Palmitate Inhibits SIRT1-Dependent BMAL1/CLOCK Interaction and Disrupts Circadian Gene Oscillations in Hepatocytes

Adenoviruses were generated using pAdViraPower system (Life Technologies) for Ad-Clock, pAdEasy system (Agilent) for Ad-Bmal1. Over-expression of BMAL1 or CLOCK in primary hepatocytes was achieved by transducing primary hepatocytes with the concentrated adenovirus at 1x10⁹ pfu/mL. SIRT1 inhibitor EX527 was purchased from Selleckchem. Resveratrol and CAY10591 were from Cayman. FK866 was from Sigma.

Mouse hepatoma Hepa1c1c-7 (Hepa1) cells were used in the synchronization study [45]. Confluent cells were serum-shocked (50% horse serum) for 2 hr and then treated with either fatty acid-free bovine serum albumin (BSA) or palmitate (50 μM) in 0.5% BSA. Cells then were collected for both mRNA and protein analysis at 4-hr intervals between 24 hr and 48 hr time points.

Primary mouse hepatocytes were isolated from 8–10 wk old wild type C57BL6 mice using a two-step collagenase digestion with 100 U/mL collagenase in HBSS at pH 7.4. After dissection, the liver was placed in DMEM and carefully pulled apart to release hepatocytes. Hepatocytes in DMEM were passed through a 100 μM cell strainer and then spun at 50 X g for 1 min. The pellet was resuspended in DMEM and then spun at 50 X g for 10 min in a Percoll gradient to remove dead hepatocytes. After washing with DMEM at 50 X g for 10 min and checking by Trypan Blue staining, hepatocytes were cultured on collagen-coated plates. All animal experiments were approved by the Institutional Animal Care and Research Advisory Committee at the University of Michigan Medical School.

Cell pellets were lysed in ice-cold RIPA (50 mM Tris, pH 8.0, 150 mM NaCl, 2 mM EDTA, 0.5% NP-40, 0.1% SDS, 0.5% sodium deoxycholate) buffer supplemented with 1x protease inhibitor and 50 mM NaF and incubated on ice for 20 min. Protein lysates were cleared by centrifugation at 14,000 rpm at 4°C for 10 min. Supernatants were collected and quantified using BioRad protein assay kit. For isolation of clock proteins in nuclear fractions, cells or tissues were first exposed to hypotonic buffer (10 mM Tris-HCl, pH 8.0, 10 mM NaCl, 3 mM MgCl₂, 0.2% NP-40, and 300 mM Sucrose) and cytosolic fractions were separated by low-speed centrifugation. Nuclear pellets were then re-suspended in RIPA buffer and centrifuged at 14,000 rpm at 4°C to obtain nuclear fractions. Blots were probed with the following primary antibodies: BMAL1, CLOCK AKT, p38, SIRT1, p53 (Santa Cruz Biotechnology); β-tubulin (Sigma); AKT-phospho (Cell Signaling, AKT phosphorylation at Ser473), p38-phospho (Cell Signaling, p38 dual phosphorylation at Thr180/Tyr182), and p53-Acetyl (Cell signaling, Lys382)

The standard immunoprecipitation method was described previously (65). For detecting the protein interaction between BMAL1 and CLOCK, cells were harvested after hr treatment with either BSA or palmitate following 2 hr of serum shock. Cell lysates were then incubated with specific antibodies (rabbit IgG control or anti-BMAL1) overnight at 4°C. The immune-complexes were captured by adding 30 μL of Protein A-Sepharose beads and incubating at 4°C for 1 hr. Beads were washed 5 times in RIPA buffer and eluted in 30 μl 2X SDS loading buffer. Western blotting was performed to detect the presence of targeted proteins with specific antibodies. For affinity purification, cells were lysed in lysis buffer and 1 mg of protein lysate incubated with 20 μL of Streptavidin magnetic beads (GE Heathcare) for 6 hr. The beads were then washed with lysis buffer for 5 times and eluted in 20 μL of 2X SDS loading buffer.

Total cellular RNA extraction was performed with TRIzol reagent (Invitrogen, Carlsbad, CA) and chloroform. cDNA was synthesized using Verso cDNA kit (Thermo Fisher Scientific, Surrey, UK) and subjected to qPCR using Absolute Blue SYBR Green ROX Mix (Thermo Fisher Scientific, Surrey, UK) on an ABI 7900 HT thermal cycler (Applied Biosystems, Foster City, CA). The value of each cDNA was calculated using the ΔΔCt method and normalized to the value of housekeeping gene control (18 S RNA). Data were plotted as fold change. The primer sequences for mouse clock and control genes were reported previously [46]. The primers for other genes for this study are listed in Table 1. These qPCR primers were designed to span exon-exon junctions by using the NCBI on-line primer design software. The selected primer shows no homology to other sequences in the mouse genome at least at last 3 nucleotides at 3’ end. The amplification efficiency of primer set were monitored by delta CT method to make sure at least 2 cycle difference in a series of diluted samples. The PCR products were verified to be the right size by gel electrophoresis.

Cells were plated in a 24-well plate overnight before Lipofectamine 2000 (Invitrogen)-mediated transfection with the mPer2 promoter-driven luciferase reporter alongside expression vectors for BMAL1 and CLOCK. 24 hr post transfection, cells were synchronized with 50% horse serum for 2 hr and switched back to serum-free medium supplemented with either BSA or palmitate. 24 hr later, cells were lysed for luciferase activity assay measurement on a BioTek Synergy 2 microplate reader. β–galactosidase construct was also co-transfected in each well for normalizing luciferase activity.

Student’s t-test was used to determine the significance of difference between two groups with p value < 0.05. One-way ANOVA (Prism software) along with post-hoc Tukey’s test was used to test the difference between more than two groups. For cosinor analysis of oscillations of circadian genes, JTK software was used to generate values of amplitude and p-values [47].

Several studies have already shown high-fat diet (HFD) feeding impairs circadian rhythmicity [48, 49] and dampens chromatin recruitment of BMAL1-CLOCK transcriptional complex [33]. We also observed that an 8-wk-HFD is sufficient to dampen diurnal oscillations of core clock genes in the liver and the spleen (data not shown). However, it remains unclear about whether disruption in these peripheral clocks is a primary response to HFD or secondary to obesity and insulin resistance following HFD feeding. It has been well documented that HFD causes elevated levels of serum free saturated fatty acids [50]. Among these fatty acids, the long-chain (C16:0) fatty acid palmitate is particularly detrimental to hepatocytes by inducing lipotoxicity such as insulin resistance, oxidative stress, and cell death at high dose and for prolonged treatment [11, 51, 52]. However, its direct effects on the molecular clock in hepatocytes have not been examined.

We first determined the appropriate concentration of palmitate for our in vitro circadian study because high-dose of palmitate (over 400 μM) triggers cell apoptosis [53]. As insulin resistance is one of the hallmarks of palmitate-induced lipotoxicity, we tested a wide range of palmitate dosages for its effects on insulin-induced AKT phosphorylation (AKT-P at Ser473) (data not shown). In freshly isolated primary mouse hepatocytes (PMHs), we observed that treatment of palmitate at as low as 50 μM overnight is sufficient to reduce AKT-p^S473 without decreasing the total AKT expression compared with BSA (Fig 1A). In addition, palmitate treatment at this concentration does not cause a significant decrease in cell viability, determined by dimethylthiazolyl diphenyltetrazolium bromide (MTT) assay (data not shown).

To test whether low dose of palmitate (50 ~ 100 μM) affects the molecular clock function in hepatocytes, we examined the effect of palmitate treatment on the activity of a mouse Per2 promoter-driven luciferase reporter (Per2-luc) in a mouse hepatoma Hepa1c1c-7 (Hepa1) cell line. Per2-luc shows a low basal activity but is potently induced by adenovirus-mediated overexpression of both BMAL1 and CLOCK, the key transcription complex of the molecular clock in hepatocytes (Fig 1B). However, addition of palmitate significantly reduces Per2-luc in the presence or absence of Bmal1 and Clock over-expression. Consistent with the Per2-luc reporter assay, palmitate treatment for 24 hr significantly reduces the mRNA levels of two classical circadian genes Dbp and Per2 in PMHs (Fig 1C). However, the expression of Bmal1 and Clock is not affected by palmitate. Thus, we showed evidence that low dose palmitate treatment can suppress the Per2-luc activity induced by the BMAL1-CLOCK complex and reduce the expression of a subset clock output genes in mouse hepatocytes.

PMHs, once isolated, maintain their hepatocyte morphology for only a short period of time, making them a less ideal model for studying circadian oscillations in a longer duration. To determine the effect of palmitate on the oscillations of key clock genes, we turned to Hepa1 cells, a previously reported in vitro hepatocyte model for studying the molecular clock in our lab [45]. After serum shock, Hepa1 cells were exposed to either BSA or palmitate for 48 hr. The mRNA oscillations of Bmal1, Clock, Cry1, Cry2, Dbp, Per2, Nr1d1, and Pgc-1α were measured between 24 hr and 48 hr post serum shock. During the entire circadian cycle, Dbp, Per2, and Nr1d1 mRNA levels show robust oscillations with about 5 to 8-fold difference between CT (Circadian Time) 32 and CT 44 in the BSA control cells. However, in the presence of palmitate, oscillations of clock genes including Dbp, Nr1d1 and Per2 are severely dampened (Fig 2A–2C & Table 2). Palmitate treatment also reduces oscillations of Pgc-1α, a key metabolic regulator known to be a target of the molecular clock [54] (Fig 2D). Meanwhile, palmitate does not seem to dampen the cyclic expression of Bmal1, Cry1 and Cry2 (Fig 2E and 2F). Taken together, our results demonstrated a potent but selective inhibition of oscillations of a panel of clock genes and clock-controlled genes by palmitate in synchronized hepatocytes.

Functioning as a transcriptional-translational feedback loop, the molecular clock is resistant to perturbations by adjusting the expression or activity of its components [55]. Since palmitate treatment does not seem to impact the Bmal1 mRNA expression in PMHs (Fig 1B), we suspect that palmitate might alter the BMAL1 and CLOCK protein expression and subsequently suppress their transcriptional function. To our surprise, increasing dosage of palmitate treatment fails to affect the protein level of CLOCK but slightly increases overall levels of BMAL1 protein (about 1.2 fold) (Fig 3A). Furthermore, in the presence of palmitate treatment, both of proteins are still predominantly localized inside the nucleus (Fig 3B), where these two proteins form a transcriptional complex to activate their targets. In the meanwhile, we detected a robust p38-P in acute palmitate-treated hepatocytes, a positive response to acute palmitate treatment [51]. Next, we tested whether chronic palmitate exposure affects protein oscillations of BMAL1 and CLOCK in synchronized Hepa1 cells. As shown in Fig 3C, levels of CLOCK protein remain constant, while BMAL1 protein displays robust oscillations in BSA-treated synchronized Hepa1 cells. However, palmitate treatment renders BMAL1 protein level constant throughout the cycle without showing effects on CLOCK. Meanwhile, chronic palmitate treatment modestly elevates p38 phosphorylation but greatly dampens AKT phosphorylation in those cells, consistent with recent findings that impairment of the hepatic circadian clock reduces AKT phosphorylation [56, 57]. Our results indicate that palmitate represses the hepatic molecular clock without reducing the overall protein abundance of nuclear BMAL1 and CLOCK in hepatocytes. Rather, we observed that palmitate blunts the oscillations of BMAL1 protein in synchronized hepatocytes.

The BMAL1:CLOCK protein complex is the major driving force for the oscillations of Per1 and Dbp [34–36]. So far, we have shown that palmitate suppresses BMAL1-CLOCK transcriptional activity without altering the abundance or localization of BMAL1 and CLOCK. This prompted us to examine whether palmitate may influence the BMAL1:CLOCK complex formation in hepatocytes. To this end, we treated Hepa1 cells with palmitate in a dose- and time-dependent manner after co-transfection with CBP (calmodulin-binding protein)- and SBP (streptavidin-binding protein)-tagged Bmal1 and Clock-Flag. We then performed immune-precipitation with streptavidin beads to capture CBP-SBP-tagged BMAL1 and its interacting proteins. As shown in Fig 4A and 4B, palmitate treatment as low as 100 μM or as short as 2 hr is sufficient to abolish BMAL1-CLOCK interaction. To confirm this finding at the endogenous level, we performed another immunoprecipitation assay with anti-BMAL1 antibody in Hepa1 cells following 24-hr palmitate treatment. Consistent with over-expression conditions, protein interaction between the endogenous BMAL1 and CLOCK is largely abolished in palmitate-treated Hepa1 cells even though the protein levels for both proteins are comparable in inputs (Fig 4C). All these results suggest that the dampened clock gene expression during palmitate treatment may be mediated through a loss of BMAL1:CLOCK interaction and subsequent inactivation of BMAL1:CLOCK-dependent transcription.

SIRT1 has been shown to be a regulator of BMAL1: CLOCK transcriptional activity [41, 58]. SIRT1 expression or activity has been shown to be reduced or inhibited in obesity or after palmitate exposure [59–61]. To test whether SIRT1 might be involved in palmitate-induced repression of BMAL1-CLOCK interaction, we overexpressed Sirt1 in 293T cells in which we detected a very low level of the endogenous SIRT1. As shown in Fig 5A, streptavidin beads are unable to pull down CLOCK-FLAG in cells cotransfected with both CBP-SBP-Bmal1 and Clock-Flag. However, significant amount of CLOCK-FLAG is present in immuno-precipitated CBP-SBP-Bmal1 in the presence of Sirt1 overexpression. We next used SIRT1-specific inhibitor EX527 [61–63] to test how inhibition of SIRT1 activity affects BMAL1-CLOCK interaction in PMH hepatocytes transduced with Ad-Bmal1-Flag. While elevated levels of acetylated p53 after SIRT1 suppression by EX527 is consistent with the literature [63], a lower amount of the endogenous CLOCK is shown to interact with FLAG-tagged BMAL1 after immuno-precipitation with anti-FLAG (Fig 5B). Next, we used FK866, an NAD biosynthesis inhibitor [64] to further test how blocking SIRT1 affects the clock protein interaction and activity in comparison with palmitate treatment. In PMH cells transduced with Ad-Bmal1-Flag, both palmitate and FK866 are able to disrupt the interaction between the endogenous CLOCK and BMAL1-FLAG (Fig 5C). Consistently, treatment with either palmitate (200 μM) or FK866 (500 nM) for 6 hr blocks the Per2-luc activation in Hepa1 cells transfected with both Bmal1 and Clock overexpression constructs (Fig 5D). Thus, we generated evidence that SIRT1 suppression has similar effects on the circadian clock as palmitate treatment, in support of the notion that SIRT1 might be targeted by palmitate to inhibit the clock function in hepatocytes.

So far, we have shown that palmitate attenuates the molecular clock function in hepatocytes likely via disrupting the BMAL1:CLOCK complex formation. We also showed that in hepatocytes SIRT1 plays a critical role in enhancing BMAL1:CLOCK protein-protein interaction and inhibition of SIRT1 mimics the palmitate suppression of BMAL1:CLOCK interaction and function. Conceivably, SIRT1 activation might counteract the effects of palmitate on BMAL1:CLOCK protein-protein interaction. To test this possibility, we used two different SIRT1 activators, CAY10591 [65, 66] and resveratrol [67, 68]. To validate their effects on SIRT1 activation, we measured the levels of acetylated p53 (p53-Ac), a direct intracellular SIRT1 substrate, in hepatocytes. Indeed, Inhibition of SIRT1 by EX527 elevates the level of p53-Ac, whereas co-treatment with either CAY10591 or resveratrol reduces p53-Ac levels to the basal (Fig 6A). Specifically, we examined the effects of treatment with either SIRT1 activator on the BMAL1-CLOCK complex during palmitate treatment. In PMH cells transduced with Ad-Bmal1-Flag, palmitate treatment alone consistently reduces BMAL1-FLAG interaction with the endogenous CLOCK protein, whereas either CAY10591 or resveratrol treatment restores interaction of both proteins (Fig 6B). We next tested whether both activators could reverse the palmitate effects on the BMAL1:CLOCK-mediated transcriptional activity. Activation of SIRT1 with either CAY10591 or resveratrol following palmitate treatment reverses suppression of the Per2-luc activity driven by overexpression of both BMAL1 and CLOCK in Hepa1 (Fig 6C). Furthermore, CAY10591 and resveratrol block suppression of the endogenous Dbp and Per2 expression in palmitated-treated Hepa1 cells (Fig 6D and 6E). Take all together, our data showed that pharmacological activation of SIRT1 mitigates the negative impact of palmitate on the molecular clock in hepatocytes.