Reversible silencing of pacemaker cells in the fly circadian network.

J BERNI; E ARANOVICH; A DEPETRIS CHAUVIN; MF CERIANI

Cold Spring Harbor Laboratories

Congreso; Neurobiology of Drosophila; 2009

CSHL

Rhythmicity in rest-activity cycles in Drosophila is under control of the circadian clock, which is based on self-sustaining, cell-autonomous transcriptional negative feedback loops. Several neuronal clusters have been shown to include a molecular oscillator. The small ventral lateral neurons (s-LNv) are at the top of the hierarchy because they direct rhythmic behavior in the absence of any other brain oscillators and command the pace of most of the remaining clusters. The s-LNvs transmit this time information releasing a neuropeptide known as pigment dispersing factor (PDF), and likely changing synaptic partners by remodeling their axonal terminals in a circadian fashion. It has been proposed that electrical activity in the PDF expressing subset of pacemaker neurons is required for the generation of circadian locomotor rhythms. In those experiments, a K+ channel (kir2.1) was over-expressed in the s-LNvs throughout the entire lifetime inducing behavioral arrhythmicity as well as blocking the free-running molecular oscillations. Using a new tool that allows the temporal control of gene expression in the s-LNvs (pdf-GeneSwitch), we revisited the previous results activating the expression of kir2.1 specifically in the adult, once the circuit has achieved the final configuration. We observed that silencing of the s-LNvs induced arrhythmicity in free-running conditions (constant darkness, DD) as previously described. Interestingly, if we later stop the expression of kir2.1, still in DD, the flies recovered rhythmicity, in a phase reminiscent of the initial entrainment, suggesting that the molecular oscillation had persisted even in the silenced neurons. Subsequent analyses of PERIOD oscillation in the s-LNvs in both conditions confirmed this was the case and strongly support the idea that the arrhythmicity was indeed a consequence of the incapability of these neurons to transmit information rather then a direct impact on the molecular clock.  Rhythmic release of PDF was abolished during electrical silencing, further supporting such conclusions. Altogether these results suggest that the changes in membrane potential of the s-LNvs are responsible of the rhythmic release of PDF and might not directly influence the molecular oscillation.