We have isolated three alleles of a novel Drosophila clock gene, double-time (dbt). Short- (dbtS) and long-period (dbtL) mutants alter both behavioral rhythmicity and molecular oscillations from previously identified clock genes, period and timeless. A third allele, dbtP, causes pupal lethality and eliminates circadian cycling of per and tim gene products in larvae. In dbtP mutants, PER proteins constitutively accumulate, remain hypophosphorylated, and no longer depend on TIM proteins for their accumulation. We propose that the normal function of DOUBLETIME protein is to reduce the stability and thus the level of accumulation of monomeric PER proteins. This would promote a delay between per/tim transcription and PER/TIM complex function, which is essential for molecular rhythmicity.

The Drosophila circadian clock consists of two interlocked transcriptional feedback loops. In one loop, dCLOCK/CYCLE activates period expression, and PERIOD protein then inhibits dCLOCK/CYCLE activity. dClock is also rhythmically transcribed, but its regulators are unknown. vrille (vri) and Par Domain Protein 1 (Pdp1) encode related transcription factors whose expression is directly activated by dCLOCK/CYCLE. We show here that VRI and PDP1 proteins feed back and directly regulate dClock expression. Repression of dClock by VRI is separated from activation by PDP1 since VRI levels peak 3-6 hours before PDP1. Rhythmic vri transcription is required for molecular rhythms, and here we show that the clock stops in a Pdp1 null mutant, identifying Pdp1 as an essential clock gene. Thus, VRI and PDP1, together with dClock itself, comprise a second feedback loop in the Drosophila clock that gives rhythmic expression of dClock, and probably of other genes, to generate accurate circadian rhythms.

The cloning of double-time (dbt) is reported. DOUBLETIME protein (DBT) is most closely related to human casein kinase Iepsilon. dbtS and dbtL mutations, which alter period length of Drosophila circadian rhythms, produce single amino acid changes in conserved regions of the predicted kinase. dbtP mutants, which eliminate rhythms of per and tim expression and constitutively overproduce hypophosphorylated PER proteins, abolish most dbt expression. dbt mRNA appears to be expressed in the same cell types as are per and tim and shows no evident oscillation in wild-type heads. DBT is capable of binding to PER in vitro and in Drosophila cells, suggesting that a physical association of PER and DBT regulates PER phosphorylation and accumulation in vivo.

Electrical silencing of Drosophila circadian pacemaker neurons through targeted expression of K+ channels causes severe deficits in free-running circadian locomotor rhythmicity in complete darkness. Pacemaker electrical silencing also stops the free-running oscillation of PERIOD (PER) and TIMELESS (TIM) proteins that constitutes the core of the cell-autonomous molecular clock. In contrast, electrical silencing fails to abolish PER and TIM oscillation in light-dark cycles, although it does impair rhythmic behavior. On the basis of these findings, we propose that electrical activity is an essential element of the free-running molecular clock of pacemaker neurons along with the transcription factors and regulatory enzymes that have been previously identified as required for clock function.

We identified a novel regulatory loop within Drosophila's circadian clock. A screen for clock-controlled genes recovered vrille (vri), a transcription factor essential for embryonic development. vri is expressed in circadian pacemaker cells in larval and adult brains. vri RNA levels oscillate with a circadian rhythm. Cycling is directly regulated by the transcription factors dCLOCK and CYCLE, which are also required for oscillations of period and timeless RNA. Eliminating the normal vri cycle suppresses period and timeless expression and causes long-period behavioral rhythms and arrhythmicity, indicating that cycling vri is required for a functional Drosophila clock. We also show that dCLOCK and VRI independently regulate levels of a neuropeptide, pigment dispersing factor, which appears to regulate overt behavior.

Summary Neural systems controlling the vital functions of sleep and feeding in mammals are tightly inter-connected: sleep deprivation promotes feeding, while starvation suppresses sleep. Here we show that starvation in Drosophila potently suppresses sleep suggesting that these two homeostatically regulated behaviors are also integrated in flies. The sleep suppressing effect of starvation is independent of the mushroom bodies, a previously identified sleep locus in the fly brain, and therefore is regulated by distinct neural circuitry. The circadian clock genes Clock (Clk) and cycle (cyc) are critical for proper sleep suppression during starvation. However, the sleep suppression is independent of light cues and of circadian rhythms because starved period mutants sleep like wild type flies. By selectively targeting subpopulations of Clk-expressing neurons we localize the observed sleep phenotype to the dorsally located circadian neurons. These findings show that Clk and cyc act during starvation to modulate the conflict of whether flies sleep or search for food.

We have studied the abilities of different transactivation domains to stimulate the initiation and elongation (postinitiation) steps of RNA polymerase II transcription in vivo. Nuclear run-on and RNase protection analyses revealed three classes of activation domains: Sp1 and CTF stimulated initiation (type I); human immunodeficiency virus type 1 Tat fused to a DNA binding domain stimulated predominantly elongation (type IIA); and VP16, p53, and E2F1 stimulated both initiation and elongation (type IIB). A quadruple point mutation of VP16 converted it from a type IIB to a type I activator. Type I and type IIA activators synergized with one another but not with type IIB activators. This observation implies that synergy can result from the concerted action of factors stimulating two different steps in transcription: initiation and elongation. The functional differences between activators may be explained by the different contacts they make with general transcription factors. In support of this idea, we found a correlation between the abilities of activators, including Tat, to stimulate elongation and their abilities to bind TFIIH.Stimulation of eukaryotic gene expression requires sequence-specific factors with DNA binding domains and activation domains that interact with the general transcription factors (GTFs) and recruit RNA polymerase II (pol II) to the promoter (3, 65). In vivo the transcriptionally active form of pol II is probably a holoenzyme complex which contains a number of the GTFs as well as other polypeptides (36). Different activation domains interact with different GTFs, including TFIIB (44), TFIID (17,19,63), and TFIIF (76). These interactions are thought to recruit, stabilize, and/or modify the activity of the pol II holoenzyme.We recently demonstrated that the activation domains of p53, VP16, and E2F1 bind directly to TFIIH (50a, 72). TFIIH is a multisubunit factor, different forms of which are required for both transcription and nucleotide excision repair of DNA (9). TFIIH is the only GTF which has enzymatic activities: it has two helicase subunits and a cyclin-dependent protein kinase subunit which phosphorylates the pol II large-subunit C-terminal domain (CTD) (12,52,58,60). Both of these enzymatic activities are implicated in steps in transcription which occur shortly after initiation of the RNA chain. First, a helicase is required for efficient formation of open complexes and for promoter clearance on linear templates in vitro (20). Second, when paused polymerases resume elongation on several Drosophila genes in vivo, the CTD becomes phosphorylated, suggesting a possible role for TFIIH kinase in regulating elongation (50, 70). Furthermore, inhibitors of the TFIIH kinase inhibit elongation under activated (74), but not basal (57), transcription conditions.In vivo, rate-limiting steps after initiation have been well documented for a number of genes. For example, polymerases stall 20 to 40 bases downstream of the start sites in the Drosophila hsp70 and human c-myc genes (38, 51, 64) and terminate...

Genome biology approaches have made enormous contributions to our understanding of biological rhythms, particularly in identifying outputs of the clock, including RNAs, proteins, and metabolites, whose abundance oscillates throughout the day. These methods hold significant promise for future discovery, particularly when combined with computational modeling. However, genome-scale experiments are costly and laborious, yielding “big data” that are conceptually and statistically difficult to analyze. There is no obvious consensus regarding design or analysis. Here we discuss the relevant technical considerations to generate reproducible, statistically sound, and broadly useful genome-scale data. Rather than suggest a set of rigid rules, we aim to codify principles by which investigators, reviewers, and readers of the primary literature can evaluate the suitability of different experimental designs for measuring different aspects of biological rhythms. We introduce CircaInSilico, a web-based application for generating synthetic genome biology data to benchmark statistical methods for studying biological rhythms. Finally, we discuss several unmet analytical needs, including applications to clinical medicine, and suggest productive avenues to address them.