In Drosophila, locomotor activity is sexually dimorphic and the brain area controlling this dimorphism has been mapped. The neurons of the pars intercerebralis (PI) have been suggested to participate in such differences between males and females. However, the precise physical nature of the dimorphism, the identity of the PI neurons involved, and the nature of the neuronal signal coding the dimorphism remain unknown. In this study, we used a video-tracking paradigm to characterize further the pattern of locomotor activity in Drosophila. We show that the number of activity͞inactivity periods (start͞stop bouts) is also sexually dimorphic, and that it can be genetically feminized in males. Moreover, the transplantation of PI neurons from a female, or of feminized PI neurons from a donor male into a receiver wild-type male is sufficient to induce the feminization of locomotor behavior, confirming that this tiny cluster of Ϸ10 neurons is directly responsible for the sexual dimorphism in locomotor activity. Finally, feeding males with fluvastatin, a juvenile hormone (JH) inhibitor, also led to start͞stop feminization, and this effect is reversible by the simultaneous application of methoprene, a JH analog, suggesting the existence of a neuroendocrine control, by JH, of such behavioral dimorphism.
Invertebrates and vertebrates show sexually dimorphic behaviors, and it has been suggested that males and females differ in the brain structures controlling such behavioral differences (1-4). Such sex differences have often been used as a way of studying how brain structure contributes to brain function. However, although many sexually dimorphic neural structures have been described (2-4), very few studies have yet demonstrated the direct involvement of a brain structure in a sexually dimorphic behavior (5, 6) and still less a behavior expressed in a nonsexual context. We recently showed that Drosophila locomotor activity is sexually dimorphic (7) outside any sexual context. We could abolish this behavioral dimorphism and create males with a female-like activity profile by genetic manipulations (8). Briefly, we used a UAS-transformer transgene (UAS-tra; ref. 9), the feminizing action of which has been shown to be cell-autonomous (10-12), under the control of different brainspecific enhancer trap P[Gal4] lines. This first study suggested strongly that a few neurons located in the mid-anterior part of the pars intercerebralis (PI) might participate in the control of this sexually dimorphic behavior (8). However, because in this previous study the fly was observed in a small tube, the locomotor activity was indirectly quantified when the fly crossed an infra-red light-gate. As a result, the precise locomotor parameters that accounted for the sexual difference could not be elucidated. Furthermore, the involvement of the PI neurons was indirectly deduced on the basis of a genetic ablation, and the molecular mechanisms linking those few PI neurons to the sexually dimorphic behavior remained to be identified. In this study, we develo...