface of one half of fly diet in a 2-choice cage, and ovipositional preference for bacterial cells was compared to fly food with water. n = 12, one of three replicates. (f,g) Fruit flies were averse to laying eggs on the media with LAB metabolites or lactate. n = 12, one of three replicates. Supernatant or lactate was added to one half of fly diet in a 2-choice cage, and the ovipositional preference for them was compared to fly food with water (H₂O), respectively. (h) Sucrose was a more robust factor that suppressed the oviposition of females than EF metabolites or lactate. Females were allowed to choose between 0.5 ml LAB metabolites (Supernatant) or 1% lactate (Lactate) and dosage-dependent sucrose (Sucrose%) using 2-choice food preference assays. High concentration sucrose reversed the avoidance to LAB metabolites and sucrose. n = 12, one of three replicates. (i) The aversion to 5% sucrose (Sucrose) was affected by lactate in a dosage-dependent manner (Lactate%). Significance was calculated by ANOVA tests with LSD post hoc analysis for Fig. 3e and ANOVA tests with LSD post hoc analysis for others. Mean ± SEM. Symbols: NS p > 0.05; *p < 0.05; **p < 0.01; ***p < 0.001.

Figure 4

Drosophila preference for fermentation correlated with sucrose consumption and bacterial population. (a) The concentration of sucrose and viable bacterial cells in fly food. The concentrations of sucrose reduced over time, while bacterial density increased. n = 12, one of three replicates. (b) Drosophila temporal oviposition preference for fermentation. n = 16, one of three replicates. (c,d) The relationship between sucrose consumption, bacterial density and Drosophila oviposition preference for fermentation. A linear standard curve with an unconstrained slope was generated and compared to a null model with slope = 0. Each data point represents sucrose concentration or viable cell number of fly food along with the mean oviposition index value toward fermentation. A semilog standard curve with an unconstrained slope was generated and compared to a null model with slope = 0. The data fit to an unconstrained slope better than to the null model (For sucrose consumption: p < 0.0001, slope = −0.95; for bacterial density: p < 0.0001, slope = −0.95). ANOVA tests with LSD post hoc analysis. Mean ± SEM. Symbols: NS p > 0.05; **p < 0.01; ***p < 0.001.

Figure 5

The composition and function of rotting fruit-associated microbiota. (a) Two-choice oviposition assay for fruit purée. Fruit purée (20%w/v) replaced sucrose in the behavior-testing media, and oviposition preference was assayed between fresh and fermented choices. The one-sample t-test was used to assess the mean deviance of each column from 0. n = 4. (b) Composition and distribution of the dominant bacterial taxa within rotting apples and grapes. (c) Venn diagram showing the presence of bacterial taxa within two fruits. The number of bacteria in rotting apples and grapes was in the circles. (d,e) PICRUSt predicted microbiota function based on inferred metagenomes of rotting fruit-associated bacteria at the primary (d) and upper (e) level using the PICRUSt algorithm. (f) Venn diagram showing the distribution of predicted KEGG genes within two fruit samples. The one-sample t-test, Mean ± SEM, Symbols: ***p < 0.001.

Figure 6

Commensal bacteria were essential for survival and fitness of Drosophila. (a) Sucrose was essential for Drosophila survival. 30 eggs were placed in the casein-cornmeal-agar media with (+Sucrose) or without sucrose (−Sucrose) in 6-mm Petri dishes. The eclosed adults were counted for survival ratio. (b) Fermentation was not required to confer protection against endoparasitoid wasps. Schematic drawing of the Y maze olfactory assay used for behavioral experiments with the wasp Leptopilina boulardi. Thirty wasps were placed at the bottom of the Y maze with a choice of fresh or fermented grape juice and wasp counts from each branch were made after 20 min. The response index of L. boulardi in the Y maze olfactory assay. (c) Microbiota facilitated the timing of adult emergence. Germ free (GF) eggs were transferred to autoclaved vials to generate GF flies, while GF eggs were replenished with mixed bacteria to conventionally reared (CR) flies. The timing of adult emergence was recorded in a cornmeal media containing casein over time (the cutoff for GF flies was arbitrarily assigned as 25-days). (d) Microbiota promoted the larval growth. The length of larval bodies was measured at day 1, 3, 5 ALE. (e) The source of microbes solely supported Drosophila survival. Thirty eggs were placed in the agar media with (Blank) or without microbes (Bacteria, Yeast), and the eclosed adults were counted for survival ratio. (f) The timing of pupa formation and adult emergence of flies in grape vials was recorded. CR flies developed from eggs without sterilization. GF eggs were transferred to vials with sterile grapes, while Acetobacter and Enterococcus (AO + EF) were replenished in vials. The timing of pupa formation and adult emergence was recorded, respectively. ANOVA tests with LSD post hoc analysis. Mean ± SEM. Symbols: NS p > 0.05; **p < 0.01; ***p < 0.001.

Figure 7

Conservation of egg-laying preference for commensal bacteria. (a) The oviposition index of D. melanogaster for a diet fermented by commensal and pathogenic microbes. The one-sample t-test was used to assess the mean deviance of each column from 0, n = 4–7. (b) The oviposition preference for EF fermentation was conserved in drosophila species. The one-sample t-test was used to assess the mean deviance of each column from 0, n = 5. Mean ± SEM; Symbols: NS p > 0.05; *p < 0.05; **p < 0.01; ***p < 0.001.