倉林 伸博

The circadian clock drives daily rhythms of gene expression and physiology. Advances in next-generation DNA sequencing have provided extensive insights into RNA expression, but more functional information at the protein level with sufficient depth has been limited by technical challenges. In this study, we generated a comprehensive mouse circadian proteome atlas (https://chronoproteinology.org/circadian_atlas) by analyzing 32 tissues, including the suprachiasmatic nucleus (SCN), using the next-generation mass spectrometer Orbitrap Astral. Data-independent acquisition of 584 samples, including developmental samples, revealed the spatiotemporal profiles of about 19,000 proteins. Proteome and phospho-proteome analyses of whole-cell and nuclear proteins in the liver revealed circadian changes in protein quantity and quality, as well as global changes in hPER2-S662G mutant mice, a genetic model of human familial advanced sleep phase (FASP). This multi-tissue circadian proteome atlas provides a fundamental resource for understanding when, where, and which proteins are expressed and function.

D-site binding protein, DBP, is a clock-controlled transcription factor and drives daily rhythms of physiological processes through the regulation of an array of genes harboring a DNA binding motif, D-box. DBP protein levels show a circadian oscillation with an extremely robust peak/trough ratio, but it is elusive how the temporal pattern is regulated by post-translational regulation. In this study, we show that DBP protein levels are down-regulated by the ubiquitin-proteasome pathway. Analysis using 19 dominant-negative forms of E2 enzymes have revealed that UBE2G1 and UBE2T mediate the degradation of DBP. A proteomic analysis of DBP-interacting proteins and database screening have identified Tumor necrosis factor Receptor-Associated Factor 7 (TRAF7), a RING-type E3 ligase, that forms a complex with UBE2G1 and/or UBE2T. Ubiquitination analysis have revealed that TRAF7 enhances K48-linked polyubiquitination of DBP in cultured cells. Overexpression of TRAF7 down-regulates DBP protein level, while knockdown of TRAF7 up-regulates DBP in cultured cells. Knockout of TRAF7 in NIH3T3 cells have revealed that TRAF7 mediates the time-of-the-day-dependent regulation of DBP levels. Furthermore, TRAF7 has a period-shortening effect on the cellular clock. Together, TRAF7 plays an important role in circadian clock oscillation through destabilization of DBP.

Sleep is regulated by homeostatic processes, yet the biological basis of sleep pressure that accumulates during wakefulness, triggers sleep, and dissipates during sleep remains elusive. We explored a causal relationship between cellular synaptic strength and electroencephalography delta power indicating macro-level sleep pressure by developing a theoretical framework and a molecular tool to manipulate synaptic strength. The mathematical model predicted that increased synaptic strength promotes the neuronal "down state" and raises the delta power. Our molecular tool (synapse-targeted chemically induced translocation of Kalirin-7, SYNCit-K), which induces dendritic spine enlargement and synaptic potentiation through chemically induced translocation of protein Kalirin-7, demonstrated that synaptic potentiation of excitatory neurons in the prefrontal cortex (PFC) increases nonrapid eye movement sleep amounts and delta power. Thus, synaptic strength of PFC excitatory neurons dictates sleep pressure in mammals.

In mammals, CLOCK and BMAL1 proteins form a heterodimer that binds to E-box sequences and activates transcription of target genes, including Period (Per). Translated PER proteins then bind to the CLOCK-BMAL1 complex to inhibit its transcriptional activity. However, the molecular mechanism and the impact of this PER-dependent inhibition on the circadian clock oscillation remain elusive. We previously identified Ser38 and Ser42 in a DNA-binding domain of CLOCK as phosphorylation sites at the PER-dependent inhibition phase. In this study, knockout rescue experiments showed that nonphosphorylatable (Ala) mutations at these sites shortened circadian period, whereas their constitutive-phospho-mimetic (Asp) mutations completely abolished the circadian rhythms. Similarly, we found that nonphosphorylatable (Ala) and constitutive-phospho-mimetic (Glu) mutations at Ser78 in a DNA-binding domain of BMAL1 also shortened the circadian period and abolished the rhythms, respectively. The mathematical modeling predicted that these constitutive-phospho-mimetic mutations weaken the DNA binding of the CLOCK-BMAL1 complex and that the nonphosphorylatable mutations inhibit the PER-dependent displacement (reduction of DNA-binding ability) of the CLOCK-BMAL1 complex from DNA. Biochemical experiments supported the importance of these phosphorylation sites for displacement of the complex in the PER2-dependent inhibition. Our results provide direct evidence that phosphorylation of CLOCK-Ser38/Ser42 and BMAL1-Ser78 plays a crucial role in the PER-dependent inhibition and the determination of the circadian period.