Background

Light is the most powerful environmental signal for entraining the biological clock to external day and night. The central brain of adult Drosophila contains approximately 240 circadian neurons, with over 25 subtypes identified through connectome analysis. While this small number of cells regulates sleep, wakefulness, and activity timing, the specific genes expressed by each neuron throughout the day and their responses to light have not been sufficiently resolved.

Existing single-cell RNA sequencing faced difficulties in distinguishing technical batch effects from actual circadian changes because samples from different time points were processed separately. There was also the issue of losing large, fragile neurosecretory neurons during the cell dissociation process. Specifically, large ventral lateral neurons (l-LNv) and some dorsal lateral neurons (LNd) were underrepresented compared to their actual abundance in the brain. Consequently, there remained limitations in separating transcription directly induced by light from the rhythms generated by the internal cellular clock.

Key Findings

Researchers at Brandeis University optimized EL-INTACT, which isolates labeled nuclei from frozen Drosophila heads, for single-nucleus RNA sequencing (snRNA-seq). Here, single nucleotide polymorphisms from the Drosophila Genetic Reference Panel (DGRP) were used as sample barcodes. This design reduces batch effects in 10x Genomics experiments by processing nuclei from different time zones together and using genotypes to trace back the original conditions.

The researchers investigated 12 time points at 2-hour intervals under both light-dark cycles and constant darkness. In the light-dark cycle, 9,899 circadian neuron nuclei were obtained, and 9,835 were obtained on the second day of constant darkness; both datasets were classified into 24 transcriptomic clusters. Rare neurosecretory neurons, including l-LNv, which was nearly missing in previous analyses, were recovered. This significantly increased the sample density for detecting time-dependent changes.

In several neuron types, the same cell groups clearly diverged depending on the collection time, and numerous transcripts, including core clock genes timeless and Pdp1, oscillated even in constant darkness. This provides evidence that most observed changes stem from intrinsic circadian transcriptional regulation rather than simple light-dark transitions. The researchers applied the JTK_cycle algorithm to determine 24-hour rhythms per gene, using strict criteria reflecting maximum expression and amplitudes of more than twofold.

Transcripts dependent solely on light were fewer than expected. However, Hr38 and stripe (sr), which correspond to immediate-early genes in mammals, showed short, intense bursts of expression in LNv at the onset of light. No signal was detected 2 hours before or after light onset, nor under constant darkness conditions at the same time. Additional analysis at 15, 30, and 60 minutes after light onset supported that this response is a transient transcriptional wave rather than a sustained increase. A key finding was that not all circadian neurons respond in the same manner.

Significance and Outlook

This study demonstrates that the Drosophila brain's biological clock is not a uniform oscillator but a network where each neuron type possesses its own transcriptional rhythm and light response. In particular, it suggests that external light may adjust the internal clock by transiently activating activity-dependent genes like Hr38 and sr in specific neurons, rather than uniformly altering widespread gene expression. This reaction is a candidate molecular link for circadian entrainment or phase shifts.

The multiplexing of EL-INTACT and DGRP is significant as it allows for the analysis of time-series transcriptomes of rare cells with minimized batch effects. The use of frozen samples facilitates the collection of various conditions and time points in advance, allowing them to be processed together. However, it cannot be concluded from temporal correlations of transcripts alone that Hr38 and sr are the causes of behavioral phase regulation. The causal roles of each neuron must be confirmed by combining gene-specific functional inhibition, neural activity measurement, and behavioral rhythm analysis. Whether the cell-type-specific responses observed in Drosophila are conserved in the mammalian suprachiasmatic nucleus remains a follow-up task.