Genetic synchronization of the brain and liver molecular clocks defend against chrono-metabolic disease Authors: Lauren N. Woodie, Ahren J. Alberto, Brianna M. Krusen, Lily C. Melink, Mitchell A. Lazar Published: 2024 ABSTRACT SignificanceMaintaining synchrony between the external environment and the precise orchestration of circadian rhythmicity among tissues within an organism is crucial for overall health and well-being. Here, we report that the REV-ERBα/β nuclear receptors coordinate rhythmicity between the suprachiasmatic nucleus (SCN) and liver to protect against metabolic disturbances caused by dissonance with the external environment. The loss of REV-ERBs in both the SCN and hepatocytes prevents the metabolic defects present in male mice lacking REV-ERBs in the SCN or hepatocytes alone. These findings suggest that synchrony between the liver clock and the SCN clock could be an important target for the treatment of metabolic disorders associated with circadian disruption. Circadian rhythmicity is a widely conserved biological mechanism that synchronizes endogenous biological processes to external cues, playing a critical and anticipatory role in controlling various physiological functions (1, 2). Disruptions in the synchrony between circadian rhythms and the environment are associated with an increased risk of developing metabolic abnormalities, including obesity, type 2 diabetes mellitus, and cardiovascular disease (1, 3–8). Thus, maintaining synchrony between the external environment and endogenous clocks is imperative for promoting general health and wellness. In mammals, circadian rhythms are synchronized to a ~24-h light:dark (LD) cycle by the suprachiasmatic nucleus (SCN) which is considered the preeminent circadian regulator of the body (9, 10). The SCN regulates a transcription/translation feedback loop (TTFL) that consists of the brain and muscle Arnt-like 1 (BMAL)1 protein, which rhythmically activates its own inhibitors, Period (PER), Cryptochrome (CRY), and the REV-ERBα/β nuclear receptors (11, 12). The interlocking nature of the TTFL creates the characteristic oscillatory pattern observed in rhythmic genes, proteins, metabolites, and behaviors (13). Under optimal conditions, the circadian TTFL machinery oscillates within every cell of the body, thereby maintaining synchronization between the SCN and peripheral organs throughout an organism (14–16). However, circadian desynchrony, the misalignment between the central circadian clock in the SCN and peripheral clocks in various tissues, can disrupt the coordinated timing of physiological processes and potentially lead to adverse health effects (17–20). Therefore, major questions in the field are how synchrony is maintained and how desynchrony between the external environment and the various tissues within an organism is mitigated. REV-ERBs are of particular interest because they are the main transcriptional repressors of BMAL1 (21–23) and lie at the intersection of rhythmicity and metabolism (21, 24–26). Deletion of hepatocyte REV-ERBs (HepDKO) exacerbated the metabolic liver dysfunction induced by diet-induced obesity (DIO) (26). Furthermore, loss of REV-ERBs in the SCN (SCN DKO) shortened the period of the SCN to ~21 h and led to metabolic dysfunction when the mice were entrained to a 24 h LD cycle and fed a high-fat diet (HFD) (27). The metabolic disturbances in these animals were corrected by matching environmental lighting to the shortened endogenous clock period, thereby synchronizing external and internal clocks (27). In addition to the dissonance between the 24-h day and their shortened endogenous period, SCN DKO mice also exhibited desynchrony between the liver and SCN (27). To assess the impact of internal desynchrony, we deleted REV-ERBs in hepatocytes of adult SCN DKO mice (Hep-SCN DDKO). Remarkably, the deletion of REV-ERBs in both the SCN and hepatocytes synchronizes the central and hepatic clock mechanisms and improves DIO outcomes despite the continued desynchrony with the 24-h day. These results demonstrate the critical importance of synchrony between central and peripheral clocks in controlling normal metabolic homeostasis. Results To investigate the loss of REV-ERBα and β in both the SCN and liver, we crossed Nr1d1fl/fl/Nr1d2fl/fl and Six3-Cre mice to generate SCN DKO mice as previously described (27). Six3-Cre is robustly expressed in the SCN with minimal expression in other brain areas (27, 28), and knockout of REV-ERBα (Nr1d1) and REV-ERBβ (Nr1d2) in the SCN was confirmed by RT-qPCR (Fig. 1 B and C). Adult SCN DKO and control littermate mice were injected with adeno-associated virus 8 (AAV8) in which the thyroxine binding globulin promoter (TBG) drove hepatocyte-specific expression (29) of either enhanced green fluorescent protein (EGFP) (control) or Cre recombinase, to generate HepDKO and Hep-SCN DDKO mice and their respective controls (Fig. 1A). Expression of both REV-ERBs was diminished in the livers of the HepDKO and Hep-SCN DKO groups (Fig. 1 D and E) and REV-ERBs expression was reduced in the SCN of the SCN DKO and Hep-SCN DDKO groups (Fig. 1 B–E). BMAL1 (Arntl) and NPAS2 (Npas2), direct targets of repression by REV-ERBs, were found to be overexpressed in the SCN (Fig. 1 F and G) and liver (Fig. 1 H and I) of the DKO groups, further confirming the knockout efficiency. REV-ERBα/β double knockout in the liver and SCN. (A) Experimental scheme. (B and C) RT-qPCR for Nr1d1 and Nr1d2 in the SCN at ZT10 (n = 5 to 8, mean ± SEM). (D and E) RT-qPCR for Nr1d1 and Nr1d2 in the liver at ZT10 (n = 5 to 7, mean ± SEM). (F and G) RT-qPCR for Arntl and Npas2 in the SCN at ZT10 (n = 5 to 8, mean ± SEM). (H and I) RT-qPCR for Arntl and Npas2 in the liver at ZT10 (n = 5 to 7, mean ± SEM). Results were compared by one-way ANOVA with Tukey’s post hoc test. *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001. We next performed wheel-running behavioral phenotyping in LD prior to release into dark:dark (DD) conditions. Control animals indeed exhibited a typical ~24 h free-running period while, as expected, the period of SCN DKO mice was considerably shorter (Fig. 2 A, B, and E). Loss of REV-ERBs in hepatocytes did not affect behavioral rhythms, as HepDKO animals maintained a free-running period of ~24 h (Fig. 2 B and E). However, Hep-SCN DDKO mice in DD exhibited a shortened behavioral period comparable to that of the SCN DKO (Fig. 2 D and E). Loss of REV-ERBs in the SCN is sufficient to shorten the circadian period of locomotor activity. (A–D) Representative actograms in LD and DD for Control, HepDKO, SCN DKO, and Hep-SCN DDKO. (E) Lomb–Scargle periodogram analysis for LD and DD actograms (n = 3, mean ± SEM). Results were compared by two-way ANOVA with Šidák’s post hoc test. ****P < 0.0001. The effect of Hep-SCN DDKO on metabolic rhythms was assessed by metabolic cage measurements. Oxygen consumption (VO2) and respiratory exchange ratio (RER) were measured in LD before release into DD (Fig. 3). Both control and HepDKO animals maintained a ~24 h period in VO2 and RER in LD and DD and developed ~21 h rhythm of VO2 and RER after release into DD, indicating that the period of these metabolic outputs is largely regulated by the SCN (Fig. 3 C and F). We also performed a series of metabolic substrate tolerance tests at ZT13. Lipid tolerance was not affected by deletion of REV-ERBs in the SCN or liver (SI Appendix, Fig. S1A), but like VO2 and RER, glucose tolerance was regulated by the SCN as SCN DKO animals exhibited impaired glucose handling (SI Appendix, Fig. S1B). Pyruvate tolerance, however, was impaired in the HepDKO group indicating that gluconeogenesis is largely driven by hepatocyte REV-ERBs (SI Appendix, Fig. S1C). Neither of these phenotypes presented in the Hep-SCN DDKO group which suggests that synchronizing molecular clock deficits in the SCN and liver prevents an aberrant response to metabolic stimuli. Loss of REV-ERBs in the SCN is sufficient to shorten the circadian period of metabolism. (A and B) LD and DD measurements for VO2 (n = 3, mean ± SEM). (C) Lomb–Scargle periodogram analysis for VO2 in LD and DD (n = 3, mean ± SEM). (D and E) LD and DD measurements for RER (n = 3, mean ± SEM). (F) Lomb–Scargle periodogram analysis for RER in LD and DD (n = 3, mean ± SEM). Periodogram results were compared using a two-way ANOVA with Šidák’s post hoc test. *P < 0.05, **P < 0.01. Circulating levels of beta hydroxybutyrate, carnitine, and T3 were not impacted by loss of SCN or liver REV-ERBs (SI Appendix, Fig. S2 A–C). Corticosterone was elevated in both the, HepDKO and SCN DKO groups, but not in the Hep-SCN DDKO group (SI Appendix, Fig. S2D). Genes associated with mitochondrial dynamics, oxidative metabolism, and carbohydrate metabolism were almost exclusively increased in livers of the HepDKO group (SI Appendix, Fig. S3 A–C). This increase was corrected by Hep-SCN DDKO indicating SCN influence over metabolic gene expression in the liver (SI Appendix, Fig. S3 A–C). Given the ameliorating effect of Hep-SCN DDKO on the tolerance tests and metabolic end points under normalchow diet (NCD) feeding, we tested the effect of synchronizing liver and SCN clock deficits on HFD. Remarkably, despite their abnormal VO2 and RER, when fed a HFD Hep-SCN DDKO animals were protected from exacerbated weight gain induced by the loss of REV-ERBs in the SCN alone (Fig. 4A). Hep-SCN DDKO also prevented an increase in food intake (Fig. 4B) that was induced by HepDKO and SCN DKO. There was no difference in circulating triglycerides among the groups (Fig. 4C), but Hep-SCN DDKO prevented an increase in hepatic triglyceride content (Fig. 4D) that was induced by HepDKO and SCN DKO alone. Thus, synchronization between the SCN and liver molecular clocks was protective against the effects of metabolic dissonance with the external environment. Matching hepatocyte and SCN clocks rescues SCN DKO mice from exacerbation of DIO. (A) Running body weight gain over 12 wk of 60% HFD feeding (n = 4 to 5, mean ± SEM). (B) kcal of food consumed over 24 h (n = 4 to 5, mean ± SEM). (C) Serum triglyceride content (n = 4 to 6, mean ± SEM). (D) Liver triglyceride content (n = 6 to 9, mean ± SEM). Body weight gain results were compared by repeated measures ANOVA with Tukey’s post hoc test. Food intake, serum triglycerides, and liver triglycerides were compared by one-way ANOVA with Tukey’s post hoc test. *P < 0.05, **P < 0.01, ***P < 0.001. Discussion Our findings have uncovered a role of REV-ERBs in coordinating circadian synchrony between the SCN and liver and observed the effect of this coordination on metabolic disease exacerbated by dissonance with the external environment. As previously shown, the loss of REV-ERBs in the SCN resulted in shorter endogenous behavioral and metabolic rhythms that, when at odds with the external environment, exacerbated DIO (27). While deletion of REV-ERBs in the hepatocytes increased food intake and hepatic triglyceride accumulation on HFD. Remarkably, loss of REV-ERBs in both the SCN and hepatocytes prevented SCN DKO- and HepDKO-specific effects. These findings indicate a role for REV-ERB nuclear receptors in maintaining internal circadian homeostasis and emphasizes the importance of synchrony between the SCN and liver clocks. SCN DKO animals developed a ~21 h behavioral and metabolic period upon release into DD, and this phenotype was not affected by the induction of hepatocyte REV-ERBs deletion in Hep-SCN DDKO animals, reinforcing the dominant role of the SCN clock. Although recent evidence has suggested that manipulations of the liver clock can affect behavioral rhythmicity in mice (30), the HepDKO group maintained a ~24 h circadian period indicating that loss of hepatocyte REV-ERBs does not impact the SCN. Most striking was the prevention of DIO exacerbation in the Hep-SCN DDKO group. The effect of molecular clock gene deletion on metabolic health is often tissue and gene dependent (31–34). However, in a number of cases, molecular clock gene deletion increases sensitivity to HFD which has led to the thought that dissonance between the external and internal environments was the root of chrono-metabolic disease (3, 20, 25–27, 35–37). Here, we show that matching REV-ERBs-mediated SCN and liver clock defects improves the metabolic profile of mice in the face of dissonance with light cues. Our study has several limitations. We focused on the role of REV-ERBs because their deletion causes a short period rather than arrthythmicity, but whether matching the deletion of other clock genes in the SCN and liver would offer similar protection against DIO exacerbation requires further study. We also focused on synchronizing deficits only between the brain and liver. The liver is peripheral hub for intermediary metabolism, but it is unclear whether similar dual knockouts between other metabolically active tissues (i.e., the pancreas, muscle, or adipose tissue) would confer similar effects on the development of DIO. Moreover, while we have shown that the hepatic vagus nerve relays chrono-metabolic signals between the brain and liver, other works have indicated that endocrine factors signal metabolic status from the periphery to the brain and time-of-day information from the brain to the periphery (38–43). How synchrony and desynchrony between the brain and liver are communicated remains to be determined in the Hep-SCN DDKO model. Finally, our studies were restricted to male mice, and whether our findings can be extrapolated to female mice will be a study of future investigation. Synchrony with the external environment and among tissues within an organism is critical for general health and metabolic disturbances develop due to desynchrony with the environment. Here, we report that matching peripheral and central clock periods prevents metabolic perturbations even in a desynchronous environment. Our findings suggest that synchrony between the liver clock and the SCN clock could be an important target for the treatment of metabolic disorders associated with circadian disruption such as those arising from shift work or jet lag. Materials and Methods Animals, Injections, and Diets. NR1D1fl/fl-NR1D2fl/fl mice were generated in the Lazar Lab (44) and maintained on a C57Bl6/J background (Jackson Labs, stock 008661). Six3-Cre, a SCN-specific Cre driver, mice were provided by S. Blackshaw (28) and maintained in the Lazar Lab on a C57Bl6/J background (Jackson Labs, stock 008661). Six3-Cre and NR1D1/2fl/fl mice were bred to generate NR1D1/2fl/fl × Six3-Cre+ and NR1D1/2fl/fl × Six3-Cre- littermates. Six to eight-week-old male littermates were split into four groups: 1. Six3-Cre- mice that were injected with AAV8-TBG-EGFP as a control, 2. Six3-Cre- mice that were injected with AAV-TBG-Cre, selectively knocking out REV-ERBα/β in hepatocytes (29) to generate HepDKO, 3. Six3-Cre+ mice that were injected with AAV8-TBG-EGFP to generate SCN DKO, and 4. Six3-Cre+ mice that were injected with AAV8-TBG-Cre to generate a Hep-SCN DDKO group. All viruses were injected with a titer of 2.0 × 1012 genome copies. For NCD experiments, animals were provided with ad libitum LabDiet 5010 (20% kcal protein, 10% kcal fat, 70% kcal carbohydrates). For HFD studies, animals were provided with ad libitum Research Diets D12492 HFD (20% kcal protein, 60% kcal fat, 20% kcal carbohydrates) beginning at 10 wk. All animals were given ad libitum access to autoclaved water. Studies followed the guidelines of the Institutional Animal Care and Use Committee of the University of Pennsylvania in accordance with the guidelines of the NIH. RNA Extraction and RT-qPCR. SCN punches were homogenized with 1 mL of QIAzol Lysis Reagent (Qiagen) using Pellet-Pestle (Kimble, 749515). Total RNA was purified and collected from RNeasy Mini spin columns according to the RNeasy Lipid Mini Kit (Qiagen, 74804). Liver samples were homogenized with Trizol Lysis Reagent (Qiagen) and total RNA was purified and collected from RNeasy Mini spin columns according to the RNeasy Mini Kit (Qiagen, 74106). RNA quality and quantity were assessed by NanoDrop (Thermo Scientific). Purified total RNA was used for complementary DNA (cDNA) synthesis (High Capacity cDNA Reverse Transcription Kit, Applied Biosystems, 43-688-14) and RT-qPCR using Power SYBR Green PCR Master Mix (Applied Biosystems, 43-091-55). The QuantStudio 6 Flex Real-Time PCR software and system were used. RT-qPCR results were analyzed by standard curve method. Liver RT-qPCR results were normalized to Arbp. SCN RT-qPCR results were normalized to Tbp. Primer sequences can be found in SI Appendix, Table S1. Wheel-Running Activity. Mice were housed individually with a running wheel and provided with ad libitum food and water in circadian cabinets. After a week of acclimation, photoentrainment was tested under a 12:12-h LD cycle, followed by an evaluation of free-running rhythms in constant darkness (DD) for 14 d. Wheel-running activity was recorded and analyzed using ClockLab software v 6.1.02. The default settings of ClockLab Analysis software (Actimetrics) were used to determine the onset and offset of wheel-running activity, circadian periods, and to generate Lomb–Scargle periodograms and actograms. Comprehensive Laboratory Animal Monitoring System (CLAMS). Twelve-week-old mice from each experimental group were individually placed in CLAMS (Columbus Instruments, Columbus, OH) metabolic cages and monitored over a 4-wk period. Two weeks of measurements were dedicated LD measurements before switching over to a DD light schedule for two more weeks of measurement. Every 7 min recordings of VO2 (volume of oxygen consumed per kilogram per hour) and RER were collected for each animal. Glucose Tolerance Test (GTT). GTT were performed in NCD-fed animals 2 to 4 wk after viral injection. Animals were fasted for 12 h starting at ZT1 for GTTs to begin at ZT13. A blood glucose measurement was taken with Contour Next (Ascensia Diabetes Care) blood glucometers at baseline before intraperitoneal injection of 2 g/kg glucose. Blood glucose measurements were then taken at 15, 30, 45, 60, 90, and 120 min after injection. Lipid Tolerance Test (LTT). LTT were performed in NCD-fed mice 2 to 4 wk after viral injection. Animals were fasted for 12 h starting at ZT1 for LTTs to begin at ZT13. A baseline blood sample (30 μL) was taken from the tail vein before oral gavage of 10 μL/g body weight of olive oil (Sigma-Aldrich). Subsequent blood samples (30 μL) were taken 1, 2, 3, and 4 h after gavage. Blood samples were then centrifuged at 4 °C at 5,000 × g for 15 min and serum was collected. Serum triglycerides were determined in undiluted serum samples using a triglyceride assay kit (StanBio). Pyruvate Tolerance Test (PTT). PTT were performed in NCD-fed mice 2 to 4 wk after viral injection. Animals were fasted for 12 h starting at ZT1 for PTTs to begin at ZT13. A baseline blood glucose measurement was taken on Contour Next (Ascensia Diabetes Care) blood glucometer before intraperitoneal injection of 1 g/kg sodium pyruvate. Blood glucose measurements were then taken at 10, 20, 30, 45, 60, 90, and 120 min after injection. BIOlogical Data AcQuisition system (BioDAQ). BioDAQ (Research Diets, New Brunswick, NJ) is a collection of cages that allows for automated and nonintrusive monitoring of precise feeding amounts and behaviors. Twelve-week-old mice from each experimental group were put into a BioDAQ system under a 12:12 LD to measure total food intake continuously. Liver Triglyceride Assay. To assess triglyceride levels, livers were homogenized in cell lysis buffer (140 mM NaCl, 50 mM Tris, pH 7.4, and 1% Triton X-100) with a TissueLyser equipped with steel beads. Next, a Bicinchoninic Acid Assay Kit (Thermo Scientific) was utilized to standardize triglyceride concentrations to 250 mg/mL for each sample. Subsequently, triglyceride concentrations in liver tissue lysates and serum were determined using a Triglyceride Assay Kit (StanBio), adhering to the manufacturer’s protocol. Serum Assays. To assess circulating factors, serum was collected at ZT10 sacrifice by letting whole trunk blood clot before 4 °C centrifugation at 5,000 × g for 15 min. Serum was drawn off whole blood and then stored at −80 °C until testing. Beta hydroxybutyrate was determined by running a ketone body assay (Sigma-Aldrich) on undiluted serum samples according to the manufacturer’s instructions. To assess serum carnitine, serum was collected as above and diluted 1:10 before running on an L-carnitine assay according to the manufacturer’s instructions (Sigma-Aldrich). To assess serum T3, serum was collected as above and diluted 1:2 before running on a T3 enzyme linked immunosorbant assay (ELISA) (MyBioSource) according to the manufacturer’s instructions. To assess serum corticosterone, serum was collected as above and diluted 1:100 before running on a corticosterone ELISA according to the manufacturer’s instructions (Cayman Chem). To assess serum triglycerides, serum was collected as above and run, undiluted on a triglyceride assay kit (StanBio), according to the manufacturer’s instructions. Statistical Analysis. Statistical analyses were performed using GraphPad Prism 6 (La Jolla, CA). The results were presented as means ± SEM. Statistical significance of gene expression, food intake, and hepatic triglyceride content were analyzed by one-way ANOVA with Tukey’s post hoc test. Statistical significance of behavioral and metabolic period was assessed with one-way ANOVA with Šidák’s post hoc test. Statistical significance of running body weight gain was assessed by repeated measures ANOVA with Tukey’s post hoc test. Supplementary Material Appendix 01 (PDF) Dataset S01 (XLSX) Dataset S02 (XLSX) Dataset S03 (XLSX) Dataset S04 (XLSX) Dataset S05 (XLSX)