Lipid droplet is a dynamic organelle that undergoes periods of biogenesis and degradation under environmental stimuli. The excessive accumulation of lipid droplets is the major characteristic of non-alcoholic fatty liver disease (NAFLD). Moderate aerobic exercise is a powerful intervention protecting against the progress of NAFLD. However, its impact on lipid droplet dynamics remains ambiguous. Mice were fed with 15 weeks of high-fat diet in order to induce NAFLD. Meanwhile, the mice performed 15 weeks of treadmill exercise. Our results showed that 15 weeks of regular moderate treadmill exercise alleviated obesity, insulin intolerance, hyperlipidemia, and hyperglycemia induced by HFD. Importantly, exercise improved histological phenotypes of NAFLD, including hepatic steatosis, inflammation, and locular ballooning, as well as prevented liver fat deposition and liver injury induced by HFD. Exercise reduced hepatic lipid droplet size, and moreover, it reduced PLIN2 protein level and increased PLIN3 protein level in the liver of HFD mice. Interestingly, our results showed that exercise did not significantly affect the gene expressions of DGAT1, DGAT2, or SEIPIN, which were involved in TG synthesis. However, it did reduce the expressions of FITM2, CIDEA, and FSP27, which were major involved in lipid droplet growth and budding, and lipid droplet expansion. In addition, exercise reduced ATGL protein level in HFD mice, and regulated lipophagy-related markers, including increasing ATG5, LAMP1, LAMP2, LAL, and CTSD, decreasing LC3II/I and p62, and promoting colocalization of LAMP1 with LDs. In summary, our data suggested that 15 weeks of moderate treadmill exercise was beneficial for regulating liver lipid droplet dynamics in HFD mice by inhibiting abnormal lipid droplets expansion and enhancing clearance of lipid droplets by lysosomes during the lipophagic process, which might provide highly flexible turnover for lipid mobilization and metabolism. Abbreviations: β-actin: actin beta; ATG5: autophagy related 5; LAMP2: lysosomal-associated membrane protein 2; LAMP1: lysosomal-associated membrane protein 1; SQSTM1/p62: sequestosome 1; GAPDH: glyceraldehyde-3-phosphate dehydrogenase; MAP1LC3/LC3: microtubule associated protein 1 light chain 3; ATGL: adipose triglyceride lipase; CSTD: cathepsin D; LAL: lysosomal acid lipase; DGAT1: diacylglycerol-o-acyltransferase 1; DGAT2: diacylglycerol-o-acyltransferase 2; CIDEA: cell death inducing dffa-like effector a; CIDEC/FSP27: cell death inducing dffa-like effector c; FITM2: fat storage-inducing transmembrane protein 2; PLIN2: adipose differentiation related protein; PLN3: tail-interacting protein 47; HSP90: heat shock protein 90; SREBP1c: sterol regulatory element binding protein-1c; chREBP: carbohydrate response element binding protein.

Although the benefits of aerobic exercise on obesity and type 2 diabetes are well-documented, the pathogenesis of type 2 diabetes and the intervention mechanism of exercise remain ambiguous. The correlation between mitochondrial quality and metabolic diseases has been identified. Disruption of the central or peripheral molecular clock can also induce chronic metabolic diseases. In addition, the interactive effects of the molecular clock and mitochondrial quality have attracted extensive attention in recent years. Exercise and a high-fat diet have been considered external factors that may change the molecular clock and metabolic state. Therefore, we utilized a DB/DB (BSK.Cg-Dock7m +/+ Leprdb/JNju) mouse model to explore the effect of chrono-aerobic exercise on the metabolic state of type 2 diabetic mice and the effect of timing exercise as an external rhythm cue on liver molecular clock-mitochondrial quality. We found that two differently timed exercises reduced the blood glucose and serum cholesterol levels in DB/DB mice, and compared with night exercise (8:00 p.m., the active period of mice), morning exercise (8:00 a.m., the sleeping period of mice) significantly improved the insulin sensitivity in DB/DB mice. In contrast, type 2 diabetes mellitus (T2DM) increased the expression of CLOCK and impaired the mitochondrial quality (mitochondrial networks, OPA1, Fis1, and mitophagy), as well as induced apoptosis. Both morning and night exercise ameliorated impaired mitochondrial quality and apoptosis induced by diabetes. However, compared with morning exercise, night exercise not only decreased the protein expression of CLOCK but also decreased excessive apoptosis. In addition, the expression of CLOCK was negatively correlated with the expression of OPA1 and Fis1. In summary, our research suggests that morning exercise is more beneficial for increasing insulin sensitivity and promoting glucose transport in T2DM, whereas night exercise may improve lipid infiltration and mitochondrial abnormalities through CLOCK–mitophagy–apoptosis in the liver, thereby downregulating glucose and lipid disorders. In addition, CLOCK-OPA1/Fis1–mitophagy might be novel targets for T2DM treatment.