How food controls aggression in Drosophila

Abstract

How animals use sensory information to weigh the risks vs. benefits of behavioral decisions remains poorly understood. Inter-male aggression is triggered when animals perceive both the presence of an appetitive resource, such as food or females, and of competing conspecific males. How such signals are detected and integrated to control the decision to fight is not clear. For instance, it is unclear whether food increases aggression directly, or as a secondary consequence of increased social interactions caused by attraction to food. Here we use the vinegar fly, Drosophila melanogaster, to investigate the manner by which food influences aggression. We show that food promotes aggression in flies, and that it does so independently of any effect on frequency of contact between males, increase in locomotor activity or general enhancement of social interactions. Importantly, the level of aggression depends on the absolute amount of food, rather than on its surface area or concentration. When food resources exceed a certain level, aggression is diminished, suggestive of reduced competition. Finally, we show that detection of sugar via Gr5a+ gustatory receptor neurons (GRNs) is necessary for food-promoted aggression. These data demonstrate that food exerts a specific effect to promote aggression in male flies, and that this effect is mediated, at least in part, by sweet-sensing GRNs.

Figure 1. Food is necessary for normal levels of male-male aggression but not male-male courtship and its effects are independent from locomotion or encounter duration.

(a) Flies performed more lunges during the observation period in the presence of a 22×22 mm food patch. n = 171, 92 male-male pairs tested for apple juice food patch and agarose patch, respectively. (b) Flies performed more lunges in the presence of arena, which was entirely covered with food. n = 113 and 44, for uniform food and uniform agarose, respectively. (c) Top: Schematic diagram of the aggression assay arenas used. Left side shows the food patch configuration and right side shows the uniform food configuration. A pair of male flies is illustrated at scale for comparison. Bottom: Position heat map shows the average amount of time flies spend in a particular position in the arena. The data shown are averages of multiple pairs of flies (same sample numbers as Figures 1a and 1b). It uses a red-blue color map from MATLAB where deep red is high frequency (60 frames, which is roughly 2 seconds, are the deepest-red) and blue is 0. Every subsequent position heat map is presented in the same manner. On the left, flies are attracted to the patch of food while on the right the uniform food does not lead to attraction to a specific spot in the arena. (d) Uniform food does not change the amount of time flies spend at various distances from each other. The inter-fly distance histogram shows amount of time flies spend (y-axis) at a given distance from each other (x-axis). The distribution is not affected by the presence of food (1-way ANOVA). There is a very prominent peak around 3–4 mm, which ranges from 2 mm (less than 1 body length of flies) to 10 mm (3–4 body lengths), and accounts for around 50% of the 20 minutes assay. The area under the curve from 0 to 10 mm is hereafter referred to as “encounter duration”. The trace is the median trace from 72 and 44 male-male pairs for food and agarose, respectively. (e) Uniform food does not increase encounter duration. Assay is 20 minutes long (1200 seconds). Same number of samples as Figure 1d. (f) Locomotion (distance traveled) in a pair of flies is increased in the presence of food. Same number of samples as Figure 1d. (g) Normalization of aggression by locomotion by dividing the number of lunges by travel distance shows that food significantly increases aggression. Same number of samples as Figure 1d. (h) Number of one-wing extensions is not changed by the presence of uniform food. Manually scored data consisting of n = 17 and 18 pairs for food and agarose conditions, respectively. (i) Normalization of courtship (number of circling bouts) by locomotion shows that food decreases male-male courtship. Same number of samples as Figure 1d. (j) In the first three minutes, food progressively increases aggression (blue circle). In contrast, one-wing extension decreases (red circle). In the absence of food, lunges do not increase or decrease (blue box); courtship decreases (red box). See Table S1 for statistics. Manually scored data of lunges and 1-wing extensions. n = 33 and 33 for food and agarose conditions for lunges. n = 34 and 31 for food and agarose conditions for one-wing extensions.

Figure 2. Flies measure the level of total nutrients to increase the level of aggression, rather than the area of food.

(a) Aggression increases as the size of food patch increases. See Figure S4 for schematic diagrams of the arena used. n = 41, 39, and 52 male-male pairs for 0, 79, 707 mm², respectively. Same pairs are further analyzed for Figures 2b-d. (b) Locomotion also increases in some cases (0 vs. 707 mm²) as the size of food increases. (c) Aggression normalized by locomotion is significantly increased in the presence of food. (d) Male-male courtship normalized by locomotion is not changed by the presence of food. (e) Left: Increasing the concentration of food while keeping the size of food constant (707 mm²) increases aggression. Right: Increasing the size of food while keeping the concentration constant also increases aggression. The concentration-dependent increase in aggression is quantitatively similar to the size-dependent increase in aggression. The absolute nutritional content remains the same between the left and the right (1∶235 = 3 mm², 1∶54 = 13 mm², etc). Some of the data in E are the same as those used in A and are replotted here for comparison purposes. n = 41, 22, 16, 29, 28, 31, 36, 37, 39, 27, and 52 male-male pairs from left to right.

Figure 3. Flies decrease the level of aggression as the availability of food resource increases.

(a) The relationship between aggression (y-axis) and the amount of food (x-axis). Aggression initially increases from 0 mm² to 707 mm² and decreases as the size of food increases further. In particular, aggression observed with the largest size tested 2376 mm² is significantly lower than 707 mm² after correcting for multiple comparisons. Some of the data are the same as those used Figure 2 and are replotted here for comparison purposes. n>28 male-male pairs for each condition tested. Pairs are further analyzed for Figures 3b and 3c. (b) Aggression normalized by locomotion shows the same initial increase and subsequent decrease (See Table S3 for pair-wise comparison statistics). (c) Male-male courtship normalized by locomotion shows no increase or decrease (See Table S4 for statistics). (d) The decrease in aggression seen in the largest food patch tested (left, 2376 mm²) can be reversed by decreasing the concentration of food to 30% (middle). Calorically, this condition is equivalent to 707 mm² food patch with 100% concentration of food (right) and the amount of aggression is indistinguishable. The 707 mm² food patch data replotted for comparison purposes. n = 32, 31, 86 male-male pairs from left to right. (e) The increase in aggression by dilution of food is significant after normalization for locomotion. n = 32, 31. (f) There is no change in courtship caused by the dilution of food. n = 32, 31.

Figure 4. Flies display territorial behavior.

(a) Top row: Schematic diagrams show the arenas with different size of food being used. Bottom row: Position heat-map of a pair of flies presented with different sizes of food. The heat-maps display two features: 1) flies spend a lot of time on top of food and 2) they spend a lot of time near the border of the food area. n = 41, 29, 86 and 41 male-male pairs from left to right. (b) Position heat map compares the distribution of flies on 30 mm and 45 mm diameter food when there is only 1 fly in the arena (left) and when there are two flies (right). 2-fly data from one experiment are individually averaged. n = 30 and 52 for 30 mm diameter food, single and pairs of flies, respectively. n = 25 and 41 for 45 mm diameter food, single and pairs of flies, respectively. The pairs are further analyzed in Figures 4c – 4f. (c and d) These histograms show the amount of time flies spend at different distances from the border of 30 mm (c) food and 45 mm (d) patch. The schematic diagrams of the behavioral setups are overlaid for visualization. Briefly, the x-axis is aligned so that 0 denotes the border of food patch while negative values indicate the distance inward from food border (inside the food patch) and positive values indicate the distance outward from the food border (outside of food patch). The blue line denotes when there is a single fly in the arena while the orange line denotes when there is a pair of flies. Lines indicate the median while shaded area denotes the interquartile range. (e) Presence of another fly increases the amount of time flies spend in Zone B (“interaction zone”) for both 30 mm and 45 mm food patches. (f) Presence of another fly does not change the amount of time flies spend on the food patch (Zone A).

Figure 5. Flies use sweet-sensing Gr5a⁺ GRNs to detect the concentration of sucrose in the food and tune the level of aggression accordingly.

(a) 100 mM sucrose is sufficient to increases aggression. (b) Sucrose does not cause attraction, as it does not lead to an apparent change in the position heat map. n = 100 and 60 for 100 mM sucrose and agarose, respectively. Pairs are further analyzed from Figures 5b-5g. (c) Presence of sucrose does not change the amount of time flies spend near each other. (d) Encounter duration does not change in the presence of sucrose. (e) Sucrose increases locomotion. (f) Sucrose increases the number of lunges per meters traveled, which implies that the increase in aggression is not merely due to increased locomotion. (g) Sucrose does not change the number of circling per meters traveled. (h) Changing sucrose concentration increases and decreases aggression. The level of aggression is increased from 0 to 200 mM but becomes indistinguishable from no food condition at 800 mM. (*): 100 to 200 mM difference is significant when individually compared (P<0.05) but not after corrections for multiple comparisons. n = 32, 23, 10 and 26 from left to right. (i) Inhibiting the sugar-sensing Gr5a⁺ GRNs by expressing TNT decreases sucrose sensitivity (n = 3 and 3 for both genotypes. Each replicate has 10 male flies to calculate fraction of responders). (j) Inhibiting the sugar-sensing Gr5a⁺ GRNs by expressing TNT decreases food-promoted aggression compared to genetic controls. n = 36, 41, and 32 from left to right.