Dietary Conjugated Linoleic Acid Modulates the Hepatic Circadian Clock Program via PPARα/REV-ERBα-Mediated Chromatin Modification in Mice

Scope: Disruptions of circadian rhythm cause metabolic disorders and are closely related to dietary factors. In this study, we investigated the interplays between the dietary conjugated linoleic acid (CLA)-induced hepatic steatosis and the circadian clock regulation, in association with lipid homeostasis. Methods and Results: Exposure of mice to 1.5% dietary CLA for 28 days caused insulin resistance, enlarged livers, caused hepatic steatosis, and increased triglyceride levels. Transcriptional profiling showed that hepatic circadian clock genes were significantly downregulated with increased expression of the negative transcription factor, REV-ERBα. We uncovered that the nuclear receptor (NR) PPARα, as a major target of dietary CLA, drives REV-ERBα expression via its binding to key genes of the circadian clock, including Cry1 and Clock, and the recruitment of histone marks and cofactors. The PPARα or REV-ERBα inhibition blocked the physical connection of this NR pair, reduced the cobinding of PPARα and REV-ERBα to the genomic DNA response element, and abolished histone modifications in the CLA-hepatocytes. In addition, we demonstrated that CLA promotes PPARα driving REV-ERBα transcriptional activity by directly binding to the PPAR response element (PPRE) at the Nr1d1 gene. Conclusions: Our results add a layer to the understanding of the peripheral clock feedback loop, which involves the PPARα-REV-ERBα, and provide guidance for nutrients optimization in circadian physiology.


INTRODUCTION
The circadian clock is a time program created endogenously in all organisms. It is controlled by genetically encoded molecular components, which generate cyclic changes in activities during a period of about a day (1,2). A wide array of metabolic processes displays circadian rhythms, which are linked with sleep-wake cycles, feeding behavior, and nutrient availability (3). Disruptions of the circadian clock may cause metabolic diseases in humans and animals. For example, the incidence rates of obesity and cancer are high in long-shift workers (4,5). The Clock mutant mice are described to develop a metabolic syndrome of hyperlipidemia, hepatic steatosis, and hypoinsulinemia (6).
At the heart of the molecular circadian clock lies a transcription-translation feedback loop (TTFL), containing the key proteins CLOCK, BMAL1 (Arntl1), PERIOD (PER homologs 1-3), and cryptochrome (CRY homologs 1 and 2) (3, 7). The CLOCK-BMAL1 heterodimer is responsible for activating the transcription of target genes in their promoter and/or enhancer regions, including the genes encoding PER and CRY (8,9), whereas the PER and CRY act as repressors to antagonize the CLOCK-BMAL1 through posttranslational inhibition (10)(11)(12). Alternatively, REV-ERBα and REV-ERBβ mediate transcriptional repression of the clock genes in the negative feedback loop (13,14). The nuclear receptors (NRs), REV-ERBs are ligand-regulated transcription factors (15) that link circadian rhythms with metabolism. REV-ERBα has been shown to regulate the expression of Arntl1 and affects the lipid signaling protein, apolipoprotein CIII in the liver cells (10). Dual depletion of REV-ERBα/β function by the creation of doubleknockout mice (DKOs) results in multiple phenotypes including altered circadian wheel-running behavior and lipid dysregulation (13). The severity of the circadian disruption was comparable to that was seen in Bmal1 −/− mice (16). Bmal1 deficiency also reprogram lipid metabolism in the livers of mice (17). Finally, the TTFL is also involved in chromatin modifications of the central circadian oscillators in mammals and plants (18). Dynamic histone modifications can change the chromatin structure of genes, and affect the accessibility of clock regulators and/or RNA polymerase II (Pol-II); thus, promote the rhythmic expression of target genes (19,20). However, it is still unclear which diverse manners of the molecular clocks are functionally integrated and exhibit distinct tissue-specific transcription regulations (3).
Reciprocally, changes in metabolism can regulate circadian physiology (21). The circadian clock responds to zeitgebers (the so-called time givers) differently (22), depending on both the strength of the stimuli and the circadian phase during which the zeitgebers are applied. In particular, the nutritional challenges, Abbreviations: ChIP, chromatin immunoprecipitation; CT, circadian time; CLA, conjugated linoleic acid; CRY, cryptochrome; DKOs, double-knockout mice; FA, fatty acid; GSEA, gene set enrichment analysis; GTT, glucose tolerance test; H&E, hematoxylin and eosin; HDAC3, histone deacetylase 3; i. p., intraperitoneal; ITT, insulin tolerance test; mRNA, messenger RNA; NAFLD, nonalcoholic fatty liver disease; NRs, nuclear receptors; NCOR1, nuclear receptor corepressor 1; PER, PERIOD; RNA polymerase II (Pol-II); siRNAs, Small interfering RNAs TTFL, transcription-translation feedback loop; WAT, white adipose tissue; ZT, zeitgeber time. such as high-fat feeding can serve as a strong zeitgeber for the liver, which results in clock reprogramming in mice (21). On administering a high protein/low carbohydrate diet to mice, the expression of BMAL1 and Cry1 were seen to be increased in both the livers and the kidneys (23). Interestingly, it is reported that conjugated linoleic acid (CLA) addition contributes to the alteration of the circadian clock gene expressions in Caco-2 cells in vitro, in association with the increased expression of miR-107, while the in vivo effects are unknown (24). The CLA is a natural polyunsaturated fatty acid (FA), derived from meat and dairy products (25). The consumption of CLA is related to the reduction of body weight in humans, thereby becomes popular (26). In the western world, the estimated CLA intake ranges from 90 mg per day in the US to 1,500 mg per day in Australia. However, the suggested beneficial intake of CLA is about 3-6 g per day (26,27). Notably, studies using experimental animals often employ higher doses of CLA than in humans, and produce a considerable variation, even severe side effects including lipoatrophy in rodents (25,26). Therefore, from the perspective of health science, it is important to better understand the action mechanism of the CLA.
In this study, we hypothesized that a high dose of CLA supplementation may cause changes in the circadian clock program and the associated lipid metabolism in mice. We found that dietary CLA induced hepatic steatosis and insulin resistance in mice during a period of 28 days, and downregulated the clock gene expressions in the liver. By targeting the associated NRs with small-molecule drugs, we aimed to investigate the nature of interactions between the CLA and the molecular clock in animal physiology.

Animals and Experimental Design
All experiments involving animals were reviewed and approved by the Animal Care and Use Committee of Yangzhou University (YZUDWSY 2017-09-06). Wild-type male Kunming mice at 5 weeks old were purchased from the Jiangsu Laboratory Animals Science Center, housed at 22 • C with 50% of humidity on a 12/12-h light/dark cycle and were allowed free access to water and food. Mice were fed a normal chow diet for 1 week for acclimation and were randomized into two groups (n = 6): the control group/vehicle received a chow diet with 1.5% (w/w) linoleic acid supplementation, whereas the CLA group received a chow diet with CLA supplementation (1.5% w/w). For in vivo studies, the CLA used was a mixture of an equal amount of c9t11 and t10c12 (26). The experiment lasted for 28 days, during which the body weight and food intake were recorded. At the end of the experiment, the animals were euthanized with anesthesia followed by pelltobarbitalum natricum (GENIA Biotech, Beijing, China) (80 mg kg −1 ) treatment. The sampling time was chosen at zeitgeber time (ZT) 8 (8 h after lights are turned on) according to previous descriptions (28). Whole blood was collected via cardiac draw with a capillary tube (100 µl, coated with K3 EDTA, Sarstedt, Nümbrecht, Germany), and plasma was prepared by centrifugation. Together with the liver tissues, the samples were taken and were snap-frozen in liquid nitrogen, thereafter stored at −80 • C until further analysis.
For the circadian rhythm measurement, a total of 72 mice in two groups were fed as described above for 28 days. The livers were collected from control (n = 6) or CLA (n = 6) mice at ZT0, ZT4, ZT8, ZT12, ZT16, and ZT20 for 24 h. The liver tissues were snap-frozen in liquid nitrogen for further analysis.
Lipid, Blood Glucose, Insulin Level Analyses, and Insulin Tolerance Test/Glucose Tolerance Test Liver triglyceride levels were determined on a Roche Integra 400 Plus analyzer (Roche Diagnostics, Basel, Switzerland). Plasma insulin and glucose levels were measured using a 125Ilabeled insulin radioimmunoassay kit (Beijing North Institute of Biological Technology, Beijing, China) and OneTouch Ultra Easy (LifeScan, Shanghai, China) following the instructions of the manufacturers, respectively. For insulin tolerance test/glucose tolerance test (ITT/GTT), the mice had been fasted for 6 h and then gave an intraperitoneal (i.p.) injection of insulin (Novolin R; Novo Nordisk, Bagsvaerd, Denmark) or glucose at a dose of 1 unit kg −1 or 2 g kg −1 body weight. The measurement of blood glucose was conducted by tail bleeding at 0, 30, 60, 90, and 120 min, using the One Touch Ultra Easy.

Histology and Oil Red O Staining
Five µm thick sections from the liver tissues were stained with hematoxylin and eosin (H&E) and were visualized using a light microscope (Leica DFC 420C; Leica Microsystems, Wetzlar, Germany) at x20 magnification. In parallel, sections were fixed with 10% formalin. After fixation, 60% of isopropanol was added for 30 s and was subsequently removed. Oil Red O working solution (Sunshinebio, Nanjing, China) was then applied for 1 h. After the removal of the Oil Red O working solution, 100% of isopropanol was added to extract Oil Red O and was finally visualized. Duplicate sections of each mouse were used for histological analysis.

Primary Murine Hepatocytes Isolation and Cell Culture
Hepatocytes were isolated from the control and the CLA-treated mice by liver perfusion using a two-step method with collagenase as described previously (29). Peripheral blood and cells were flushed from the liver in Hank's balanced salt solution followed by perfusion using the collagenase digestion solution. Thereafter, the liver was dissected and mechanically dissociated. The acquired cell suspension was filtered through a 100 µm-cell mesh followed by centrifugation at 50 x g for 5 min at 4 • C to obtain hepatocytes and were maintained in OptiCulture hepatocyte media (Sekisui XenoTech LLC, KS, USA).

Real-Time Quantitative Reverse Transcription PCR and Western Blotting Analysis
Total RNA was isolated from mice livers or hepatocytes. The complementary DNA (cDNA) was prepared, amplified, and measured in the presence of SYBR Green as previously described (30). The fluorescent values were collected and a melting curve analysis was performed. Gapdh was used as the internal reference to normalize the relative level of each transcript. The primers used are shown in Supplementary Table 1. Hepatocytes were analyzed by immunoblotting with PPARα (Millipore, MA, USA; MAB3890), REV-ERBα (cell signaling, #13418), and Bmal1 (cell signaling, #14020).

RNA-Sequencing Analysis
The liver tissue RNA was prepared for RNA-sequencing (RNAseq) analysis using Illumina Tru-Seq RNA Sample Prep Kit (Illumina Inc., CA, USA), according to the instructions of the manufacturer. Libraries were constructed and validated with an Agilent Bioanalyzer (Agilent Technologies, Palo Alto, CA, USA). Sequencing was performed on an Illumina HiSeq 2000 sequencer at BGI Tech (Hong Kong, China). The FASTQ-formatted sequence data were obtained and analyzed using a standard BWA-Bowtie-Cufflinks workflow. Sequence reads were mapped to mm9 assembly with BWA and Bowtie software. The Cufflinks package was used for transcripts assembly, quantification of normalized gene and isoform expression, and the analysis of different expressions. Gene set enrichment analysis (GSEA v.3.0) was applied to rank the genes based on the shrunken limma log2 fold-changes. The GSEA tool was used in a "pre-ranked" model with default parameters.

ChIP-Re-ChIP Assays
Chromatin immunoprecipitation was performed as described previously (33) with the following modifications. The crude chromatin solutions of the CLA-treated hepatocytes were first cleared with protein A beads (Invitrogen, MA, USA) that had been precoated with preimmune serum for 2 h at 4 • C. Then, the precleared chromatin solutions were incubated at 4 • C overnight with the other antibodies listed in the Supplementary Material, prior to precipitation with protein A beads that had been preblocked with bovine serum albumin and sonicated salmon sperm DNA. For re-ChIP, the immunoprecipitated complexes from the first ChIP were eluted with 20 mM dithiothreitol at 37 • C for 30 min with brief vortexing, diluted with the ChIP buffer for 50 times, and cleared by centrifugation, before they were incubated with antibodies for the secondary ChIP overnight at 4 • C. The immunoprecipitated DNA was analyzed by realtime PCR. Enrichment of genomic DNA was presented as the percent recovery relative to the input. The primers are listed in Supplementary Table 1.

Reporter Constructs and Reporter-Gene Assay
Transient transfection and reporter-gene assays were performed according to our previous study (30). For reporter-gene assays, the oligonucleotides encompassing four copies of PPAR element (PPRE) motifs found in the murine Nr1d1 promoter and enhancer were inserted into the tk-luciferase reporter vector. The mutant form, tk-Nr1d1-promoter-mu contains sequences mutated from TCAGGGTGAGGTCA to TCAGGGTGCAAGGA. The mutant form, tk-Nr1d1-enhancermu contains sequences mutated from GGATTAACTAGGTCA to GGATTAACTCAAGGA. Cells (HepG2) were cotransfected with CMX-mPPARα with wild-type or mutant forms of Nr1d1 promoter and enhancer reporter constructs. The Renilla plasmid was co-transfected for normalization. After 12 h of incubation, the cells were treated with vehicle or CLA or cotreated with compounds as indicated for another 24 h. The luciferase was then analyzed with a Dual-Luciferase Assay system (Promega, WI, USA) on a luminometer according to the instructions of the manufacturer. All transfections were performed in sextuplicate, and each experiment was repeated at least three times.

Statistical Analysis
Statistical analysis was performed with GraphPad Prism software 8.0. The data are presented as mean ± SD/SEM from at least three independent experiments. Statistical analysis was conducted using two-tailed Student's t-tests or ANOVA with Tukey's post hoc test to compare the means. The value of p < 0.05 was considered significant.

Dietary Conjugated Linoleic Acid Induces Nonalcoholic Fatty Liver Disease in Mice
With 1.5% dietary conjugated linoleic acid (CLA) supplementation for 28 days, the changes in animal physiology and lipid metabolism were investigated. This CLA treatment did not affect the body weight or food intake of the mice during the whole period of the experiment compared to the control group (Figures 1A,B). In contrast, the insulin tolerance test/glucose tolerance test (ITT/GTT) showed impaired insulin resistance in the CLA-exposed mice compared to the control (Figure 1C), in parallel with the elevated blood insulin levels ( Figure 1D). Meanwhile, the CLA treatment dramatically increased the liver weight in mice compared to those of the control group ( Figure 1E), along with a significant increase of triglycerides content in the mice fed with CLA ( Figure 1F). Histological analysis and Oil Red O staining confirmed the hepatic lipid accumulation in the CLA-treated mice, but not in the control group (Figures 1G,H). These results indicated that dietary CLA supplementation impairs animal body fitness, characterized by a heavier liver, hepatic steatosis, and insulin resistance.

The Hepatic Circadian Clock Gene Program Is Disrupted in the CLA-Treated Mice
Transcriptional profiling of livers was performed to provide the molecular understanding of the CLA-induced nonalcoholic fatty liver disease (NAFLD) in mice. Intriguingly, gene set enrichment analysis (GSEA) revealed that the hallmarks of the circadian clock gene program were strongly downregulated, whereas the fatty acid metabolism pathway was significantly elevated by the CLA treatment compared to the control group (Figures 2A-C). The pathway-focused analysis confirmed that the genes of the circadian program were significantly downregulated in the livers of the CLA-treated mice ( Figure 2D). Furthermore, quantitative real-time PCR (qRT-PCR) analysis verified that the key Clock genes including Per1, Per2, Clock, Arntl, and Cry1 were downregulated in the livers of the CLA-treated mice compared to the control group ( Figure 2E). To further examine whether the CLA-treatment impairs the liver circadian clock genes rhythmicity, we studied the temporal gene expression in the livers of mice, collected at each indicated zeitgeber time (ZT) (sampled every 4 h) under a 12-h light/dark cycle. The majority of the measured hepatic circadian clock genes lost their rhythmicity in the CLA-treated mice compared to the control, including NR1D1, Clock, Cry1, Arntl, and Per2 ( Figure 2F). In contrast, the primary negative transcription factor of the circadian clock NR REV-ERBα (Nr1d1) was significantly upregulated at both the transcription and translation levels (Figures 2G-I). In Immunoblotting of REV-ERBα, PPARα, and Bmal1 protein expression normalized to GAPDH levels and quantification of relative protein density. The data are shown as the means ± SEM, n = 6 per group, *p < 0.05, ***p < 0.001, using two-tailed student's t-test.
association with the downregulated Arntl transcripts, the protein level of Bmal1 was dramatically reduced in the CLA group. We also examined the expression of another NR PPARα, which is a master lipid sensor in the liver. However, there was no difference in the PPARα expression of mRNA or protein between the treatments (Figures 2G-I).

REV-ERBα and PPARα Drive Circadian Gene Transcription in the Hepatocytes From the CLA-Treated Mice
The action of the transcription factors on the circadian gene expression was further investigated. Small interfering RNAs (siRNAs) were used to knockdown REV-ERBα and PPARα in the hepatocytes derived from the CLA-treated mice, respectively ( Figure 3). As expected, REV-ERBα knockdown by siREV-ERBα-1/-2 showed that the messenger RNA (mRNA) levels of the key circadian genes Per1, Per2, Crem, Clock, Arntl, Bhlhe41, and Cry1 were restored to that of the control group, compared to the CLAalone treatment (Figures 3A,B). Interestingly, the downregulated expression of the circadian genes was also switched to the control levels when PPARα was silenced (Figures 3C,D). Similarly, the inhibition of PPARα or REV-ERBα using small-molecule antagonists showed a comparable capacity to increase the key circadian genes expression in the hepatocytes from the CLA-treated mice, as in the siRNA approach ( Figure 3E). This was linked to the alteration of lipid output. Thus, the antagonists significantly decreased the CLA-induced high cellular triglycerides levels in hepatocytes ( Figure 3F). Together, these results indicate that both REV-ERBα and PPARα have vital functions in driving the circadian gene transcription in response to CLA, even though the endogenous expression of PPARα was unaltered.

REV-ERBα Recruits Histone Repressors on the Circadian Genes in the CLA-Treated Mice
Next, bioinformatic analysis of a previous PPARα and REV-ERBα dataset was performed (34). The key transcription factor, REV-ERBα, was observed to directly bind to the target loci of major circadian genes Clock, Cry1, Per1, and Arntl. Surprisingly, we found that PPARα also binds to the same loci of the abovementioned four genes in the mice. The binding enrichments were decreased, when the REV-ERBα DNA binding element was mutated or PPARα was knocked out (Figure 4A). Using the genome-wide analysis and ChIP-qPCR analysis, we confirmed the binding sites of REV-ERBα and PPARα on the gene Clock and Cry1. The CLA supplementation strongly increased the enrichment of REV-ERBα on both the gene promoters (Figure 4B), and enhanced the PPARα binding on the Clock gene ( Figure 4C, upper panel) but not on the Cry1 locus ( Figure 4C, lower panel). Subsequently, the repressors of histone deacetylase 3 (HDAC3) and nuclear receptor corepressor 1 (NCOR1) also exhibited higher enrichment at the promoters of Clock and Cry1, respectively, whereas the occupancy of the coactivator histone acetylase p300 was decreased ( Figure 4D). In addition, the transcriptional activation-linked histone marks H3K27ac, H3K4me2, and the RNA polymerase II (Pol-II) recruitment to the target genes were significantly reduced by the CLA treatment. Therefore, our results suggest that CLA enhances the REV-ERBα recruitment and transcriptional repression of the circadian gene program, which results in the inhibition of their transcriptional responses.

PPARα Dominates REV-ERBα-Mediated Transcription Repression of the Circadian Genes
To gain a deeper understanding of the interactions between the PPARα and REV-ERBα at the circadian gene promoter, ChIP-re-ChIP was performed in hepatocytes from the CLAtreated mice. The PPARα-chromatin complexes were enriched by the first ChIP with anti-PPARα antibody and were subject to the second ChIP. A strong REV-ERBα association was then detected by the REV-ERBα-specific antibody, but not by the control IgG at the Clock and Cry1 gene promoters, respectively ( Figure 5A). Moreover, both the PPARα inhibitor, GW6471 and the REV-ERBα inhibitor, SR8278 were able to reduce the associated enrichments, which indicate that PPARα binds concurrently with REV-ERBα to the circadian genes in the CLA-derived hepatocytes. Importantly, we have provided the proof that PPARα may directly control the REV-ERBα transcription activity by the inhibition of PPARα using the siRNAs approach (Figure 5B), and small-molecule antagonists ( Figure 5C). Indeed, the classic PPARα binding motif was found at the promoter and the enhancer of the REV-ERBα gene (Nr1d1) in a ChIP-seq genome-wide browser analysis ( Figure 5D). To verify the function of this putative PPAR response element (PPRE)-containing site, we performed reporter gene assays with the wild-type or mutated PPRE of Nr1d1 gene enhancer and promoter. Cell-based reporter assays with constructs including PPRE confirmed the transactivation. The CLA addition was further induced and the antagonists combined with the CLA suppressed the luciferase expression, which was initially higher in the CLA-alone group relative to the basal transactivation (Figures 5E,F). Similar results were observed where the mutation of the Nr1d1-PPRE blocked all responses to CLA and/or antagonists. This suggests that PPARα directly binds to PPRE at the Nr1d1 gene, thus determining the REV-ERBα-mediated transcriptional activity in response to CLA.

REV-ERBα Recruits Histone Repressors on the Circadian Genes in the CLA-Treated Mice
We performed ChIP-qPCR to examine the chromatin modification at the Nr1d1 gene promoter and enhancer. In association with the gain of Nr1d1 mRNA occupancy, the CLA treatment enhanced PPARα enrichments both at the Nr1d1 promoter and enhancer ( Figure 6). As expected, tool inhibited compounds, such as GW6471 and MK866 against PPARα or SR8278 and GSK1462 against REV-ERBα, exhibited a remarkable efficiency to reduce the binding enrichments at the specific locus ( Figures 6A,B). Another evidence that supported the finding is the similar profile we observed when PPARα was inhibited by siRNAs-inducible knockdown ( Figure 6C). The histone marks, H3K27ac and H3K4me1 were significantly reduced at the enhancer regions, while H3K27ac and H3K4me3 were reduced at the promoter regions (Figures 6D,E). In addition, the antagonists of PPARα or REV-ERBα also reduced the promoter occupancies of Pol-II, its transcription elongation-associated CTD-ser 2 phosphorylated form (Ser2P Pol-II) and transcription initiation-associated CTD-ser 5 phosphorylated form (Ser5P Pol-II) at the enhancer and the promoter of Nr1d1, compared to the CLA alone-treated hepatocytes (Figures 6D,E). Finally, The experiments were repeated three times. *p < 0.05, ***p < 0.001, using the two-tailed student's t-test or ANOVA with Tukey's post hoc test.
our results showed that the PPARα antagonist quenched the activation of the transcription repressor, REV-ERBα by blocking the PPARα binding to its gene and the associated local histone modification induced by the CLA. Therefore, the results suggest that in CLA-treated hepatocytes, the PPARα plays a dominant role in directing the REV-ERBα expression via the recruitment of histone marks and cofactors (Figures 7A,B).

DISCUSSION
Defining the molecular mechanisms of the circadian clock control may help to better understand how dietary lipids affect animal metabolic health. In this study, we found that the addition of polyunsaturated conjugated linoleic acid (CLA) disrupts the hepatic circadian clock program including the reduced expression of the genes, Clock and Cry1. It was downregulated by the well-defined negative transcription factor, REV-ERBα, but also requires the PPARα activation and their physical association at the locus of target genes. In particular, we demonstrated that CLA induces the activating PPARα and repressive REV-ERBα transcriptional complexes, simultaneously enhance the corepressors, histone deacetylase 3 (HDAC3) and nuclear receptor corepressor 1 (NCOR1), and reduces the coactivators, p300, RNA polymerase II (Pol II), and the associated local histone modification. During this regulation, the PPARα facilitates the REV-ERBα transcriptional activity by directly binding to PPRE at the Nr1d1 gene in the CLAtreated hepatocytes. Consequently, we showed that the CLAinduced hepatic triglyceride accumulation can be rescued by small-molecule drugs targeting either PPARα or REV-ERBα with the modified circadian clock program; this may be an alternative strategy to manage the associated metabolic disorders. It is commonly accepted that excessive amounts of intrahepatic triglyceride accumulation are a prominent feature of nonalcoholic fatty liver disease (NAFLD) (35). In this study, we observed that the dietary CLA supplementation caused enlarged livers with increased triglyceride contents, increased fat depot, and insulin resistance in young mice. In agreement, several studies have demonstrated that feeding mice with CLA results in lipodystrophy, hepatic steatosis, and dyslipidemia (36)(37)(38). In contrast, the CLA has been intensively studied for its antiobesity, antidiabetic, and antiatherogenic effects (26,39). The proposed mechanisms include the CLA-regulation of lipid metabolism, adipogenesis, adipocyte apoptosis, and energy expenditure, whereas the differential effects of CLA are likely to depend on dose (low-or high-duration), (short-or long-term), and animal body phenotype (lean or obese) (26,40). As CLA supplements that contain isomers cis-9, trans-11, and trans-10, cis-12 are widely used for weight management in humans, it is of importance to improve our knowledge on molecular mechanisms of the CLA action (40) and other polyunsaturated fatty acid (FA). Notably, an earlier study reported that a 6-week CLA (1%) feeding regimen disrupts the normal day-night rhythm of respiratory quotients in mice and results in increased FA oxidation at night and impairment of white adipose tissue (WAT) metabolism (41). We discovered that dietary CLA supplementation for 28 days already caused alterations of the circadian rhythm process, including the key clock genes, Per1, Per2, Crem, Clock, Arntl, Bhlhe41, and Cry1 in the liver of mice with the onset of steatosis, while high-fat diet has been shown to induce metabolic abnormalities involving the NR RORα and REV-ERBα, and affects the expression of Clock and Bmals during 12 weeks of experiment (42).
The maintenance of a proper circadian clock program is important that perturbations of the circadian rhythm predispose individuals to metabolic disorders, cardiovascular diseases, and tumor development (43,44). In turn, changes in physiological processes, such as food variations including CLA supplementation in this study could reprogram the circadian clock, likely through transcriptional mechanisms. As CLA-induced elevation of the REV-ERBα expression resulted in reduced clock gene messenger RNA (mRNA) expression, this confirmed the role of REV-ERBα as a transcriptional repressor in the liver (45). We further demonstrated that multiple components are involved in the molecular clock modulation by CLA. On top of the increased binding of REV-ERBα on the Clock and Cry1 genes, the CLA enhanced the recruitment of histone repressors, NCOR1, and HDAC3 at the binding sites. Alternatively, the transcriptional activationlinked histone marks H3K27ac and H3K4me2, and coactivators, p300 and Pol II were reduced by CLA, suggesting a twoarmed regulation apparatus. Although like other nuclear receptor (NR), REV-ERBα is activated by ligands, such as naturally occurring heme (16), it lacks the canonical NR-activation domain (46). It is unclear whether ingested CLA can directly regulate REV-ERBα expression in the liver. Studies of REV-ERB deletion in mice (REV-ERBα and/or REV-ERBβ) have shown a mild rhythmic phenotype and partially impaired metabolisms in various tissues (13,14,47). Using genetic and pharmacological approaches to inhibit REV-ERBα expression in hepatocytes, we revealed an altered clock gene program and reduced triglyceride accumulation in a CLA-specific context. We and other researchers thereby suggest that REV-ERBα is an essential component of the clock machinery, but maybe partially redundant (13,48). Notably, overexpression of REV-ERBα in the liver of the mouse leads to the suppression of 90% of cycling transcripts, implying a broader impact on circadian rhythm than deletion (49). Therefore, additional or complementary roles of REV-ERBα and its paralog, REV-ERBβ needs to be addressed.
Dietary CLA when ingested is readily assimilated into multiple cell types, which directly influences their metabolism. For 6 weeks of CLA administration in young mice, ∼1% of CLA isomer has been shown to be incorporated into the heart, the liver, and the peripheral tissues (50). We hypothesized that in this study, CLA modulates the circadian gene program, in part, by a mechanism depending on the activation of PPARα. This nuclear receptor (NR) is well-known as the master lipid sensor (51,52), and several isomers of CLA are known as high-affinity ligands and activators of PPARα (53,54). By using small-molecule antagonists of PPARα and siPPARα-1/-2, we provide evidence that PPARα participates in the REV-ERBα recruitment of histone repressors on the circadian genes and associated triglyceride production increment in response to the CLA. In addition to cooccupying the binding region with REV-ERBα on the clock genes, the PPARα controls the transcriptional activity of REV-ERBα in the CLAhepatocytes by binding to the Nr1d1 gene at PPAR response element (PPRE), at both the promoter and the enhancer. We thereby propose a unique "NR crosstalk" of PPARα-REV-ERBα in the regulation of the circadian clock gene program in the CLA-induced hepatic steatosis. Both REV-ERBα and PPARα could function as sensors for the metabolic state of the cell and may entrain the clock to metabolic cues (45). Under specific nutritional challenges, non-cyclic genes including PPARγ are reported to become circadian through the cyclic activation of surrogate transcription pathways, which underline the plasticity of the clock system (28). We suggest that under CLA treatment, the NR pair recruited coregulator complexes to specific regions of the Clock and Cry1 genes, can, in turn, impact the epigenome, that is, the nuclear corepressor 1 (NCOR1) may interact with NR proteins, as well as HDAC3, which deacetylates histone protein tails to create a repressive chromatin environment (8). This is supported by the chromatin immunoprecipitation sequencing (ChIP-seq) analysis of mouse liver that revealed a significant overlap of HDAC3, NCOR, and REV-ERBα on target genes, which is inversely related to histone acetylation and Pol II recruitment (5,8).
The process of circadian rhythm draws increasing attention due to the growing epidemic of NAFLD and associated metabolic disorders. In this study, we revealed that the NR REV-ERBα serves as a key link between the dietary CLAinduced hepatic steatosis and the downregulation of the circadian gene program in the liver, while another NR PPARα serves as the major CLA respondent and facilitates the REV-ERBαmediated transcription repression of the circadian genes and its recruitment of histone repressors in the CLA-treated mice. Our results suggest that both the REV-ERB antagonist and the PPARα antagonist may be used as therapeutics to reset dietassociated disruption of the circadian rhythm and reestablish the metabolic balance.

DATA AVAILABILITY STATEMENT
All data supporting the findings of this study are available within the article, in Supplementary Information, or from the corresponding author upon request. All RNA-seq data generated in this study are deposited in the Gene Expression Omnibus (GEO) database under the accession numbers "GSE159219" (https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi? acc=GSE159219).

ETHICS STATEMENT
All experiments involving animals were reviewed and approved by the Animal Care and Use Committee of Yangzhou University (YZUDWSY 2017-09-06).