Circadian Clock Interaction with HIF1α Mediates Oxygenic Metabolism and Anaerobic Glycolysis in Skeletal Muscle

Abstract

Circadian clocks are encoded by a transcription-translation feedback loop that aligns energetic processes with the solar cycle. We show that genetic disruption of the clock activator BMAL1 in skeletal myotubes and fibroblasts increased levels of the hypoxia-inducible factor 1α (HIF1α) under hypoxic conditions. Bmal1^-/- myotubes displayed reduced anaerobic glycolysis, mitochondrial respiration with glycolytic fuel, and transcription of HIF1α targets Phd3, Vegfa, Mct4, Pk-m, and Ldha, whereas abrogation of the clock repressors CRY1/2 stabilized HIF1α in response to hypoxia. HIF1α bound directly to core clock gene promoters, and, when co-expressed with BMAL1, led to transactivation of PER2-LUC and HRE-LUC reporters. Further, genetic stabilization of HIF1α in Vhl^-/- cells altered circadian transcription. Finally, induction of clock and HIF1α target genes in response to strenuous exercise varied according to the time of day in wild-type mice. Collectively, our results reveal bidirectional interactions between circadian and HIF pathways that influence metabolic adaptation to hypoxia.

Keywords: BMAL1; HIF; circadian; clock; exercise; glycolysis; hypoxia; lactate; mitochondria; skeletal muscle.

Figure 1. Circadian clock controls oxygen consumption and anaerobic glycolysis through regulation of HIF1α

(A) OCR from WT and Bmal1^−/− C2C12 myotubes treated sequentially with oligomycin and FCCP (carbonyl cyanide-p-trifluoromethoxyphenylhydrazone) (n=9–10). (B) ECAR from intact WT and Bmal1^−/− C2C12 myotubes treated sequentially with glucose and oligomycin (n=10). (C) Immunoblots of HIF1α and β-actin in WT vs Bmal1^−/− (left) and Cry1/2^+/− vs Cry1/2^−/− (right) MEFs exposed for 14 hours to increasing doses of CoCl₂ (0 μM and a dilution curve from 7.8 to 125 μM). (D) Immunoblots of HIF1α and β-actin following exposure to 1% O₂ for indicated times in WT vs Bmal1^−/− C2C12 myotubes. (E) Expression of HIF target genes in WT vs Bmal1^−/− C2C12 myotubes exposed to 1% O₂ for 6 hours (n=7–14). (F) Expression of HIF target genes in gastrocnemius muscle from adult life-inducible skeletal muscle Bmal1^−/− mice (ACTA-rtTA-TRE-Cre;Bmal^fx/fx) and controls (ACTA-rtTA-TRE-Cre and Bmal^fx/fx) (n=4–5). (G) Relative luciferase activity of C212 myoblasts transfected with HRE-LUC and plasmids expressing circadian and HIF TFs (n=3). Data are represented as mean ± SEM. *p<0.05, **p<0.01, ***p<0.001. See also Figures S1–S2.

Figure 2. HIF regulates circadian clock function

(A) Normalized photomultiplier detection of PER2:LUC reporter oscillations in synchronized C2C12 myotubes following exposure to 125 μM dimethyloxalylglycine (DMOG), a HIF1/2α stabilizer (green line) or untreated controls (black line) (n=3). (B) ChIP of HIF1α and occupancy of the E-box-containing promoter regions of Per2 and Cry1, the HRE-containing promoter regions of Ldha and Vegfa, and the Slc2a2 gene promoter which has no E-box or HRE (n=3). (C) Relative luciferase activity of PER2:LUC in the presence of indicated BMAL1 and HIF combinations in C212 myoblasts (n=4). (D–E) Relative gene expression of clock target genes in (D) non-synchronized C2C12 myotubes exposed to 1% O₂ for 6 hours vs normoxic conditions (n=7–14) and (E) MEFs isolated from Cag-CRE-ER;Vhl^fx/fx vs control mice (n=4). Data are represented as mean ± SEM. *p<0.05, **p<0.01, ***p<0.001. See also Figure S3.

Figure 3. Clock and HIF transcriptional response to strenuous exercise varies according to time of day in skeletal muscle

WT mice were exercised by treadmill running to exhaustion at ZT0 (start of light period), ZT6 (mid-light period), ZT12 (start of dark period), or ZT18 (mid-dark period) prior to immediate extraction of gastrocnemius muscle and subsequent quantification of HIF and clock target mRNAs (n=7). Data are represented as mean ± SEM. *p<0.05, **p<0.01, ***p<0.001 using 2-way ANOVA statistical analysis.