Generalization of courtship learning in Drosophila is mediated by cis-vaccenyl acetate

Abstract

Reproductive behavior in Drosophila has both stereotyped and plastic components that are driven by age- and sex-specific chemical cues. Males who unsuccessfully court virgin females subsequently avoid females that are of the same age as the trainer. In contrast, males trained with mature mated females associate volatile appetitive and aversive pheromonal cues and learn to suppress courtship of all females. Here we show that the volatile aversive pheromone that leads to generalized learning with mated females is (Z)-11-octadecenyl acetate (cis-vaccenyl acetate, cVA). cVA is a major component of the male cuticular hydrocarbon profile, but it is not found on virgin females. During copulation, cVA is transferred to the female in ejaculate along with sperm and peptides that decrease her sexual receptivity. When males sense cVA (either synthetic or from mated female or male extracts) in the context of female pheromone, they develop a generalized suppression of courtship. The effects of cVA on initial courtship of virgin females can be blocked by expression of tetanus toxin in Or65a, but not Or67d neurons, demonstrating that the aversive effects of this pheromone are mediated by a specific class of olfactory neuron. These findings suggest that transfer of cVA to females during mating may be part of the male's strategy to suppress reproduction by competing males.

Figure 1. Courtship suppression learned with a mated female trainer is generalized to all types of females

A) Training with an immature decapitated virgin female produces suppression with only when the tester is also an immature female. B) Training with a mature decapitated female produces suppression of courtship with mature (virgin and mated) testers. C) Training with a decapitated mated female produces a generalized suppression to all female testers. Data was analyzed as described in Experimental Procedures; * means P < 0.05 for the indicated comparison. Trainer and tester types are abbreviated as follows: Md, decapitated mated female; mVd, decapitated mature virgin; iVd, decapitated immature virgin. Histogram bars for data in which males were trained with mated females are black; with mature virgins, dark gray, immature virgins, light gray.

Figure 2. The generalizing cue is a volatile pheromone

Histogram bars for data in which males were trained with mature virgins, dark gray, immature virgins, light gray; no courtship objects, white. * indicates P < 0.05 for comparison to control. A) Males were trained in a courtship arena with two chambers separated by a mesh as shown in the diagram. The male was placed in the top chamber with or without a courtship object. The bottom chamber contained a decapitated fly on a filter paper soaked with either water or cVA. Mesh position vis-à-vis cues is indicated over each histogram bar in panels B and C by hatching. B) Training with a mature decapitated female (mVd) produces very little suppression of immature virgin (iVd) courtship (left bar, labeled ‘control’). Addition of a decapitated mated female (Md), a decapitated male (male-d) or pure cVA to the bottom chamber causes the male to generalize his unsuccessful courtship experience and suppress courtship of the immature tester. Presentation of a mated female or a male across the mesh in the bottom chamber in the absence of courtship object does not alter courtship of the immature tester. C) Training with an immature decapitated female produces very little suppression of mature virgin courtship (left bar, labeled ‘control’). Addition of a decapitated mated female (Md), a decapitated male (male-d) or a high dose of cVA to the bottom chamber causes the male to generalize his unsuccessful courtship experience and suppress courtship of the mature tester. Presentation of a mated female or a male across the mesh in the absence of courtship object does not alter courtship of the mature tester.

Figure 3. Cuticular hydrocarbons of mature virgins and mated females differ in levels of cis-vaccenyl acetate

Cuticular hydrocarbons were extracted in hexane and analyzed by gas chromatography-flame ionization detection and mass spectrometry. A) Total hydrocarbon profiles for mature virgin (top), mated female (middle) and male (bottom) extracts. Each profile is from a hexane extract of 20 age-matched virgin or mated (24 h before extraction) females. B) Quantification of major compounds for mature virgin (light bars, n = 28) and mated female (black bars, n = 29) single fly extracts. Mated females were extracted 24 h after copulation. * indicates P < 0.05 for comparison between mated and virgin female. C) 7-tricosene (left panel) and cVA (right panel) levels from decapitated mature virgins (mV, n = 28), decapitated mVs that were courted, but not mated, and extracted immediately (mVm, n = 22), mated females who copulated > 14 min (M, n = 29) and mated females who had copulation interrupted after 2 min (M2, n = 16). Mated females were extracted 24 h after copulation. * indicates P < 0.05 for comparison of the marked female type to all other types. Other types were not significantly different from one another (P > 0.05).

Figure 4. Modulation of courtship by cVA requires Or65a-GAL4, but not Or67d-GAL4 olfactory sensory neurons

Initial courtship level of naive males was measured with a mature virgin courtship object in a two chambered courtship arena with a filter containing either cVA or water in the lower chamber. * means P < 0.05 for indicated comparison; N.S. means not significant. A) Wild type males court mated females with less vigor than they court mature virgins. This courtship decrement can be reproduced by the odor of cVA. UAS-TNT/+ and Or65a-GAL/+ males are indistinguishable in cVA sensitivity from wild type. Expression of tetanus toxin under control of two independent Or67d promoter GAL4 drivers does not eliminate cVA-dependent modulation of initial courtship. B) Animals expressing active tetanus toxin under control of our Or65a-GAL4 or a published Or65a line [25] (V-Or65a-GAL4) fail to suppress initial courtship in the presence of cVA. The difference in strength of blockade is likely to be due to the fact that Or65a-GAL4 is a much stronger driver than V-Or65a-GAL4 (see Supplemental Figure 2). Expression of mutant tetanus toxin (TNT-VA) does not block the courtship suppressing effects of cVA with either driver.