Increased sugar valuation contributes to the evolutionary shift in egg-laying behavior of the fruit pest Drosophila suzukii

Abstract

Behavior evolution can promote the emergence of agricultural pests by changing their ecological niche. For example, the insect pest Drosophila suzukii has shifted its oviposition (egg-laying) niche from fermented fruits to ripe, non-fermented fruits, causing significant damage to a wide range of fruit crops worldwide. We investigate the chemosensory changes underlying this evolutionary shift and ask whether fruit sugars, which are depleted during fermentation, are important gustatory cues that direct D. suzukii oviposition to sweet, ripe fruits. We show that D. suzukii has expanded its range of oviposition responses to lower sugar concentrations than the model D. melanogaster, which prefers to lay eggs on fermented fruit. The increased response of D. suzukii to sugar correlates with an increase in the value of sugar relative to a fermented strawberry substrate in oviposition decisions. In addition, we show by genetic manipulation of sugar-gustatory receptor neurons (GRNs) that sugar perception is required for D. suzukii to prefer a ripe substrate over a fermented substrate, but not for D. melanogaster to prefer the fermented substrate. Thus, sugar is a major determinant of D. suzukii's choice of complex substrates. Calcium imaging experiments in the brain's primary gustatory center (suboesophageal zone) show that D. suzukii GRNs are not more sensitive to sugar than their D. melanogaster counterparts, suggesting that increased sugar valuation is encoded in downstream circuits of the central nervous system (CNS). Taken together, our data suggest that evolutionary changes in central brain sugar valuation computations are involved in driving D. suzukii's oviposition preference for sweet, ripe fruit.

Fig 1. Sugar is valued more highly by D. suzukii in two-choice oviposition assays on natural fruit substrates.

(A) Oviposition substrate preference in a two-choice assay in a large chamber (see Methods for details) opposing a ripe strawberry purée (indicated by the red box above) and the same purée fermented for 3 d under controlled conditions (brown box below; see S1 Fig for additional information and controls). Preference is quantified by a preference index (see Methods). Filled circles in this and the following graphs indicate a significant preference for one of the 2 substrates (Mann–Whitney paired test); open circles indicate no significant preference for either substrate. Shaded bars: mean, error bars: standard deviation. D. melanogaster (blue) and D. suzukii (red) show opposite preferences for these substrates. Species preferences are significantly different from each other; Mann–Whitney U test, p-values indicated on the graph (n = 35, 30 replicates). (B) Stimulation (no-choice) assays with these substrates (indicated by red and brown boxes, respectively) show that neither is repulsive to the 2 species, both stimulate oviposition to a similar extent when presented alone (n = 20, 20, 20, 20). (C) Oviposition assay with ripe vs. fermented substrates for several Drosophila species (for some species, multiple wild-type strains were used; see Methods for species names; n = 50, 20, 20, 30, 20, 20, 20, 18, 20, 20, 20, 15, 15, 20, 20, 49, 17, 20, 20, 18, 20). (D) Two-choice assays opposing sugar alone (glucose + fructose at the concentration found in the ripe substrate, i.e.: 1.6%, indicated by the pink box above the graph) versus agar or agar with acetic acid (1%, similar to fermented substrates, orange boxes below the graph). D. melanogaster and D. suzukii show opposite preferences for sugar vs. acetic acid (n = 36, 39, 40, 40). (E, F) Relative value of acetic acid and sugar in the ripe and fermented substrates. (E) Adding 1% acetic acid to the ripe substrate shifts oviposition preferences toward acetic acid to a similar extent in both species (n = 40, 40, 40, 39). (F) Adding 1.6% sugar to the fermented substrate abolishes the preference of D. suzukii for the ripe substrate but does not shift D. melanogaster’s preference for the fermented substrate (n = 30, 30, 30, 30). The data underlying this figure can be found in S1 Dataset.

Fig 2. D. suzukii responds to lower concentrations of sugar than D. melanogaster in oviposition assays.

(A) Two-choice oviposition assays with sugar at the concentration of the ripe strawberry substrate (1.6% glucose + fructose) versus plain agar (empty box at the bottom) in 3 different experimental setups (see Methods for details). Both wild-type D. melanogaster and D. suzukii prefer sugar in each experimental setup, but the preference is more pronounced in D. melanogaster (n = 12, 20, 16, 30, 31, 48; see raw data in S2A Fig). (B) Two-choice assays opposing different concentrations of glucose. D. suzukii does not discriminate between concentrations when both are higher than approximately 1% glucose, whereas D. melanogaster always shows preference for the higher concentration (n = 35, 45, 23, 30, 43, 44, 29, 30). (C, D) Oviposition stimulation assays (no-choice) on increasing concentrations of (C) glucose, fructose, and sucrose and (D) glucose with 2 other wild-type strains of D. melanogaster (iso1) and D. suzukii (AM). Data are shown as the mean (dots + lines) +/− standard deviation (shaded areas). Sugar concentration (x-axis) is on a logarithmic scale. The estimated EC50s are shown at the bottom and with the dashed lines. The EC50 is consistently lower for D. suzukii compared to D. melanogaster (n = 30 for each condition). (E) Two-choice assays with low concentrations of glucose versus plain agar. The proportion of replicates choosing glucose (defined as preference index >0.2) is shown by blue/red bars, the proportion of replicates choosing agar (preference index <−0.2) in dark gray, and the proportion with no oviposition response (egg-laying rate <5 eggs) or no choice in light gray. A greater proportion of D. suzukii replicates choose sugar at low concentrations (0.05% to 0.5%) compared to D. melanogaster (n = 45 for each condition; see raw data in S2C Fig). (F) Two-choice assay opposing sugar concentrations corresponding to those of the ripe and fermented substrates (1.6% vs 0.2% glucose + fructose, respectively). D. suzukii shows a significant preference for the higher concentration substrate (n = 33, 45). The data underlying this figure can be found in S2 Dataset.

Fig 3. Sugar sensing is required for ripe substrate preference in D. suzukii.

DsuzGr64af-Gal4 expression in female proboscis (A) and legs (B) observed with a UAS-GCaMP7s-T2A-Tomato reporter. The mean number of positive neurons ± standard deviation is indicated for each organ on the upper side and each tarsus (n = 14 proboscises, forelegs, midlegs, and hindlegs imaged; see also S3A Fig for additional characterization of the DsuzGr64af-Gal4 line). (C) Comparative sugar GRN counts in D. melanogaster (blue) and D. suzukii (red) females determined from Gr64af-Gal4 reporter expression and electrophysiogical recordings of sensilla showing a consistent reduction in D. suzukii proboscis and forelegs compared to D. melanogaster (^athis study; ^b[22]; ^c[33]; ^d[34]; ^e[35]; nd: not determined). (D) Two-choice oviposition assay with 1.6% glucose + fructose vs. plain agar. Sugar-GRN inhibition via UAS-Kir2.1 expression reduces—but does not completely abolish—the response to sugar to a similar extent in both species (n = 25, 18, 34, 56, 35, 46, 30). (E) Oviposition choice assay on ripe vs. fermented substrate. Sugar-GRN inhibition does not alter the preference of D. melanogaster but significantly reduces the preference of D. suzukii for the ripe substrate (n = 30, 33, 33, 34, 33, 35, 26, 34). The data underlying this figure can be found in S3 Dataset.

Fig 4. Comparative calcium imaging of sugar responses in the PNS.

(A) GCaMP7s imaging in the synaptic terminals of sugar-GRNs in the SEZ upon stimulation of the proboscis with glucose. Example images are shown for the 2 species 15 s before stimulation, at the peak response, and 15 s after stimulation. Fluorescence intensity is color coded (scales on the side). ROIs used for quantification are indicated (dashed circles). (B, C) Stimulation with water, glucose, fructose, sucrose, and KCl. (B) ΔF/F₀ traces of individual ROIs plotted as heat maps (color scale at bottom right) for the different conditions. The period of water or sugar stimulation is indicated by gray rectangles at the bottom. (C) Distribution of peak ΔF/F₀ for each condition (shaded area: mean, error bars: standard deviation). The magnitude of the calcium response is generally higher for D. melanogaster at high sugar concentrations (Mann–Whitney U test, p-values indicated on the graph). p-Values in color below the graph indicate statistical comparisons with water controls (n = 9, 11, 10, 10, 8, 9, 8, 10, 10, 9, 8, 7, 8, 6, 7, 6, 8, 8, 22, 23 brains). See S4 Fig for additional information. (D, E) Manipulation of sugar perception in D. melanogaster shifts its oviposition preference in ripe versus fermented substrate assays. (D) Increasing sugar input to the CNS via UAS-NaChBac expression in sugar GRNs significantly increases the value of the ripe substrate. Compared to parental controls (gray and black), Gr64af>NaChBac females (green) show a shift in preference toward the ripe substrate (n = 30, 30, 30, 30). (E) Increasing the sugar concentration of the ripe substrate increases its value relative to the fermented substrate for D. melanogaster. One dose of glucose + fructose (1.6%) added to the ripe substrate (which already contains 1.6% sugar) eliminates the preference for the fermented substrate, whereas the addition of 2 doses (3.2%) reverses the preference in favor of the sweeter ripe substrate (n = 30 for each condition). The data underlying this figure can be found in S4 Dataset.