By selectively speeding up the brain or the PG clock we show that the brain clock exerts a dominant action on the PG clock. Assuming that sNPF is temporally co-released with PDF, our findings imply that the central clock’s contribution to the gating of eclosion would be mediated through an inhibition of steroidogenesis starting in the early morning, when PDF release is believed to be maximal47. This signal would then cause the ecdysone titer to drop below the required threshold30,31, resulting in the commitment to emerge during the next gate to be made around lights-off48 and causing the gate to open starting at dawn of the following day.



　　Ecdysone feeds into the molecular clockwork by modulating clockwork orange and clock/period expression via the ecdysone receptor and the downstream targets, Eip75 (ecdysone-induced protein 75) and microRNA let-7 (refs 54, 56, 57). Interestingly, Eip75B is the fly homologue of REV-ERB α and REV-ERB β, which are key regulatory elements of the core molecular clock of mammals6,29. This notable conservation between mammals and Drosophila in the functioning and steroid feedback of the circadian clock extends the previously noted conservation at the level of the pacemaker itself58 and at the level of the intra-pacemaker coordination by peptides activating the same Family B1 GPCR signalling pathway vasoactive intestinal polypeptide (VIP), and PDF, in mice and flies, respectively26,59). This could mean that the genetically tractable circadian system of the fly could also be used to understand how the different mammalian clocks are coordinated. It will also be interesting to determine whether similar deep homologies extend to the actions of the sNPF signalling pathway LED High Bay Light and those of its vertebrate homologue, prolactin-releasing peptide60, which has been implicated in the control of mammalian circadian rhythmicity and/or sleep regulation61.

　　Methods

　　Fly rearing and fly stocks

　　All flies were reared at room temperature (21–25 °C) on standard cornmeal/molasses/yeast food medium. All stocks with the exception of phm-gal80 and LexAop-Channelrhodopsin2-XXL (see below) have previously been described and are listed in Supplementary Methods. All genotypes involving the use of UAS-RNAi transgenes included a copy of a UAS-dcr2 transgene. When performing RNAi knockdown, several UAS-RNAi lines were tested and the one producing the most extreme phenotype was used (see Supplementary Methods).

　　phm-gal80

　　A 1.2 Kb DNA fragment containing the phm promotor was PCR-amplified using primers (see Supplementary Table 1) that included BglII (5′) and SpeI (3′) restriction sites, and cloned into the pGEM-T easy vector (Promega, WI, USA). It was then sequenced for verification, cut out using BglII and SpeI, and cloned into pPelican P-element vector62 cut with BamHI and SpeI (first purified away from the fragment containing the HSP70 TATA and eGFP DNA); this last step was completed by Genewiz (NJ, USA). DNA from sequence-verified clone was then sent for transformation to BestGene (Chino Hills, CA, USA).

　　LexAop-Channelrhodopsin2-XXL