The BMAL1/HIF2A heterodimer modulates circadian variations of myocardial injury

Abstract

Acute myocardial infarction stands as a prominent cause of morbidity and mortality worldwide^1-6. Clinical studies have demonstrated that the severity of cardiac injury following myocardial infarction exhibits a circadian pattern, with larger infarct sizes and poorer outcomes in patients experiencing morning onset myocardial infarctions^7-14. However, the molecular mechanisms that govern circadian variations of myocardial injury remain unclear. Here, we show that BMAL1^14-20, a core circadian transcription factor, orchestrates diurnal variability in myocardial injury. Unexpectedly, BMAL1 modulates circadian-dependent cardiac injury by forming a transcriptionally active heterodimer with a non-canonical partner, hypoxia-inducible factor 2 alpha (HIF2A)^6,21-23, in a diurnal manner. Substantiating this finding, we determined the cryo-EM structure of the BMAL1/HIF2A/DNA complex, revealing a previously unknown capacity for structural rearrangement within BMAL1, which enables the crosstalk between circadian rhythms and hypoxia signaling. Furthermore, we identified amphiregulin (AREG) as a rhythmic transcriptional target of the BMAL1/HIF2A heterodimer, critical for regulating circadian variations of myocardial injury. Finally, pharmacologically targeting the BMAL1/HIF2A-AREG pathway provides effective cardioprotection, with maximum efficacy when aligned with the pathway's circadian trough. Our findings not only uncover a novel mechanism governing the circadian variations of myocardial injury but also pave the way for innovative circadian-based treatment strategies, potentially shifting current treatment paradigms for myocardial infarction.

Extended Data Fig. 1.. Circadian variation of myocardial injury in C57BL/6J mice.

a, Schematic of the experimental setup for evaluating myocardial injury and cardiac function in C57BL/6J mice subjected to myocardial IRI at different ZTs (ZT2, ZT8, ZT14, and ZT20). b-e, b, Representative heart slices subjected to Evan’s blue and TTC double staining following 2 h of reperfusion: infarct area (green line) and area at risk (AAR; blue line); scale bar, 1 mm. c, Percentage of the AAR relative to the size of LV. d, Infarct sizes represented as the percentage of the AAR. e, Serum troponin I level. b-c: n = 7 mice/group/time point. Statistical analysis was performed using one-way ANOVA. f-k, Cardiac function was measured by speckle-tracking echocardiography analysis on day 14 post-MI in C57BL/6J mice subjected to myocardial IRI at ZT8 or ZT20. f, Left ventricular systolic function in ejection fraction (EF), fractional shortening (FS), and global longitudinal strain (GLS); left ventricular end-diastolic volume (EDV) and end-systolic volume (ESV); end-diastolic left ventricular mass (EDLVM) and end-systolic left ventricular mass (ESLVM). Statistical analysis was performed using unpaired Student’s t-tests. g, Representative left ventricular 3D longitudinal strain (48 points) and 6-segment longitudinal strain images demonstrating wall motion abnormalities (Color-coded six segments: Ant. Base, Anterior Base; Ant. Mid, Anterior Middle; Ant. Apex, Anterior Apex; Post. Apex, Posterior Apex; Post. Mid, Posterior Middle; Post. Base, Posterior Base). h, Left ventricular segmental wall contractility detected by peak longitudinal strain. Statistical analysis was performed using unpaired two-way ANOVA. i, Representative left ventricular 3D long-axis radial strain (48 points) and 6-segment radial strain images demonstrating wall motion abnormalities. j, Left ventricular segmental wall contractility detected by peak radial strain. Statistical analysis was performed using unpaired two-way ANOVA. k, Left ventricular mechanical dyssynchrony as measured by intra-ventricular delay in time-to-peak strain and standard deviation of time-to-peak strain based on peak longitudinal strain. Statistical analysis was performed using unpaired Student’s t-tests. f-k, n = 7 mice/group/time point. l, Protein levels of cleaved-caspase3, caspase3, and Bax were measured by Western blot analysis in the AAR of mouse hearts on day 3 post-MI. m, Quantification of protein levels in (l). n = 5 mice/group/time point. Statistical analysis was performed using unpaired two-way ANOVA. All data are mean ± s.e.m, *p < 0.05, **p < 0.01, ***p < 0.001, and ****p < 0.0001, ns: not significant.

Extended Data Fig. 2.. BMAL1 is a key transcription factor in regulating circadian variation of myocardial injury.

a-d, RNA-seq analysis was conducted on the AAR from C57BL/6J mice subjected to myocardial IRI at ZT8 or ZT20. a, PCA of mRNA expression profiles. b, DEGs were detected in mouse hearts when comparing ZT8 vs. ZT20 (fold change > 1.5, adjusted p < 0.05). c, The top five enriched KEGG pathways for DEGs in the ZT8 vs. ZT20 groups. d, Gene dysregulation network constructed using DEGs. Triangle represents transcription factor (TF), and circle denotes gene. Green nodes represent downregulated TF/genes, while red nodes denote upregulated TF/genes in the ZT8 mouse hearts compared to the ZT20 group. The size of the node represents the degree in the network, and the width of the edges represents the strength of correlations. n =3 mice/group/time point. e-g, Human RNA-seq analysis using left ventricular biopsies from cardiac surgical patients in the morning (AM) or afternoon (PM) groups. e, PCA demonstrating the distinct transcriptional signatures across different patient groups. f, DEGs when comparing AM with PM patient samples (log2 fold change > 0.5 and p < 0.01). g, The top three KEGG pathways enriched by DEGs. The colors and size of nodes represent the p value and fold change of each gene. n = 56/morning and n = 17/afternoon. h, The peak transcriptional activity of BMAL1 in the ischemic mouse hearts, demonstrated by the maximum expression of its target genes (Dbp and Nr1d1) at ZT8, aligns with the least myocardial injury. Data modified from Fig. 1c and Extended Data Fig. 2d. n = 7/infarct size measurement and n = 3/gene expression analysis. All data are mean ± s.e.m.

Extended Data Fig. 3.. Cardiomyocyte-specific deletion of Bmal1 in the Bmal1^loxP/loxPMyosin Cre+ mice.

Bmal1^loxP/loxP Myosin Cre+ mice and Myosin Cre+ mice were administered tamoxifen at 8–12 weeks of age, and the transcript levels of Bmal1 and its target gene Dbp were measured one week later by real-time PCR analysis in isolated cardiomyocytes (a), lungs (b), and kidneys (g). n = 3–4 for cardiomyocytes/Myosin Cre+ mice, n = 4–5 for cardiomyocytes/Bmal1^loxP/loxP Myosin Cre+ mice, n = 3–4 for lungs/Myosin Cre+ mice, n = 3–4 for lungs/Bmal1^loxP/loxP Myosin Cre+ mice, n = 4–5 for kidneys/Myosin Cre+ mice, n = 3–5 for kidneys/Bmal1^loxP/loxP Myosin Cre+ mice. BMAL1 protein levels were measured by Western blot analysis in isolated cardiomyocytes (b), lungs (e), and kidneys (h). Quantification of BMAL1 protein levels in cardiomyocytes (c), lungs (f), and kidneys (i). n = 5 for cardiomyocytes/Myosin Cre+ mice, n = 5 for cardiomyocytes/Bmal1^loxP/loxP Myosin Cre+ mice, n = 4 for lungs/Myosin Cre+ mice, n = 4 for lungs/Bmal1^loxP/loxP Myosin Cre+ mice, n = 4 for kidneys/Myosin Cre+ mice, n = 4 for kidneys/Bmal1^loxP/loxP Myosin Cre+ mice. Statistical analysis varied depending on the context: Welch’s t test was used for assessing Dbp transcript levels in cardiomyocytes and lungs, the Mann-Whitney test was employed for evaluating Bmal1 transcript levels in the lungs, and the Student’s t test was applied in the analysis of the remaining figures. All data are mean ± s.e.m, *p < 0.05, and ***p < 0.001, ns: not significant.

Extended Data Fig. 4.. Baseline cardiac structure and function are unaffected by induced cardiomyocyte-specific Bmal1 deletion.

a-f, Bmal1^loxP/loxP Myosin Cre+ mice and Myosin Cre+ mice were administered tamoxifen at 8–12 weeks of age, and cardiac function was measured by speckle-tracking echocardiography analysis one week later. a, LV systolic function (EF, FS, and GLS), LV volume (EDV and ESV) and mass (EDLVM and ESLVM). Statistical analysis was performed using unpaired Student’s t-tests. b, Representative left ventricular 3D longitudinal strain (48 points) and 6-segment longitudinal strain images demonstrating wall motion patterns (Color-coded six segments: Ant. Base, Anterior Base; Ant. Mid, Anterior Middle; Ant. Apex, Anterior Apex; Post. Apex, Posterior Apex; Post. Mid, Posterior Middle; Post. Base, Posterior Base). c, Left ventricular segmental wall contractility detected by peak longitudinal strain. Statistical analysis was performed using two-way ANOVA. d, Representative left ventricular 3D long-axis radial strain (48 points) and 6-segment radial strain images demonstrating wall motion patterns. e, Left ventricular segmental wall contractility detected by peak radial strain. Statistical analysis was performed using unpaired Student’s t-tests. f, Left ventricular mechanical dyssynchrony as measured by intra-ventricular delay in time-to-peak strain and standard deviation of time-to-peak strain based on peak longitudinal strain. n = 9/Myosin Cre+ mice and n = 11/Bmal1^loxP/loxP Myosin Cre+ mice. Statistical analysis was performed using unpaired Student’s t-tests. g, Representative images of wheat germ agglutinin (WGA) showing cardiomyocyte size and quantification of myocyte cross-sectional area in mice; scale bar, 20 μm. Each quantification value dot represents the average value of three fields in one section, n = 5. Statistical analysis was performed using unpaired Student’s t-tests. h, Representative images of HE staining; scale bar, 20 μm. n = 5. i, Representative images of terminal deoxynucleotidyl transferase dUTP nick end labeling (TUNEL). Sections were co-stained with cardiomyocyte marker a-sarcomeric and DAPI. Incubation of sections with DNase I served as the positive control. Scale bar. 20 μm. n = 5. All data are presented as box-plots. ns: not significant.

Extended Data Fig. 5.. Cardiomyocyte-specific Bmal1 deletion diminishes the circadian variation of cardiac function impairment.

Cardiac function was evaluated by speckle-tracking echocardiography analysis on day 14 post-MI in Bmal1^loxP/loxP Myosin Cre+ mice and Myosin Cre+ mice subjected to IRI at ZT8 or ZT20. a, LV volume (EDV and ESV) and mass (EDLVM and ESLVM). b, Representative B-mode images from the left ventricular long-axis view with 2D longitudinal strain analysis, showing 6 segments (1. Ant. Base, Anterior Base; 2. Ant. Mid, Anterior Middle; 3. Ant. Apex, Anterior Apex; 4. Post. Apex, Posterior Apex; 5. Post. Mid, Posterior Middle; 6. Post. Base, Posterior Base). c, Left ventricular segmental wall contractility detected by peak longitudinal strain. d, Left ventricular mechanical dyssynchrony as measured by intra-ventricular delay in time-to-peak strain and standard deviation of time-to-peak strain based on peak longitudinal strain. a-d: n = 7/Bmal1^loxP/loxP Myosin Cre+ mice/time point and n = 9 for Myosin Cre+ mice/time point. All data are mean ± s.e.m. Statistical analysis was performed using two-way ANOVA. *p < 0.05, **p < 0.01, ***p < 0.001, and ****p < 0.0001, ns: not significant,

Extended Data Fig. 6.. Interaction of BMAL1 and HIF2A.

a, GO enrichment analysis was conducted to elucidate the molecular functions (MF), biological processes (BP), and cellular components (CC) associated with proteins predicted by the HuRI to potentially interact with BMAL1. b, Western Blot analysis of reciprocal co-IP with HIF2A in hypoxia-treated (1% O₂, 4h) or normoxia-treated HEK293 cells. Cytosolic and nuclear protein extracts were immunoprecipitated with HIF2A and blotted with anti-BMAL1, anti-HIF2A, anti-HIF1B, anti-α-tubulin, and anti-Lamin B antibodies. An IgG control affirmed procedure specificity. H indicates HIF2A. n = 3. c, Size-exclusion chromatography analysis of the BMAL1/HIF2A heterodimer. The purified BMAL1/HIF2A complex was loaded onto a Superdex 200 Increase 10/300 GL column. n = 3. The molecular weights of makers are as indicated. d, SDS-PAGE analysis of the recombinant BMAL1/HIF2A heterodimer purified by size-exclusion chromatography. n = 3.

Extended Data Fig. 7.. Cardiomyocyte-specific Hif2a deletion diminishes the circadian variation of cardiac function impairment.

Cardiac function was evaluated by speckle-tracking echocardiography analysis on day 14 post-MI in Hif2a^loxP/loxP Myosin Cre+ mice and Myosin Cre+ mice subjected to IRI at ZT8 or ZT20. a, Left ventricular volume (EDV and ESV) and mass (EDLVM and ESLVM). b, Representative B-mode images from the left ventricular long-axis view with 2D longitudinal strain analysis, showing 6 segments (1. Ant. Base, Anterior Base; 2. Ant. Mid, Anterior Middle; 3. Ant. Apex, Anterior Apex; 4. Post. Apex, Posterior Apex; 5. Post. Mid, Posterior Middle; 6. Post. Base, Posterior Base). c, Left ventricular segmental wall contractility detected by peak longitudinal strain. a-c: n = 8/Hif2a^loxP/loxP Myosin Cre+ mice/time point, n = 9/Myosin Cre+ mice/time point. All data are mean ± s.e.m. Statistical analysis was all performed using two-way ANOVA. *p < 0.05, **p < 0.01, ***p < 0.001, and ****p < 0.0001, ns: not significant,

Extended Data Fig. 8.. Cardiomyocyte-specific Hif1a deletion does not influence circadian-dependent myocardial injury.

a, Experimental setup for evaluation of myocardial injury and cardiac function in Hif1a^loxP/loxP Myosin Cre+ mice and Myosin Cre+ mice subjected to IRI at ZT8 or ZT20. b-e, b, Representative heart slices subjected to Evan’s blue and TTC double staining following 2h of reperfusion: infarct area (green line) and AAR (blue line); scale bar, 1 mm. c, Percentage of the AAR relative to the size of the LV. d, Infarct sizes represented as the percentage of the AAR. e, Serum troponin I levels. b-e: n = 7 mice/group/time point. Data are mean ± s.e.m. Statistical analysis was performed using two-way ANOVA. f-k, Cardiac function was evaluated by speckle-tracking echocardiography analysis on day 14 post-MI. f, EF and FS. g, GLS, EDV, ESV, EDLVM, and ESLVM. h, Representative B-mode images from the left ventricular long-axis view with 2D longitudinal strain analysis, showing 6 segments (1. Ant. Base, Anterior Base; 2. Ant. Mid, Anterior Middle; 3. Ant. Apex, Anterior Apex; 4. Post. Apex, Posterior Apex; 5. Post. Mid, Posterior Middle; 6. Post. Base, Posterior Base). i, Left ventricular segmental wall contractility detected by peak longitudinal strain. j, Representative left ventricular 3D (48 points) longitudinal strain and 6-segment longitudinal strain images demonstrating wall motion abnormalities. k, Left ventricular mechanical dyssynchrony as measured by intra-ventricular delay in time-to-peak strain and standard deviation of time-to-peak strain based on peak longitudinal strain. f-k: n = 9 mice/group/time point. Data are presented as box-plots. Statistical analysis was performed using two-way ANOVA. *p < 0.05, **p < 0.01, ***p < 0.001, and ****p < 0.0001, ns: not significant.

Extended Data Fig. 9.. Increased induction of AREG in cardiomyocytes in the border zone from mouse hearts on day 1 post-MI.

a-b, a, Representative immunostaining of AREG (red), α-sarcomeric (cardiomyocyte marker; green), and nuclei (DAPI; blue) on day 1 post-MI in the border zone, infarct area, and viable zone of hearts from C57BL/6J mice subjected to myocardial IRI at ZT8 or ZT20, scale bar, 50 μm. Arrows indicate α-sarcomeric⁺/AREG⁺ cells. b, Quantification of fluorescence intensity of AREG in (a). Normalized to the AREG levels in the viable zone at ZT8. Each quantification value dot represents the average value of three fields in one section. n = 4 mice/group/time point. Data are mean ± s.e.m. Statistical analysis was performed using two-way ANOVA. ***p < 0.001, and ****p < 0.0001, ns: not significant. c-d, Representative immunostaining of AREG (red), vimentin (fibroblast marker; green) (c), α-smooth muscle actin (α-SMA, smooth muscle cell marker; green) (d) and nuclei (DAPI; blue) on day 1 post-MI in the border zone of hearts from C57BL/6J mice subjected to myocardial IRI at ZT8 or ZT20, scale bar, 50 μm. n = 3–5 mice/group/time point.

Extended Data Fig. 10.. DNA-binding analysis of the BMAL1/HIF2A heterodimer.

The electrophoretic mobility shift assay (EMSA) was used to investigate the binding of the BMAL1/HIF2A heterodimer to the biotinylated-HRE dsDNA. Five different protein concentrations were used, as specified. The 22 bp biotin end-labeled DNA duplex was detected using HRP-conjugated streptavidin with chemiluminescent substrates.

Extended Data Fig. 11.. Areg deletion diminishes the circadian variation of cardiac function impairment.

Cardiac function was measured by speckle-tracking echocardiography analysis on day 14 post-MI in Areg^−/− and wild-type (WT) mice subjected to IRI at ZT8 or ZT20. a, FS. b, EDV, ESV, EDLVM, and ESLVM. c, Left ventricular segmental wall contractility detected by peak longitudinal strain. d, Representative left ventricular 3D (48 points) longitudinal strain and 6-segment longitudinal strain images demonstrating wall motion abnormalities (Color-coded six segments: Ant. Base, Anterior Base; Ant. Mid, Anterior Middle; Ant. Apex, Anterior Apex; Post. Apex, Posterior Apex; Post. Mid, Posterior Middle; Post. Base, Posterior Base). e, Left ventricular mechanical dyssynchrony as measured by intra-ventricular delay in time-to-peak strain and standard deviation of time-to-peak strain based on longitudinal strain. n = 8/animals/group/time point. All data are mean ± s.e.m. Statistical analysis was all performed using two-way ANOVA. *p < 0.05, **p < 0.01, ***p < 0.001, and ****p < 0.0001, ns: not significant.

Extended Data Fig. 12.. AREG administered before ischemia dampens the myocardial injury at ZT20.

a, Schematic of the experimental setup for evaluating myocardial injury in AREG-treated or vehicle (Veh)-treated (saline) mice subjected to myocardial IRI at ZT8 or ZT20. AREG (10 μg) was administered 30 minutes before IRI. b, Representative heart slices subjected to Evan’s blue and TTC double staining after 2h of reperfusion: infarct area (green line) and AAR (blue line); scale bar, 1 mm. c, Percentage of the AAR relative to the size of the LV. d, Infarct sizes represented as the percentage of the AAR. e, Serum troponin I levels. n = 7 mice/group/time point. All data are mean ± s.e.m. Statistical analysis was performed using two-way ANOVA. ***p < 0.001, and ****p < 0.0001, ns: not significant.

Extended Data Fig. 13.. Timed AREG treatment improves cardiac function when administered at ZT20.

Cardiac function was evaluated in timed AREG treatment or Vehicle treatment (saline) in mice subjected to IRI at ZT20. To mimic the clinical scenario where treatment begins after myocardial infarction onset, AREG (10 μg) or vehicle was administered at the start of reperfusion and subsequently given daily post-MI at either ZT8 or ZT20. a, FS. b, EDV, ESV, EDLVM and ESLVM. c, Representative B-mode images from the left ventricular long-axis view with 2D longitudinal strain analysis, showing 6 segments (1. Ant. Base, Anterior Base; 2. Ant. Mid, Anterior Middle; 3. Ant. Apex, Anterior Apex; 4. Post. Apex, Posterior Apex; 5. Post. Mid, Posterior Middle; 6. Post. Base, Posterior Base). d, Left ventricular segmental wall contractility detected by peak longitudinal strain. e, Left ventricular mechanical dyssynchrony as measured by intra-ventricular delay in time-to-peak strain and standard deviation of time-to-peak strain based on longitudinal strain. n = 7 mice/group/time point. Data are mean ± s.e.m. Statistical analysis was performed using two-way ANOVA. *p < 0.05, **p < 0.01, ***p < 0.001, and ****p < 0.0001, ns: not significant.

Extended Data Fig. 14.. NOB treatment provides circadian-dependent cardioprotection at ZT20.

NOB-treated (200 mg/kg, i.p., every other day) or vehicle-treated C57BL/6J mice were subjected to myocardial IRI at ZT8 or ZT20. a, Transcript levels of Per1, Per2, Cry2, and Dbp in the AAR from mouse hearts were measured by real-time PCR analysis following 2h of reperfusion. n = 3 mice/group/time point. b-e. Cardiac function was evaluated by speckle-tracking echocardiography analysis on day 14 post-MI. b, EDV, ESV, EDLVM and ESLVM. c, Representative B-mode images from the left ventricular long-axis view with 2D longitudinal strain analysis, showing 6 segments (1. Ant. Base, Anterior Base; 2. Ant. Mid, Anterior Middle; 3. Ant. Apex, Anterior Apex; 4. Post. Apex, Posterior Apex; 5. Post. Mid, Posterior Middle; 6. Post. Base, Posterior Base). d, Left ventricular segmental wall contractility detected by peak longitudinal strain. e, Left ventricular mechanical dyssynchrony as measured by intra-ventricular delay in time-to-peak strain and standard deviation of time-to-peak strain based on longitudinal strain. b-e: n = 7 mice/group/time point. Data are mean ± s.e.m. Statistical analysis was performed using two-way ANOVA. *p < 0.05, **p < 0.01, ***p < 0.001, and ****p < 0.0001, ns: not significant.

Extended Data Fig. 15.. NOB failed to alleviate myocardial injury in Bmal1^loxP/loxP Myosin Cre+ mice.

NOB-treated (200 mg/kg, i.p., every other day) or Veh-treated Bmal1^loxP/loxP Myosin Cre+ mice were subjected to myocardial IRI at ZT8 or ZT20. a-d, a, Representative heart slices subjected to Evan’s blue and TTC double staining after 2 h of reperfusion: infarct area (green line) and AAR (blue line); scale bar, 1 mm. b, Percentage of the AAR relative to the size of the LV. c, Infarct sizes represented as the percentage of the AAR. d, Serum troponin I levels. n = 7 mice/group/time point. e-j, Cardiac function was evaluated by speckle-tracking echocardiography analysis on day 14 post-MI. e, EF and FS. f, GLS, EDV, ESV, EDLVM, and ESLVM. g, Representative B-mode images from the left ventricular long-axis view with 2D longitudinal strain analysis, showing 6 segments (1. Ant. Base, Anterior Base; 2. Ant. Mid, Anterior Middle; 3. Ant. Apex, Anterior Apex; 4. Post. Apex, Posterior Apex; 5. Post. Mid, Posterior Middle; 6. Post. Base, Posterior Base). h, Left ventricular segmental wall contractility detected by peak longitudinal strain. i, Representative left ventricular 3D (48 points) longitudinal strain and 6-segment longitudinal strain images demonstrating wall motion abnormalities. j, Left ventricular mechanical dyssynchrony as measured by intra-ventricular delay in time-to-peak strain and standard deviation of time-to-peak strain based on longitudinal strain. n = 8/Veh-treated Bmal1^loxP/loxP Myosin Cre+/time point, n = 7/NOB-treated Bmal1^loxP/loxP Myosin Cre+/time point. All data are mean ± s.e.m. Statistical analysis was all performed using two-way ANOVA. ns: not significant.

Extended Data Fig. 16.. NOB failed to alleviate myocardial injury in Areg^−/− mice.

NOB-treated (200 mg/kg, i.p., every other day) or Veh-treated Areg^−/− mice were subjected to myocardial IRI at ZT8 or ZT20. a-c, a, Representative heart slices subjected to Evan’s blue and TTC double staining after 2h of reperfusion: infarct area (green line) and AAR (blue line); scale bar, 1 mm. b, Percentage of the AAR relative to the size of the LV. c, Infarct sizes represented as the percentage of the AAR (n = 6 animals/group/time point). d-i Cardiac function was evaluated by speckle-tracking echocardiography analysis on day 14 post-MI. d, EF, FS, GLS. e, EDV, ESV, EDLVM, and ESLVM. f, Representative B-mode images from the left ventricular long-axis view with 2D longitudinal strain analysis, showing 6 segments (1. Ant. Base, Anterior Base; 2. Ant. Mid, Anterior Middle; 3. Ant. Apex, Anterior Apex; 4. Post. Apex, Posterior Apex; 5. Post. Mid, Posterior Middle; 6. Post. Base, Posterior Base). g, Left ventricular segmental wall contractility detected by peak longitudinal strain. h, Representative left ventricular 3D (48 points) longitudinal strain and 6-segment longitudinal strain images demonstrating wall motion abnormalities. i, Left ventricular mechanical dyssynchrony as measured by intra-ventricular delay in time-to-peak strain and standard deviation of time-to-peak strain based on longitudinal strain. n = 8/Veh-treated Areg^−/−/time point and n = 6/NOB-treated Areg^−/−/time point. Data are mean ± s.e.m. Statistical analysis was performed using two-way ANOVA. ns: not significant.

Extended Data Fig. 17.. Cryo-EM structure determination of the BMAL1/HIF2A/DNA complex.

a, A cryo-EM micrograph of the BMAL1/HIF2A/DNA complex. b, Representative 2D class averages of the BMAL1/HIF2A/DNA complex. c, Flowchart of image processing of the BMAL1/HIF2A/DNA cryo-EM data. d, Cryo-EM density map of the BMAL1/HIF2A/DNA complex colored by local resolution. e, FSC curves (left) and angular distribution plot (right) of the BMAL1/HIF2A/DNA complex. The average resolution for the BMAL1/HIF2A/DNA complex is 3.7 Å according to the gold-standard FSC = 0.143 criterion. f, Local structures with their corresponding densities from bHLH, PAS-A and PAS-B domains are shown.

Extended Data Fig. 18.. Structural analysis of the BMAL1/HIF2A/DNA complex.

a, Four domain interfaces (I to IV) between BMAL1 and HIF2A. Each of these is indicated by a dashed ellipse. b, Zoom-in views of the interfaces (I to IV) between HIF2A and BMAL1. c, Comparison of the DNA-binding by two bHLH domains in HIF1B/HIF2A (left, PDB ID 4ZPK), BMAL1/HIF2A (middle), and BMAL1/CLOCK (right, PDB ID 4H10). The DNA contacted by the bHLH domains is highlighted in yellow. The PAS domains are omitted for clarity. d, Detailed interactions between two bHLH domains of BMAL1/HIF2A and the HRE DNA (yellow). e, Structural comparison of BMAL1/CLOCK (left, PDB ID 4F3L), BMAL1/HIF2A (middle), and HIF1B/HIF2A (right, PDB ID 4ZPK). For clarity, DNA was omitted. f, BMAL1 undergoes structural rearrangements upon binding with various partners. Superimposing the BMAL1/HIF2A and BMAL1/CLOCK complexes by aligning their bHLH domains reveals that BMAL1 undergoes a substantial conformational change, with the two PAS domains bending in nearly opposite directions. BMAL1 exhibits a compact overall architecture when bound with CLOCK (green) and a distinctly separated conformation when interacting with HIF2A (purple).

Fig. 1.. Identification of BMAL1 in modulating the circadian variation of myocardial injury.

a, Heatmap with expression patterns of DEGs in the AAR from mouse hearts subjected to myocardial IRI at ZT8 or ZT20. n = 3 mice/time point. b, Volcano plot showing DEGs when comparing ZT8 to ZT20. n = 3 mice/time point. c, Circadian expression patterns of Bmal1 and its target genes, Per2, Nr1d1, and Dbp, by real-time PCR analysis in the AAR of mouse hearts, subjected to IRI at different times of the day (ZT0, ZT4, ZT8, ZT12, ZT16, and ZT20). n = 3 mice/time point. d, Top 10 enriched GO biological process terms in IRI mouse hearts. e, Schematic illustration of human RNA-seq analysis using LV biopsies from cardiac surgical patients in the morning (AM) or afternoon (PM) groups. n = 56/morning and n = 17/afternoon. f, Heatmap showing clustered Pearson’s correlation matrix of patient LV samples. Major covariates, including treatment status, smoking history, pulmonary disease, and renal disease, were shown at the top. g, Volcano plot showing DEGs when comparing AM to PM patient LV samples. h, Enriched GO biological process terms for upregulated, down-regulated, and differentially expressed genes. i, Normalized read count for BMAL1, NR1D1, and PER2 in patient LV samples. j, Experimental setup for evaluation of myocardial injury and cardiac function in Bmal1^loxP/loxP Myosin Cre+ mice and Myosin Cre+ mice subjected to IRI at ZT8 or ZT20. k, Representative heart slices subjected to Evan’s blue and TTC double staining after 2 h of reperfusion: infarct area (green line) and AAR (blue line); scale bar, 1 mm. l, Percentage of the AAR relative to the size of the LV. n = 7 mice/group/time point. Statistical analysis was performed using two-way ANOVA. m, Infarct sizes represented as the percentage of the AAR. n = same number as above. Statistical analysis was performed using two-way ANOVA. n, Serum troponin I levels were evaluated at 2h after reperfusion. n = 7/Bmal1^loxP/loxP Myosin Cre+ mice/time point and n = 8/Myosin Cre+/time point. Statistical analysis was performed using two-way ANOVA. o and p, Cardiac function was evaluated by speckle-tracking echocardiography on day 14 post-MI. o, LV systolic function in EF, FS, and GLS. n = 7/Bmal1^loxP/loxP Myosin Cre+ mice/time point and n = 9/Myosin Cre+ mice/time point. Statistical analysis was performed using two-way ANOVA. p, Representative left ventricular 3D (48 points) longitudinal strain and 6-segment longitudinal strain images demonstrating wall motion abnormalities (Color-coded six segments: Ant. Base, Anterior Base; Ant. Mid, Anterior Middle; Ant. Apex, Anterior Apex; Post. Apex, Posterior Apex; Post. Mid, Posterior Middle; Post. Base, Posterior Base). All data are mean ± s.e.m, *p < 0.05, **p < 0.01, ***p < 0.001, and ****p < 0.0001, ns: not significant.

Fig. 2.. HIF2A interacts with BMAL1 and regulates circadian variation of myocardial injury.

a, Predicted protein-protein interactions between BMAL1 and other bHLH-PAS transcription factors using the Human Reference Interactome (HuRI) in the left ventricles of human hearts. EPAS1, endothelial PAS domain protein 1 (also known as HIF2A); NPAS, neuronal PAS domain protein; AHR, aryl hydrocarbon receptor; HI-Union, a dataset specifically represents a combined set of protein-protein interactions identified through various high-throughput Yeast two-Hybrid screens performed by the HuRI project; Literature, protein-protein interactions that have been identified and reported in scientific literature. b, Western blot analysis of co-IP in HEK293 cells overexpressing BMAL1-Flag. Cytosolic and nuclear protein extracts from hypoxia- (1% O₂, 4 h) or normoxia-treated HEK293 cells were immunoprecipitated with Flag and blotted with anti-HIF2A, anti-HIF1A, anti-BMAL1, anti-CLOCK, anti-α-tubulin, and anti-Lamin B antibodies. An IgG control affirmed procedure specificity. n = 3 independent experiments. c, Western blot analysis of co-IP with HIF2A in HCMs following 1% O₂ treatment for indicated times (0h, 0.5h, 1h, 2h, 4h, and 8h). n = 3 independent experiments. d and e, HCMs were treated with 1% O₂ for indicated times (0h, 1h, 4h, and 8h), and PLA was performed between BMAL1 and HIF2A, as well as BMAL1 and HIF1A. Representative confocal images (d) and corresponding quantification of PLA signals per nucleus (e) are presented; Scale bar, 25 μm. White arrows indicate the close interaction between BMAL1:HIF2A or BMAL1:HIF1A in the nuclei. n = 20/BMAL1:HIF2A, n = 20/BMAL1:HIF1A, and n = 10/BMAL1 negative control, independent experiments. Statistical analysis was performed using two-way ANOVA. ####p < 0.0001 compared to BMAL1:HIF1A; ****p < 0.0001 compared to BMAL1 negative control. f, Schematic illustration of the HIF2A (purple) and BMAL1 (orange) proteins with functional domains indicated. TAD, transactivation domain. g, Flag pull-down analysis. Recombinant Flag-tagged BMAL1 immobilized on anti-Flag M2 beads was used to pull down purified His-tagged HIF2A or His-tagged HIF1A. The elutes were analyzed by Western blot analysis. n = 3 independent experiments. h and i, h, C57BL/6J mice were subjected to myocardial IRI at ZT8 or ZT20, and protein levels of BMAL1 and HIF2A in the nuclear extracts from the AAR following 2h of reperfusion were measured by Western blot analysis. i, Quantification of the protein levels in (h). n = 3 mice/group/time point. Statistical analysis was performed using two-way ANOVA. j, Representative images of fluorescence immunostaining for BMAL1 (red) and HIF2A (green) in the border zone of the hearts of C57BL/6J mice on day 1 post-MI; Scale bar, 50 μm, white arrows indicate the colocalization of BMAL1 and HIF2A in the nuclei. n = 3 mice/group/time point. k, Representative heart slices subjected to Evan’s blue and TTC double staining after 2 h of reperfusion: infarct area (green line) and AAR (blue line); scale bar, 1 mm. l, Percentage of the AAR relative to the size of the LV. n = 7 mice/group/time point. Statistical analysis was performed using two-way ANOVA. m, Infarct sizes represented as the percentage of the AAR. n = same as above. Statistical analysis was performed using two-way ANOVA. n, Serum troponin I levels were evaluated at 2 hours after reperfusion. n = 8 mice/group/time point. Statistical analysis was performed using two-way ANOVA. o-r, Cardiac function was evaluated by speckle-tracking echocardiography on day 14 post-MI. LV systolic function, including EF, FS (o) and GLS (p), was measured. q, Representative left ventricular 3D (48 points) longitudinal strain and 6-segment longitudinal strain images demonstrating wall motion abnormalities (Color-coded six segments: Ant. Base, Anterior Base; Ant. Mid, Anterior Middle; Ant. Apex, Anterior Apex; Post. Apex, Posterior Apex; Post. Mid, Posterior Middle; Post. Base, Posterior Base). r, Left ventricular mechanical dyssynchrony as measured by intra-ventricular delay in time-to-peak strain and standard deviation of time-to-peak strain based on peak longitudinal strain. n = 8/Hif2a^loxP/loxP Myosin Cre+ mice/time point and n = 9/Myosin Cre+ mice/time point. Statistical analysis was performed using two-way ANOVA. All data are mean ± s.e.m, *p < 0.05, **p < 0.01, ***p < 0.001, and ****p < 0.0001, ns: not significant.

Fig. 3.. Areg is a circadian-dependent target of the BMAL1/HIF2A heterodimer.

a, Heatmap with the expression patterns of the top 20 potential HIF2A target genes (reanalyzed our previously published microarray data). n = 4 mice/group. b, Potential HIF2A target genes upregulated at ZT8 in the AAR from C57BL/6J mice following 2h of reperfusion. n = 3 mice/group/time point. Statistical analysis was performed using two-way ANOVA. c, AREG protein levels in the cytosolic extracts were measured by Western blot analysis. n = 3 mice/group/time point. d, Quantification of the AREG protein levels in (c). n = same as above. Statistical analysis was performed using two-way ANOVA. e, Representative immunostaining of AREG (red), α-sarcomeric (cardiomyocyte marker; green), and nuclei (DAPI; blue) on day 1 post-MI in the border zone of C57BL/6J mouse hearts subjected to myocardial IRI at ZT8 or ZT20; Scale bar, 50 μm. Arrows indicate α-sarcomeric⁺/AREG⁺ cells. n = 4 mice/time point. f, Quantification of AREG fluorescence intensity in (e). n = same as above. Statistical analysis was performed using unpaired Student’s t-tests. g-i, HCMs were synchronized by dexamethasone (200 nM, 1h) and then exposed to normoxia or hypoxia for 4h (1% O₂) at different intervals post-synchronization (CT4 to CT40). g, AREG transcript levels were measured by real-time PCR analysis. h, Protein levels of BMAL1, HIF2A or AREG were measured by Western blot analysis. i, Quantification of protein levels in (h). n = 3 independent experiments. j, HEK293 cells were transfected with BMAL1 siRNA, CLOCK siRNA, HIF2A siRNA, HIF1A siRNA, HIF1B siRNA (50 nM), or Scrambled siRNA for six hours and exposed to hypoxia for 4h (1% O₂). AREG transcript levels were then measured by real-time PCR analysis. n = 3 independent experiments. Statistical analysis was performed using one-way ANOVA. k, Transcript levels of AREG were evaluated by real-time PCR analysis following 2h of reperfusion in the AAR of hearts from Myosin Cre+ mice, Hif2a^loxP/loxP Myosin Cre+ mice and Bmal1^loxP/loxP Myosin Cre+ mice. n = 5 mice independent experiments. Statistical analysis was performed using unpaired Student’s t-tests. l, AREG protein levels were evaluated by Western blot analysis following 2h of reperfusion in the AAR of hearts from Bmal1^loxP/loxP Myosin Cre+ mice, Hif2a^loxP/loxP Myosin Cre+ mice and Myosin Cre+ mice. n = 5/Myosin Cre+, n = 4/Hif2a^loxP/loxP Myosin Cre+ mice, and n = 4/Bmal1^loxP/loxP Myosin Cre+ mice independent experiments. m, Quantification of AREG protein levels in (l). n = same as above. Statistical analysis was performed using one-way ANOVA. n and o, SPR analysis of the interaction between BMAL1/HIF2A and DNA. Biotinylated HRE DNA (n) or E-box DNA (o) was immobilized on a sensor. Protein concentrations of the BMAL1/HIF2A heterodimer used for measurements are indicated. n = 3 independent experiments. p, A highly conserved putative common binding site (CAGGTG) for both BMAL1 and HIF2A on the AREG promoter was predicted by JASPAR (https://jaspar.elixir.no/). q, HCMs were synchronized by dexamethasone (200 nM, 1h) and exposed to normoxia or hypoxia for 4h (1% O₂) at CT20 or CT32. ChIP-qPCR assays were conducted using HIF2A to examine the binding of hypoxia-induced HIF2A to the common binding site on the human AREG promoter. IgG was used as a negative control. n = 5 independent experiments. Statistical analysis was performed using two-way ANOVA. r, Synchronized HCMs after hypoxia treatment (1% O₂, 4h) were collected at CT20 or CT32, nuclear protein extracts were immunoprecipitated with HIF2A and blotted with anti-HIF2A and anti-BMAL1 antibodies. n = 3 independent experiments. s, HEK293 cells were transfected with BMAL1-Flag and exposed to normoxia or hypoxia (1% O₂) for 4h. ChIP-qPCR assays were conducted by HIF2A or BMAL1 antibodies to examine their binding to the human AREG promoter. n = 10–13 independent experiments. Statistical analysis was performed using two-way ANOVA. Outliers have been identified and removed using the ROUT (Q = 1%) method in GraphPad Prism. t, HEK293 cells were transfected with HIF2A-HA, BMAL1-Flag, or both, and luciferase assays were conducted to examine the transcription activation activity of the BMAL1/HIF2A complex on human AREG promoter bearing the common binding site. n = 4 independent experiments. Statistical analysis was performed using one-way ANOVA. All data are mean ± s.e.m. All t-tests were two-tailed. *p < 0.05, **p < 0.01, ***p < 0.001, and ****p < 0.0001, ns: not significant.

Fig. 4.. AREG drives circadian-dependent cardioprotection.

a, Experimental setup for evaluation of myocardial injury and cardiac function in Areg^−/− mice and WT mice subjected to IRI at ZT8 or ZT20. b, Representative heart slices subjected to Evan’s blue and TTC double staining after 2 h of reperfusion: infarct area (green line) and AAR (blue line); scale bar, 1 mm. c, Percentage of the AAR relative to the size of the LV. n = 7 mice/group/time point. Statistical analysis was performed using two-way ANOVA. d, Infarct sizes represented as the percentage of the AAR. n = same as above. Statistical analysis was performed using two-way ANOVA. e, Serum troponin I level, evaluated at 2h after reperfusion. n = same as above. Statistical analysis was performed using two-way ANOVA. f and g, Cardiac function was evaluated by speckle-tracking echocardiography on day 14 post-MI, f, EF and GLS. g, Representative B-mode images from the left ventricular long-axis view with 2D longitudinal strain analysis, showing 6 segments (1. Ant. Base, Anterior Base; 2. Ant. Mid, Anterior Middle; 3. Ant. Apex, Anterior Apex; 4. Post. Apex, Posterior Apex; 5. Post. Mid, Posterior Middle; 6. Post. Base, Posterior Base). n = 8 mice/group/time point. Statistical analysis was performed using two-way ANOVA. h and i, h, Representative photomicrographs of TUNEL and nuclear DAPI staining of cardiomyocyte marker α-sarcomeric-positive cardiomyocytes in the border zone of hearts obtained from WT and Areg^−/− mice on day 1 post-MI. White arrows point out TUNEL-positive (green) cardiomyocyte (red) nuclei (blue); scale bar, 50 μm. i, Percentage of TUNEL-positive cardiomyocytes after MI. n = 4 mice/group/time point. Statistical analysis was performed using two-way ANOVA. j, Experimental setup for evaluating myocardial injury and cardiac function in AREG-treated or vehicle (Veh)-treated (saline) mice subjected to IRI at ZT8 or ZT20. To mimic the clinical scenario, AREG (10 μg) was administered at the start of reperfusion and subsequently given daily for the first three days following injury at either ZT8 or ZT20. k, AREG protein level was measured by Western blot analysis following 2 h of reperfusion in the AAR from AREG-treated or Veh-treated mouse hearts. n = 3 mice/group/time point. l, Quantification of AREG protein levels in (k). n = same as above. Statistical analysis was performed using two-way ANOVA. m, Representative heart slices subjected to Evan’s blue and TTC double staining after 2h of reperfusion: infarct area (green line) and AAR (blue line); scale bar, 1 mm. n, Percentage of the AAR relative to the size of the LV. n = 7 mice/group/time point. Statistical analysis was performed using two-way ANOVA. o, Infarct size represented as the percentage of the AAR. n = same as above. Statistical analysis was performed using two-way ANOVA. p, Serum troponin I level, evaluated at 2h after reperfusion. n = same as above. Statistical analysis was performed using two-way ANOVA. q and r, Cardiac function was evaluated by speckle-tracking echocardiography analysis on day 14 post-MI in the ZT20 IRI mice following timed AREG treatment administered either at ZT8 or ZT20. q, EF and GLS. r, Representative left ventricular 3D (48 points) longitudinal strain and 6-segment longitudinal strain images demonstrating wall motion abnormalities (Color-coded six segments: Ant. Base, Anterior Base; Ant. Mid, Anterior Middle; Ant. Apex, Anterior Apex; Post. Apex, Posterior Apex; Post. Mid, Posterior Middle; Post. Base, Posterior Base). n = 7 mice/group/time point. Statistical analysis was performed using two-way ANOVA. s and t, s, Percentage of TUNEL-positive cardiomyocytes. t, Representative photomicrographs of TUNEL and nuclear DAPI staining of cardiomyocyte marker α-sarcomeric-positive cardiomyocytes in the border zone of hearts obtained from timed AREG-treated (ZT8 or ZT20) or Veh-treated mice on day 1 post-MI. White arrows point out TUNEL-positive (green) cardiomyocyte (red) nuclei (blue); scale bar, 50 μm. n = 5 mice/Veh-treated/ZT8, n = 4 mice/Veh-treated/ZT20, n = 5 mice/AREG-treated/ZT8, n = 5 mice/AREG-treated/ZT20. Statistical analysis was performed using two-way ANOVA. All data are mean ± s.e.m. *p < 0.05, **p < 0.01, ***p < 0.001, and ****p < 0.0001, ns: not significant.

Fig. 5.. Targeting BMAL1 using NOB for circadian-dependent cardioprotection.

a, Experimental setup for evaluating myocardial injury and cardiac function in NOB-treated (200 mg/kg, i.p., every other day) or Veh-treated C57BL/6J mice subjected to myocardial IRI at ZT8 or ZT20. b-d, Transcript (b) and protein levels (c) of BMAL1, RORα and AREG were measured following 2h of reperfusion in the AAR from NOB-treated or Veh-treated C57BL/6J mouse hearts. d, Quantification of protein levels in (c). n = 3 mice/group/time point. Statistical analysis was performed using two-way ANOVA. e, Representative images, along with quantification of BMAL1 immunofluorescence staining (red) in cell nuclei (DAPI: blue), on day 1 following IRI in the border zone of hearts from NOB-treated or Veh-treated mice subjected to myocardial IRI at ZT8 or ZT20; Scale bar, 50 μm. White arrows indicate BMAL1 within the nuclei. n = 5 mice/Veh-treated/ZT8, n = 4 mice/Veh-treated/ZT20, n = 5 mice/NOB-treated/ZT8, and n = 5 mice/NOB-treated/ZT20. Statistical analysis was performed using two-way ANOVA. f, Representative immunostaining of AREG (red), α-sarcomeric (green), and nuclei (DAPI: blue) on day 1 post-MI in the border zone of hearts from NOB-treated or Veh-treated mice subjected to myocardial IRI at ZT8 or ZT20; Scale bar, 50 μm. Arrows indicate α-sarcomeric⁺/AREG⁺ cells. The quantification of the fold change in AREG intensity is also presented. n = 4 mice/Veh-treated/ZT8, n = 5 mice/Veh-treated/ZT20, n = 4 mice/NOB-treated/ZT8, and n = 5 mice/NOB-treated/ZT20. Statistical analysis was performed using two-way ANOVA. g, Representative heart slices subjected to Evan’s blue and TTC double staining after 2h of reperfusion: infarct area (green line) and AAR (blue line); scale bar, 1 mm. h, Percentage of the AAR relative to the size of the LV. n = 7 mice/group/time point. Statistical analysis was performed using two-way ANOVA. i, Infarct size represented as the percentage of the AAR. n = same as above. Statistical analysis was performed using two-way ANOVA. j, Serum troponin I levels were evaluated at 2h after reperfusion. n = same as above. Statistical analysis was performed using two-way ANOVA. k and l, Cardiac function was evaluated by speckle-tracking echocardiography on day 14 post-MI in NOB-treated or Veh-treated mice subjected to myocardial IRI at ZT8 or ZT20. k, EF, FS, and GLS. l, Representative left ventricular 3D (48 points) longitudinal strain and 6-segment longitudinal strain images demonstrating wall motion abnormalities (Color-coded six segments: Ant. Base, Anterior Base; Ant. Mid, Anterior Middle; Ant. Apex, Anterior Apex; Post. Apex, Posterior Apex; Post. Mid, Posterior Middle; Post. Base, Posterior Base). n = 7 mice/group/time point. Statistical analysis was performed using two-way ANOVA. m and n, m, Representative photomicrographs of TUNEL and nuclear DAPI staining of cardiomyocyte marker α-sarcomeric-positive cardiomyocytes in the border zone of hearts obtained from Veh-treated or NOB-treated mice on day 1 post-MI. White arrows point out TUNEL-positive (green) cardiomyocyte (red) nuclei (blue); scale bar, 50 μm. n, Percentage of TUNEL-positive cardiomyocytes in (m). n = 5 mice/group/time point. Statistical analysis was performed using two-way ANOVA. All data are mean ± s.e.m.*p < 0.05, **p < 0.01, ***p < 0.001, and ****p < 0.0001, ns: not significant.

Fig. 6.. Structural analysis of the BMAL1/HIF2A heterodimer in complex with DNA.

a, Cryo-EM density map of the BMAL1/HIF2A/DNA complex. The HIF2A, BMAL1, and HRE DNA are colored in purple, red, and yellow, respectively. The bHLH, PAS-A, and PAS-B domains are as indicated. b, Two views of the overall structure of the BMAL1/HIF2A/DNA complex. c, Individual structures of HIF2A and BMAL1 within the DNA-bound BMAL1/HIF2A heterodimer. d, Detailed interactions of the four interfaces (I to IV) between HIF2A and BMAL1. Residues involved in the interaction between BMAL1 (red) and HIF2A (purple) are indicated. e, Pull-down analysis showing impaired interaction of GST-HIF2A with Flag-tagged BMAL1 mutants. Mutations in bHLH, PAS-A, and PAS-B domains of the BMAL1 are indicated. n = 3 independent experiments. f, HEK293 cells overexpressing WT or mutated Flag-tagged BMAL1 were exposed to ambient hypoxia (1% O₂) for 4h, then followed by immunoprecipitation with Flag and blotted with anti-HIF2A and anti-Flag antibodies. n = 3 independent experiments. g, HEK293 cells were transfected with either WT or mutated BMAL1, along with the HIF2A vector. Luciferase reporter assay was conducted to evaluate the transcription activation activity of the BMAL1/HIF2A complex on human AREG, which contains the common binding site. n = 4 independent experiments. Data are mean ± s.e.m. Statistical analysis was performed using one-way ANOVA. ####p < 0.0001 compared to cells transfected with HIF2A only, ****p < 0.0001 compared to cells transfected with WT BMAL1. h, Schematic representation illustrating the substantial structural rearrangement of BMAL1 (red) when accommodating various partners to be involved in different pathways. The PAS domains of BMAL1 (red) bend toward nearly opposite direction and position separately when intertwining with HIF2A (purple).