The circadian clock comprises transcriptional feedback loops of clock genes. Cryptochromes are essential components of the negative feedback loop in mammals as they inhibit CLOCK-BMAL1-mediated transcription. We purified mouse CRY1 (mCRY1) protein complexes from Sarcoma 180 cells to determine their roles in circadian gene expression and discovered that Myb-binding protein 1a (Mybbp1a) interacts with mCRY1. Mybbp1a regulates various transcription factors, but its role in circadian gene expression is unknown. We found that Mybbp1a functions as a co-repressor of Per2 expression and repressed Per2 promoter activity in reporter assays. Chromatin immunoprecipitation (ChIP) assays revealed endogenous Mybbp1a binding to the Per2 promoter that temporally matched that of mCRY1. Furthermore, Mybbp1a binding to the Per2 promoter correlated with the start of the down-regulation of Per2 expression and with the dimethylation of histone H3 Lys9, to which it could also bind. These findings suggest that Mybbp1a and mCRY1 can form complexes on the Per2 promoter that function as negative regulators of Per2 expression.

The naphthoquinone pigment, shikonin, is a natural product derived from Lithospermum erythrorhizon and an active component of a Chinese traditional herbal therapeutic. We identified shikonin as a candidate for shortening the circadian period using real-time reporter gene assays based on NIH3T3-derived stable reporter cells. Period length that became shortened in cells incubated with shikonin or etoposide reverted to that of control cells after continued incubation without these compounds. These findings indicated that shikonin and etoposide shorten the circadian period reversibly and through similar mechanisms. Topoisomerase II (Top2)-specific decatenation assays confirmed that shikonin, liker etoposide, is a Top2 inhibitor. Shikonin was incorporated into the nucleus and Top2 was located in the Bmal1 promoter, suggesting the relationship between Bmal1 transcription and Top2 inhibition. Top2a siRNA also shortened period length, suggesting that Top2 is involved in this process. Promoter assays showed that Top2a siRNA, etoposide and shikonin reduce Bmal1 promoter activity. These findings indicated that Top2 is involved in Bmal1 transcription and influences the circadian period, and that shikonin is a novel contributor to the control of period length in mammalian cells.

The control of the circadian rhythm is important for health because it regulates physiological functions and is associated with health hazards. We aimed to identify a circadian biomarker of health status in human saliva, since collecting saliva is non-invasive, straightforward, and cost-effective. Among 500 genes potentially controlled by the salivary clock identified using chromatin immunoprecipitation (ChIP) assays, 22 of them showed reasonable transcriptional responses according to a DNA array in a salivary model system. Among these 22 genes, ARRB1, which is expressed in human salivary glands, was also expressed in model HSG cells at the transcriptional and translational levels. The profile of ARRB1 expression in human saliva was circadian, suggesting that ARRB1 could serve as a candidate circadian biomarker in saliva. We compared ARRB1 with other biomarkers in salivary samples from jet-lagged individuals. The circadian profile of ARRB1 reflected the time lag more than the profile of melatonin, whereas the profiles of cortisol and α-amylase did not reflect the time lag. Overall, these findings suggest that salivary ARRB1 could serve as a candidate biomarker that could be used to monitor the internal body clock.

Period2 (Per2) is an essential component of the mammalian clock mechanism and robust circadian expression of Per2 is essential for the maintenance of circadian rhythms. Although recent studies have shown that the circadian E2 enhancer (a non-canonical E-box) accounts for most of the circadian transcriptional drive of mPer2, little is known about the other cis-elements of mPer2 oscillatory transcription. Here, we examined the contribution of E4BP4 to Per2 mRNA oscillation in the cell-autonomous clock. Knockdown experiments of E4BP4 in both Northern blots and real-time luciferase assays suggested that endogenous E4BP4 negatively regulates Per2 mRNA oscillation. Sequence analysis revealed two putative E4BP4-binding sites (termed A-site and B-site) on mammalian Per2 promoter regions. Luciferase assays with mutant constructs showed that a novel E4BP4-binding site (B-site) is responsible for E4BP4-mediated transcriptional repression of Per2. Furthermore, chromatin immunoprecipitation assays in vivo showed that the peak of E4BP4 binding to the B-site on the Per2 promoter almost matched the trough of Per2 mRNA expression. Importantly, real-time luciferase assays showed that the B-site in addition to the E2 enhancer is required for robust circadian expression of Per2 in the cell-autonomous clock. These findings indicated that E4BP4 is required for the negative regulation of mammalian circadian clocks.

We have been investigating transcriptional regulation of the BMAL1 gene, a critical component of the mammalian clock system including DNA methylation. Here, a more detailed analysis of the regulation of DNA methylation of BMAL1 proceeded in RPMI8402 lymphoma cells. We found that CpG islands in the BMAL1 and the PER2 promoters were hyper- and hypomethylated, respectively and that 5-aza-2'-deoxycytidine (aza-dC) not only enhanced PER2 gene expression but also PER2 oscillation within 24 h in RPMI8402 cells. That is, such hypermethylation of CpG islands in the BMAL1 promoter restricted PER2 expression which was recovered by aza-dC within 1 day in these cells. These results suggest that the circadian clock system can be recovered through BMAL1 expression induced by aza-dC within a day. The RPIB9 promoter of RPMI8402 cells, which is a methylation hotspot in lymphoblastic leukemia, was also hypermethylated and aza-dC gradually recovered RPIB9 expression in 3 days. In addition, methylation-specific PCR revealed a different degree of aza-dC-induced methylation release between BMAL1 and RPIB9 These results suggest that the aza-dC-induced recovery of gene expression from DNA methylation is dependent on a gene, for example the rapid response to demethylation by the circadian system, and thus, is of importance to clinical strategies for treating cancer.

The Bmal1 gene is essential for the circadian system, and its promoter has a unique open chromatin structure. We examined the mechanism of topoisomerase I (Top1) to understand the role of the unique chromatin structure in Bmal1 gene regulation. Camptothecin, a Top1 inhibitor, and Top1 small interfering RNA (siRNA) enhanced Baml1 transcription and lengthened its circadian period. Top1 is located at an intermediate region between two ROREs that are critical cis-elements of circadian transcription and the profile of Top1 binding indicated anti-phase circadian oscillation of Bmal1 transcription. Promoter assays showed that the Top1-binding site is required for transcriptional suppression and that it functions cooperatively with the distal RORE, supporting that Bmal1 transcription is upregulated by Top1 inhibition. A DNA fragment between the ROREs, where the Top1-binding site is located, behaved like a right-handed superhelical twist, and modulation of Top1 activity by camptothecin and Top1 siRNA altered the footprint profile, indicating modulation of the chromatin structure. These data indicate that Top1 modulates the chromatin structure of the Bmal1 promoter, regulates Bmal1 transcription and influences the circadian period.

The human Kank gene encodes an ankyrin repeat domain-containing protein which regulates actin polymerization. There are at least two types of Kank protein depending on cell type, likely due to differences in transcription. Here, to examine the transcriptional initiation and genomic organization of the human Kank gene, we performed 5'-RACE (rapid amplification of cDNA ends) using total RNA from normal kidney and a human kidney cancer cell line, VMRC-RCW cells. The results suggest that the human Kank gene has several alternative first exons. While mRNA from VMRC-RCW cells encoded Kank protein (referred to as Kank-S) as reported previously, mRNA from the normal kidney tissue encoded a novel type of Kank protein (referred to as Kank-L), which contained an additional N-terminal sequence 158 amino acids long. Promoter activity and the expression of the Kank variants in normal and cancer tissues were examined.