Liver–gut axis signaling regulates circadian energy metabolism in shift workers

Abstract

Circadian rhythm is critical to maintaining the whole‐body metabolic homeostasis of an organism. Chronic disruption of circadian rhythm by shift work is an important risk factor for metabolic diseases. Fibroblast growth factor 15/19 (FGF15/19), a key component in the liver–gut axis, potently suppresses bile acid (BA) synthesis and improves insulin sensitivity. FGF15/19 emerges as a novel pharmaceutical target for prevention and treatment of metabolic diseases. The nicotinamide adenine dinucleotide (NAD⁺)‐dependent sirtuin 1 (SIRT1) deacetylase plays an important role in the maintenance of hepatic homeostasis by linking hepatic metabolism to circadian rhythm. Here, our clinical study identified that circadian rhythmicity and levels of plasma FGF19 and BA profiling, and cellular NAD⁺‐dependent SIRT1 signaling were disturbed in night shift (NS, n = 10) compared to day shift (DS, n = 12) nurses. Our in vitro data showed that recombinant FGF19 protein rescued cellular circadian rhythm disrupted by SIRT1 inhibitors. Furthermore, we determined the effect of FGF15 on circadian rhythm and hepatic metabolism in wild‐type (WT), Fgf15 knockout (KO), and Fgf15 transgenic (TG) mice. The expressions of circadian‐controlled genes (CCGs) involved in SIRT1 signaling, BA and lipid metabolism, and inflammation were disrupted in Fgf15 KO compared to WT and/or Fgf15 TG mice. Moreover, systemic FGF15 deficiency led to the circadian disturbance of NAD⁺‐dependent SIRT1 signaling and significant reduction during nighttime in mice. These findings suggest that FGF15/19 regulates the circadian energy metabolism, which warrants further studies as a putative prognostic biomarker and pharmaceutical target for preventing against metabolic diseases associated with chronic shift work.

Chronic disruption of circadian rhythm by shift work is an important risk factor for metabolic diseases. Our studies suggest that fibroblast growth factor 15/19 (FGF15/19), a key component in the liver–gut axis, regulates the circadian energy metabolism through the increase of NAD⁺‐dependent SIRT1 signaling, leading to prevention against liver metabolic diseases associated with chronic shift work.

Abbreviations

BA

bile acid

CCG

clock‐controlled gene

CDCA

chenodeoxycholic acid

CG

clock gene

DS

day shift

FGF15/19

fibroblast growth factor 15/19

GCDCA

glycochenodeoxycholic acid

GDCA

glycodeoxycholic acid

HCC

hepatocellular carcinoma

HNF4α

hepatocyte nuclear factor 4 alpha

KO

knockout

MASH

metabolic dysfunction‐associated steatohepatitis

MASLD

metabolic dysfunction‐associated steatotic liver disease

NAD⁺

nicotinamide adenine dinucleotide

NS

night shift

PBMCs

peripheral blood mononuclear cells

PGC1α

peroxisome proliferator‐activated receptor gamma coactivator 1α

SCN

suprachiasmatic nucleus

SIRT1

sirtuin 1

TG

transgenic

WT

wild‐type

1. INTRODUCTION

The circadian clock is an endogenous biological timekeeping system that generates circadian rhythms, representing ∼24‐h periodic oscillations in behavioral (e.g., rest/activity and feeding/fasting) and biological (e.g., biochemical) processes. Endogenous circadian rhythms can be synchronized to day/night cycles. Circadian rhythms are essential to living beings by enhancing their ability to adapt to changing environmental conditions. In mammals, biological clocks comprise a set of biochemical timing systems, including the central clock, as a master pacemaker, within the suprachiasmatic nucleus (SCN) of the hypothalamus, and peripheral clocks present in the cells of essential tissues and organs. Together, these timing systems enable organisms to drive circadian rhythms at multiple levels, ranging from the micro‐level of gene expressions and cell metabolism to the macro‐level of biological and metabolic processes in a temporally appropriate manner. ¹ , ²

An important function of the SCN pacemaker is to hierarchically synchronize metabolic rhythms in peripheral clocks (e.g., liver clock) through hormonal cues (e.g., glucocorticoids) and autonomic nervous system signals to coordinate environmental and endogenous cues, maintaining metabolic homeostasis. Liver clock, one of the most important peripheral clocks, regulates a variety of biochemical processes including synthesis and metabolism of glucose, lipid, and bile acid (BA), nutrient homeostasis, and detoxification by controlling cyclic expressions of master regulators and rate‐limiting enzymes involved in key hepatic metabolic processes. ³ Global metabolic profiling studies have shown that various liver metabolites involved in metabolic pathways exhibit different degrees of circadian oscillations throughout a 24‐h cycle. ⁴ Kristin et al. created a “CircadiOmics” database to demonstrate the coordination of transcriptome, proteasome, and metabolome to maintain liver metabolic homeostasis under circadian control. Sirtuin 1 (SIRT1), a member of the sirtuin protein family, is a highly conserved nicotinamide adenine dinucleotide (NAD)⁺‐dependent histone and protein deacetylase in mammals. NAD⁺‐dependent SIRT1 enzyme, as a metabolic/energy sensor, plays an important role in regulating glucose, lipid and BA metabolism to maintain metabolic homeostasis in response to energy availability.

The molecular machinery of circadian clocks in mammals is characterized by the transcription–translation feedback loops, which consist of a set of clock and clock‐controlled genes (CGs and CCGs) to generate the rhythmic outputs of gene expressions. The core transcription factors, BMAL1 and CLOCK, form the heterodimer complex, which binds to the E‐box elements in the promoters of CGs including PERs, CRYs, REV‐ERBs, and RORs to activate their transcription. PER and CRY proteins form the negative arm of the circadian feedback loops by preventing BMAL1:CLOCK complex from binding to E‐box elements, leading to the inhibition of their transcriptional activity. ⁵ , ⁶ Additionally, BMAL1:CLOCK complex activates nuclear receptors REV‐ERBs and RORs, which serve as repressors and activators respectively, forming another feedback loop to regulate BMAL1:CLOCK complex. ⁷ , ⁸ The core driver BMAL1:CLOCK can transcriptionally regulate downstream CCGs, of which circadian expressions mediate a number of rhythmic biological and physiological processes through the body.

Circadian disruption is defined as the desynchrony of internal circadian network and/or their misalignment with environmental cues including light/dark and eating/fasting cycles. Accumulating evidence from both animals and humans suggests that long‐term disruption of circadian rhythm by shift work, social jetlag, or light‐at‐night is associated with increased risks of chronic diseases, including metabolic diseases and cancers. ⁹ , ¹⁰ Emerging evidence indicates dysregulation of hepatic metabolic homeostasis caused by shift work was shown to increase the risk of liver diseases such as metabolic dysfunction‐associated steatotic liver disease (MASLD), which includes a range of disease conditions from hepatic steatosis, metabolic dysfunction‐associated steatohepatitis (MASH), or even to cirrhosis or hepatocellular carcinoma (HCC). ¹¹ , ¹² , ¹³ , ¹⁴

BA homeostasis is well known to be regulated by the circadian system. Disruption of circadian rhythm contributes to dysregulated BA metabolism, which has been shown to be strongly associated with the risk of developing MASLD. ¹⁵ , ¹⁶ Fibroblast growth factor 15 (FGF15), the mouse homologue of human FGF19, is an endocrine FGF secreted from intestines responding to postprandial increased BAs. ¹⁷ FGF15/19 is a late fed‐state gut‐derived hormone known to mediate postprandial hepatic responses, including suppressing BA synthesis and improving insulin sensitivity. ¹⁸ FGF19 transgenic (TG) mice fed with a high‐fat diet showed decreased body weight and fat content, and increased white adipose tissue browning, ¹⁹ whereas Fgf15 knockout (KO) mice presented increased body weight and fat mass when treated with a high‐fat diet. ²⁰ Treatment with FGF19 or engineered FGF19 analogue improved BA metabolism, lipid metabolism, oxidative stress, insulin sensitivity, hepatic inflammation and fibrogenesis within MASLD and MASH development in mice. ²¹

Hence, it is important to investigate the potential impact of shift work on BA metabolism and examine the role of FGF15/19 in regulating circadian energy metabolism in the pathogenesis and development of metabolic diseases, especially MASLD. Here, we determined the circadian rhythms and levels of plasma FGF19 and BA profiling, as well as critical components in NAD⁺‐dependent SIRT1 signaling pathway in peripheral blood mononuclear cells (PBMCs) of day shift (DS) and night shift (NS) nurses over a 24‐h period. We also determined the effect of FGF15/19 on circadian rhythm, BA and lipid metabolism and inflammation, and NAD⁺‐dependent SIRT1 signaling in vitro and in vivo.

2. RESULTS

2.1. Disrupted circadian regulation of FGF19‐mediated BA metabolism in NS versus DS nurses

DS nurses presented significant rhythmicity in plasma FGF19 levels over 24 h, with the peak at 15:30 (p = .04). By contrast, the rhythmicity was disrupted in NS nurses (p = .17), with a 4‐h delayed peak time in NS compared to DS nurses (Figure 1A). Specifically, compared to DS, NS decreased plasma FGF19 levels during the daytime with significant difference at two time points (11:00 and 15:30), but not during night.

FIGURE 1.

24‐h profiles of plasma FGF19 and total BAs in DS versus NS nurses. (A) Plasma FGF19 and (B) total BA levels in DS (Blue) and NS (Orange) nurses. Y‐axis presents means ± SEM; X‐axis presents sample collection time points in average. # indicates statistical significance at p < .05 for the circadian rhythmicity. * or **indicates statistical significance at p < .05 or <0.01 at specified time in/between DS and NS, respectively. Gray and black bars above X‐axis represent the average sleep time of DS and NS nurses, respectively. (C) BA composition in DS and NS nurses over 24 h. BA, bile acid; DS, day shift; NS, night shift.

As observed for plasma FGF19, plasma total BA levels presented diurnal oscillations with the peak at 15:30 in response to food intake at lunch in DS nurses. The coordinated patterns of plasma total BA and FGF19 oscillation are consistent with FGF19 secretion from the intestine in response to the postprandial increase in transintestinal BA flux after food intake. ²² However, the oscillation of plasma total BA was disrupted with overall higher levels in NS compared to DS nurses over 24 h, although no significance identified (Figure 1B). For individual BA species, glycochenodeoxycholic acid (GCDCA) is the most prominent BA species in the plasma of either DS or NS nurses over 24 h (except chenodeoxycholic acid (CDCA) at 04:00 for DS and glycodeoxycholic acid (GDCA) at 16:00 for NS) (Figure 1C).

2.2. Disrupted circadian rhythms of NAD ⁺‐dependent SIRT1 signaling in NS versus DS

To investigate the impact of shift work on NAD⁺‐dependent SIRT1 signaling, we assessed NAD⁺ level, NAD⁺/NADH ratio, and SIRT1 activity as well as related mRNA expressions of NAMPT and SIRT1 genes in the same sets of PBMC samples from DS and NS nurses.

Although there were variations among individual levels at each time point, the overall data of NAD⁺/NADH ratio presented circadian rhythms over 24 h in DS (p = .037) with the highest level in the early night (20:00) and the lowest level before the early morning (04:00), while the rhythmicity was abolished in NS nurses, with bimodal peaks at 08:30 and 16:00 and bimodal nadirs at 12:30 and 18:30 (Figure 2B). Similarly, SIRT1 activity also showed significant rhythmicity (p = .049) with increases from the lowest level in the late morning (11:00) to the highest level at midnight (24:00) in DS; however, this circadian rhythm was disrupted with 4‐h delay of the highest peak at 04:00 (Figure 2C). However, neither DS nor NS present rhythmicity in NAD⁺ level, although DS showed two peaks of NAD+ levels at 15:30 and 04:00, respectively, but not in NS nurses (Figure 2A). No significant difference was identified for NAD⁺ level, NAD⁺/NADH ratio or SIRT1 activity in DS versus NS at each specified time.

FIGURE 2.

24‐h profiles of NAD⁺‐dependent SIRT1 signaling in PBMCs of DS versus NS nurses. (A) NAD⁺ level, (B) NAD⁺/NADH ratio, (C) SIRT1 activity, (D) NAMPT mRNA expression, and (E) SIRT1 mRNA expression in PBMCs in DS (Blue) and NS (Orange) nurses. Y‐axis presents means ± SEM; X‐axis presents sample collection time points in average. # indicates statistical significance at p < .05 for the circadian rhythmicity. **indicates statistical significance at p < .01 at specified time in/between DS and NS, respectively. Gray and black bars above X‐axis represent the average sleep time of DS and NS nurses, respectively. PBMCs, peripheral blood mononuclear cells; DS, day shift; NS, night shift.

In addition, we determined the circadian levels and rhythmicity of circadian regulatory genes, NAMPT and SIRT1 involved in NAD⁺‐dependent SIRT1 signaling, in PBMCs from DS and NS nurses (Figure 2D,E). The overall mRNA expression levels of NAMPT and SIRT1 genes were significantly higher in NS compared to DS nurses over 24 h (except early morning for SIRT1), with a significant difference at 16:00. However, neither NAMPT nor SIRT1 expression presented circadian rhythmicity in DS or NS. Interestingly, NAMPT and SIRT1 exhibited similar expression patterns over 24 h in either DS or NS nurses. The result is consistent with previous findings that NAMPT and SIRT1 form a positive feedback loop within the NAD⁺ salvage pathway, in which NAMPT is the rate‐limiting enzyme of NAD⁺ synthesis, thus enhancing SIRT1 activity by increasing NAD⁺ levels. Moreover, their expression levels and patterns are inversely correlated with the change of NAD⁺ levels over 24 h as reported in the liver previously. ²³

2.3. Restoration of cellular circadian rhythm by recombinant FGF19 protein

In our previous studies, we developed an in vitro cellular circadian reporter system to characterize the cellular circadian rhythm and their response to external cues for up to 7 days. ²⁴ Specifically, human mammary epithelial cells, MCF10A cells with a PER2 promoter‐driven luciferase reporter vector, are considered as circadian responsive reporter cells in this in vitro model to identify the environmental disruptors and chemopreventive enhancers for the circadian system via a bioluminescence assay. Additionally, MCF10A cells endogenously express FGF receptor 1 (FGFR1) and overactivation of FGFR1 has been shown to induce cell transformation in this cell line, ²⁵ , ²⁶  and thus it is considered an appropriate model to determine the potential effect of FGF19 on circadian rhythm in this study. In order to investigate the effect of FGF19 on the cellular circadian rhythm, the recombinant FGF19 protein was added to the in vitro cellular circadian reporter system, in which the cellular circadian rhythm had been disrupted by a highly selective and potent SIRT1‐specific inhibitor, EX527, and a SIRT1/2 inhibitor, cambinol.

The control, consisting of stably transfected cells with this circadian reporter system, generated at least two complete cycles of circadian luminescence signaling following 2‐h synchronization with 50% horse serum (Figure 3). The SIRT1/2 inhibitor, cambinol, disrupted the cellular circadian rhythm by abolishing the initial cycles and disturbing the subsequent cycles at concentrations of either 1 or 2 μM compared to the control group. The addition of recombinant FGF19 protein at 5 ng/mL restored the disrupted circadian rhythm in cells treated with cambinol at 2 μM (Figure 3A), whereas FGF19 did not exert any effect on the normal control cells.

FIGURE 3.

The cellular circadian rhythm in MCF10A/PER2‐dLuc reporter cells identified via in vitro bioluminescence assay. (A) Yellow: control; red: 1 μM cambinol; green: 2 μM cambinol; blue: 2 μM cambinol +5 ng/mL recombinant FGF19; brown: 5 ng/mL recombinant FGF19 only. (B) Yellow: control; red: 20 nM EX527; green: 40 nM EX527; blue: 40 nM EX527 + 5 ng/mL recombinant FGF19; brown: 5 ng/mL recombinant FGF19 only. X‐axis, time (days); y‐axis, amplitude.

Similar to results seen with another SIRT1 specific inhibitor, EX527 disrupted the cellular circadian rhythm with dampened circadian cycles and caused the advanced appearance of luminescence peaks. Significantly, addition of FGF19 into the recording medium at 5 ng/mL rescued circadian rhythms in cells treated with EX527 at 40 nM (Figure 3B). These results indicate that the disrupted cellular circadian rhythm caused by SIRT1 inhibitors can be restored by recombinant FGF19.

2.4. Effects of FGF15 on mRNA expressions of genes involved in circadian rhythm, liver metabolism, and inflammation in mouse liver

To further investigate the effect of FGF15 on the hepatic circadian clock in vivo, we compared the circadian expressions of CGs and CCGs involved in circadian rhythm, liver metabolism, and inflammation in the liver in wild‐type (WT), Fgf15 TG, and Fgf15 KO mice at 4‐h intervals over 24 h (Figure 4). As expected, hepatic CGs (Per1, Per2, Clock, Bmal1, Rev‐erbα, and Rev‐erbβ) expressions exhibited 24‐h rhythmicity in WT mice (Figure 4A). However, ablation of FGF15 enhanced the circadian amplitude of Per2, Rev‐erbα, and Rev‐erbβ, while reducing the amplitude of Clock and Bmal1 expressions in mouse liver over 24 h. Neither overexpression nor knockout of Fgf15 disrupted the circadian rhythmicity of these CGs.

FIGURE 4.

The effect of FGF15 on circadian mRNA expressions of CGs and CCGs. (A) 24‐h circadian rhythms and expression levels of CGs (Per1, Per2, Clock, Bmal1, Rev‐erbα, and Rev‐erbβ), and CCGs involved in (B) NAD⁺‐dependent SIRT1 signaling (Sirt1, Nampt, and Nmnat1), (C) BA (Fgf15, Cyp7a1, and Fxr) metabolism, and (D) lipid (Cyp4a10, Lcn2, and Mtp) metabolism and inflammation (Cd36) in the liver or in the intestine in WT, Fgf15 TG, and Fgf15 KO mice. Y‐axis presents means ± SEM; X‐axis presents sample collection time points. ZT: Zeitgeber time; ZT0: 06:00, lights on; ZT12: 18:00, lights off; black bar below X‐axis represents light off time. N = 5/time point/group. * indicates statistical significance at p < .05 at specified time between WT, Fgf15 TG or Fgf15 KO mice. # indicates statistical significance at p < .05 for the circadian rhythmicity.

Hepatic Sirt1, Nampt, and Nmnat1 expressions also presented significant rhythmicity in WT and Fgf15 TG mice, whereas knockout of Fgf15 abolished their circadian rhythmicity (Figure 4B). Moreover, Fgf15 KO mice showed significantly higher levels of Nampt at ZT16 and Nmnat1 at ZT0, ZT4, ZT16, and ZT20.

The systemic deficiency of FGF15 was associated with reduced expressions and disrupted rhythms of intestinal Fgf15 in Fgf15 KO mice, while the overexpression of hepatic and intestinal Fgf15 was observed in Fgf15 TG mice compared to WT and/or Fgf15 KO mice (Figure 4C). CCGs, Cyp7a1 and Fxr, involved in BA signaling pathway showed significant 24‐h circadian expression patterns in the liver in WT mice over 24 h. Interestingly, the circadian amplitude of Cyp7a1 and Fxr mRNA expressions were significantly enhanced in Fgf15 KO mice, whereas no circadian rhythmicity was observed in Fgf15 TG mice.

Hepatic Cyp4a10, Lcn2, Mtp, and Cd36 expressions showed significantly higher levels in Fgf15 TG compared to WT and/or Fgf15 KO mice (Figure 4D). However, no significant rhythmicity was identified for these gene expressions among WT, Fgf15 KO, and Fgf15 TG mice over 24 h.

2.5. Effects of FGF15 on circadian regulation of NAD ⁺ metabolism and SIRT1 activity in mouse liver

To investigate the impact of FGF15 on the circadian regulation of NAD⁺‐dependent SIRT1 signaling, we examined the circadian levels and rhythms of intracellular NAD⁺ level, NAD⁺/NADH ratio and relevant SIRT1 protein activity in liver tissues collected from WT, Fgf15 TG, and Fgf15 KO mice, respectively, over 24 h (Figure 5). WT and Fgf15 TG mice presented similar circadian patterns of liver NAD⁺ level with bimodal peaks at ZT8 and ZT16. By contrast, Fgf15 KO mice showed 24‐h circadian rhythmicity with significantly lower levels at ZT16 compared to WT and Fgf15 TG mice (Figure 5A). NAD⁺/NADH ratio presented circadian rhythmicity with significantly higher level at ZT12 in WT mice, however, either overexpression or ablation of FGF15 abolished the rhythmicity (Figure 5B). WT and Fgf15 TG mice exhibited circadian rhythms of liver SIRT1 activity with the highest level at ZT16, indicating the circadian control of NAD⁺‐dependent SIRT1 pathway (Figure 5C). Liver SIRT1 activity also showed 24‐h oscillation rhythmicity with decreased levels in the nighttime and a significant difference at ZT16 in Fgf15 KO mice compared to WT and Fgf15 TG mice.

FIGURE 5.

Effects of FGF15 on circadian profiles of key components involved in NAD⁺‐dependent SIRT1 signaling in mouse liver. (A) NAD⁺ level, (B) NAD⁺/NADH ratio, (C) SIRT1 activity in the liver in WT, Fgf15 TG, and Fgf15 KO mice. Y‐axis presents means ± SEM; X‐axis presents sample collection time points. ZT: Zeitgeber time; ZT0: 06:00, lights on; ZT12: 18:00, lights off; black bar above X‐axis represents light off time. N = 5/time point/group. * indicates statistical significance at p < .05 at specified time between WT, Fgf15 TG or Fgf15 KO mice. # indicates statistical significance at p < .05 for the circadian rhythmicity.

3. DISCUSSION

The increasing prevalence of unhealthy lifestyles including social jetlag, light switch, and shift work in modern societies contributes to decreased health through frequent disruption of our biological timing systems. Disruption of circadian rhythm by persistent shift work is strongly associated with the increased risk of chronic diseases including metabolic diseases and cancers. ²⁷ , ²⁸ Human epidemiological studies revealed that misalignment of biological clocks is associated with the occurrence of liver metabolic diseases. A prospective human study found that individuals who worked NS (usual/permanent frequency) were more likely to develop MASLD compared to those who never or rarely worked NS. ¹³ Moreover, NS workers who experienced longer duration, higher frequencies, more consecutive NS or longer work time per shift are at higher risk of developing MASLD. Chronic circadian dysfunction caused by occupational jetlag (8‐h shift work) was identified as an independent risk factor of MASLD‐related hepatocarcinogenesis in mouse hepatocytes in a mouse model ²⁹ and in human hepatocytes in a humanized mouse model. ³⁰ These studies indicated that chronic jetlag induced glucose intolerance, hepatic fat accumulation and damage, inflammation and fibrosis accompanied by histologic features of MASLD, MASH, and fibrosis, which further progressed to HCC in both mouse and humanized mouse models.

In this study, we observed significant circadian rhythmicity of plasma FGF19 in DS nurses, with the peak at 15:30 in response to postprandial increases in BAs. This observation was consistent with previous findings on the circadian patterns of serum FGF19 in five healthy volunteers, ²² whereas our studies did not observe the second peak of plasma FGF19 after dinner, which might be attributed to the food administration in our subjects different from the standardized meals given in their study. In addition, the possibility of missing a postprandial peak around 22:00 between our sample collection times (20:00 and 24:00) in DS nurses in our study cannot be excluded. However, the circadian rhythm was disrupted with a significant decrease during the day and a 4‐h delayed peak of plasma FGF19 in NS compared to DS nurses, which might be due to their sleep schedule and food deprivation during the day. Moreover, we found that NS presented higher plasma total BA levels over 24 h compared DS. These results indicate that disrupted circadian rhythm, especially disrupted eating/fasting signal, significantly dysregulates the level and rhythmicity of FGF19, along with increased BA levels in plasma, which may contribute to MASLD development in NS workers.

Circadian rhythm is tightly interrelated with energy metabolism. Liver, as the central metabolic organ, is one of the most important peripheral clocks. The liver clock plays an important role in maintaining energy and metabolism homeostasis by timely regulating cyclic expressions of master regulators and rate‐limiting enzymes involved in anabolic and catabolic processes. The feeding/fasting cycle can entrain clocks in peripheral tissues independently of the central clock in SCN, with synchronization occurring in a tissue‐specific manner. As a key metabolic center responsible for nutrient processing, the liver's internal clock is most sensitive to the feeding/fasting signal compared to other tissues. In addition, the liver clock is also influenced by the central clock through hormonal and neuronal signals and behavioral rhythms (e.g., sleep/wake and rest/active) adapted to environmental light/dark cycles through the central clock directly influence feeding/fasting rhythm. Therefore, the liver clock coordinates with other peripheral clocks to modulate peripheral circadian function in response to light/dark and feeding/fasting signals as an organized network. ³¹ , ³² Moreover, clocks in the liver can operate as tissue‐autonomous oscillators to achieve 10 ~ 20% circadian function in metabolism, including NAD+ salvage pathway and glycogen metabolism. ³² The liver clock regulates the circadian rhythm of important components in BA metabolism. Numerous transcription factors including D‐binding protein (DBP), differentiated embryo chondrocyte (DEC), E4 promoter‐binding protein 4 (E4BP4), hepatocyte nuclear factor 4 alpha (HNF4α), peroxisome proliferator‐activated receptor alpha (PPARα), liver X receptor alpha (LXRα), and REV‐ERBs regulate the circadian expression of Cyp7a1 and other genes involved in BA metabolism. ³³ RORα serves as the positive regulator of Cyp8b1 expression. ³⁴ Interestingly, feeding signals including glucose and insulin stimulate Cyp7a1 expression, whereas fasting induces C yp8b1 expression. ³⁴ , ³⁵ CLOCK directly activates the expression of SHP by binding its E‐box element. ³⁶ Moreover, FGF15/19 provided insights into the involvement of intestinal clocks in the circadian regulation of BA metabolism.

However, impaired feeding/fasting signaling including nutrient challenge and metabolic stress affects the circadian synchrony. Studies have shown that high‐fat diet (HFD) extended free‐running circadian period, disrupted behavioral rhythms, altered CG expressions and impaired circadian regulation of nuclear receptor network in mice. ³⁷ Moreover, HFD induced the circadian reprogramming of transcriptome and metabolome within the liver clock. ³⁸ Our study found that FGF15 deficiency caused only weak shifts or slightly dampened or increased patterns of oscillation of most CGs. Interestingly, systemic ablation of FGF15 did not perturb the feedback loops of core components within the transcription/translation molecular circadian clock, nor disrupted the circadian rhythmicity of hepatic CGs. However, the circadian rhythms and levels of CCGs involved in NAD⁺‐dependent SIRT1 signaling, BA and lipid metabolism, and inflammation were disrupted in Fgf15 KO mice. The potential mechanism by which FGF15/19 influences the circadian clock is by modulating the expression of key transcription factors, which subsequently regulate the transcriptional activity of CGs. FGF15/19 dephosphorylates the transcription factor cAMP regulatory element‐binding protein (CREB), leading to the inhibition of CREB‐peroxisome proliferator‐activated receptor gamma coactivator 1α (PGC1α) signaling, which is identified to coactivate orphan nuclear receptor ROR family to promote the expression of CGs. ³⁹ , ⁴⁰ In addition, FGF15/19 may suppress the activity of HNF4α by promoting the expression of SHP. HNF4α modulates circadian clock through transrepressing CLOCK:BMAL1 activity in HepG2 cells. ⁴¹

SIRT1, a member of the sirtuin protein family, is a highly conserved NAD⁺‐dependent histone and protein deacetylase in mammals. NAD⁺‐dependent SIRT1 enzyme, as a metabolic/energy sensor, plays an important role in regulating glucose, lipid and BA metabolism to maintainmetabolic homeostasis in response to energy availability. Energy deprivation including fasting, caloric restriction or exercise activates SIRT1, which adaptively regulates transcriptional activity to mediate various metabolic responses such as increased gluconeogenesis, glycolysis, β‐oxidation, and insulin sensitivity in the liver. ⁴² SIRT1 deacetylase also plays an important role in modulating deacetylation of CLOCK:BMAL1 heterodimers and PER2 protein, thus regulating rhythmic expressions of CGs and CCGs. ⁴³ , ⁴⁴ In turn, NAD⁺‐dependent SIRT1 signaling is under circadian control, with key components in the pathway exhibiting circadian expressions. NAD + ‐dependent SIRT1 signaling plays a key role in integrating the circadian clock to liver metabolism, immune function, as well as DNA damage response and repair. ⁴⁴ , ⁴⁵ NAMPT/NAD⁺/SIRT1 network functions an important role in the prevention of fatty liver diseases via regulation of hepatic lipid metabolism, oxidative stress, and inflammation. ⁴⁶ SIRT1 agonists such as resveratrol and SRT1720, and NAD⁺ with its precursor nicotinamide riboside have been identified as potential therapeutic agents which shown great protective effects against MASLD or MASH. ⁴⁷ , ⁴⁸ , ⁴⁹ , ⁵⁰ The dysregulation of circadian control of NAMPT/NAD⁺/SIRT1 is associated with MASLD development. ⁴² Downregulated intracellular NAD⁺ and NAMPT have been identified in MASLD development, whereas increased extracellular NAMPT is associated with MASLD and cancer progression. ⁵¹ , ⁵² The circadian disruption of NAD⁺‐dependent SIRT1 signaling in NS nurses may suggest the metabolic dysregulation caused by shift work, which contributes to the development of liver metabolic diseases.

The BA‐FXR‐FGF15/19 signaling pathway has provided a potential strategy to treat liver metabolic diseases. ⁵³ Treatment with FGF19 or engineered FGF19 analogues improved BA and lipid metabolism, oxidative stress, insulin sensitivity, cholestasis, hepatic inflammation and fibrogenesis, and inhibited BA synthesis within MASLD and MASH development in animal studies. ⁵⁴ , ⁵⁵ , ⁵⁶ FGF19 analogue NGM282 improved liver steatosis and fibrosis, and histological features in patients with MASH in preclinical studies. ⁵⁷ , ⁵⁸ We found that FGF15/19 enhanced the circadian regulation of NAD⁺‐dependent SIRT1 signaling, indicating that FGF19 plays an important role in maintaining circadian homeostasis of energy metabolism. However, the direct evidence linking FGF15/19 and SIRT1 interactions is limited. SIRT1 plays an important role in regulating FXR transactivation by direct deacetylation or indirect regulation of hepatocyte nuclear factor 1α (HNF1α) and PGC‐1α. FXR can directly control the transcriptional activity of FGF15/19, which in turn post‐translationally modulates FXR function. ⁵⁹ , ⁶⁰ It may indicate the potential relationship between SIRT1 and FGF15/19. Other transcription factors involved in overlapping pathways may link FGF15/19 and SIRT1. In addition, as a key metabolic sensor and mediates homeostatic responses to nutrient availability, SIRT1 may response to changes in nutritional states caused by FGF15 overexpression or deficiency in our mouse models. Further studies are needed to explore the interactions between SIRT1 and FGF15/19 in response to metabolic changes.

Our studies suggest that FGF15/19 may be able to restore circadian expressions of genes involved in circadian rhythms and BA and lipid metabolism through the increase of NAD⁺‐dependent SIRT1 signaling, leading to prevention against liver metabolic diseases associated with circadian disruption of energy metabolism. In addition, this study provides a molecular perspective to develop FGF15/19 as a potential prognostic biomarker and therapeutic target for metabolic diseases in shift workers by enhancing the circadian regulation of liver metabolism.

4. MATERIALS AND METHODS

4.1. Human study design and participant recruiting

The human study design has been described in detail previously. ⁶¹ Briefly, nonsmoking, nonobese, nonpregnant, premenopausal (age 21–45 years) healthy female nurses having normal sleep habits were recruited in this study. Volunteers were working at day shift (DS) or night shift (NS) on a regular basis for at least 3 months. Twelve DS and 10 NS nurses were recruited in the study. Average blood sample collection times in DS nurses were 06:30, 11:00, 15:30, 20:00, 24:00, and 04:00, and in NS nurses were 08:30, 12:30, 16:00, 18:30, 23:00, and 04:00. The blood sample was collected in a BD Vacutainer CPT Mononuclear Cell Preparation tube (containing 0.1 M sodium citrate and Ficoll medium), and PBMCs were isolated from plasma by centrifugation for 30 min at 370g RCF, then aliquoted plasma and PBMCs samples were snapped into liquid nitrogen and frozen in −80°C for future analysis.

4.2. Animal treatment and tissue collection

Male mice (8–12 weeks old) with WT, systemic Fgf15 deletion (Fgf15 KO), ⁶² or Fgf15 overexpressed in both hepatocytes and enterocyte (Fgf15 TG) ⁶³ under similar genetic background were housed in Rutgers pathogen‐free facility with standard environment of 12 h/12 h light/dark cycle (light off from 6 pm to 6 am). For the circadian rhythm, food was taken out at 06:00 when light on (ZT0), and standard rodent chow was put back at 18:00 when light off (ZT12) in order to maintain the regular/consistent eating/fasting rhythm in mice. Mice (n = 3‐5/group) were euthanized every 4 h over a 24‐h period (ZT0, ZT4, ZT8, ZT12, ZT16, and ZT20), and blood, liver, and ileum tissues were collected. This animal study was approved by Rutgers University Institutional Animal Care and Use Committee (IACUC) and conformed to all relevant federal guidelines and institutional policies.

4.3. Determination of human plasma FGF19 levels

Human plasma FGF19 levels were determined by the Human FGF‐19 Quantikine ELISA Kit (R&D Systems) following the manufacture's protocol. Each sample was run in triplicate, and FGF19 concentration was presented as pg/mL.

4.4. Cellular circadian rhythm assay

Cellular circadian rhythms were determined in well‐established circadian responsive reporter MCF10A/hPER2‐dLuc cells using in vitro bioluminescence assay as described previously. ²⁴ Specifically, cells were treated with SIRT1 inhibitor EX527 at 20 or 40 nM or Cambinol at 1 or 2 μM for 1 h in mammary epithelial basal medium (MEBM) following 2‐h synchronization by 50% horse serum. Then the cells were incubated in the circadian recording medium containing 5 ng/mL FGF19, either alone or in combination with 40 nM EX527 or 2 μM cambinol. The luminescence signals were collected from circadian reporter cells with a luminometer LumiCycle 32 (Actimetrics) and analyzed by LumiCycle Analysis Software (Actimerics). The circadian parameters, including period length, phase, rhythm amplitude, and damping rate were obtained by detrending raw data with a running average, and analyzing the best fits to a sine wave using the Levenberg–Marquardt algorithm as described previously. ⁶⁴

4.5. Real‐time quantitative PCR

The mRNA expression level of circadian regulatory genes in PBMCs, NAMPT, and SIRT1, was determined using real‐time quantitative PCR (RT‐qPCR) with Taqman‐based technology as described previously. ⁶¹ The primer information is listed in the Table S1. The housekeeping gene RPS13 was used as an endogenous control. The comparative Ct (2‐∆∆Ct) analysis was used for calculating the relative mRNA expression levels.

The mRNA expression level of genes expressed in mouse liver and ileum was determined using Sybr green‐based technology as described previously. ⁶² , ⁶³ Briefly, total RNA was extracted from liver or ileum by using trizol from Thermofisher. The primers using for the qPCR are listed in the Table S1. The housekeeping gene b‐actin was used as an endogenous control, and relative mRNA expression levels of target genes were determined with comparative delta CT (2^−∆∆Ct) method.

4.6. Quantification of NAD ⁺/NADH in human PBMCs and mouse liver

The nicotinamide nucleotides in the cell or tissue, including NADH, NAD+, and NAD+/NADH ratio were determined by NAD+/NADH Quantification Kit (BioVision) as previously described. ⁶⁵

4.7. Determination of SIRT1 activity in human PBMCs and mouse liver

The SIRT1 activity was quantified by the SIRT1 fluorometric drug discovery assay kit (Enzo Life Sciences) by determining the lysyl deacetylase activity. Total protein from human PBMCs was extracted using extraction buffer and SIRT1 deacetylase activity in the protein extract was determined by the complete fluorescent assay system as previously described. ⁶⁵

For determining the SIRT1 activity in mouse liver, total protein extracts were prepared from frozen liver from WT, Fgf15 TG, and Fgf15 KO mice with RIPA buffer (Sigma Life Science) containing 1× proteinase inhibitor and phosphatase inhibitor cocktail (Thermofisher Scientific). Hepatic SIRT1 deacetylase activity in mice was determined as previously described. ⁶⁵

4.8. BA profiling in human plasma

BAs in the human plasma were extracted and 17 major human BA species were quantified by an Ultra Performance Liquid Chromatography/Electrospray/Ion Trap Mass Spectrometer (UPLC/ESI/ITMS, Thermo Fisher Scientific). The detailed protocol for BA extraction and profiling has described in our previous papers. ⁶⁶ , ⁶⁷ BA profiling is presented in ng/mL as mean ± SEM.

4.9. Statistical analysis

The statistical analysis of human study has been described previously. ⁶¹ The in vitro experiments were performed in triplicate and average results were provided by the software. The circadian rhythmicity of circadian indicators was analyzed by single cosinor analysis using Time Series Analysis (TSA)—Cosinor Software (Expert Soft Tech.). For animal studies, two‐way analysis of variance (ANOVA) followed by Tukey's multiple comparison test was used to test the statistical differences between time points and genotype groups. These data were presented as means ± SEM, and p‐values <.05 was considered statistically significant.

AUTHOR CONTRIBUTIONS

GLG, MF, HZ, and HK conceived and designed the study; BK, ZY, RT, KB, and VB collected and analyzed samples; ZY conducted the statistical analyses; GLG, MF, and ZY wrote the draft; GLG, MF, HZ, BK, and ZY interpreted the data and critically reviewed.

DISCLOSURES

All authors disclose no conflict of interest.

Supporting information

Table S1.

Yang Z, Zarbl H, Kong B, et al. Liver–gut axis signaling regulates circadian energy metabolism in shift workers. The FASEB Journal. 2024;38:e70203. doi: 10.1096/fj.202402102R

DATA AVAILABILITY STATEMENT

The data that support the findings of this study are available within this article.