Sympathetic activation links light exposure and rest-activity rhythm to liver injury and bile acid dysregulation

Abstract

Prolonged light exposure and rest-activity alteration are increasingly prevalent in modern society, yet their health consequence and the underlying mechanism remain poorly characterized. The liver is an organ central to metabolic and circadian regulation. We hypothesized that extended wakefulness, which was modeled as long photoperiod in humans and short photoperiod in nocturnal mice, would affect liver integrity and function. Actigraphic watch, wheel running, morphological, biochemical, transcriptomic, and chromatographic techniques were employed to examine the impact of light exposure alteration. To reverse the circadian disorders mediated by the sympathetic nervous system (SNS) activation, SNS inhibitor or time-restricted feeding was applied. In both humans and mice, extended wakefulness increased daily activity and reduced rest, as well as inducing tissue injury exclusively in the liver and predominantly affecting bile acid metabolism. Mechanistically, short photoperiod enhanced hepatic sympathetic innervation and activity in mice, activating a norepinephrine–β2-adrenergic receptor–protein kinase A signaling axis that reprogrammed the hepatic circadian clock and bile acid metabolism. Chemical sympathectomy not time-restricted feeding restored bile acid rhythmicity and alleviated liver injury. These findings identified the liver as a primary target of circadian disruption and established elevated sympathetic tone as a key mediator linking extended light exposure to metabolic dysfunction.

Introduction

Circadian rhythm is orchestrated by a hierarchical system comprising the central pacemaker in the suprachiasmatic nucleus (SCN) and peripheral clocks in other tissues^1,2. Light serves as a dominant entraining cue for the SCN, which synchronizes with the peripheral clocks through neural and hormonal outputs³. Among these, the autonomic nervous system plays a critical role, with sympathetic signaling modulating peripheral clock function via rhythmic release of norepinephrine (NE)^4,5. In contrast, feeding behavior acts as a potent zeitgeber for peripheral tissues, particularly the liver^6,7. Besides circadian rhythm, animals also respond to seasonal changes in light exposure, i.e., photoperiod. Photoperiod-dependent light changes alter SCN activity and downstream signaling, thereby influencing metabolic processes and clock gene expression in peripheral tissues⁸.

In modern societies, sleep deficiency and prolonged light exposure, as a result of extended artificial lighting and behavioral factors, are increasingly prevalent and are key drivers of circadian misalignment, which increases the incidence of physical and mental diseases⁹. The liver, a central hub for metabolic regulation, exhibits robust circadian oscillations that coordinate lipid, glucose, and bile acid homeostasis. Disruption of hepatic circadian rhythms triggers liver diseases including nonalcoholic steatohepatitis and hepatocellular carcinoma¹⁰. Whether moderate conditions like photoperiodic changes also affect liver integrity and metabolism remains unknown.

The cellular clock circuit consists of a positive (Clock/Bmal1) arm and a negative (Per/Cry) arm and operates by transcriptional-translational negative feedback to generate and maintain rhythmicity^11,12. Clock and Bmal1 activate the transcription of Per and Cry, while Per and Cry accumulate to form a repressor complex, which in turn inhibits the transcriptional activity of Clock and Bmal1. Per and Cry eventually get degraded by the ubiquitin-proteasome system and release the transcriptional repression of Clock/Bmal1, allowing the transcription of Per and Cry to restart and hence another circadian cycle to resume. In a secondary feedback loop, Clock and Bmal1 induce the expression of Rev-Erbα/β, which repress Bmal1 transcription by competing with activator Rorα/γ¹³. About 10% of genes in the cell are controlled by the core clock circuit and exhibit cyclic expression¹⁴.

Bile acids are synthesized in the liver, aiding lipid absorption and acting as signaling molecule in multiple physiological and pathological processess^15,16. Dysregulation of bile acid homeostasis is closely associated with liver injury and metabolic disease. Cholesterol 7α-hydroxylase (Cyp7α1) and cholesterol 27α-hydroxylase (Cyp27α1) are the rate limiting enzymes in the classic and nonclassic bile acids synthesis pathways. The sodium taurocholate co-transporter (Ntcp) and bile salt export pump (Bsep) mediate the sinusoidal reabsorption and canalicular secretion of bile acids. The bile acid metabolism is primarily regulated by farnesoid-X receptor (Fxr), a nuclear receptor. The levels of bile acids as well as the expressions of bile acid synthetic enzymes and transporters exhibit circadian oscillation governed in part by the hepatic clock^17–20. While circadian disruption by restricted feeding or knockout of Per1/Per2 increased the levels of bile acids, alanine aminotransferase (ALT) and aspartate aminotransferase (AST), markers of liver damage¹⁷, it is unknown whether light exposure affects bile acid metabolism.

In this study, we first examined the association between major metabolites/sympathetic nervous system (SNS) transmitter and rest-activity rhythm in a human study, and identified liver injury and bile acid metabolic pathway as the primary targets of extended light exposure. We then created a short photoperiod (SP) mouse model to recapitulate the rest-activity changes in human study and investigated the underlying mechanisms. We found the elevated SNS tone and downstream signaling as a major mechanism underpinning the SP-induced liver injury and bile acid dysfunction. We finally tested the therapeutic benefits of SNS inhibition to ameliorate the SP-induced circadian disorders.

Results

Extended light exposure associates with sympathetic activation and liver injury in humans

We first examined whether differential light exposure and rest-activity rhythm are associated with any tissue injury in humans by combining actigraphy analysis with comprehensive metabolic profiling. Individuals exposed to prolonged light exhibited markedly altered rest-activity duration and rhythm, accompanied by delayed sleep onset time, advanced out-bed time, and increased irregular activity during daytime, when compared with those under normal light exposure (Fig. 1A, B). This was further characterized by reduced total sleep time (TST), increased wake after sleep onset (WASO) and sleep fragmentation index (SFI), collectively indicative of impaired sleep and circadian alignment disorders (Fig. 1C–E).

Fig. 1. Prolonged light exposure is associated with liver injury and sympathetic activation in humans.

A Representative actograms of individuals with normal light exposure (NLE) + normal sleep (NS), long light exposure (LLE) + short sleep (SS), and NLE + SS for 24 h. Blue bars indicate activity with the same scale for each actogram. Black dotted vertical lines indicate 24:00/00:00, red dotted vertical lines denote 06:00, and salmon colored areas represent rest/sleep duration. B Sleep onset time and out-bed time, C TST, D WASO, and E SFI. F Correlation of NE level with ALT or TBA level in the individuals under NLE + NS and LLE + SS conditions. G Correlation of NE level with ALT or TBA level in the individuals under NLE + NS and NLE + SS conditions. Data are presented as mean ± SD.*P < 0.05, **P < 0.01, ***P < 0.001. Blue: NLE + NS, n = 16; Red: LLE + SS, n = 16; Magenta: NLE + SS, n = 8. Nonparametric Spearman correlation was used to determine the correlation between NE level and the level of ALT or TBA.

Metabolic profiling revealed changes specific for hepatocellular injury and disrupted bile acid homeostasis in individuals with extended light exposure, such as elevated blood levels of ALT, AST, and total bile acids (TBA), all markers of liver damage (Table 1). The level of lactate dehydrogenase (LDH) was also increased, which is less specific to liver damage. Notably, the plasma level of NE, a SNS neurotransmitter, was also increased in individuals exposed to extended light and showed a positive correlation with both ALT and TBA levels (Fig. 1F).

Table 1.

Participant characteristics

	Normal light exposure		Long light exposure Short sleep	P value
	Normal sleep	Short sleep
Group characteristics
Male/female	9/7	4/4	8/8	0.933
Age, years	62 (53.5−71.5)	68 (63−74.5)	61.5 (54.5−69)	0.384
Plasma tests
ALT, U/L	30.00 (23.50−38.00)a	62.00 (49.00−71.25) a	136.50 (80.25−173.00)b	<0.0001
AST, U/L	29.00 (21.75−38.50)a	42.50 (36.75−60.25)a	153.00 (111.75−211.00)b	<0.0001
TBA, mmol/L	1.50 (0.98−2.35)a	1.55 (1.33−3.05)a	6.05 (4.95−6.85)b	<0.0001
LDH, IU/L	237.00 (194.00−256.50)a	224.50 (172.00−292.25)a	638.00 (542.75−717.25)b	<0.0001
ALP, U/L	97.00 (81.00−124.00)	74.5 (56.75−103.25)	102.00 (85.50−125.00)	0.072
Urea, mmol/L	4.30 (2.73−5.98)	4.00 (2.98−5.55)	4.55 (3.00−6.20)	0.760
Cr, μmol/L	61.50 (49.25−74.00)	46.00 (37.75−63.25)	54.25 (47.00−74.00)	0.057
TBil, IU/L	0.31 (0.29−0.35)a	0.56 (0.40−0.72)b	0.23 (0.19−0.27)c	<0.0001
DBil, U/L	0.30 (0.26−0.35)ba	0.27 (0.19−0.37)a	0.18 (0.15−0.22)b	<0.0001
HDL, mmol/L	1.05 (1.02−1.11)	1.04 (0.96−1.15)	1.07 (0.91−1.16)	0.905
LDL, mmol/L	0.74 (0.56−0.97)	0.80 (0.59−1.14)	0.71 (0.57−0.84)	0.873
Glucose, mmol/L	4.75 (3.93−5.53)ab	5.65 (4.83−6.25)a	4.50 (3.75−5.35)b	0.05
Tp, g/L	73.20 (65.28−75.45)	74.50 (73.00−81.00)	66.70 (59.40−74.40)	0.204
ALB, g/L	46.50 (39.25−49.50)ab	48.00 (46.25−50.50)a	42.50 (37.50−46.75)b	0.038
TG, mmol/L	1.00 (0.77−1.35)a	1.12 (0.69−1.59)b	1.40 (0.50−1.80)ab	0.016
TC, mmol/L	3.97 (3.08−4.87)	4.75 (3.83−5.25)	3.70 (2.98−4.73)	0.519
NE, nmol/L	2.98 (2.25−3.45)a	2.25 (1.73−2.58)a	8.40 (6.13−10.58)b	<0.0001

Data are presented as number or median (quartile 1 – quartile 3). Superscript letters (a, b, c) indicate significant differences at P < 0.05; groups with the same letter are not significantly different. ALT alanineamino transferase, AST aspartate amino transferase, TBA total bile acid, LDH lactate dehydrogenase, ALP alkaline phosphatase, BUN blood urea nitrogen, Cr creatinine, DBil direct bilirubin, TBil total bilirubin, HDL-C high-density lipoprotein cholesterol, LDL-C low-density lipoprotein cholesterol, Tp total protein, ALB albumin, TG triglycerides, TC total cholesterol, NE norepinephrine.

The level of TBA, not ALT or NE, was correlated with TST (Fig. S1A), suggesting that bile acids (not other metabolites) likely directly regulate circadian rhythmicity and locomotor activity. This is consistent with our previous report of bile acid’s role in sleep disorders induced by chronic liver disease²¹. Because sexual dimorphism exists in SNS activity²² and bile acid metabolism²³, we further analyzed our human data based on sex and found the correlation between NE and ALT or TBA virtually identical between males and females (Fig. S1B). This suggests that the relationship between sympathetic activation and liver injury in extended light exposure is regardless of sex.

To validate that prolonged light exposure rather than increased activity per se is associated with liver injury and disrupted bile acid metabolism, we also included a group of individuals who had normal light exposure (LD: 16:8) but reduced rest and increased activity at night (Fig. 1A). The individuals under this yoked short sleep condition showed comparable changes in activity and sleep (TST, WASO and SFI) but no significant difference in liver injury (ALT, AST and TBA) and sympathetic activation (NE) which showed no correlation either (Table 1 and Fig. 1A–E, G). This result suggests that the “yoked” short sleep condition does not uniformly affect the liver or SNS as extended light exposure. Together, these data demonstrate that prolonged light exposure and more precisely nighttime light exposure accounts for sleep disruption, sympathetic activation, liver injury and disrupted bile acid metabolism in humans.

Short photoperiod recapitulates circadian disruption and liver injury in mice

To model extended wakefulness in a controlled setting, we exposed nocturnal mice to a short photoperiod (SP; LD 4:20), which prolongs the activity phase (corresponding to light exposure phase in diurnal animals) and mimics extended light exposure in humans. As the liver and sympathetic phenotypes showed no sexual dimorphism, we chose male mice for the rest of the investigation. Compared with mice under normal photoperiod (NP; LD 12:12), SP mice exhibited increased locomotor activity and duration, accompanied by irregular running patterns (Fig. 2A–C and S2). Phase angle of entrainment as well as onset and offset variability were also increased, indicating disrupted rhythmicity (Fig. 2D, E). Consistent with elevated energy expenditure associated with increased activity and reduced rest, SP mice displayed reduced body weight and fat mass, despite unchanged food intake and unaltered levels of appetite-regulating hormones such as leptin and ghrelin (Fig. S3A–E). Hematological parameters, including red blood cell count (RBC) and hemoglobin (HGB) levels, were increased, reflecting physiological adaptation to heightened activity (Fig. S3F–H).

Fig. 2. Short photoperiod induces liver injury and higher NE level in mice.

A Representative actograms of the wheel-running experiments for two consecutive days, n = 3 for each group. B, C Wheel revolution and running/rest time, n = 4. D Onset at ZT4 and offset at ZT24 phase angle of entrainment, n = 4. E Onset at ZT4 and offset at ZT24 variability of entrainment, n = 4. F H&E staining of the liver tissues from the mice placed under the NP and SP schedule for 1, 2, and 3 months, n = 3. G, H TUNEL staining and quantitation of the mouse liver sections from the mice placed under the NP and SP schedule for 1, 2, and 3 months, n = 3. I, J Immunohistochemical staining and quantitation of F4/80, NF-κB, Zo-1, Claudin2, Mrp2 in the liver from the mice under the NP and SP schedule for 2 months, n = 3. K Correlation of NE with TBA and ALT. Data are presented as mean ± SD. *P < 0.05, **P < 0.01, ***P < 0.001. One-way ANOVA with Tukey’s multiple-comparisons test.

Metabolic profiling revealed a selective increase in liver injury markers, including ALT, AST, and TBA, whereas markers of renal function, lipid metabolism, and hepatic synthetic capacity remained largely unchanged (Table 2). Histological analysis confirmed progressive liver injury, characterized by disrupted lobular architecture, inflammatory infiltration, and increased hepatocyte apoptosis (Fig. 2F–H). In the SP mice, the LDH level was also increased, which is not specific to liver damage. This prompted us to examine the H&E staining of other tissues, including heart, kidney, spleen, and stomach, which showed no alteration of tissue structure or inflammatory response, i.e., little sign of injury (Fig. S4A). Immunohistochemistry confirmed inflammatory response (increased F4/80 and NF-κB staining) and disrupted cell polarity (altered staining of tight junction/canalicular proteins ZO-1, Claudin 2, and Mrp2) in the SP livers (Fig. 2I, J).

Table 2.

The metabolites levels in NP and SP mice

	NP			SP
	Month 1	Month 2	Month 3	Month 1	Month 2	Month 3
ALT (U/L)	28.33 ± 1.96	29.00 ± 2.82	30.57 ± 8.06	34.57 ± 6.77	47.23 ± 9.26*	76.20 ± 13.46**
AST (U/L)	21.64 ± 0.01	18.52 ± 0.01	27.64 ± 0.03	116.47 ± 13.97**	260.77 ± 37.41**	178.43 ± 10.97**
LDH (IU/L)	349.67 ± 68.63	381.4 ± 37.60	369.27 ± 62.50	807.63 ± 210.87*	605.33 ± 97.33*	584.17 ± 94.41*
ALP (U/L)	92.00 ± 7.00	87.33 ± 11.93	90.00 ± 4.58	100.33 ± 15.86	90.56 ± 10.42	73.67 ± 10.97
Urea(mmol/L)	12.85 ± 1.77	12.15 ± 0.90	13.09 ± 0.50	11.67 ± 0.99	12.64 ± 0.99	12.24 ± 1.23
Cr (μmol/L)	14.73 ± 3.56	15.52 ± 2.28	14.98 ± 3.75	13.93 ± 3.17	14.87 ± 3.86	15.13 ± 5.21
TBA (mmol/L)	2.86 ± 0.71	2.88 ± 2.07	3.15 ± 2.85	7.29 ± 2.06*	11.77 ± 2.33*	4.84 ± 0.86
TBil (IU/L)	0.33 ± 0.22	0.28 ± 0.08	0.26 ± 0.14	0.22 ± 0.14	0.27 ± 0.18	0.14 ± 0.11
DBil (U/L)	0.30 ± 0.70	0.31 ± 0.1	0.28 ± 0.09	0.17 ± 0.10	0.10 ± 0.08	0.09 ± 0.04
HDL (mmol/L)	1.00 ± 0.35	0.99 ± 0.36	1.11 ± 0.25	1.03 ± 0.18	1.07 ± 0.41	1.14 ± 0.42
LDL (mmol/L)	0.14 ± 0.02	0.12 ± 0.02	0.12 ± 0.03	0.19 ± 0.04	0.20 ± 0.06	0.17 ± 0.02
Glucose (mmol/L)	4.40 ± 0.45	4.37 ± 0.27	4.16 ± 0.36	3.23 ± 0.24*	4.21 ± 0.24	5.14 ± 0.76
Tp (g/L)	42.7 ± 0.89	44.83 ± 4.26	42.27 ± 9.92	47.93 ± 47.40	43.40 ± 1.51	44.8 ± 4.30
ALB (g/L)	24.6 ± 0.61	29.17 ± 4.30	27.27 ± 6.67	27.67 ± 1.02	26.60 ± 0.61	26.07 ± 1.77
TG (mmol/L)	0.55 ± 0.05	0.51 ± 0.19	0.58 ± 0.09	0.66 ± 0.11	0.49 ± 0.09	0.46 ± 0.13
TC (mmol/L)	2.27 ± 0.13	2.31 ± 0.19	2.30 ± 0.23	2.05 ± 0.39	1.60 ± 0.23	2.07 ± 0.15

Data are shown as mean ± SD. * P < 0.05 and ** P < 0.01. In each group, n = 6.

Importantly, hepatic NE levels were also elevated and strongly correlated with ALT and TBA levels (Fig. 2K), mirroring the observations in humans. To investigate if the SP exposure might induce stress in mice, we checked the levels of glucocorticoid and triiodothyronine, two stress-triggered hormones, which remained unchanged (Fig. S4B and C), suggesting distinct mechanisms transcending the signal from SCN and underlying the liver injury in the SP mice. Together, these results demonstrate that SP selectively induces liver injury and bile acid dysregulation in association with increased sympathetic activity.

Short photoperiod induces hepatic sympathetic remodeling and activation

Given the observed increase in NE and unaltered levels of stress hormones, we next examined whether SP alters hepatic autonomic innervation. SP markedly increased the expression of tyrosine hydroxylase (TH) and growth-associated protein 43 (GAP43), markers of sympathetic innervation and nerve growth, while reducing the level of choline acetyltransferase (ChAT), a parasympathetic marker (Fig. 3A, B).

Fig. 3. Short photoperiod enhances hepatic SNS activity in mice.

A, B Western blots and quantitation of TH, ChAT, and GAP43 in the NP and SP liver at ZT2, ZT8, ZT14, and ZT20, n = 3. C, D Immunohistochemical staining and quantitation of TH in the NP and SP liver. E, F Immunohistochemical staining and quantitation of GAP43 in the NP and SP liver. All the images are at original magnification of 400 X, n = 3. G ELISA analysis of NE levels in the NP and SP liver, n = 3. H Image of electrochemical recording of NE in liver slice using electrical stimulation (E-stim). I, J Representative amperometric currents and quantitation from the liver slices from NP and SP mice. n = 5 slices from 3 mice, ***P < 0.001, two-tailed paired t-test. K Image of electrochemical recording of NE using high K⁺ as a stimulus. L, M Representative amperometric currents evoked by treatment with 80 mM KCl and quantitation. n = 5 slices from 3 mice, ** P < 0.01, two-tailed paired t-test. Data are presented as mean ± SD.

Immunohistochemical analyses revealed that sympathetic fibers were primarily localized to the portal regions and exhibited circadian oscillation under the NP condition (Fig. 3C–F). SP induced a robust increase in TH and GAP43 expression without altering their oscillating peaks, i.e., enhancing sympathetic tone rather than reprogramming rhythmic phase. These immunohistochemical data were consistent with the results from Western blotting. Consistent with the changes in TH and GAP43, the hepatic NE content was also higher in SP mice and exhibited a similar circadian oscillating pattern (Fig. 3G). Furthermore, carbon fiber amperometry revealed that hepatic NE release was increased in SP mice, as evoked by either electrical field stimulation (Fig. 3H–J) or higher potassium (Fig. 3K–M). These data demonstrate that SP induces hepatic remodeling of autonomic innervation, favoring sympathetic dominance.

Short photoperiod reprograms the hepatic circadian clock and transcriptome

Transcriptomic profiling revealed widespread gene expression changes in the livers of the mice placed under SP after two months. Notably, the expression of many genes remained unchanged one month after being placed under SP but changed dramatically after two month, and a small number of genes changed their expression in opposite direction from one month to two months (Fig. 4A). This collectively results in the overlapping DEGs between SP1 and SP2 to be rather small (Fig. 4B). Gene ontology analysis highlighted enrichment in pathways related to circadian rhythm, bile acid metabolism and inflammatory process (Fig. 4C, D). The progressive divergence of DEGs from NP to SP1 and SP2 could reflect the course of adaptation to SP: first physiological adaptation and then progressing to pathological condition, which is corroborated by increased inflammatory response seen in Fig. 4D. Owing to such differential expression, we thereafter chose 2 months as a standard time point for the rest of our studies.

Fig. 4. Transcriptomic profiling of the NP and SP livers.

A Heat map of the DEGs identified by microarray analysis from the mice placed under NP or SP for one (SP1M) or two months (SP2M), n = 3. B The number of DEGs in the SP1M and SP2M groups and overlapping genes. C The DEG-enriched GO-biological process identified from over-representation analysis. D The heat maps show the representative DEGs between SP1M or SP2M group and NP group involved in circadian rhythm, bile acid metabolism and inflammatory response, n = 3.

Analysis of core clock components revealed a profound imbalance between the positive and negative arms of the molecular clock. In the NP liver, the mRNA expressions of Clock and Bmal1, the positive regulators, showed a circadian oscillation with peaks at (Zeitgeber time) ZT14 - ZT22, while Per1 and Per2, the negative regulators, exhibited an opposite oscillation with the peaks at ZT6 - ZT14 (Fig. 5A). In the SP liver, the expression of Clock and Bmal1 was markedly reduced and phase shifted, whereas the levels of Per1 and Per2 were increased with their oscillation unchanged. In the NP liver, the expression of Rev-erbα and RORα, the transcriptional repressor and activator of Bmal1, oscillated with a peak around ZT6 (Fig. 5A). SP also increased and reduced the mRNA levels of Rev-erbα and RORα, respectively, with oscillation unchanged. The levels and oscillation of other clock genes including Bmal2, Per3, Cy1, Cy2, Dbp2, Npas2 and Nrld2 remained unchanged (Fig. S5). Importantly, these qPCR data confirmed the DEG changes between SP2 and NP observed in the microarray study. The changes at mRNA level in Clock, Per1 and Rev-erbα were further confirmed by Western blotting (Fig. 5B, C) and immunohistochemistry (Fig. 5D, E). Together, these findings indicate that SP disrupts the amplitude and phase of the hepatic clock circuit, favoring a repressive transcriptional state.

Fig. 5. Circadian oscillation of the clock genes in the NP and SP livers.

A qPCR analysis of Clock, Bmal1, Per1, Per2, Rev-erbα and RORα from the NP and SP mice at ZT2, ZT6, ZT10, ZT14, ZT18 and ZT22, with the normal (LD 12:12) and short (LD 4:20) photoperiod indicated at the top and bottom, respectively. B, C Western blots and quantitation of Clock, Per1 and Rev-erbα in the NP and SP liver at ZT2, ZT6, ZT10, ZT14, ZT18 and ZT22. β-actin serves as a loading control. D, E Immunohistochemical staining and quantitation of Clock, Per1 and Rev-erbα. All the images are at original magnification of 400 X. * P < 0.05, ** P < 0.01, *** P < 0.001. n = 3 for each group. Data are presented as mean ± SD.

Short photoperiod disrupts bile acid homeostasis and its circadian rhythmicity

Consistent with the elevated circulating bile acids, SP significantly increased the bile acid pool size across the liver, gallbladder, and intestine (Fig. 6A). Circadian analysis revealed both increased amplitude and phase shifts in serum bile acid rhythms, with a ~4-h advance in peak levels (Fig. 6B). Hepatic bile acids were similarly elevated, although their oscillatory pattern was preserved (Fig. 6C).

Fig. 6. Bile acid profiles and circadian oscillation in the NP and SP livers.

A The bile acid pool size and bile acid levels in the liver, gallbladder, and intestines in the NP and SP mice (kept for 1, 2, and 3 months). B The levels of serum TBA measured at ZT2, ZT6, ZT10, ZT14, ZT18 and ZT22 in the NP and SP mice (kept for 2 months). C The levels of liver TBA measured at ZT2, ZT6, ZT10, ZT14, ZT18 and ZT22. D The liver bile acid composition analyzed by mass spectrometry. E The levels of conjugated and unconjugated primary and secondary bile acids. F qPCR analysis of Cyp7α1, Cyp27α1, Fxr, Shp, Bsep and Ntcp. Western blots (G) and quantitation (H) of Cyp7α1, Cyp27α1, Fxr, Bsep and Ntcp. β-actin serves as a loading control. Immunohistochemical staining (I) and quantitation (J) of Cyp7α1, Cyp27α1 and Fxr. All the images are at original magnification of 400 X. *P < 0.05, **P < 0.01, ***P < 0.001, n = 3. Data are presented as mean ± SD.

Compositional analysis demonstrated a marked shift toward conjugated bile acids and a reduction in unconjugated species (Fig. 6D, E), indicative of enhanced synthesis and altered enterohepatic circulation. At the molecular level, SP upregulated the expression of Cyp7a1 and downregulated Fxr and small heterodimer protein (Shp), a transcription factor induced by Fxr and directly activating Cyp7a1 and Bsep transcription, all of which are consistent with increased bile acid synthesis (Fig. 6F–J). Transporters were also dysregulated, with decreased Bsep and increased Ntcp expression. These coordinated changes promote bile acid accumulation and impair bile acid excretion, providing a mechanistic basis for cell damage, apoptosis, and eventually liver injury.

Sympathetic activation drives hepatic clock and bile acid reprogramming via β2-adrenergic signaling

To directly test the role of sympathetic signaling, we isolated and stimulated the fourth cervical branch of the left splanchnic nerve hepatic branch. The liver mRNA levels of Rev-erbα, Per1, Ntcp and Cyp7a1 were increased after electrical stimulation treatment (EST), while the expression of Clock, Bmal1, Bsep and Fxr was decreased (Fig. 7A). Western blotting confirmed increases of Rev-erbα, Per1, NTCP and Cyp7α1 and decreases of Clock, Bmal1, Bsep and Fxr after EST (Fig. 7B, C). Immunohistochemistry also confirmed the increase of Rev-erbα and Per1 (Fig. 7D, E). These changes upon EST recapitulated the alterations of these genes in the SP liver.

Fig. 7. NE-β2AR-PKA pathway mediates sympathetic activation and hepatic metabolic reprograming.

A qPCR analysis of Clock, Bmal1, Rev-erbα, Per1, Cyp7α1, Fxr, Bsep and Ntcp from the sham and EST liver. Western blots (B) and quantitation (C) of Clock, Bmal1, Rev-erbα, Per1, Cyp7α1, Fxr, Bsep and NTCP. β-actin serves as a loading control. Immunohistochemical staining (D) and quantitation (E) of Rev-erbα and Per1. Scale bar = 20 μm. Western blots (F) and quantitation (G) of Clock, Bmal1, Rev-erbα, P-Creb, Per1, Cyp7α1, Fxr, Bsep, Alas1 and NTCP from the HepG2 cells treated with NE, Propranolol, Bisoprolol or KT5720 as well as untreated negative control (NC). *P < 0.05, **P < 0.01, ***P < 0.001, n = 3. One-way ANOVA with Tukey’s multiple-comparisons test. Data are presented as mean ± SD.

NE is known to activate β-adrenergic receptors (β-ARs), leading to cAMP production, which in turn activates protein kinase A (PKA). PKA can phosphorylate cAMP-response element binding protein (CREB), a transcription factor binding to the target gene promoter and activating transcription. To characterize the downstream signaling of NE, we treated the hepatoma HepG2 cells with NE alone or in combination with Propranolol (a nonselective antagonist for β1- and β2-adrenergic receptors), Bisoprolol (a selective blocker for β1-adrenergic receptors), or KT5720 (a selective inhibitor for PKA). NE alone increased the phosphorylation of Creb and recapitulated the SP-induced changes in Clock, Bmal1, Per1, Rev-erba, Суp7a1, Fxr, Bsep and Ntcp in the HepG2 cells, thus validating this cell model (Fig. 7F, G). Propranolol or KT5720 abolished these effects, whereas Bisoprolol had no impact (Fig. 7F, G). These results demonstrate a NE–β2-AR–PKA axis as the mediator linking sympathetic activation to hepatic clock and metabolic reprogramming.

Compared with the sham condition, EST also increased the HGB content and RBC count (Fig. S6A), mirroring the changes in the SP mice. Per1 can regulate heme biosynthesis through delta-aminolevulinate synthase 1 (Alas1), a mitochondrial enzyme controlling both heme synthesis and bile acid metabolism. Indeed, EST also increased the expression of Alas1 (Fig. S6B). Treatment with NE, Propranolol, Bisoprolol and KT5720 in the HepG2 model also showed that Alas1 acted downstream of NE through the β2-AR–PKA axis (Fig. 7F, G).

Sympathetic inhibition, not time-restricted feeding, rescues liver injury and metabolic disruption

Finally, we tested whether repressing sympathetic nerve could reverse SP-induced liver damage and disruption of bile acids homoeostasis using chemical sympathectomy with 6-OHDA. This intervention effectively reduced hepatic sympathetic markers TH and GAP43 (Fig. 8A–D) and NE release (Fig. 8E, F), significantly ameliorated liver injury, apoptosis (Fig. 8G–I), and higher levels of ALT, AST, and TBA, restored circadian oscillation of serum TBA (Fig. 8J, L) and circadian oscillation of the liver clock genes, Clock, Bmal1, Per1 and Rev-erbα (Fig. 8K). In order to better dissect the role of 6-OHDA alone on circadian rhythm and liver biology, we also included a condition whereby the mice exposed to NP were treated with 6-OHDA. Treatment with 6-OHDA alone did reduce NE release as well as the level of TH and GAP43 (Fig. 8A–D). The impact of 6-OHDA on the liver integrity (e.g., ALT, AST, and apoptosis) and its clock circuit was rather modest, consistent with the previous reports^24,25.

Fig. 8. The impact of 6-OHDA and time-restricted feeding on the liver damage and bile acid disruption induced by SP.

A, B Western blots and quantitation of TH and GAP43 in liver tissues at ZT14 from the NP mice, NP mice treated with 6-OHDA, SP mice, SP mice with restricted feeding (RF), and SP mice treated with 6-OHDA. β-actin serves as a loading control. C, D Immunohistochemical staining and quantitation of liver TH and GAP43 at ZT14 from the above mice. E, F Representative amperometric currents and quantitation from the liver slices from the above mice. n = 3 slices from 3 mice, **P < 0.01, ***P < 0.001, one-way ANOVA with Tukey’s multiple-comparisons test. G H&E liver staining of the above mice. H, I TUNEL staining and quantitation of the liver sections from the above mice. J The serum levels of ALT and AST in the above mice. K qPCR analysis of Clock, Bmal1, Per1, and Rev-erbα at ZT2, ZT6, ZT10, ZT14, ZT18, and ZT22. L The serum and liver TBA levels measured at ZT2, ZT6, ZT10, ZT14, ZT18 and ZT22. *P < 0.05, **P < 0.01, ***P < 0.001 by one-way ANOVA with Tukey’s multiple-comparisons test, n = 3. Data are presented as mean ± SD.

As a comparison, we also tested the benefit of time-restricted feeding on SP-induced liver injury. We fed the SP mice exclusively between ZT12 and ZT24, i.e., in their normal active phase, to restore the rhythmicity of liver clock. Restricted feeding partially restored bile acid rhythmicity but failed to reduce sympathetic activation or improve liver injury, nor fully the oscillation of liver clock circuit (Fig. 8A–K). These findings indicate that sympathetic activation is a primary driver in liver injury and bile acid dysfunction induced by extended wakefulness, whereas feeding cues only partially modulate rhythmic phase but not amplitude.

Discussion

This study identified the liver as a primary target of circadian disruption induced by extended wakefulness and established sympathetic activation as a central mechanism linking environmental light exposure to metabolic dysfunction. Using complementary human and mouse models, we demonstrated that prolonged light exposure selectively impairs liver integrity and bile acid homeostasis, with minimal effects on other organs. Mechanistically, we show that enhanced sympathetic tone reprograms the hepatic circadian clock and bile acid metabolism via a NE–β2-AR–PKA signaling axis. Although our findings suggest similar mechanisms underpinning the pathologies in humans exposed to long photoperiod, we haven’t directly determined causality using sympathetic inhibition. Therefore, it is plausible that cross-species differences may exist.

The photoperiodic light influences the circadian system via complex mechanisms. Distinct components of seasonal light, e.g., photoperiod length and total day-night cycle (T-cycle), differentially affect the activity duration (α), free-running period (τ), and phase angle of entrainment (ψ)²⁶. The photoperiods drive changes in α, whereas the photoperiod length combined with T-cycle length influence τ. Both photoperiod and T-cycle length influence ψ. In our study, we kept the T-cycle kept constant as 24 h and the SP schedule increased α and ψ, which is in line with the previous report²⁶. The increased irregular activities in the SP mice were also reported in the Djungarian and Syrian hamsters, which exhibited irregular, unstable running after transfer from long days to short days^27,28.

Notably, the liver and bile acid homoeostasis seem most sensitive to extended light exposure. A previous study also reported that constant lighting or nocturnal light disruption in rats induced necrosis and apoptosis of hepatocytes, altered the liver rhythmicity of Per2 and Bmal1, and stimulated hepatocyte proliferation²⁹. In humans, increased nocturnal light exposure is associated with higher risk of obesity, metabolic syndrome, and metabolic dysfunction–associated steatotic liver disease, in addition to poor sleep quality³⁰. Furthermore, individuals with brighter night light had higher all-cause mortality risk^31,32.

It should be noted that the SP condition in our study represents a moderate circadian dysregulation, compared with constant light, jet lag, or shift work. Nevertheless, SP in mice or long light exposure (LLE) in humans, which basically represents nighttime light pollution, is different from the seasonal response to photoperiod changes in several key ways: (1) nighttime light pollution was passively enforced on mice or humans, which cannot be fully adapted and induce pathological conditions after exceeding organism’s physiological adaptation capacity; (2) seasonal photoperiodism is a response to regular and gently progressing light change that organisms can fully anticipate and actively adapt to; (3) owing to differential adaptability seasonal light change induce oscillating alteration to much lesser extent, while nighttime light pollution brings about rhythmic changes not only in phase but also in amplitude. In short, nighttime light pollution fundamentally disrupts circadian rhythm as it acts as zeitgebers in the opposite phase, while seasonal light change provides adaptable signal allowing organisms to live in better harmony with the changing light-dark environment. Similarly, the polar residents who live in the Arctic or Antarctic experience extreme photoperiodism, including midnight sun and polar night. However, most polar residents do not manifest physical or mental disorders, largely owing to active light and sleep adaptation, despite that rhythmic alteration in their liver gene transcription and metabolic profiles were reported^33,34.

A key conceptual advance of this work is the identification of sympathetic signaling as an upstream regulator of hepatic circadian dysfunction. Light exposure is known to increase sympathetic tone and suppress vagal nerve output, and these changes can be abolished by SCN lesion^35,36. While previous studies tend to emphasize the role of intrinsic clock disruption or feeding behavior in metabolic disease, our findings of remodeled hepatic autonomic innervation, characterized by increased sympathetic and reduced parasympathetic inputs, position neural output as a dominant driver triggering metabolic dysregulation. Elevated sympathetic activity was also reported in the other circadian misalignment models, like chronic jet lag, restricted feeding, or Per1/Per2 or Cry1/Cry2 knockout mice^37,38.

Our data further reveal a mechanistic link that elevated SNS acted via the NE-β2-AR-PKA-Creb axis to regulate the liver clock gene expression and bile acid metabolism (Fig. 9). A previous study showed that SCN lesion abolished the rhythms of Per1, Per2, and Bmal1 as well as the NE level in mouse liver, which was restored by epinephrine³⁹. NE signaling has been proposed to modulate the clock gene expression via cAMP and PKA^40–42. Under the SP schedule, downregulation of positive regulators and upregulation of negative regulators of the liver clock circuit dampened its rhythmicity, which can alter the expression of the clock-controlled gene, like those in bile acid metabolism (Fig. 9).

Fig. 9. Sympathetic activation imbalances the liver clock circuit and disrupts bile acid metabolism in extended wakefulness.

The diagram illustrates how prolonged light exposure in humans enhances the hepatic release of NE, which changes the balance of the liver clock circuit via two transcription-translation feedback loops and disrupts bile acid metabolism. SNS activation also activates Alas1 and affects heme synthesis. SNS, sympathetic nervous system; TTFL, transcription-translation feedback loop.

PKA also directly activates Fxr⁴³, which in turn controls the expression of Cyp7A1, Cyp8B1, Cyp27A, NTCP and BSEP. Rev-Erbα also regulates bile acid synthesis via Shp and Cyp7α1⁴⁴ (Fig. 9). Accumulated bile acids are toxic in hepatocytes and eventually trigger apoptosis. In addition to the clock circuit-mediated regulation described in this study, SNS also modulates bile acid metabolism through several other mechanisms. SNS activation via stimulation of the splanchnic nerves or NE infusion can reduce the bile acid-independent bile secretion, leading to bile acid accumulation in the liver and blood^45,46. In certain scenario, high levels of bile acids can impair and cross blood-brain permeability barrier, directly impacting on SCN to modulate circadian rhythm²¹.

Finally, our intervention studies demonstrate that targeting sympathetic activity is sufficient to restore both circadian rhythmicity and metabolic homeostasis, whereas restricted feeding primarily corrects phase without improving overall metabolic integrity. These findings highlight a hierarchical relationship in which neural signals dominate over feeding cues in shaping hepatic responses to circadian disruption.

From a translational perspective, our findings provide a mechanistic framework linking modern lifestyle factors—such as light pollution and sleep deficiency—to liver disease. The selective vulnerability of the liver suggests that circadian misalignment may contribute to early, subclinical metabolic injury before the onset of systemic disease. Therapeutically, modulation of sympathetic activity or downstream β2-adrenergic signaling may represent promising strategies for mitigating circadian-related metabolic disorders.

Methods

Human study

Forty volunteers were selected based on their response to sociodemographic and sleep questionnaires and were recruited by the Third People’s Hospital of Taiyuan for this study. Among those, 16 individuals maintained a normal rest-activity (close to LD 16:8) rhythm, 16 individuals had extended light exposure and sleep reduction, and 8 individuals had normal light exposure (LD 16:8) but with greater nighttime activity. The light exposure time and rest-activity patterns are the key inclusion criteria of this study. In addition, we only selected similarly aged individuals (age 53.5–75.5), as we are aware of the prevalence of sleep disorders in aging population. We had no sex preference in volunteer inclusion. Exclusion criteria included diagnosed significant cardiovascular, pulmonary, renal, or neurological disease; diabetes; significant untreated pain; severe depression, anxiety; diagnosed insomnia, sleep apnea; shift work; medication use; drug abuse and heavy alcohol use. The human study followed the Declaration of Helsinki Principles and the Guideline issued by the Ethics Committee of Taiyuan Third People’s Hospital (Approval No. 2023-21) with informed consent forms completed by all participants. The rest-activity (sleep-wake) cycle was monitored using actigraphic watch equipped with ActiGraph GT3X+ accelerometer (ActiGraph, Pensacola, FL) and worn by participants for 3 consecutive days. The ActiLife 6 software (ActiGraph, Pensacola, FL) was used to analyze the rest-activity parameters as described previously²¹. This is a at-home study with no blinding involved. The blood samples were taken for metabolite analysis from the fasting individuals at 8 A.M. on the third day of this study.

Wheel-running experiment in mice

Two hundred and forty-six 6-week-old male C57BL/6J mice purchased from the Experimental Animal Center, Chinese Academy of Medical Sciences were housed in a temperature-controlled facility (25 ± 1 °C) and maintained under a 12:12 h light-dark (LD 12:12: 12 h under 200 lux bright light, ZT0 – ZT12; 12 h darkness, ZT12 – ZT24) cycle for 2 weeks before the experiments, with chow and water provided ad libitum. The animal experiments were approved by the Ethics Committee of Shanxi Medical University (Approval ID SYDL2023014), following the criteria outlined in the “Guide for the Care and Use of Laboratory Animals” prepared by the National Institutes of Health (86–23 revised 1985). Efforts were made to minimize the number of animals and animal suffering.

The mice performed voluntary wheel running, and the locomotor activities were continuously monitored by the VitalView Data Acquisition System (Mini-Mitter Inc., USA), with data analyzed as previously described⁴⁷. After the wheel-running experiments, mice were sacrificed by cervical dislocation. On the last day of the NP and SP experiments, the mice were kept fasting, with blood samples taken and tissues harvested from the sacrificed mice immediately after the experiments. Tissues were harvested at ZT2, ZT6 and ZT10 under room light and at ZT14, ZT18 and ZT22 under dim red light (4 lux), fixed in 4% formalin or stored in liquid nitrogen for subsequent analyses.

Short photoperiod, time-restricted feeding and chemical sympathectomy

After adaptation in LD 12:12 environment, some C57BL/6J mice were placed under the LD 4:20 (4 h under 200 lux bright light, ZT0 – ZT4; 20 h darkness, ZT4 – ZT24) SP schedule for 1–3 months, while other mice were continued under the LD 12:12 normal photoperiod (NP). Such a longitudinal study was to characterize how SP induces phenotypic changes over time. For time-restricted feeding (RF), the mice were fed at the “normal” nighttime between ZT12 and ZT24 for one month. For chemical sympathectomy, hydroxydopamine hydrochloride (6-OHDA), which is a dopamine inhibitor and neurotoxin selectively destroying sympathetic nerves, was prepared in 0.1% of ascorbic acid (w/v) and intraperitoneally injected at 100 mg/kg bodyweight once a day for 3 days. The number of mice for NP, SP, RF, and 6-OHDA groups is 80. 80. 45 and 45, respectively.

Histology and in situ apoptosis detection

The tissues of heart, liver, spleen, stomach, and kidney were fixed with 4% formalin and embedded in paraffin. Five-μm tissue sections were cut with a microtome. Hematoxylin and eosin (H&E) staining were performed, and images were captured under an Olympus BX51 microscope. Detection of apoptosis in mouse liver tissues was performed using in situ Apoptosis Detection Kit (Abcam, ab206386). Briefly, liver sections were deparaffinized, rehydrated, and reacted with a 3% hydrogen peroxide/methanol solution to inactivate endogenous peroxidase. Apoptotic nuclei were labeled with biotinylated terminal deoxynucleotidyl transferase and detected using a streptavidin-horseradish peroxidase conjugate. Apoptotic cells were visualized by diaminobenzidine.

Laboratory tests

The levels of ALT, AST, LDH, alkaline phosphatase (ALP), blood urea nitrogen (BUN), creatinine, TBA, direct bilirubin (DBil), total bilirubin (TBil), high-density lipoprotein cholesterol (HDL-C), low-density lipoprotein cholesterol (LDL-C), fasting glucose, total protein and albumin, triglycerides (TG), total cholesterol (TC), haemoglobin and serum norepinephrine (NE) were measured by the Department of Laboratory Medicine, Third People’s Hospital of Taiyuan.

Enzyme linked immunosorbent assay (ELISA)

The serum and liver hormonal levels were determined using ELISA kits, including those for triiodothyronine (T3) (Runyu Biotechnology, Shanghai, China), glucocorticoid (Feiya Biotechnology, Jiangsu, China), leptin and ghrelin (Meimian Biology, Shanghai, China). Briefly, 500 μL of blood was taken from mouse eyes under anesthesia, and serum was prepared after centrifugation at 2500 rpm for 15 min. Optical absorbance values were measured at 450 nm and hormone levels were determined using a standard curve.

Bile acid extraction and liquid chromatography-mass spectrometry (LC-MS)

Mouse liver samples were extracted by protein precipitation using ice-cold alkaline ACN (5% NH₄OH in ACN). Approximately 100 mg of liver tissues was homogenized in 200 μL of 50% methanol. The bile acids were analyzed as previously described²¹.

Microarray and pathway analysis

Microarray analysis was performed using 2 μg total liver RNAs collected at ZT2 from the NP and SP mice with Mouse OneArray Plus MOA 2.1 (Zhuoli Biotech, Shanghai, China). The differentially expressed genes (DEGs) were identified and grouped into the up- and down-regulated genes, with fold change >2 fold and p-values < 0.05, which were adjusted for multiple testing using the Benjamini–Hochberg false discovery rate procedure. A false discovery rate of 5% (q-value < 0.05) was considered as significant. Gene ontology (GO) (http://www.geneontology.org) was used to predict the molecular functions and biological processes of the DEGs. DEG-enriched pathways were analyzed by KEGG (Kyoto Encyclopedia of Genes and Genomes Mapper, http://www.genome.jp/kegg/pathway.html). Heat maps of the representative genes were generated using HeatmapViewer with the GenePattern analysis suite.

Immunohistochemistry

Liver sections were first deparaffinized, rehydrated, and washed with phosphate-buffered saline (PBS). A two-step immunohistochemical protocol was used. Briefly, tissue sections were first treated with an EDTA-containing antigen retrieval buffer (ZSGB-BIO Co., Ltd., China) for 15 min at 98 °C to unmask the antigenic epitope and washed with PBS. Then the sections were blocked with PBS containing 5% normal goat serum, and incubated with the primary antibodies respectively against F4/80 (Santa Cruz, # sc-6927), ZO-1 (Life Technologies, # 402200), Claudin-2 (Life Technologies, # 325600), MRP2 (Santa Cruz, # sc-5770), NF-κB (Cell Signaling Technology, # 8242), Clock (Santa Cruz, # sc-6927), Per1 (Santa Cruz, # sc-6927), Rev-erbα (Santa Cruz, # sc-100910), FXR (Santa Cruz, # sc-13063), CYP7A1 (Santa Cruz, # sc-14423), CYP27A1 (Abcam, # ab126785), tyrosine hydrolase (Abcam, # ab112), and GAP43 (Abcam, # ab12274). After rinse with PBS, the sections were incubated with biotinylated secondary antibody (Jackson ImmunoResearch Laboratories Inc., USA). The immunoreactive proteins were visualized using the liquid DAB substrate-chromogen system (ZSGB-BIO Co, Ltd., Beijing, China). Control experiments were conducted without adding the primary antibody. To quantify the expression of immunolabelled protein, staining images were captured with a digital camera, and the signals were analyzed as previously described⁴⁷.

Western blotting

The liver tissues were lysed in a radioactive immunoprecipitation buffer, and the protein concentration was determined as previously described⁴⁷. Fifty protein of lysate was analyzed by 10% sodium dodecyl-sulfate polyacrylamide gel electrophoresis. After electrophoretic transfer, the nitrocellulose membranes were incubated with respective primary antibodies, including Clock (Santa Cruz, # sc-6927), Per1 (Merck, # AB2201), Rev-erbα (Santa Cruz, # sc-100910), BmalI (Santa Cruz, # sc-8550), FXR (Santa Cruz, # sc-13063), CYP7A1 (Santa Cruz, # sc-14423), CYP27A1 (Abcam, # ab126785), tyrosine hydrolase (Abcam, # ab112), choline acetyltransferase (Abcam, # ab 178850), GAP43 (Abcam, # ab12274), BSEP and NTCP (gift antibodies from Dr Bruno Stieger, University of Zürich Hospital) and β-actin (Molecular Biological Laboratories, PM053). After incubation with peroxidase-conjugated secondary antibody anti-Rabbit IgG (KPL, 074-1506), anti-Mouse IgG (Molecular Biological Laboratories, 330) and anti-Goat IgG (Sigma-Aldrich, A5420), the immunolabelled proteins were visualized with Clarify Western ECL Substrate (Bio-Rad).

Real time quantitative polymerase chain reaction (qPCR)

Total RNAs were extracted from liver tissues using TRIZOL reagent (Invitrogen). Single-strand cDNAs were prepared from 2 μg total RNA by reverse transcription using TransScript First-Strand cDNA Synthesis SuperMix (Transgen, Beijing, China). qPCR was performed using primers specific for the mouse target genes and β-actin, which served as a control, with the StepOnePlus Real-time PCR system (Applied Biosystems) and SYBR Green PCR Master Mix (Transgen, Beijing, China). The mRNA levels of each gene were normalized against the β-actin mRNA level and expressed as relative change over control. The qPCR primers were designed using the Primer Express V1.5 software (Applied Biosystems, CA, USA) and validated using Primer-BLAST (National Centre for Biotechnology Information, NCBI), and their sequences are: Bmal1-F: CCAAGAAAGTATGGACACAGA CAAA, Bmal1-R: GCATTCTTGATCCTTCCT TGGT; Clock-F: TTGCTCCACG GGAATCCTT, Clock-R: GGAGGGAAAGTGCTCTGTTGTAG; Per1-F: GACCTT GGCCACACTGCAGTA, Per1-R: CTCCAGACTCCACTGCTGGTAA; Per2-F: ATGCTCGCC ATCCACAAGA, Per2-R: GCGGAATCGAATGGGAGAAT; Rev-erbα-F: ATGCCCATGACAAGTTAGGC, Rev-erbα-R: GGGCTACCTGATGCAT GATT; Rorα-F: AC GCCCACCTACAACATCTC, Rorα-R: TGCCCATCCATATA GGTGCT; Fxr-F: CCAACCTGGGTTTCTACCC, Fxr-R: CACACAGCTCATCCCC TTT; Ntcp-F: ATCT GACCAGCATTGAGGCTCT, Ntcp-R: CCGTCGTAGA TTCCTTTGCTGT; Cyp7α1-F: CGGGCGTCCCTATTCACTT, Cyp7α1-R: CAGA GACCTGGTCCCTCATGTAC; Cyp27α1-F: GAGAGTGAATCAGGGGACCA, Cyp27α1-R: TCAGGAATGGAGGGTTTCAG; Shp-F: CGATCCTCTTCAACCC AGATG, Shp-R: AGGGCTCCAAGACTTCACACA; Bsep-F: GCTCGGGCCATT GGTACGAGAT, Bsep-R: CACCACTCCTTGTG ACATGACG; Actb (β-actin)-F: 5′-TCAAGATCATTGCTCCTCCTGAG; Actb-R: 5′-CTGCTTGCTGATCCACAT CTG.

Electrical stimulation treatment (EST) of sympathetic nerve

Mice were anesthetized by intraperitoneal injection with 10% chloral hydrate solution (0.3 mL/100 g body weight) and underwent an open-chest surgery, with the fourth cervical (C4) branch of the left splanchnic nerve hepatic branch isolated and cut distally. The splanchnic nerve was placed on a bipolar silver hook electrode and electrically stimulated (5 V, 5 msec, 3 Hz) for 30 min as previously described⁵. The mice were finally sacrificed, and liver tissues were harvested.

Liver slice electrochemistry

Following anesthesia with urethane (1.3 g/kg body weight, i.p.), the mouse abdominal cavity was excised to remove the liver, which was rinsed in saline. Using a Vibratome (Leica, Germany), 250 μm thick liver slices were prepared. The measurement of the in situ release of NE was performed as previously described⁴⁸. Briefly, the slices were placed in a recording buffer (118 mM NaCl, 4.7 mM KCl, 2.5 mM CaCl₂, 1.2 mM KH₂PO₄, 1.2 mM MgSO₄, 10 mM glucose), with a bipolar stimulating electrode placed to induce NE release using single square-wave pulses (2 mA, 2 ms) with ISO-100 stimulator (Taimeng Co., Ltd., China) and a 250 μm long carbon fiber electrode inserted to record NE release with an EPC10/2 amplifier (HEKA, Germany), as well as a drug puff system for reagent delivery. When the carbon-fiber electrode was held at +780 mV, an electrochemical signal characteristic of NE oxidation can be detected upon a single depolarizing pulse or treatment with high-K⁺ solution.

High-performance liquid chromatography with electrochemical detection (HPLC-ECD)

The NE content in liver tissue was determined by HPLC-ECD. Liver tissues were homogenized in 30 mM acetic acid containing 0.1% EDTA, and the lysates were centrifuged 18,000 × g for 30 min. The resulting supernatant was diluted two-fold, and a 50 µL aliquot was used for analysis. The NE extraction was all performed on ice and under light-protection. For HPLC, the mobile phase was prepared in mM KH₂PO₄, 0.11 mM EDTA, and 0.72 mM SDS with pH adjusted to 4.3 and chromatography-grade methanol added at 5% (v/v). The solution was thoroughly mixed, filtered, and sonicated prior to injection to HPLC.

Cell experiments

Human hepatoma HepG2 cells were purchased from China Center for Type Culture Collection (Wuhan, China) and cultured in Dulbecco modified Eagle medium (Hyclone) supplemented with 10% fetal bovine serum (Hyclone), 100 U/mL penicillin, and 100 μg/mL streptomycin in a 37 °C CO₂ incubator. NE (A7257) was purchased from Merck KGaA (Darmstadt, Germany) and added at 0.1 μM. To inhibit β1-ARs, β2-ARs, and PKA, 100 μM Propranolol hydrochloride (HY-B0573), 100 μM Bisoprolol (HY-129029), and 3 μM KT5270 (HY-N6789), purchased from MedChemExpress (Monmouth Junction, USA), were added.

Statistical analysis

Statistical difference between two groups was evaluated by paired or unpaired Student’s t-tests. Statistical differences among three or more groups were analyzed using one-way or two-way ANOVA followed by Tukey’s multiple comparisons test. Nonparametric Spearman correlation was used to determine the correlation between NE level and the level of ALT or TBA. P value less than 0.05 was considered statistically significant. Analyses were performed using GraphPad Prism 9, and data were presented as mean ± standard deviation (SD).

Author contributions

L.W. and L.Z. conceived and designed the experiments; M.Y. performed most of the studies and analyzed the data; J.H.Y., J.L., M.N., J.W.L., Y.J.L. and C.L. performed some real-time qPCR, immunoblotting and immunoprecipitation experiments; Q.L., X.J.Z., X.W.W. and H.B.Y. helped with the animal experiments; Q.W., J.R.H. and T.S. helped with Fig. preparation; X.Q.G. collected human specimens; L.W. wrote the final manuscript; L.Z., J.M.C., and L.W. reviewed the manuscript. All authors read and approved the final version.

Data availability

The microarray data can be retrieved from GEO under the name of GEO_metadata_ template_GPL22909_mouse_liver_microarray.xls. Other datasets of this study are available from the corresponding author on reasonable request.

Competing interests

The authors declare no competing interests.

Supplementary information

The online version contains supplementary material available at https://doi.org/10.1038/s44323-026-00103-2.