Michael Rosbash

The Drosophila adult central brain contains 240 circadian neurons, of which there are more than 25 different neuron subtypes based on connectomic data. Recent single-cell RNA-seq characterization of these neurons “around the clock” also indicates a similar number of molecular subtypes of circadian neurons, but other conclusions from these transcriptomic studies warranted verifying and extending with other approaches. To address these limitations, we used three complementary approaches: 1) We used a genetic multiplexing strategy to profile the transcriptomes of circadian neurons from multiple time points in a single experiment, reducing confounding technical variation between timepoints; 2) Large numbers of single nuclei were sequenced (snRNA-seq), which was enabled because the method El-INTACT purifies nuclei from frozen heads; 3) We assayed 12 time points under both light–dark and constant darkness conditions. These approaches showed dramatic transcriptional differences between time points in many circadian neuron types and enhanced time-of-day gene expression analysis. The data indicate that most of this regulation is transcriptional and circadian. There were however a small number of light-dependent transcripts, including some that correspond to mammalian immediate-early genes. They probably play a role in the light-regulation of gene expression and behavior in specific neurons, perhaps circadian entrainment or phase-shifting. The results taken together provide a more comprehensive picture of gene expression heterogeneity within adult Drosophila circadian neurons including how intrinsic clock mechanisms and light cues are integrated across circadian neuron subtypes.

Neurons engage compensatory pathways that promote survival when confronted with mitochondrial dysfunction. Indeed, we recently showed that Drosophila neurons upregulate Ldh transcription to help survive the loss of the key mitochondrial fusion gene Opa1. Here, we further characterize this metabolic flexibility and show that it reflects a more general increase in glycolytic activity. A distinct mitochondrial perturbation, TFAM overexpression, similarly induces glycolytic gene expression including Ldh and also elevates lactate levels. LDH is also required to maintain neuronal function under TFAM overexpression. Notably, raising NAD ⁺ /NADH ratio by expressing the bacterial NADH oxidase LbNOX does not substitute for LDH function. On the contrary, it further compromises neuronal function in Opa1 -deficient and TFAM -overexpressing neurons. Moreover, mitochondria-targeted LbNOX expression alone induces mitochondrial dysfunction and the compensatory glycolytic response. Together, these findings indicate that LDH-mediated rescue does not reflect an increase in NAD ⁺ /NADH ratio but is part of a broader neuroprotective metabolic reprogramming which enables neurons to withstand diverse forms of mitochondrial impairment.

The Drosophila adult central brain contains 240 circadian neurons, of which there are more than 25 different neuron subtypes based on connectomic data. Recent single cell RNA-seq (scRNAseq) characterization of these neurons “around the clock” also indicates a similar number of molecular subtypes of circadian neurons, but other conclusions from these transcriptomic studies warranted verifying and extending with other approaches. To this end: 1) We used a genetic multiplexing strategy to profile the transcriptomes of circadian neurons from multiple time points in a single experiment, reducing confounding technical variation between timepoints; 2) Large numbers of single nuclei were sequenced (snRNA-seq), which was enabled because the new method EL-INTACT purifies nuclei from frozen heads; 3) We assayed 12 time points under both light–dark (LD) and constant darkness (DD) conditions. These approaches showed dramatic transcriptional differences between time points in many circadian neuron types and enhanced time-of-day gene expression analysis. The data indicate that most of this regulation is transcriptional and circadian. There were however a small number of light-dependent transcripts, including a few that correspond to mammalian immediate-early genes. They probably play a role in the light-regulation of gene expression and behavior in specific neurons, perhaps circadian entrainment or phase-shifting. The results taken together provide a more comprehensive picture of gene expression heterogeneity within adult Drosophila circadian neurons including how intrinsic clock mechanisms and light cues are integrated across circadian neuron subtypes.

The Drosophila central brain is remarkably heterogeneous. Molecular as well as connectomic data indicate that many of its ca. 65,000 neurons are present in only a few copies per brain. This makes molecular characterization of individual cell types challenging. To address this issue, we developed Elution-based INTACT (El-INTACT), a nuclei-based purification technique that offers much higher purity, yield and efficiency compared to existing methods. El-INTACT successfully isolated two subsets of circadian neuron nuclei, one with ∼120 nuclei per brain and the other with only ∼16 nuclei per brain; the latter are the well-known circadian LNvs or PDF neurons. The method facilitated ATAC-Seq ( A ssay for T ransposase A ccessible C hromatin) assays, which revealed temporally regulated enhancers that cycle throughout the day within clock neurons. The ATAC-seq assays also identified cell-type-specific enhancers that correlate with adjacent genes, which are also only expressed in certain neuron types. To test the functional relevance of these enhancers, we developed a second approach, a CRISPR/Cas9 and multiplexed guide-RNA strategy and used it to disrupt individual enhancers of the core circadian gene Clock (Clk) . Our results show that these two methods, El-INTACT to identify enhancers and a CRISPR-based strategy to perturb these enhancers, significantly improves the understanding of gene regulation within discrete neuron types.

While many sexually dimorphic behaviors exhibit distinct time-of-day preferences, our understanding of how sex shapes the molecular and circuit properties of central brain neurons remain limited. Here, we uncover the transcriptomic and circuit basis of sexual dimorphism within the Drosophila circadian network. By leveraging single-cell RNA sequencing of male and female clock neurons, we identify specific subsets of dorsal lateral neurons (LNds), dorsal neurons 1p (DN1ps), and dorsal neurons 3 (DN3s) with dramatic dimorphic gene expression profiles. These sex differences are primarily characterized by cell-type-specific expression of genes involved in neural connectivity, particularly cell adhesion molecules (CAMs). Focusing on the dimorphic Cry-negative E3 LNds, we show that they form functionally active, synaptic connections with downstream doublesex -expressing pC1 and pCd-1 neurons, which serve as central regulators of dimorphic behaviors. Moreover, we demonstrate that formation and maintenance of these connections are mediated at least in part by sex-enriched CAMs, dpr9 in males and dpr3 in females. Thus, our work reveals sexual differentiation mechanisms at both the molecular and circuit levels, identifying specific molecules that sculpt sex-specific pathways to link the circadian clock to dimorphic outputs.

Circadian timekeeping is orchestrated by multiple clock proteins, but mechanistic understanding is limited. AlphaFold predicts novel interactions between two helical domains of the Drosophila PERIOD (PER) protein and its major kinase Doubletime (DBT). These PER domains, SYQ and LT, are conserved with mammals and are predicted to interact identically with the DBT ortholog casein kinase 1 (CK1). SYQ and LT mutations designed to disrupt this interface result in severe period and temperature-sensitive fly phenotypes, and two LT mutations cause unprecedented free-running periods ≥ 44 hours and temperature-sensitive changes of ∼20 hours. Further characterization of a LT mutant strain shows a severe reduction in both PER phosphorylation and PER degradation. A human PER equivalent of the same LT mutation is a much less efficient in vitro substrate for CK1-mediated phosphorylation. These observations indicate that this conserved PER-DBT interaction is critical for PER phosphorylation, circadian timing and its enigmatic temperature compensation phenomena.

Circadian neurons within animal brains orchestrate myriad physiological processes and behaviors, but the contribution of these neurons to the regulation of sleep is not well understood. To address this deficiency, we leveraged single-cell RNA sequencing to generate a comprehensive census of transcriptomic cell types of Drosophila clock neurons. We focused principally on the enigmatic DN3s, which constitute most fly brain clock neurons and were previously almost completely uncharacterized. These DN3s are organized into 12 clusters with unusual gene expression features compared to the more well-studied clock neurons. We further show that previously uncharacterized DN3 subtypes promote sleep through a G protein–coupled receptor, TrissinR . Our findings indicate an intricate regulation of sleep behavior by clock neurons and highlight their remarkable diversity in gene expression and functional properties.

Sleep is essential and highly conserved, yet its regulatory mechanisms remain largely unknown. To identify sleep drive neurons, we imaged Drosophila brains with calcium-modulated photoactivatable ratiometric integrator (CaMPARI). The results indicate that the activity of the protocerebral bridge (PB) correlates with sleep drive. We further identified a key three-layer PB circuit, EPG-SpsP-PEcG, in which the four SpsP neurons in the PB respond to ellipsoid body (EB) signals from EPG neurons and send signals back to the EB through PEcG neurons. This circuit is strengthened by sleep deprivation, indicating a plasticity response to sleep drive. SpsP neurons also receive inputs from the sensorimotor brain region, suggesting that they may encode sleep drive by integrating sensorimotor and navigation cues. Together, our experiments show that the four SpsP neurons and their sleep regulatory circuit play an important and dynamic role in sleep regulation.

Sleep is an essential and conserved behavior, yet the mechanisms underlying sleep regulation remain largely unknown. To address the neural mechanisms of sleep drive, here we carry out whole brain calcium-modulated photoactivatable ratiometric integrator (CaMPARI) imaging of Drosophila and show that the activity of the protocerebral bridge (PB), a part of the central complex, correlates with sleep drive. Through a neural activation screen followed by anatomical and functional connectivity assays, we further narrow down the key player of sleep regulation in the PB to a three-layer circuit composed of 4 SpsP neurons and their upstream and downstream synaptic partners: the 4 SpsP neurons act as an integrating hub by responding to ellipsoid body (EB) signals from EPG neurons, and by sending signals back to the EB through PEcG neurons. Moreover, sleep deprivation enriches the presynaptic active zones of SpsP neurons and strengthens the connections of the EPG-SpsP-PEcG circuit, indicating plasticity gating in the circuit in response to sleep drive change. As the SpsP neurons also receive input from the sensorimotor brain region and given their known role in navigation, these neurons potentially further integrate sleep drive with other sensorimotor cues. The data taken together indicate that the four SpsP neurons and their sleep regulatory circuit play an important and dynamic role in sleep regulation.