Regulation of circadian clock transcriptional output by CLOCK:BMAL1

Abstract

The mammalian circadian clock relies on the transcription factor CLOCK:BMAL1 to coordinate the rhythmic expression of 15% of the transcriptome and control the daily regulation of biological functions. The recent characterization of CLOCK:BMAL1 cistrome revealed that although CLOCK:BMAL1 binds synchronously to all of its target genes, its transcriptional output is highly heterogeneous. By performing a meta-analysis of several independent genome-wide datasets, we found that the binding of other transcription factors at CLOCK:BMAL1 enhancers likely contribute to the heterogeneity of CLOCK:BMAL1 transcriptional output. While CLOCK:BMAL1 rhythmic DNA binding promotes rhythmic nucleosome removal, it is not sufficient to generate transcriptionally active enhancers as assessed by H3K27ac signal, RNA Polymerase II recruitment, and eRNA expression. Instead, the transcriptional activity of CLOCK:BMAL1 enhancers appears to rely on the activity of ubiquitously expressed transcription factors, and not tissue-specific transcription factors, recruited at nearby binding sites. The contribution of other transcription factors is exemplified by how fasting, which effects several transcription factors but not CLOCK:BMAL1, either decreases or increases the amplitude of many rhythmically expressed CLOCK:BMAL1 target genes. Together, our analysis suggests that CLOCK:BMAL1 promotes a transcriptionally permissive chromatin landscape that primes its target genes for transcription activation rather than directly activating transcription, and provides a new framework to explain how environmental or pathological conditions can reprogram the rhythmic expression of clock-controlled genes.

Fig 1. Mouse liver CLOCK:BMAL1 transcriptional output is heterogeneous.

A, B. Mouse liver BMAL1 (blue, A) and CLOCK (green, B) ChIP-Seq peaks from Koike et al., 2012 were mapped to their target genes and parsed based on their transcriptional output (Nascent-Seq from Menet et al., 2012, C). Each dot represents the phase of maximal DNA binding, and the ChIP-Seq signal is displayed using different shades of color to illustrate differences in binding intensity. C. Heatmap representation of the Nascent-Seq signal of direct CLOCK:BMAL1 target genes classified based on their transcriptional output in the mouse liver. Each lane represents the Nascent-Seq signal of a gene corresponding to the CLOCK and BMAL1 peaks in A and B. Nascent-Seq signal was ordered based on the phase of nascent RNA oscillations for the in-phase and out-of-phase transcriptional cyclers. Ordering of arrhythmically transcribed genes is based on the peak time of maximal expression; the lack of a distinctive 24-hr rhythm profile of nascent RNA expression over the 48-hr time-scale is indicative of arrhythmic transcription. No heatmap could be generated for the non-expressed genes because of the lack of nascent RNA expression. D. Nascent RNA expression, calculated as reads/bp, for each of the 4 CLOCK:BMAL1 transcriptional output group (Rinφ: in-phase transcriptional cyclers; Ro/φ: out-of-phase transcriptional cyclers; AR: arrhythmically transcribed target genes; NE: non-expressed target genes). Groups with different letters are significantly different (Kruskal-Wallis test; p < 0.05). E. Phase of maximal BMAL1 (left) and CLOCK (right) rhythmic DNA binding for each of the 4 CLOCK:BMAL1 transcriptional output categories. Groups with different letters are significantly different (Kruskal-Wallis test; p < 0.05). F. BMAL1 (top) and CLOCK (bottom) ChIP-Seq signal for each of the 4 CLOCK:BMAL1 transcriptional output groups. Signal is also displayed for the Rinφ and Ro/φ groups after removal of the ChIP-Seq signal at peaks targeting core clock genes. Groups with different letters are significantly different (Kruskal-Wallis test; p < 0.05). G. Location of CLOCK:BMAL1 peaks within gene loci for each of the 4 CLOCK:BMAL1 transcriptional output groups. TSS (transcriptional start site; +/- 1kb from annotated TSS); Gene body: + 1kb from TSS to + 1kb from transcription termination site; Extended promoter: - 10 kb to—1 kb from the annotated TSS. Numbers correspond to the percentage and numbers of peaks (outside and inside the pie chart, respectively) within each location for each group. * denotes a significant difference in the distribution of peaks between the AR and NE groups (chi square test; p < 0.05).

Fig 2. REV-ERBα and REV-ERBβ ChIP-Seq signal is higher at CLOCK:BMAL1 DNA binding sites targeting genes transcribed at night.

A-C (Left). Average ChIP-Seq signal for REV-ERBα (A), REV-ERBβ (B) and E4BP4 (C) at CLOCK:BMAL1 DNA binding sites (center ± 1kb) for each of the 4 CLOCK:BMAL1 transcriptional output group. A-C (Right). Distribution of REV-ERBα (A), REV-ERBβ (B) and E4BP4 (C) ChIP-Seq signal at CLOCK:BMAL1 peaks for each of the 4 CLOCK:BMAL1 transcriptional output groups (signal for each peak was averaged at CLOCK:BMAL1 peak center ± 250bp). Groups are labeled as in Fig 1. Those with different letters are significantly different (Kruskal-Wallis test; p < 0.05). REV-ERBα and REV-ERBβ ChIP were performed from mice liver collected at ZT10, while E4BP4 ChIP was performed from mice liver collected at ZT22. ChIP-Seq datasets were retrieved from Cho et al., 2012 [35], and E4BP4 ChIP-Seq datasets from Fang et al., 2014 [36].

Fig 3. CLOCK:BMAL1 rhythmic DNA binding is associated with rhythmic nucleosome signal, but not with rhythmic histone post-translational modifications and eRNA transcription.

A-D. Rhythmic nucleosome signal at CLOCK:BMAL1 DNA binding sites for each of the 4 CLOCK:BMAL1 transcriptional output groups: (A) Rhythmic-in-phase (Rinφ, red); (B) Rhythmic out-of-phase (Ro/φ, orange); (C) arrhythmic (AR, black); (D) non expressed (NE, grey) target genes). Nucleosome signal was retrieved from mouse liver MNase-Seq datasets [20], which consists of 6 time points each separated by 4 hours with n = 4 mice for each time point. (Left): 6-time points rhythm of nucleosome signal at CLOCK:BMAL1 binding sites (calculated at CLOCK:BMAL1 peak center ± 75 bp for each peak), displayed as the average ± s.e.m. of the signal (n = 4) calculated for each mouse and for each transcriptional output category. The phase of rhythm (average ± s.e.m. from 4 independent rhythm) is indicated in the bottom right. Each rhythm is double-plotted for better visualization. (Right): average nucleosome signal for each transcriptional output group at CLOCK:BMAL1 DNA-binding sites (±0.6 kb) during the light phase (ZT2, ZT6, and ZT10; green) and dark phase (ZT14, ZT18, and ZT22; red/orange) of wild-type mice and in Bmal1-/- mice (average signal for six time points; black). E. maximal and minimal nucleosome signal from the 6-time points rhythms for each of the CLOCK:BMAL1 transcriptional output groups. Groups with different letters are significantly different (2-way ANOVA; p < 0.05). F. Circadian rhythm of H3K27ac ChIP-Seq signal in the mouse liver at CLOCK:BMAL1 DNA binding sites (calculated at CLOCK:BMAL1 peak center ± 1 kb) for each of the 4 CLOCK:BMAL1 transcriptional output group. Datasets were retrieved from Koike et al., 2012 [3] and re-analyzed (see methods for more details). Values correspond to the ChIP-Seq signal median for each group. Each rhythm is double-plotted for better visualization. G. Rhythm of enhancer RNA (eRNA) signal in the mouse liver at CLOCK:BMAL1 DNA binding sites (calculated at CLOCK:BMAL1 peak center ± 500 bp) for each of the 4 CLOCK:BMAL1 transcriptional output groups. Datasets were retrieved from Fang et al., 2014 [36] and re-analyzed (see methods for more details). Values correspond to the eRNA signal median for each group. Each rhythm is double-plotted for better visualization.

Fig 4. Tissue-specific and ubiquitous transcription factors are differentially recruited at CLOCK:BMAL1 enhancers.

A-C. Enrichment for the DNA binding motif of CLOCK:BMAL1 (A), tissue-specific transcription factors (B) and ubiquitous transcription factors (C) at CLOCK:BMAL1 DNA binding sites for each of the four CLOCK:BMAL1 transcriptional output categories. Enrichment was calculated using HOMER and is reported as the ratio between the calculated enrichment over the calculated background. * q < 0.05 (Benjamini-Hochberg procedure). D-F. ChIP-Seq signal of tissue-specific transcription factors (D), ubiquitous transcription factors (E), and transcriptional co-activators / RNA Polymerase II at ZT6 (F) at CLOCK:BMAL1 DNA binding sites (peak center ± 250bp) for each of the transcriptional output categories. Groups with different letters are significantly different (Kruskal-Wallis test; p < 0.05). G. Transcription factor DNA binding variability index at CLOCK:BMAL1 DNA binding sites. The TF DNA binding variability index reflects differential TF DNA binding by calculating the variance of TF ChIP-Seq signal between the four CLOCK:BMAL1 transcriptional output groups (see methods for details). The variability index is displayed as a dot for each TF: CLOCK, BMAL1, PER1, PER2, CRY1, CRY2 (blue), seven ts-TFs (CEBPA, CEBPB, FOXA1, FOXA2, HNF1, HNF4A, and HNF6; red), thirteen u-TFs (REV-ERBα, REV-ERVβ, RORα, E4BP4, RXR, LXR, PPARα, GR-ZT12, E2F4, STAT5, BCL6, ERα, and GABPA; green), as well as for p300, CBP, and Pol II at seven time points (ZT02 to ZT26) (black). The horizontal lines represent the variability index median for the first 3 groups of TF. ChIP-Seq datasets used in this analysis are described in the method section. The variability index was calculated using all CLOCK:BMAL1 peaks analyzed in Fig 1 (left), or CLOCK:BMAL1 peaks that do not target a clock gene (removal of TF ChIP-Seq signal at peaks targeting Per1, Per2, Cry2, Dbp, Rev-erbα, and Rev-erbβ, Tef, Hlf, Gm129, and Rorγ; right).

Fig 5. CLOCK:BMAL1 regulation of clock-controlled gene expression likely relies on the cooperation of CLOCK:BMAL1 with other transcription factors.

A. Proposed model incorporating tissue-specific (ts-TFs) and ubiquitous (u-TFs) transcription factors into CLOCK:BMAL1 regulation of clock-controlled gene transcription. See text for details. B, C. Intron (B) and exon (C) signals of direct CLOCK:BMAL1 target genes classified based on their transcriptional output (Rinφ: in-phase transcriptional cyclers; Ro/φ: out-of-phase transcriptional cyclers; AR: arrhythmically transcribed target genes; NE: non-expressed target genes), in wild-type (left) and Bmal1-/- (right) mouse liver. Values correspond to the median RPKM for each transcriptional output group, and are displayed as the average ± s.e.m. of four (wild-type) or two (Bmal1-/-) independent samples for each time point. Data were retrieved from public RNA-Seq datasets [53], and are double-plotted for better visualization. D-G. Rhythm of mRNA expression in the liver of mice fed ad libitum (blue) or fasted for at least 22 hours (orange). Data were retrieved from a public dataset [56]. Mouse liver mRNA expression is displayed for Clock and Bmal1 (D), as well as CLOCK:BMAL1 target genes that are rhythmically expressed (E), arrhythmically expressed (F), or not expressed (G) in the liver of mice fed ad libitum.