HIF1α mediates circadian regulation of skeletal muscle metabolism and substrate preference in response to time-of-day exercise

Abstract

The regulation of metabolism in peripheral tissues is intricately linked to circadian rhythms, with hypoxia-inducible factor-1α (HIF1α) implicated in modulating time-of-day-specific exercise responses. To investigate this relationship, we generated a skeletal muscle-specific HIF1α knockout (KO) mouse model and performed extensive metabolic phenotyping and transcriptomic profiling under both basal conditions and following acute exercise during early rest (ZT3) and active (ZT15) phases. Our findings reveal that HIF1α drives a more robust transcriptional and glycolytic response to exercise at ZT3, promoting glucose oxidation and mannose-6-phosphate production while potentially sparing fatty acid oxidation. In the absence of HIF1α, skeletal muscle metabolism shifts toward oxidative pathways at ZT3, with notable alterations in glucose fate. These results establish HIF1α as an important regulator of time-of-day-specific metabolic adaptations, integrating circadian and energetic signals to optimize substrate utilization. This work highlights the broader significance of HIF1α in coordinating circadian influences on metabolic health and exercise performance.

Keywords: circadian; energy metabolism; exercise; metabolism; transcription factor.

Fig. 1.

Metabolic phenotyping of Hif1α KO mice. (A) Schematic of the skeletal muscle HIF1α KO model. (B) Hif1α gene expression in quadriceps muscle bulk tissue from fl/fl and HIF1α KO mice (n = 14-16; males). (C) Body weight; (n = 8). (D) Average food intake during the light phase (ZT0-12) and dark phase (ZT12-24); (n = 8). (E) Activity trace over the course of a full 24-h period; (n = 8). Data show average values over hourly intervals. (F) EE trace across the 24-h period; (n = 8). Data show average values over hourly intervals. (G) RER trace across the 24-h period; (n = 8). Data show average values over hourly intervals. (H) Liver glycogen was measured every 4 h over a 24-h period. (I) Average distance run on voluntary running wheels across the 24-h light–dark cycle and cumulative meters run during the light and dark phases; (n = 4-8). (J) EE across the 24-h light–dark cycle in mice with access to running wheels; (n = 4-8). (K) RER across the 24-h light–dark cycle in mice with access to running wheels; (n = 4-8). Values from (I–K) are an average across a 5-d period. (L) Blood glucose measurements collected across a 90 min oGTT conducted at either ZT3 or ZT15; (n = 22-24). (M) iAUC of glucose from the oGTT; (n = 22-24). (N) Blood insulin measurements occurring during an oGTT at ZT3 and ZT15; (n = 22-24). (O) AUC of insulin from oGTT; (n = 22-24). ZT refers to mice undergoing the experiment at ZT3 versus ZT15. Time refers to the subsequent samples obtained from mice undergoing experiment at either ZT3 or ZT15. Genotype refers to fl/fl versus KO mice. Light refers to the summation or average of samples occurring in light phase versus the dark phase. Data are presented as mean values ± the SEM. Statistical significance was tested by unpaired two-tailed Student’s t test or two-way ANOVA with the Holm–Šídák post hoc test; *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001. Significant P-values in bold text.

Fig. 2.

Time-of-day graded exercise test assessment. (A) Average running distance during a graded exercise test at ZT3 and ZT15; (n = 24). (B) Blood glucose measurements immediately before (pre) and after (post) an exercise test performed at ZT3 and ZT15; (n = 24). (C) Lactate measurements prior to pre- and post-graded exercise test; (n = 24). Pre versus Post refers to samples collected immediately prior versus immediately after the exercise bout. Genotype refers to fl/fl versus KO, ZT refers to ZT3 versus ZT15. Data are presented as mean values ± the SEM. Statistical significance was tested by two-way or three-way ANOVA with the Holm–Šídák post hoc test; *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001.

Fig. 3.

Metabolic response to an acute bout of exercise at ZT3 versus ZT15. (A) Schematic of experimental design. (B) Blood glucose measurements at ZT3 and (C). Blood glucose measurements at ZT15 before and after an acute exercise bout, during the rest period following the exercise, and during a 90 min oGTT; (n = 16-18). (D) Change in blood glucose from exercise (pre-exercise blood glucose; timepoint 0 min, subtracted from post exercise blood glucose; timepoint 60 min) at ZT3 and at ZT15. (E) Average blood glucose measurements during the rest period following exercise. (F) Blood glucose AUC from oGTT at ZT3 and ZT15. (G) Insulin AUC from oGTT. ZT refers to mice undergoing the experiment at ZT3 versus ZT15. Time refers to the subsequent samples obtained from mice undergoing experiment at either ZT3 or ZT15. Genotype (Gen) refers to fl/fl versus KO mice. Exercise (Ex) refers to exercised mice versus sedentary control mice. Data are presented as mean values ± the SEM. Statistical significance was tested by two-way or three-way ANOVA with the Holm–Šídák post hoc test; *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001.

Fig. 4.

Metabolic fate of a glucose load following an exercise at ZT3. (A) Overview of the experimental procedures. (B) Whole body glucose oxidation measured via exhaled ¹³CO₂ and (C). Accumulative whole body glucose oxidation following a 1 h acute exercise bout. (D) Effect of acute exercise on ¹³C fractions of metabolites gastrocnemius muscle following ¹³C-glucose administration (n = 8). Gray circles indicate no significant changes in isotopologue between sedentary (fl/fl and KO mice) and exercised (fl/fl and KO mice) and red circles indicate a significant increase while blue indicate a significant decrease. (E) Altered ¹³C fractions of metabolites in skeletal muscle between fl/fl and HIF1α KO mice in sedentary and exercised conditions. (F) Summary of significantly altered ¹³C fractions of metabolites between fl/fl and HIF1α KO mice in sedentary and exercised conditions in skeletal muscles (n = 8). (G) mRNA expression levels of genes in the glycolytic pathway differentially expressed in fl/fl and HIF1α KO mice at ZT3 in sedentary and exercised conditions (n = 8). Data are presented as mean values ± the SD. Statistical significance was tested by two-way or three-way ANOVA with the Holm–Šídák post hoc test; *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001.

Fig. 5.

Transcriptomic analysis of gastrocnemius muscle following exercise at ZT3 and ZT15. (A) Heatmap showing differentially expressed transcripts between ZT3 and ZT15. (B) Gene ontology analysis showing the top 7 biological processes up or downregulated between ZT 3 and ZT 15. Gene set enrichment analysis was performed on genes ranked by FDR. (C) Differential expression of logFC between fl/fl and HIF1α KO mice. (D) Number of differentially expressed genes between fl/fl and HIF1α KO mice and the intersection of these transcripts. (E) Gene ontology analysis showing the top 7 biological processes up or downregulated between fl/fl and HIF1α KO mice. Gene set enrichment analysis was performed on genes ranked by FDR. (F) Number of differentially expressed genes within the top 6 biological processes related to aerobic metabolism and mitochondrial function.