CLOCK/BMAL1 regulates circadian change of mouse hepatic insulin sensitivity by SIRT1

Authors

  • Ben Zhou,

    1. Key Laboratory of Nutrition and Metabolism, Institute for Nutritional Sciences, Shanghai Institutes for Biological Sciences, Chinese Academy of Sciences, Graduate University of Chinese Academy of Sciences, Shanghai, China
    Search for more papers by this author
    • These authors contributed equally to this work.

  • Yi Zhang,

    1. Key Laboratory of Nutrition and Metabolism, Institute for Nutritional Sciences, Shanghai Institutes for Biological Sciences, Chinese Academy of Sciences, Graduate University of Chinese Academy of Sciences, Shanghai, China
    Search for more papers by this author
    • These authors contributed equally to this work.

  • Fang Zhang,

    Corresponding author
    1. Key Laboratory of Nutrition and Metabolism, Institute for Nutritional Sciences, Shanghai Institutes for Biological Sciences, Chinese Academy of Sciences, Graduate University of Chinese Academy of Sciences, Shanghai, China
    • Address reprint requests to: Qiwei Zhai, Ph.D., Key Laboratory of Nutrition and Metabolism, Institute for Nutritional Sciences, Shanghai Institutes for Biological Sciences, Chinese Academy of Sciences, 320 Yueyang Road, Shanghai 200031, China. E-mail: qwzhai@sibs.ac.cn; fax: +86 21 5492 0291; or Fang Zhang, Ph.D., Key Laboratory of Nutrition and Metabolism, Institute for Nutritional Sciences, Shanghai Institutes for Biological Sciences, Chinese Academy of Sciences, 320 Yueyang Road, Shanghai 200031, China. E-mail: fzhang@sibs.ac.cn.

    Search for more papers by this author
    • These authors contributed equally to this work.

  • Yulei Xia,

    1. Key Laboratory of Nutrition and Metabolism, Institute for Nutritional Sciences, Shanghai Institutes for Biological Sciences, Chinese Academy of Sciences, Graduate University of Chinese Academy of Sciences, Shanghai, China
    Search for more papers by this author
  • Jun Liu,

    1. Key Laboratory of Nutrition and Metabolism, Institute for Nutritional Sciences, Shanghai Institutes for Biological Sciences, Chinese Academy of Sciences, Graduate University of Chinese Academy of Sciences, Shanghai, China
    Search for more papers by this author
  • Rui Huang,

    1. Key Laboratory of Nutrition and Metabolism, Institute for Nutritional Sciences, Shanghai Institutes for Biological Sciences, Chinese Academy of Sciences, Graduate University of Chinese Academy of Sciences, Shanghai, China
    Search for more papers by this author
  • Yuangao Wang,

    1. Key Laboratory of Nutrition and Metabolism, Institute for Nutritional Sciences, Shanghai Institutes for Biological Sciences, Chinese Academy of Sciences, Graduate University of Chinese Academy of Sciences, Shanghai, China
    Search for more papers by this author
  • Yanan Hu,

    1. Key Laboratory of Nutrition and Metabolism, Institute for Nutritional Sciences, Shanghai Institutes for Biological Sciences, Chinese Academy of Sciences, Graduate University of Chinese Academy of Sciences, Shanghai, China
    Search for more papers by this author
  • Jingxia Wu,

    1. Key Laboratory of Nutrition and Metabolism, Institute for Nutritional Sciences, Shanghai Institutes for Biological Sciences, Chinese Academy of Sciences, Graduate University of Chinese Academy of Sciences, Shanghai, China
    Search for more papers by this author
  • Changgui Dai,

    1. Key Laboratory of Nutrition and Metabolism, Institute for Nutritional Sciences, Shanghai Institutes for Biological Sciences, Chinese Academy of Sciences, Graduate University of Chinese Academy of Sciences, Shanghai, China
    Search for more papers by this author
  • Hui Wang,

    1. Key Laboratory of Nutrition and Metabolism, Institute for Nutritional Sciences, Shanghai Institutes for Biological Sciences, Chinese Academy of Sciences, Graduate University of Chinese Academy of Sciences, Shanghai, China
    Search for more papers by this author
  • Yanyang Tu,

    1. Key Laboratory of Nutrition and Metabolism, Institute for Nutritional Sciences, Shanghai Institutes for Biological Sciences, Chinese Academy of Sciences, Graduate University of Chinese Academy of Sciences, Shanghai, China
    Search for more papers by this author
  • Xiaozhong Peng,

    1. The State Key Laboratory of Medical Molecular Biology, Department of Molecular Biology and Biochemistry, Institute of Basic Medical Sciences, Chinese Academy of Medical Sciences and Peking Union Medical College, Beijing, China
    Search for more papers by this author
  • Yiqian Wang,

    1. Key Laboratory of Nutrition and Metabolism, Institute for Nutritional Sciences, Shanghai Institutes for Biological Sciences, Chinese Academy of Sciences, Graduate University of Chinese Academy of Sciences, Shanghai, China
    Search for more papers by this author
  • Qiwei Zhai

    Corresponding author
    1. Key Laboratory of Nutrition and Metabolism, Institute for Nutritional Sciences, Shanghai Institutes for Biological Sciences, Chinese Academy of Sciences, Graduate University of Chinese Academy of Sciences, Shanghai, China
    • Address reprint requests to: Qiwei Zhai, Ph.D., Key Laboratory of Nutrition and Metabolism, Institute for Nutritional Sciences, Shanghai Institutes for Biological Sciences, Chinese Academy of Sciences, 320 Yueyang Road, Shanghai 200031, China. E-mail: qwzhai@sibs.ac.cn; fax: +86 21 5492 0291; or Fang Zhang, Ph.D., Key Laboratory of Nutrition and Metabolism, Institute for Nutritional Sciences, Shanghai Institutes for Biological Sciences, Chinese Academy of Sciences, 320 Yueyang Road, Shanghai 200031, China. E-mail: fzhang@sibs.ac.cn.

    Search for more papers by this author

  • Potential conflict of interest: Nothing to report.

  • This work was supported by grants from the National Natural Science Foundation of China (31030022, 31200591, 31200595, and 81321062), the National Basic Research Program of China (2014CB542300 and 2009CB918403), the Program of Shanghai Subject Chief Scientist (11XD1405800), the Knowledge Innovation Program of the Chinese Academy of Sciences (KSCX2-EW-R-09), the Knowledge Innovation Program of Shanghai Institutes for Biological Sciences, the Chinese Academy of Sciences (2012KIP309), and the SA-SIBS Scholarship Program.

Abstract

The protein deacetylase, sirtuin 1 (SIRT1), involved in regulating hepatic insulin sensitivity, shows circadian oscillation and regulates the circadian clock. Recent studies show that circadian misalignment leads to insulin resistance (IR); however, the underlying mechanisms are largely unknown. Here, we show that CLOCK and brain and muscle ARNT-like protein 1 (BMAL1), two core circadian transcription factors, are correlated with hepatic insulin sensitivity. Knockdown of CLOCK or BMAL1 induces hepatic IR, whereas their ectopic expression attenuates hepatic IR. Moreover, circadian change of insulin sensitivity is impaired in Clock mutant, liver-specific Bmal1 knockout (KO) or Sirt1 KO mice, and CLOCK and BMAL1 are required for hepatic circadian expression of SIRT1. Further studies show that CLOCK/BMAL1 binds to the SIRT1 promoter to enhance its expression and regulates hepatic insulin sensitivity by SIRT1. In addition, constant darkness-induced circadian misalignment in mice decreases hepatic BMAL1 and SIRT1 levels and induces IR, which can be dramatically reversed by resveratrol. Conclusion: These findings offer new insights for coordination of the circadian clock and metabolism in hepatocytes by circadian regulation of hepatic insulin sensitivity via CLOCK/BMAL1-dependent SIRT1 expression and provide a potential application of resveratrol for combating circadian misalignment-induced metabolic disorders. (Hepatology 2014;59:2196–2206)

Abbreviations
Akt

protein kinase B

BMAL1

brain and muscle ARNT-like protein 1

bp

base pairs

Ch-IP

chromatin immunoprecipitation

CRY

cryptochrome

CT

circadian time

Foxo

forkhead box O

GAPDH

glyceraldehyde 3-phosphate dehydrogenase

GSK3β

glycogen synthase kinase 3β

GTT

glucose tolerance test

HOMA-IR

homeostatic model assessment of IR

InsR

insuiln receptor

IR

insulin resistance

ITT

insulin tolerance test

KO

knockout

LDL

low-density lipoprotein

mRNA

messenger RNA

NAD+

nicotinamide adenine dinucleotide

NAMPT

nicotinamide phosphoribosyltransferase

PCR

polymerase chain reaction

PER2

period circadian protein homolog 2 (or period 2)

PMHs

primary mouse hepatocytes

siRNA

small interfering RNA

SIRT1

sirtuin 1

WT

wild type

ZT

zeitgeber time

Growing evidence shows that circadian rhythms regulate a wide variety of metabolic processes,[1, 2] and numerous metabolites, including glucose and lipids, and some metabolism-related hormones, such as insulin, oscillate in a circadian manner in blood.[3-5] Epidemiological studies show that circadian misalignment increases the risk of a series of diseases, including obesity and type 2 diabetes. Type 2 diabetes is usually characterized by abnormal high blood glucose and insulin resistance (IR), because insulin target tissues, including the liver, respond inadequately to circulating insulin. Lifestyle factors, such as diets rich in fat and poor in dietary fiber, sedentary lifestyle, and depression, are common causes for IR.[6] Circadian misalignment, a characteristic of jet lag and shift work, has also been reported to induce IR in human.[7-9] Circadian misalignment in rats elevates blood glucose and insulin levels, suggesting development of IR.[10] Genetic disruption of clock genes perturbs metabolic functions of specific tissues in mice at distinct phases of the sleep/wake cycle.[2, 11, 12] CLOCK and brain and muscle ARNT-like protein 1 (BMAL1), two core circadian transcription factors, can form heterodimers, bind to E-box elements in promoters of genes, including two main oscillators (Per and Cry), and induce their expression.[2] Mice with Clock mutation show improved insulin tolerance at 3 months of age and normal insulin sensitivity at 8 months.[13] Mice with Bmal1 deletion show improved insulin sensitivity,[14, 15] whereas liver-specific knockout (KO) of Bmal1 leads to decreased insulin sensitivity.[14] Thus, the role of CLOCK/BMAL1 in regulating insulin sensitivity and the underlying mechanisms need further investigation. Moreover, whether CLOCK/BMAL1 is involved in development of IR induced by circadian misalignment is yet to be elucidated.

SIRT1, a nicotinamide adenine dinucleotide (NAD+)-dependent protein deacetylase, plays an important role in regulating insulin sensitivity. Transgenic mice with moderate overexpression of sirtuin 1 (SIRT1) display increased glucose tolerance and insulin sensitivity when fed with a high-fat diet.[16, 17] Adenovirus-mediated overexpression of SIRT1 in liver of insulin-resistant obese mice attenuates hepatic steatosis and alleviates systemic IR.[18] On the other hand, liver-specific Sirt1 deficiency causes hepatic glucose overproduction, chronic hyperglycemia, and oxidative stress, leading to IR.[19] Resveratrol, a plant-derived polyphenol, can induce activation of SIRT1 and protect mice from diet-induced obesity and insulin resistance.[20-22] Resveratrol also improves insulin sensitivity in obesity or type 2 diabetes patients.[23, 24] Recent studies show that SIRT1 deacetylates period circadian protein homolog 2 (or period 2; PER2) and BMAL1, thus participating in circadian regulation, and that SIRT1 protein and/or activity have a circadian oscillation.[25, 26] However, whether circadian change of SIRT1 is also involved in regulating insulin sensitivity is still unclear, and the underlying mechanisms for circadian expression of SIRT1 are yet to be elucidated.

Here, we show the important roles of CLOCK/BMAL1 in regulating circadian expression of SIRT1 and hepatic insulin sensitivity. Modulation of hepatic insulin signaling by CLOCK/BMAL1 is largely dependent on SIRT1. More important, resveratrol supplementation can efficiently reverse IR and other metabolic disorders induced by circadian misalignment.

Materials and Methods

Immunoblotting

Immunoblotting was performed with antibodies against CLOCK, BMAL1, PER2, acetyl-RelA/p65 (K310; Abcam, Cambridge, MA), glyceraldehyde 3-phosphate dehydrogenase (GAPDH), SIRT1 (Millipore, Billerica, MA), insulin receptor (InsR), Tyr1150/1151-phosphorylated InsR, protein kinase B (Akt), Ser473-phosphorylated Akt, glycogen synthase kinase 3β (GSK3β), Ser9-phosphorylated GSK3β, p53, acetyl-p53 (K379; Cell Signaling Technology Inc., Danvers, MA), and RelA/p65 (Santa Cruz Biotechnology, Santa Cruz, CA). Protein quantification was normalized to GAPDH, actin, or tubulin.

Statistical Analysis

Data are expressed as mean ± standard deviation of at least three independent experiments. Statistical significance was assessed by one-way analysis of variance for glucose tolerance test (GTT), insulin tolerance test (ITT), and messenger RNA (mRNA) levels and luciferase activity after serum shock and by the Student t test for all other experiments. Differences were considered statistically significant at P < 0.05.

Full descriptions of additional materials and methods are given in the Supporting Information.

Results

Hepatic Expression of CLOCK and BMAL1 Is Down-Regulated Under Insulin-Resistant Conditions

To investigate the correlation of CLOCK and BMAL1 with hepatic insulin sensitivity, we treated primary mouse hepatocytes (PMHs) with glucosamine or palmitate to induce IR, as described previously.[22, 27, 28] Glucosamine and palmitate impaired insulin signaling and significantly down-regulated protein and mRNA levels of CLOCK and BMAL1 (Fig. 1A,B and Supporting Fig. 1A). Consistently, both protein and mRNA levels of CLOCK and BMAL1 were also decreased in liver of db/db mice with IR, compared with those in wild-type (WT) mice (Fig. 1C,D and Supporting Fig. 1B). These data show that down-regulation of CLOCK and BMAL1 are associated with hepatic IR.

Figure 1.

Protein and mRNA levels of CLOCK and BMAL1 are down-regulated under insulin-resistant conditions. (A and B) Protein (A) and mRNA (B) levels of CLOCK and BMAL1 were down-regulated in insulin-resistant PMHs induced by glucosamine. Immunoblotting and quantitative PCR were performed to detect protein and mRNA levels. (C and D) Protein (C) and mRNA (D) levels of CLOCK and BMAL1 decreased at ZT7 in liver of db/db mice, compared to those in WT. n = 4 for each group. *P < 0.05; **P < 0.01.

CLOCK and BMAL1 Positively Regulate Insulin Sensitivity in Hepatocytes

To investigate whether down-regulation of CLOCK and/or BMAL1 leads to hepatic IR, we used small interfering RNAs (siRNAs) to knock down CLOCK or BMAL1 in PMHs. Knockdown of CLOCK or BMAL1 induced insulin resistance, as indicated by markedly impaired insulin-stimulated phosphorylation of InsR, Akt, and GSK3β (Fig. 2A,B). Consistently, ectopic expression of CLOCK and BMAL1 by adenovirus attenuated insulin resistance in PMHs induced by glucosamine (Fig. 2C). Moreover, tail vein injection of adenoviruses expressing CLOCK and BMAL1 improved insulin tolerance in db/db mice (Fig. 2D,E). These data show that increased expression of CLOCK and BMAL1 can improve hepatic insulin sensitivity under insulin-resistant conditions.

Figure 2.

CLOCK and BMAL1 regulate insulin sensitivity in hepatocytes and insulin tolerance in mice. (A and B) Knockdown of CLOCK (A) or BMAL1 (B) induced IR in PMHs, as measured by immunoblotting for insulin-induced phosphorylation of InsR, Akt, and GSK3β. si-Ctrl, control siRNA; si-CLOCK, CLOCK siRNA; si-BMAL1, BMAL1 siRNA. (C) Ectopic expression of CLOCK and BMAL1 by adenovirus attenuated IR in PMHs induced by glucosamine. (D and E) Tail vein injection of adenovirus expressing CLOCK and BMAL1 (Ad-C&B) improved insulin tolerance at ZT7 in db/db mice (n = 8-11 for each group). Protein levels in liver collected at ZT7 were measured by immunoblotting. *P < 0.05; **P < 0.01.

Taken together, these data show CLOCK/BMAL1 plays an important role in regulating hepatic insulin sensitivity.

CLOCK/BMAL1 and SIRT1 Are Required for Circadian Change of Insulin Sensitivity

To further investigate the role of CLOCK/BMAL1 in regulating insulin sensitivity in vivo, we first monitored the circadian rhythm of insulin sensitivity in mice by the ITT. Consistent with previous reports,[15, 29-31] under both fed and fasted conditions, we also observed that mice at ZT13 (the day-to-night transition; zeitgeber time [ZT]) showed significantly increased insulin sensitivity, compared with those at ZT1 (the night-to-day transition), as measured by the ITT (Fig. 3A,B and Supporting Fig. 2A). Furthermore, we found that both ClockΔ19/Δ19 mutant mice and liver-specific Bmal1 KO mice showed similar insulin sensitivity at ZT1 and ZT13 (Fig. 3A,B). These data demonstrate that CLOCK and hepatic BMAL1 are required for circadian change of insulin sensitivity. Glucose tolerance in WT mice was similar with liver-specific Bmal1 KO mice at ZT1 and showed a slight difference at ZT13 (Supporting Fig. 2B,C).

Figure 3.

CLOCK/BMAL1 and SIRT1 are required for circadian change of insulin sensitivity. (A-C) CLOCK, SIRT1, and hepatic BMAL1 were required for circadian change of insulin sensitivity. ITT and the relative areas under the curves (AUC) were measured in Clock mutant (ClockΔ19/Δ19), liver-specific Bmal1 KO (Liver-Bmal1−/−) and Sirt1 KO (Sirt1−/−) mice at ZT1 and ZT13 fasted for 4 hours. n = 7-10 for each group. (D-F) BMAL1 and SIRT1 are required for the circadian oscillation of insulin sensitivity in hepatocytes. After serum shock for the indicated times, insulin-induced phosphorylation of InsR, Akt, and/or GSK3β in WT (D), Bmal1 KO (E), and Sirt1 KO (F) hepatocytes was measured by immunoblotting. *P < 0.05; **P < 0.01.

Recently, growing evidences indicate that SIRT1 can enhance insulin sensitivity.[18, 19, 22, 32] In contrast to WT littermate control mice, Sirt1−/− mice at ZT13 even showed decreased insulin sensitivity, compared with Sirt1−/− mice at ZT1 (Fig. 3C), suggesting that SIRT1 is also required for circadian change of insulin sensitivity.

Serum shock has been demonstrated to induce rhythmic clock gene expression in hepatocytes.[33, 34] We found that insulin-induced phosphorylation of InsR, Akt, and GSK3β showed circadian oscillation over a period of 24 hours (Fig. 3D). However, the circadian oscillation pattern of insulin signaling was altered in Bmal1 or Sirt1 KO hepatocytes (Fig. 3E,F), suggesting that BMAL1 and SIRT1 are required for circadian change of hepatic insulin sensitivity.

Taken together, these results demonstrate that CLOCK/BMAL1 and SIRT1 are required for circadian change of insulin sensitivity.

CLOCK/BMAL1 Is Required for Circadian Expression of Hepatic SIRT1

In primary hepatocytes exposed to serum shock, Sirt1 mRNA exhibited circadian oscillation similar to that of Per2 and Per1, two reported transcription targets of CLOCK/BMAL1, and its circadian oscillation was altered in Bmal1 KO hepatocytes as Per2 (Fig. 4A and Supporting Fig. 3). SIRT1 protein was also expressed in a circadian manner as PER2 after serum shock (Fig. 4B). In addition, the rhythmic change of SIRT1 protein was altered in Bmal1 KO hepatocytes as PER2, and the acetylation level of RelA/p65 (a well-known substrate of SIRT1) was consistent with the SIRT1 protein level (Fig. 4B,C), indicating that BMAL1 is required for circadian expression of hepatic SIRT1.

Figure 4.

CLOCK and BMAL1 are required for circadian expression of hepatic SIRT1. (A) Bmal1-dependent circadian expression of Sirt1 and Per2. After serum shock for the indicated times, mRNA levels of indicated genes in PMHs were measured by quantitative PCR. (B and C) Bmal1 deficiency altered circadian expression of indicated proteins. After serum shock for the indicated times, the indicated proteins in WT (B) and Bmal1 KO hepatocytes (C) were measured by immunoblotting. (D) Indicated protein levels in liver of C57BL/6 mice at ZT1 and ZT13 fasted for 12 hours. n = 4 for each group. (E) SIRT1 and PER2 protein levels in liver of ClockΔ19/Δ19 mice at ZT1 and ZT13 fasted for 12 hours. n = 4 for each group. *P < 0.05; **P < 0.01.

Consistent with a previous report,[25] immunoblotting showed that SIRT1 protein levels were increased in the liver at ZT13, compared with those at ZT1 (Fig. 4D), which is in line with the increased insulin sensitivity at ZT13 (Fig. 3A,B and Supporting Fig. 2A). Of note, PER2, a well-known transcription target of CLOCK/BMAL1, showed a similar circadian change as SIRT1 (Fig. 4D). In accord with the similar insulin sensitivity of ClockΔ19/Δ19 mice at ZT1 and ZT13 (Fig. 3A), protein levels of SIRT1 and PER2 in liver of ClockΔ19/Δ19 mice were also similar at ZT1 and ZT13 (Fig. 4E), which shows that circadian expression of hepatic SIRT1 is dependent on CLOCK as PER2.

These data show that CLOCK and BMAL1 are required for circadian expression of hepatic SIRT1.

CLOCK/BMAL1 Binds to Sirt1 Promoter and Activates Its Transcription

Ectopic expression of CLOCK and BMAL1 by adenoviruses in primary hepatocytes markedly increased SIRT1 mRNA and protein levels (Fig. 5A). Consistently, knockdown of CLOCK or BMAL1 decreased both of the mRNA and protein levels of SIRT1 in hepatocytes (Fig. 5B). In liver of ClockΔ19/Δ19 mutant mice, SIRT1 mRNA and protein levels were also down-regulated (Fig. 5C). There are two E-boxes (CACGTG), 103 or 175 base pairs (bp) upstream of the Sirt1 initiation codon (Supporting Fig. 4A). CLOCK/BMAL1 has been shown to be recruited to E-box cis-element and activates transcription.[35] As expected, chromatin immunoprecipitation (Ch-IP) showed that CLOCK and BMAL1 bound to the Sirt1 promoter region containing the two E-boxes (Fig. 5D and Supporting Fig. 4B). Luciferase reporters with different lengths of Sirt1 promoter were also used to analyze their transcriptional activity. Ectopic expression of CLOCK/BMAL1 markedly increased transcriptional activity of the 629-bp Sirt1 promoter, but had no significant effect on that of the 91-bp Sirt1 promoter without the two E-boxes (Fig. 5E). Furthermore, mutation of the two E-boxes in the 629-bp Sirt1 promoter disrupted the effect of CLOCK/BMAL1 and the circadian change of the promoter activity induced by serum shock (Fig. 5F). These results show that CLOCK/BMAL1 binds to the Sirt1 promoter region containing the two E-boxes and activates its transcription.

Figure 5.

CLOCK and BMAL1 bind to Sirt1 promoter and increase its transcription and expression. (A) Ectopic expression of CLOCK and BMAL1 by adenovirus in primary hepatocytes increased SIRT1 mRNA and protein levels. Quantitative PCR and immunoblotting were performed to measure mRNA and protein levels. (B) Knockdown of CLOCK or BMAL1 in primary hepatocytes decreased SIRT1 mRNA and protein levels. (C) SIRT1 mRNA and protein levels at ZT13 in liver were decreased in ClockΔ19/Δ19 mice. n = 3. (D) CLOCK and BMAL1 bound to the Sirt1 promoter region containing the two E-boxes, as revealed by Ch-IP assay. (E) The effect of CLOCK and BMAL1 on luciferase expression under the control of the indicated Sirt1 promoter fragments. (F) Mutation of the two E-boxes diminished the activation of the 629-bp Sirt1 promoter by CLOCK and BMAL1 and its circadian change induced by serum shock. The results of three independent experiments a, b, and c for the circadian changes are presented. *P < 0.05; **P < 0.01.

CLOCK/BMAL1 Regulates Hepatic Insulin Sensitivity in a SIRT1-Dependent Manner

Based on the above-described findings, we speculated that SIRT1 is involved in CLOCK/BMAL1-regulated insulin sensitivity. Ectopic expression of CLOCK and BMAL1 in Sirt1−/− PMHs failed to improve insulin sensitivity (Fig. 6A-C). In addition, IR caused by knockdown of BMAL1 or CLOCK in PMHs was markedly attenuated by ectopic expression of SIRT1 with adenovirus (Fig. 6D-F). These results clearly indicate the important role of SIRT1 in CLOCK/BMAL1-regulated hepatic insulin sensitivity.

Figure 6.

Regulation of insulin sensitivity by CLOCK and BAML1 is dependent on SIRT1. (A-C) CLOCK and BMAL1 failed to attenuate IR induced by glucosamine in Sirt1−/− hepatocytes. (D-F) Ectopic expression of SIRT1 alleviated the IR induced by knockdown of BMAL1 or CLOCK in primary hepatocytes. *P < 0.05; **P < 0.01.

Constant Darkness Down-Regulates Hepatic BMAL1 and SIRT1 and Induces IR

To test the effect of circadian misalignment on SIRT1 and insulin sensitivity, we used light deprivation as a model of circadian misalignment. In constant darkness, there is a consistent shift in activity onset of mice, and the circadian period is shortened.[36, 37] After being kept in constant darkness for 2 weeks, both BMAL1 and SIRT1 protein levels were reduced in liver, whereas CLOCK protein level did not show significant change (Fig. 7A and Supporting Fig. 5A,B). Consistent with decreased hepatic BMAL1 and SIRT1 protein levels, constant darkness induced IR at ZT/CT 7 (circadian time; CT), as measured by the homeostatic model assessment of IR (HOMA-IR) index (Fig. 7B). Furthermore, GTT and ITT also showed that constant darkness significantly induced glucose intolerance at ZT/CT 1 and 13 and IR at ZT/CT 13 (Fig. 7C,D). Moreover, insulin-induced phosphorylation of InsR, Akt, or GSK3β were all attenuated in liver of mice under constant dark condition at ZT/CT 1, 7, and 13 (Fig. 7E and Supporting Fig. 5C,D). These data show that circadian misalignment induced by light deprivation decreases hepatic BMAL1 and SIRT1 and induces IR.

Figure 7.

Circadian misalignment induced by constant darkness decreases hepatic BMAL1 and SIRT1 protein levels and causes IR. (A) CLOCK, BMAL1, and SIRT1 protein levels in liver of mice under light/dark (LD) and constant dark (DD) conditions at ZT/CT 7. n = 5 for each group. (B) HOMA-IR index of mice under LD or DD conditions at ZT/CT 13. n = 4-5 for each group. (C and D) GTT and ITT in C57BL/6 mice under LD and DD conditions at the indicated circadian times. n = 8-11 for each group. (E) Insulin-induced phosphorylation of InsR, Akt, and GSK3β in liver of mice under LD or DD conditions at ZT/CT 7. n = 4 for each group. *P < 0.05; **P < 0.01.

Resveratrol Attenuates IR Induced by Constant Darkness

It has been reported that IR associated with decreased SIRT1 can be attenuated by resveratrol, which can induce activation of SIRT1.[22, 38, 39] We fed mice with resveratrol at a dose of 25 mg/kg/day for 36 days, including 14 days in the normal condition and then 22 days in constant darkness. Then, acetylation of p53 and RelA/p65, two known targets of SIRT1, were detected to assess SIRT1 activity. Acetylation levels of p53 and RelA/p65 at ZT/CT 7 were markedly increased in liver of mice under a constant dark condition, indicating that constant darkness decreased SIRT1 activity (Fig. 8A). Treatment with resveratrol abolished the negative effect of constant darkness on SIRT1 activity (Fig. 8A). Although NAD+ levels were decreased in liver of ClockΔ19/Δ19 and liver-specific Bmal1 KO mice, constant darkness or resveratrol did not significantly change NAD+ levels in liver (Supporting Fig. 6A-C). Furthermore, nicotinamide phosphoribosyltransferase (NAMPT) inhibitor FK866 decreased intracellular NAD+ levels to approximately 50% as serum shock for about 12-24 hours and had no significant effect on insulin signaling and SIRT1 protein level (Supporting Fig. 6D-F). These data suggest that the altered SIRT1 activity is not based on the change of NAD+ levels. Surprisingly, treatment with resveratrol dramatically reversed glucose and insulin intolerance induced by constant darkness at ZT/CT 1 and 13, respectively (Fig. 8B,C). Moreover, constant darkness induced impaired hepatic insulin signaling at ZT/CT 7, including the decreased insulin-induced phosphorylation of InsR, Akt, and GSK3β, was also reversed by resveratrol (Fig. 8D). Consistently, HOMA-IR index, blood glucose, serum total cholesterol, and low-density lipoprotein (LDL) cholesterol levels at ZT/CT 7 in mice under a constant dark condition were significantly decreased by resveratrol treatment (Supporting Fig. 7).

Figure 8.

Resveratrol increases SIRT1 activity and attenuates the IR induced by circadian misalignment. (A) Resveratrol increased SIRT1 activity at ZT/CT 7 in liver of mice under a constant dark (DD) condition close to that under a light/dark (LD) condition, as measured by acetylation levels of p53 and RelA/p65 with immunoblotting. n = 8 for each group. In this and (B-D), white, black, and gray bars indicate LD vehicle, DD vehicle, and DD resveratrol, respectively. (B and C) Resveratrol improved glucose (B) and insulin (C) tolerance in mice under DD condition. GTT and ITT were performed at ZT/CT 1 and ZT/CT 13, respectively. n = 8-9 for each group. *P < 0.05 and **P < 0.01 versus LD vehicle at the same time point; #P < 0.05 and ##P < 0.01 versus DD vehicle at the same time point. (D) Insulin-induced phosphorylation of InsR, Akt, and GSK3β at ZT/CT 7 in liver of mice treated with vehicle or resveratrol under LD or DD condition. n = 8-9 for each group. (E and F) Overexpression of SIRT1 by tail vein injection of adenovirus improved insulin tolerance (E) and hepatic insulin sensitivity (F) in mice under DD condition. n = 7-8 for each group. *P < 0.05; **P < 0.01.

Then, we tested whether a relatively low dose of resveratrol with short-time treatment could also attenuate constant darkness-caused IR and other metabolic disorders. Mice were fed with resveratrol at a dose of 2.5 mg/kg/day for 5 days in the normal condition and then in constant darkness for the indicated times. Surprisingly, we found that resveratrol at such a low dose with a 5-day pretreatment could significantly improve glucose and insulin tolerance at ZT/CT 1 and 13, and serum total triglyceride and LDL cholesterol levels at ZT/CT 7 in mice under a constant dark condition (Supporting Fig. 8). Moreover, constant darkness-induced impaired hepatic insulin signaling at ZT/CT 1 and 13 was also reversed by resveratrol (Supporting Fig. 9).

In addition, overexpression of SIRT1 by tail vein injection of adenovirus significantly improved insulin tolerance and hepatic insulin sensitivity of mice in constant darkness (Fig. 8E,F).

Taken together, these data demonstrate that feeding with resveratrol or overexpression of SIRT1 by tail vein injection of adenovirus can improve insulin sensitivity in mice under a constant dark condition.

Discussion

In this study, we show that down-regulation of CLOCK and BMAL1 induces hepatic IR, and constant darkness down-regulates hepatic BMAL1 and induces IR. Consistently, mice with liver-specific deletion of Bmal1 show decreased insulin sensitivity,[14] and circadian misalignment also induces insulin resistance in humans.[7-9] However, both Clock mutant and Bmal1 deficiency mice mainly show increased insulin sensitivity.[14, 15] The precise mechanism responsible for such a difference remains to be elucidated in the future.

SIRT1, a master metabolic regulator, has been shown to be regulated by various transcriptional factors.[40, 41] Here, we demonstrate that CLOCK/BMAL1 binds to the E-box elements in Sirt1 promoter, up-regulates its transcription and contributes to its circadian change. It has been reported that SIRT1 binds to CLOCK/BMAL1, deacetylates BMAL1 and PER2, maintains high-magnitude circadian transcription of several core clock genes, and thus regulates the circadian clock.[25, 26] Thus, CLOCK/BMAL1 and SIRT1 form a feedback loop. Similarly, some transcription factors, including p53, forkhead box O (Foxo)1, and Foxo3a, also form feedback loops with SIRT1.[42-44] It is well established that CLOCK/BMAL1 and their repressors, PER and cryptochrome (CRY), form a core feedback loop and regulate circadian rhythm in mammals.[45] The CLOCK/BMAL1-SIRT1 feedback loop should be intimately involved in regulation of CLOCK/BMAL1-PER/CRY feedback loop to orchestrate the normal circadian clock. Furthermore, although the decrease of NAD+ level had no significant contribution on insulin sensitivity in this study, NAMPT, the rate-limiting enzyme in mammalian NAD+ biosynthesis, affects SIRT1 activity[46] and is regulated by CLOCK/BMAL1-SIRT1.[34, 46] All these clues suggest that the CLOCK/BMAL1-SIRT1 feedback loop is a core connection between the circadian clock and metabolism.

SIRT1 has been reported to regulate insulin sensitivity.[22, 47] Here, we demonstrate that SIRT1 is required for the normal circadian change of hepatic insulin sensitivity and mediates CLOCK/BMAL1-regulated hepatic insulin sensitivity. Resveratrol, a natural polyphenol, can induce activation of SIRT1 in mice and has salutary effects on IR and other metabolic disorders.[20, 21] Similarly, here, we show that resveratrol supplementation, no matter whether at a relatively high or low dose, all markedly improves insulin sensitivity and glucose/lipid metabolism in mice under a constant dark condition. Recently, resveratrol has also been shown to improve insulin sensitivity in obesity or type 2 diabetes mellitus patients,[23, 24] supporting the possibility for using resveratrol to treat metabolic disorders induced by circadian misalignment.

Our findings shed new light on coordination of the circadian clock and metabolism by circadian regulation of hepatic insulin sensitivity by CLOCK/BMAL1-dependent SIRT1 expression and offer a potential drug target, SIRT1, for combating metabolic disorders under circadian misaligment conditions.

Acknowledgment

The authors thank Dr. Michael McBurney for providing Sirt1+/− mice. The authors also thank Dr. Toren Finkel for providing Sirt1 promoter luciferase reporters and Dr. Kazuhiro Yagita for CLOCK- and BMAL1-expressing plasmids.

Ancillary