Single-cell Multiomic and Spatiotemporal Dissection of the Liver Circadian Clock

Abstract

Circadian rhythms are remarkably widespread across most organisms, regulating hormonal, metabolic, physiological, and behavioral oscillations through molecular clocks that orchestrate the rhythmic expression of thousands of genes. Here, we generate the first time-series single-nucleus RNA and ATAC multiomics data to simultaneously characterize gene expression and chromatin accessibility of mouse liver cells across the 24-hour cycle. We uncover pronounced cell-type-specific circadian programs, with hepatocytes displaying the strongest and most widespread rhythmicity at both the transcriptomic and epigenomic levels. Chromatin accessibility rhythms frequently precede or coincide with transcriptional rhythms, indicating coordinated regulation between chromatin state and gene activation. Across the genome, rhythmic gene expression is driven primarily by oscillations in the fraction of transcriptionally active (“bursting”) cells rather than changes in transcript output per cell, revealing burst frequency as a dominant mechanism of circadian control. By reconstructing spatial zonation of the liver lobule, we identify both genes and cis-regulatory elements (CREs) that are temporally, spatially, or spatiotemporally variable. We infer gene regulatory relationships linking transcription factors (TFs), CREs, and target genes, recapitulating core clock feedback loops and revealing spatiotemporally resolved regulatory logic underlying metabolic gene expression. Our findings apply to existing single-cell data of mouse and Drosophila brains and are validated by time-series single-molecule fluorescence in situ hybridization and vast amounts of orthogonal omics data. Altogether, our study constructs a comprehensive map of the time-series transcriptomic and epigenomic landscapes that elucidate the function and mechanism of the liver peripheral clocks.