Slide Session I

(Speakers in Bold)

Identification of novel genes associated with CLK­CYC complexes that regulate

circadian rhythms in Drosophila

GURUSWAMY MAHESH1, Jerry H Houl2, Ravi Amunugama3, Richard Jones3, David L Allen3, Paul Hardin6

1BIOLOGY, TEXAS A&M UNIVERSITY, COLLEGE STATION, TX, UNITED STATES

2Center for Biological Clocks Research, Texas A&M University, College Station, TX, UNITED STATES

3MS Bioworks, LLC, Ann Arbor, MI, UNITED STATES

6Biology, Texas A&M University, College Station, TX, UNITED STATES

Microbes, plants and animals use endogenous circadian clocks to regulate daily rhythms in metabolism physiology and behavior. In animals, a conserved set of clock genes keep circadian time and control rhythmic outputs via transcriptional feedback loops. The feedback loop in Drosophila has been well­characterized, and is initiated when CLOCK­CYCLE (CLK­CYC) heterodimers bind E­box elements to activate period (per) and timeless (tim) transcription. Accumulating levels of PER and TIM repressors feed back to inhibit CLK­CYC, followed by PER and TIM degradation, which permits the next round of CLK­CYC transcription. The timing and operation of this feedback loop is dependent on proteins that interact with CLK­CYC and PER­TIM complexes to regulate stability/degradation, nuclear localization, DNA binding, transcriptional activation, and transcriptional repression. Some of these proteins have been identified through genetic screens, but our understanding of how this feedback loop maintains a 24 h period and drives transcriptional rhythms is incomplete. To uncover proteins that control rhythmic transcription, we will identify proteins associated with CLK­CYC complexes throughout a diurnal cycle. For this, transgenes expressing GFP­tagged CLK and CYC were generated. Both transgenes rescue molecular and behavior rhythms in their respective null mutants, and are thus functional. We used GFP nanobodies to purify GFP­CLK and GFP­CYC complexes from fly heads at different times during a light­dark cycle and analyzed them by Mass Spectroscopy. As expected, CLK and CYC were observed in complexes at times when CLK­CYC activates transcription, and CLK, CYC, PER, TIM and DBT were detected when CLK­CYC transcription is repressed. Many additional proteins are associated with GFP­ CLK and GFP­CYC complexes including chromatin remodeling factors, transcription factors, kinases, phosphatases, and proteosome pathway genes among others. Some of these proteins are known to have circadian phenotypes including the prohibitin l(2)37Cc, the ubiquitin ligase Bruce, and the JARID1a homolog lid, demonstrating that these complexes contain proteins important for clock function. Current experiments focused on confirming the presence of GFP­CLK and GFP­CYC associated proteins in CLK­CYC complexes and determining whether novel GFP­CLK and GFP­CYC associated proteins disrupt behavioral and transcriptional rhythms will be presented.

Operating circuits in the Drosophila multi­oscillator system

Abhishek Chatterjee1, Angélique Lamaze1, Elisabeth Chélot1, Joydeep De1, Béatrice Martin1, Paul Hardin6, Patrick Emery7, Francois Rouyer8

1Institut de Neurobiologie Alfred Fessard, CNRS, Université Paris Sud, Gif­sur­Yvette, FRANCE

6Biology, Texas A&M University, College Station, TX, UNITED STATES

7Department of Neurobiology, University of Massachusetts Medical School, Worcester, MA, UNITED STATES

8INAF, CNRS, Gif­sur­Yvette, , FRANCE

Inter­oscillator communication modulates and sustains the circadian locomotor rhythms in flies and rodent models. In Drosophila, the multi­oscillator network that controls sleep­wake cycles include about 150 clock neurons. A subset of lateral neurons (LNs) expressing the Pigment­dispersing factor (PDF) appears to act as a master clock, at least under constant darkness (Stoleru, 2005). However, the extent of hierarchy and how is dominance exerted remains unclear. We are assessing the DN1p dorsal neurons to probe the logic of hierarchical network operation in the clock circuit.

In DD condition, whereas PDF­negative LNs run independently, the DN1ps rapidly abandons their intrinsic period and follow the pace of the PDF neurons. While the clock of the PDF cells alone determines the behavioral period, phase regulation is conveyed via the DN1ps. In LD conditions, the bimodal activity is generated by a more complex set of interactions. Although a functional clock restricted to the LNs or to the DN1ps can give rise to bimodal activity (Grima, 2004; Stoleru, 2004; Zhang, 2010), the morning (M) output of the DN1ps requires non­circadian PDF input from the PDF­expressing LNs. In contrast, PDF shows an inhibitory effect on the DN1p­mediated evening (E) peak. This hierarchal relationship between LNvs and DN1ps is opposed by CRY signaling within the DN1p. Our data reveal that PDF­expressing LNs organize the M peak through the DN1p conduit, whereas PDF­negative LNs control the E peak through a DN1p­independent circuit. While the morning output of DN1p is reliant on LNv, the evening output of DN1p is independent of LNd underscoring the semiautonomous nature of this oscillator. Our results also indicate that this coordinated three­pronged control of diurnal behavior is crucial for the photoperiodic adjustment of the fly’s locomotor activity.

Bride of DBT is a noncanonical FK506­binding protein that forms cytosolic foci

during the night and interacts with DBT to stimulate its circadian activity towards

PER.

Jin­Yuan Fan1, Boadi Agyekum2, Anandakrishnan Venkatesan2, David Hall2, Andrew Keightley2, Edward Bjes2, Samuel Bouyain7, Jeffrey Price1

1Molecular Biology and Biochemistry, UMKC, Kansas City, MO, UNITED STATES

2School of Biological Sciences, Molecular Biology and Biochemistry, UMKC, Kansas City, MO, UNITED STATES

7School of Biological Sciences, Molecular Biology and Biochemistry, UMKC, Kansas City, MO, UNITED STATES

To identify novel regulators of Drosophila Doubletime (DBT), the principal protein kinase targeting PERIOD (PER), a proteomic analysis of DBT interactors was undertaken. CG17282 was identified as one of these and shown to interact with DBT in GST­pull­ downs, and its N­terminal domain was shown to interact with DBT in co­immunoprecipitation assays. RNAi­mediated knock­down of CG17282 produced behavioral arrhythmicity, long­periods, high levels of hypophosphorylated nuclear PER and altered DBT phosphorylation. Because these phenotypes were suppressed by overexpression of DBT in flies and overexpression of CG17282 in S2 cells enhanced DBT­dependent PER degradation, CG17282 was shown to stimulate DBT’s circadian function. CG17282 accumulates rhythmically in cytosolic foci in the middle of the night, and pero and dbt RNAi reduce these foci. Determination of the three­ dimensional structure of CG17282 by X­ray crystallography demonstrated it is a noncanonical FK506­binding protein with an inactive peptide prolyl­isomerase domain that binds DBT and tetratricopeptide repeats that may promote assembly of larger complexes. We have named CG17282 Bride of DBT (bdbt) and established that it interacts with DBT through an inactive peptide prolyl­isomerase domain to stimulate DBT’s effects on PER. We are currently determining the regions of DBT that interact with BDBT, the biochemical consequences of this interaction and whether other Drosophila FK506­binding proteins are circadian components. The nature of the cytosolic foci is also being addressed.

Rhythmic Rho1 activity regulates pacemaker neuron structural plasticity and

seasonal adaptation

Afroditi Petsakou1, Themistoklis Sapsis2, Justin Blau3

1Biology, New York University, New York, UNITED STATES

2Courant Institute for Applied Mathematics, New York University, New York, UNITED STATES

3Biology, NYU, New York, NY, UNITED STATES

Circadian pacemaker neurons in Drosophila show clock­dependent daily rhythms in the structure of their axonal termini. However, the mechanism of this structural plasticity and its behavioral relevance are not well understood. In addition, the predictable nature of this plasticity makes pacemaker neurons an excellent system to study how gene expression regulates neuronal plasticity in vivo.

We first developed a method to quantify LNv termini and found that LNv structural plasticity consists of axonal growth and reduction in addition to the previously described cycle of fasciculation and defasciculation. This lead us to hypothesize that cytoskeletal changes underlie LNv plasticity and therefore that the Rho family of GTPases are involved. We found that inducing expression of Rho1 at dusk keeps LNv termini in a retracted dusk­like state 12hr later at dawn. We also found rhythms that peak at dusk in a sensor of endogenous Rho1 activity and in myosin light chain phosphorylation, a Rho1­regulated event that leads to actomyosin contraction. Therefore we conclude that endogenous Rho1 activity is dynamic in LNv axons. What provides circadian regulation and subcellular specificity to Rho1 activity in LNvs? Examining LNv expression profiles lead us to dPuratrophin (dPura), a novel Rho­GEF that shows clock­dependent circadian expression. We found that dPura genetically interacts with Rho1 and is required for Rho1 activity rhythms.

These molecular insights gave us tools to test the importance of LNv structural plasticity on circadian rhythms. We found that keeping LNv termini in a constitutively retracted state does not affect the LNv molecular clock but alters the phase of downstream clock neurons. This makes flies arrhythmic in DD and they also fail to adapt to altered daylength. Therefore Rho­regulated LNv structural plasticity is required for LNvs to communicate time across the circadian network, for rhythms per se and for flies to adapt to different seasons.

The hierarchy of landmark and celestial cues in animal navigation: Insight through

manipulating the circadian clock

James Cheeseman1, Craig Millar2, Uwe Greggers3, Konstantin Lehmann3, Matthew Pawley5, Charles Gallistel6, Guy Warman7, Randolf Menzel3

1Anaesthesiology, The University of Auckland, Auckland, Auckland, NEW ZEALAND

2School of Biological Sciences, The University of Auckland, Auckland, NEW ZEALAND

3Institute of Biology–Neurobiology, Free University of Berlin, Berlin, GERMANY

5Institute of Information and Mathematical Sciences, Massey University, Auckland, NEW ZEALAND

6Center for Cognitive Science, Rutgers University, , UNITED STATES

7Anaesthesiology, University of Auckland, Auckland, Auckland, NEW ZEALAND

Animals use time­compensated sun­compass navigation, landmarks and other earthbound sources to navigate. Honey bees are a particularly good model for studying animal navigation because they display expertise in a variety of computationally complex navigational tasks. By shifting the circadian clock of the bee, we put sun­compass­based directions in disagreement with map­based directions in bees displaced from a feeder to an experimenter­chosen release site. In the absence of prominent landmarks, clock­ shifted and control bees first flew their sun­compass­referenced home vector, before correcting their course and flying quickly home. However, in the presence of prominent landmarks, both clock­shifted and control bees flew a map­based course directly from the release site back to the hive, without first flying their sun­compass­referenced home vector. The results are not consistent with the hypothesis that displaced bees compute a home vector by summing the remembered sun­compass vectors associated with different landmarks perceptible at the release site. The results imply that displaced bees rely initially on the sun­compass and the dead­ reckoning home vector only when they fail to establish their current location (the release site) on their map. Finally, the results imply an integrated cognitive map, which represents the metric geometric relations between the hive, the feeder and the landmarks. We discuss the implications of these and other recent results for the computational neuroscience of navigation.

Molecular mechanism of temperature input to the Drosophila circadian clock

Ozgur Tataroglu1, Ania Busza2, Patrick Emery3

1UMass Medical School, Worcester, Massachusetts, UNITED STATES

2Neurobiology, UMass Medical School, Worcester,

3Neurobiology, University of Massachusetts Medical School, Worcester, MA, UNITED STATES

Circadian clocks receive and integrate environmental inputs such as light and temperature. Although light input has been extensively studied, the molecular mechanism by which circadian clocks sense temperature remains poorly understood. We found that a physiologically relevant temperature pulse (TP) shifts the central and peripheral clocks in Drosophila by triggering the specific degradation of TIMELESS (TIM) in vitro and in vivo. This degradation is not mediated by CRY or the proteasome, and is therefore mechanistically distinct from photic TIM degradation. Pharmacologically­induced increase of intracellular calcium in isolated heads results in TIM degradation similar to that induced by TP, while buffering of Calcium with Parvalbumin blocks thermal phase­shifts and TIM degradation in vivo. We identified six redundant putative Calmodulin (CaM)­binding sites on TIM and preliminary data suggest that TIM binds to CaM in a calcium­dependent manner. Mutations of these sites result in a thermostable TIM and blocks TP­induced phase shifts in vivo. Finally, TIM’s thermal degradation is mediated by Small optic lobes (SOL), an atypical calcium­dependent protease. Reduction of SOL levels using RNAi results in reduced thermal TIM degradation and blocks phase shifts both in vitro and in vivo. In summary, we identify several novel components of a cell­autonomous, Calcium­dependent circadian thermo­sensory pathway. Our study also shows that both light and temperature converge on TIM, which provides a simple mechanism for the synergistic entrainment of the Drosophila circadian clock by these environmental cues.