Abstract

   Metabolic dysfunction-associated fatty liver disease (MAFLD) is a
   disease that causes an abnormal accumulation of fat in the liver,
   triggering inflammation and fibrosis, the mechanism of which is not
   fully understood and for which there is a lack of specific drug
   therapy. Far-infrared radiation (FIR) has demonstrated evident
   therapeutic efficacy across various diseases, and novel nanomaterial
   graphene patches can emit it through electric heating. This study aimed
   to investigate the potential protective effects of FIR against MAFLD.
   Mice were fed with a MCD diet to mimic MAFLD progression, and
   histopathology analysis, biochemical analysis, RT-qPCR, and Western
   blotting analysis were performed to assess the effect of FIR on MAFLD
   in vivo. The effect of FIR treatment on MAFLD in vitro was investigated
   by biochemical analysis and gene expression profiling of hepatocytes.
   Mice subjected to the MCD diet and treated with FIR exhibited reduced
   hepatic lipid deposition, inflammation, fibrosis and liver damage. The
   therapeutic effect exerted by FIR in mice may be caused by the
   enhancement of AMPK phosphorylation and inhibition of the TGFβ1-SMAD2/3
   pathway. Besides, FIR intervention alleviated MAFLD in hepatocytes in
   vitro and the results were verified by gene expression profiling. Our
   results revealed a promising potential of FIR as a novel therapeutic
   approach for MAFLD.

   Keywords: Far-infrared radiation, MAFLD, Therapy, AMPK, TGFβ1-SMAD

   Subject terms: Diseases, Health care, Medical research, Molecular
   medicine

Introduction

   Metabolic dysfunction-associated fatty liver disease (MAFLD) stands as
   the most prevalent chronic liver disease worldwide, with its incidence
   continuing to rise in recent years^[31]1. This increasing prevalence
   poses a growing economic burden, accompanied by a surge in patients
   suffering from cirrhosis and end-stage liver disease^[32]2–[33]4.
   Excessive accumulation of fat within the liver is one of the hallmark
   features associated with MAFLD. Histologically, MAFLD is typically
   diagnosed by the occurrence of steatosis in more than 5% of
   hepatocytes^[34]5. MAFLD encompasses a spectrum of liver disorders,
   including isolated steatosis, metabolic dysfunction-associated
   steatohepatitis (MASH), advanced fibrosis, cirrhosis, and
   hepatocellular carcinoma (HCC)^[35]6. Interestingly, the adverse
   effects of MAFLD extend beyond the liver to other organs. Mounting
   clinical evidence points to the revelation that MAFLD and
   cardiovascular disease are mutually aggravating. MAFLD serves as an
   independent risk factor for various cardiovascular conditions, which in
   turn exacerbates cardiovascular disease, and vice versa^[36]7. Despite
   the escalating impact of MAFLD, there is currently a lack of
   FDA-approved pharmaceutical drugs for the effective treatment of this
   condition^[37]8.

   The graphene patch employed in this study is a novel nanomaterial
   composed of sp2 hybrid carbon atoms arranged in a hexagonal lattice. It
   possesses the capability to be precisely adjusted to the desired
   treatment temperature and primarily emits far-infrared radiation (FIR)
   in the wavelength range of 3.0–100 microns. In therapeutic
   applications, FIR-generated energy can be absorbed as mild radiant
   heat^[38]9. The interaction between electromagnetic radiation within
   the FIR spectrum and biological structures, as well as living systems,
   exerts two noteworthy effects. Firstly, FIR can alter cell membrane
   potentials and mitochondrial metabolism^[39]10. Secondly, FIR energy
   can be absorbed by the vibrational energy levels of molecular bonds.
   Considering the high water proportion in biological systems, coupled
   with the solvation effect and the dielectric properties of water, the
   therapeutic benefits of FIR treatment stem from the vibrating water
   molecules^[40]11. FIR has demonstrated a multitude of biological
   effects with successful application in various medical
   treatments^[41]12. Nevertheless, its specific therapeutic role in
   addressing MAFLD has yet to be elucidated.

   The primary objective of this study was to explore the therapeutic
   potential of FIR in treating MAFLD and to unveil the possible
   underlying mechanism. Our investigations revealed that FIR, emitted
   from an electric graphene patch, could alleviate the MAFLD model in
   both in vivo and in vitro. The therapeutic mechanism probably involved
   the activation of the AMPK pathway and the inhibition of the
   TGFβ1-SMAD2/3 pathway.

Results

FIR treatment reduces MCD diet-induced hepatic lipid deposition

   The mice were subjected to a 2-week administration of the MCD diet,
   followed by alternate-day FIR therapy for 4 weeks, concurrently
   maintaining the continuous MCD diet regimen. In mice exhibiting MAFLD
   induced by MCD diet, body weight was marginally increased in the
   FIR-treated groups, while liver weight remained unaltered
   (Fig. [42]1a–c). Consequently, FIR treatment mitigated the elevation in
   the liver index (liver/body weight ratio) caused by the MCD diet
   (Fig. [43]1d). Evaluation of liver sections through H&E and ORO
   staining in each group indicated an evident reduction in hepatic lipid
   accumulation of FIR treatment group (Fig. [44]1e). These findings were
   further corroborated by the results of liver-specific TG and TC assays
   (Fig. [45]1f,g). Serum analysis through various biochemical assays also
   revealed a decrease in TG, TC, and LDL-C levels, along with a rise in
   HDL-C levels (Fig. [46]1h-k).

Figure 1.

   [47]Figure 1
   [48]Open in a new tab

   Effect of FIR on hepatic lipid deposition in MAFLD mouse model. Liver
   tissues and serum were harvested from mice fed with NC (normal chow) or
   MCD diet and subsequently treated with FIR1 (39 °C) or FIR2 (41 °C),
   for 1 h every alternate day. (a) Weekly body weight; (b,c) final body
   weight, liver weight; (d) liver weight to body weight ratio; (e) gross
   livers, H&E staining and oil red O staining of liver sections; (f,g)
   hepatic levels of TG, TC; (h–k) serum levels of TG, TC, HDL-C and
   LDL-C; (l–p) mRNA expression of 5 genes normalized against Hprt; (q)
   protein levels of 6 proteins with β-tubulin as internal reference;
   (r,s) normalized ratio of phosphorylated to total protein fluorescence
   intensity in three independent experiments. Asterisks indicate
   statistical significance (*p < 0.05, **p < 0.01, ***p < 0.001, and
   ****p < 0.0001) compared to the MCD diet group.

   RT-qPCR analyses revealed the downregulation of marker genes associated
   with lipogenesis (CD36, FASN, PDK4), while marker genes linked to fatty
   acid oxidation (PPARα, ACC1) were upregulated in the FIR-treated groups
   (Fig. [49]1l–p). These results were consistently supported by western
   blot analyses (Fig. [50]1q,r). Considering the pivotal role of the AMPK
   signaling pathway in lipid metabolism specific to MAFLD, we
   hypothesized that FIR treatment reduced lipid deposition through AMPK
   activation. As anticipated, increased AMPK phosphorylation in the liver
   tissue was observed following FIR treatment, thereby promoting
   metabolic activity (Fig. [51]1q,s). Notably, the temperature of FIR
   treatment did not appear to exert a discernible impact on the
   therapeutic outcome.

FIR treatment attenuates MCD diet-induced hepatic inflammation

   Inflammation has been recognized as a key process in the pathogenesis
   and development of MAFLD, often characterized by hepatic lipid
   steatosis^[52]15. To evaluate the effect of FIR treatment on liver
   inflammation in MAFLD, we conducted F4/80 immunofluorescence staining
   of liver sections and found that FIR treatment led to an attenuation of
   hepatic inflammation in the MCD-induced MAFLD mouse model
   (Fig. [53]2a). Then Real-time PCR was performed to detect the
   expression of related genes. PCR analyses of the total RNA extracted
   from the liver tissues of the 4 experimental mice groups revealed a
   notable downregulation in the expression of marker genes related to
   inflammatory response, including MCP1, TNF-α, CCR2, IL-1β, CXCL10,
   NLRP3 (Fig. [54]2b–g). Additionally, the hepatic injury was also
   reduced by FIR treatment, with evident alleviation of elevated ALT and
   AST levels in response to the MCD diet (Fig. [55]2h,i). Similar to
   hepatic lipid deposition results, the FIR treatment temperature
   exhibited minimal impact on the attenuation of hepatic inflammation.

Figure 2.

   [56]Figure 2
   [57]Open in a new tab

   Effect of FIR on hepatic inflammation in MAFLD mouse model. Liver
   tissues were harvested from mice fed with NC (normal chow) or MCD diet,
   followed by treatment with FIR1 (39 °C) or FIR2 (41 °C) for 1 h every
   other day. (a) F4/80 immunofluorescence staining of liver sections;
   (b–g) mRNA expression of 6 genes normalized by Hprt; (h,i) serum levels
   of ALT, AST. Asterisks indicate statistical significance (*p < 0.05,
   **p < 0.01, ***p < 0.001, and ****p < 0.0001) compared to the MCD diet
   group.

FIR treatment mitigates MCD diet-induced hepatic fibrosis

   As a major lesion in the middle and late stages of MAFLD, the severity
   of liver fibrosis determines the prognosis of the disease. We stained
   liver sections with PSR, and under the microscope, red collagen fibers
   were seen to be predominant in the MCD-induced model group, while
   collagen fibers were reduced in both FIR-treated groups (Fig. [58]3a).
   The results of PCR assay further verified this conclusion that the
   expression of liver fibrosis-related genes including Col1a1, Acta2,
   Ctgf, Loxl2, Timp1, Mmp13, Pai1 and TGFβ1 was down-regulated in the
   FIR-treated group (Fig. [59]3b–i). The results of western blot
   experiments showed that the increased phosphorylated SMAD2/3 proteins
   in the model group were significantly reduced in the FIR-treated group
   (Fig. [60]3j,k). The activation of the TGFβ1-SMAD2/3 pathway is one of
   the classical pathways responsible for the development of hepatic
   fibrosis in patients with MAFLD, and the FIR treatment may have
   attenuated the hepatic fibrosis in the model mice by inhibiting this
   pathway.

Figure 3.

   [61]Figure 3
   [62]Open in a new tab

   Effect of FIR on hepatic fibrosis in MAFLD mouse model. Liver tissues
   were harvested from mice fed with NC (normal chow) or MCD diet and
   subsequently treated with FIR1 (39 °C) or FIR2 (41 °C) for 1 h every
   alternate day. (a) PSR staining of liver sections; (b–i) mRNA
   expression of 8 genes normalized by Hprt; (j) protein levels of 2
   proteins with β-tubulin or β-actin as internal reference; (k)
   normalized ratio of phosphorylated to total protein fluorescence
   intensity in three independent experiments. Asterisks indicate
   statistical significance (*p < 0.05, **p < 0.01, ***p < 0.001, and
   ****p < 0.0001) compared to the MCD diet group.

FIR treatment alleviates MAFLD in hepatocytes in vitro

   Primary mouse hepatocytes and HepG2 cells were exposed to MASH cocktail
   and sodium palmitate, respectively, to investigate the effects of FIR
   treatment on hepatocytes in vitro. The examination of cytotoxicity in
   HepG2 cells through CCK-8 assays revealed that cell viability remained
   above 90% when exposed to FIR treatment at 39 °C (Fig. [63]4a). FIR
   treatment commenced 2 h after introducing the steatosis-inducing drug
   into the culture medium. Accordingly, two treatment groups were
   established, namely FIR1 and FIR2, each receiving treatment for 12 and
   24 h per day, respectively. After two days, the cells were fixed for
   ORO staining and simultaneously harvested for TG and TC assays. A
   reduction in the numbers and sizes of lipid droplets was observed
   post-treatment, suggesting the alleviation of drug-induced steatosis in
   both cells (Fig. [64]4b). Furthermore, TG content was reduced in
   primary mouse hepatocytes as well as HepG2 cells (Fig. [65]4c,e),
   whereas the TC levels did not exhibit apparent changes in response to
   FIR treatment (Fig. [66]4d,f).

Figure 4.

   [67]Figure 4
   [68]Open in a new tab

   Effect of FIR on hepatocytes in MAFLD cell model. The mouse primary
   hepatocytes and HepG2 cells were induced with MASH cocktail to be model
   group (MOD). Both were then treated with FIR1 (39 °C, 12 h) or FIR2
   (39 °C, 24 h) for 2 days. (a) Cytotoxic effects of FIR on HepG2 cells;
   (b) oil red O staining of hepatocytes; (c,d) TG and TC contents in
   mouse primary hepatocytes; (e,f) TG and TC contents in HepG2 cells; (g)
   PCA plot for all groups; (h) volcano plot for DEGs from MOD and FIR;
   (i,j) heat map for DEGs from MOD and FIR; (k) GO enrichment analysis
   for MOD versus FIR; (l) KEGG pathway enrichment analysis for MOD versus
   FIR. Asterisks refer to statistical significance (*p < 0.05,
   **p < 0.01, ***p < 0.001, and ****p < 0.0001) in comparison to MOD
   group.

   We performed gene expression profiling of mRNA sequences from primary
   mouse hepatocyte samples. The principal component analysis plot showed
   negligible differences within groups (Fig. [69]4g). From the volcano
   plot of differentially expressed genes (DEGs), more genes were
   downregulated in the FIR-treated group compared to the model group
   (Fig. [70]4h). The information in the heat map of DEGs suggests that
   the FIR treatment had a beneficial effect on reducing lipids,
   attenuating inflammation, and alleviating fibrosis (Fig. [71]4i,j).
   Similarly, GO enrichment of DEGs revealed that the DEGs were involved
   in lipid metabolism, inflammation and fibrosis biological processes or
   executed related molecular functions (Fig. [72]4k). KEGG pathway
   enrichment was also added to confirm this series of findings
   (Fig. [73]4l).

Discussion

   In recent years, metabolic dysfunction-associated fatty liver disease
   has emerged as the most prevalent chronic liver disorder, afflicting
   approximately 25% of the global adult population^[74]16. MAFLD has now
   swiftly ascended to become the leading cause of liver-related mortality
   worldwide^[75]17, emphasizing the pressing need for effective
   therapeutic measures. While lifestyle modifications, including dietary
   changes, exercise, and weight loss, represent major treatment
   strategies for mild MAFLD patients^[76]18, more advanced disease
   conditions often require pharmacological interventions. The most
   frequently employed pharmacological interventions involve drugs related
   to glucose and lipid metabolism, such as PPAR, bile acids (BAs)
   therapeutics, farnesoid X receptor (FXR) agonists, and thyroid hormone
   receptor-β (THR-β) agonists^[77]19. Additionally, anti-cellular stress
   medicines, including vitamin E, antioxidant carotenoid
   beta-cryptoxanthin, melatonin, coenzyme Q10 curcumin, green tea,
   epigallocatechin gallate, and anti-apoptosis agents are also commonly
   applied in MAFLD therapy^[78]20–[79]27. However, the long-term efficacy
   and safety of these drugs remain subjects of clarification.

   FIR represents a sub-division of the electromagnetic spectrum, residing
   between the long-wavelength red edge of the visible spectrum and the
   shorter edge of the terahertz spectral bands. The International
   Commission on Illumination (CIE) classifies infrared radiation (IR)
   into three sub-divisions: near-infrared (0.7–1.4 microns), mid-infrared
   (1.4–3.0 microns), and far-infrared (3.0–100 microns)^[80]12.
   Importantly, only FIR transmits energy in the form of pure thermal
   energy, perceptible to the thermal receptors of human skin as radiant
   heat. The graphene patch used in this study predominantly emits FIR
   following electrical heating. FIR has found extensive application in
   medical and healthcare domains. For example, FIR demonstrated the
   ability to inhibit peroxide production by macrophages and block
   reactive oxygen species (ROS)-mediated cytotoxicity^[81]28. Moreover,
   FIR enhanced the production of intracellular nitric oxide in breast
   cancer cells and inhibited the growth of melanoma cells^[82]29. FIR
   sauna therapy has gained prominence in Japan and South Korea,
   particularly in the management of cardiovascular
   diseases^[83]30,[84]31. A recent report has documented a positive
   therapeutic effect of FIR against allergic rhinitis^[85]32.

   However, there is a dearth of studies concerning the potential
   therapeutic impact of FIR on MAFLD. To investigate this prospect, we
   established MCD diet-induced MAFLD mouse models and MASH cell models.
   In previous studies, exposing mice to hyperthermia at 42 °C reduced the
   number of circulating antithrombin, causing thrombus formation to block
   blood vessels, further triggering localized tissue hypoxia and
   ultimately leading to cell damage^[86]33. Exposure to 40.5 °C probably
   decreased the cell viability of isolated mouse hepatocytes by 35%,
   possibly due to membrane instability and dysfunctional mitochondria and
   protein transport^[87]34. At temperatures exceeding 41 °C, protein
   denaturation increases, thus promoting apoptosis^[88]35. Accordingly, a
   temperature range of 39–41 °C was selected for FIR treatment. Indeed,
   FIR treatment has been found to have a series of beneficial effects on
   the MAFLD model both in vivo and in vitro, but at the same time, no
   marked difference in the FIR therapeutic properties was observed within
   this temperature range. Likewise, variations in the FIR treatment
   duration for cell models, spanning from 12 to 24 h, appeared to exert
   minimal influence on treatment outcome. Our study was limited to the
   cell model in vitro and the MCD diet-induced mouse model. Although the
   pathophysiologic manifestations of this animal model in terms of
   steatohepatitis and liver fibrosis are quite similar to those of MAFLD
   patients, it does not match the actual clinical status of most MAFLD
   patients on weight gain and fat accumulation. To further explore the
   ameliorative effects of FIR on MAFLD, more other animal models are
   currently needed to deepen our understanding of this novel therapeutic
   modality. Additionally, research on more profound mechanisms needs to
   be further explored as well.

   MAFLD is a disease with an ever-expanding impact, but no therapeutic
   drug has yet been approved by the FDA specifically for its treatment.
   Given its high prevalence and potential long-term health hazards, it
   has become critical to find more effective treatments. FIR, as a form
   of physical therapy, exhibits unique advantages in the treatment of
   MAFLD. First, FIR therapies typically have fewer side effects compared
   to medications. Second, FIR therapy has the advantage of being
   non-invasive compared to surgical therapy. In addition, compared with
   dietary modification and exercise, which are currently the most
   recognized lifestyle modification therapies, FIR physiotherapy offers a
   higher level of ease of use and comfort, and this ease of use can
   significantly improve patient adherence to treatment, thus enhancing
   the remission effect of MAFLD. In future studies, the feasibility of
   FIR therapy can be further broadened so that it can serve MAFLD
   patients as a safe, non-invasive, and comfortable new treatment option.

Conclusion

   In this study, we observed that FIR treatment resulted in the reduction
   of hepatic lipid deposition, inflammation, fibrosis and liver damage in
   mice induced by the MCD diet. It also alleviates MAFLD in hepatocytes
   in vitro. The effects may be caused by the activation of AMPK and
   inhibition of the TGFβ1-SMAD2/3 pathway. Collectively, the possible
   potential of FIR as a novel therapeutic approach for MAFLD has been
   illustrated.

Materials and methods

Graphene patch

   We obtained graphene patches with a total resistance of about 15 Ω and
   a power density of about 67 mW·cm^−2 at 5 V by uniformly printing
   graphene conductive ink on flexible polyimide foils. As an innovative
   nanomaterial, the main part of our graphene patch consists of sp2
   hybrid carbon atoms organized in a hexagonal lattice. The patch has
   many advantages, including no resistance change by deformation, a
   constant temperature over a range of voltages, and a good match of the
   far-infrared radiation emission peaks of 8–9 μm to the reported FIR
   peaks in mice and human beings^[89]12. Details about the fabrication
   method and physical properties of the graphene patches are described in
   this article we cited^[90]13. In short, it is an ideal far-infrared
   radiation tool for conducting this experiment.

Animal

   Eight-week-old male C57BL/6 mice were acquired from GemPharmatech
   (CHN), and subsequently housed and raised in the standard pathogen-free
   (SPF) environment in Xiamen University. All animal experiments
   conducted in this study were approved by the Institutional Animal Care
   and Use Committee of Xiamen University. Following a one-week adaptation
   period to the new environment, the mice were divided into four distinct
   groups: one group was fed the normal diet, and the remaining three
   groups were subjected to an MCD (Methionine-Choline Deficient) diet
   (Dyets CHN) for six weeks to establish the MAFLD mouse model.
   FIR-treated groups were categorized into two treatment groups, denoted
   as FIR1 and FIR2; mice in these two groups were generally anesthetized
   by isoflurane gas and enveloped in graphene patches electrically heated
   to 39 °C or 41 °C, respectively, around the abdominal region for one
   hour during the treatment protocol. At the end of the experimental
   period, all mice were anesthetized by isoflurane gas and then executed
   by cervical dislocation. Xiamen University’s Animal Care and Use
   Committee approved the experiment (approval number XMULAC20190070).

Histology

   Liver tissue specimens were initially fixed in a 4% paraformaldehyde
   solution for 24 h, followed by dehydration and embedding in molten
   paraffin. The resulting paraffin blocks were sliced into 5 μm thick
   sections and then subjected to hematoxylin–eosin (H&E), Sirius red
   (PSR) and F4/80 Immunohistochemical fluoresce staining. For oil red O
   (ORO) staining, liver tissue samples were first fixed in 4%
   paraformaldehyde solution for 20 min, and then dehydrated with 30%
   sucrose solution at 4 °C overnight. Consecutively, the tissue samples
   were embedded in OCT (Optimal Cutting Temperature) (SAKURA USA 4583)
   and sectioned into 10 μm frozen slices for staining.

Biochemistry

   Biochemical analysis of the serum samples involved the utilization of
   assay kits to assess various parameters, including triglycerides (TG)
   (Applygen CHN E1003), total cholesterol (TC) (Applygen CHN E1005),
   high-density lipoprotein cholesterol (HDL-C) (Applygen CHN E1017),
   low-density lipoprotein cholesterol (LDL-C) (Applygen CHN E1018),
   alanine aminotransferase (ALT) (Applygen CHN E2021) and aspartate
   aminotransferase (AST) (Applygen CHN E2023). Analysis of the liver
   tissue samples specifically employed TG and TC kits (Applygen CHN
   E1025, E1026). For quantification, the optical density (OD) values were
   measured using a SpectraMax ® Absorbance Reader (Molecular Devices
   CHN).

Cell isolation

   Collagenase D solution (Sigma USA V900893) was perfused into the
   inferior vena cava (IVC) of experimental mice under anesthesia.
   Subsequently, the liver was carefully extracted and minced gently in a
   solution containing DMEM, 5% FBS, 1% penicillin/streptomycin, 0.5%
   insulin–transferrin–sodium selenite (Procell CHN [91]PB180429) and 0.3%
   dexamethasone (selleck USA S1322). Following this, a series of steps,
   including filtration, centrifugation, and removal of dead cells, were
   carried out and finally, the cells were cultured in collagen-coated
   dishes (Solarbio CNH C8062-10 mg).

Cell culture

   Primary hepatocytes were isolated from 8-week-old male C57BL/6 mice and
   were cultivated in collagen-coated dishes containing DMEM medium
   supplemented with 1% penicillin/streptomycin, 0.5%
   insulin–transferrin–sodium selenite and 0.2% dexamethasone. Besides,
   HepG2 cells were grown in DMEM medium supplemented with 10% FBS and 1%
   penicillin/streptomycin. All cell cultures were maintained in a
   humidified incubator at 37 °C with 5% CO[2] under sterile conditions.
   To establish a MASH cell model using primary mouse hepatocytes, a MASH
   cocktail consisting of 0.3 mM sodium palmitate (Sigma USA P0500-10G),
   0.3 mM sodium oleate (Sigma USA O7501-1G), 5.5 mM sucrose (Sigma USA
   G-8644), 10 mM fructose (Sigma USA F3510), and 10 μg/μl
   lipopolysaccharide (LPS) (Invivogen FRA tlrl-3pelps) was added to the
   culture medium. Likewise, cellular lipotoxicity was induced in HepG2
   cells by adding 0.3 mM sodium palmitate into the medium^[92]14. After
   2 h, a graphene patch heated to 39 °C was placed beneath the cell
   culture dishes for 12 or 24 h corresponding to the two treatment
   groups, FIR1 and FIR2. For ORO staining, cells were fixed with a 4%
   paraformaldehyde solution. Cellular lipid accumulation was assessed
   using TG and TC assay kits (Applygen CHN E1025, E1026). Cell counting
   kit-8 (CCK-8) (GLPBIO USA GK10001) was used to evaluate cell viability.

Gene expression profiling

   Mouse primary hepatocyte samples were sent to Gene Denovo Biotechnology
   Co. (Guangzhou, China). Total RNA was extracted using Trizol reagent,
   and mRNA was enriched by Oligo (dT) beads. The enriched mRNA was
   fragmented into short fragments and reversely transcribed into cDNA.
   The purified double-stranded cDNA fragments were end-repaired, A base
   added and ligated to Illumina sequencing adapters. The ligation
   reaction was purified with the AMPure XP Beads (1.0X), and the
   polymerase chain reaction (PCR) was amplified. The resulting cDNA
   library was sequenced using Illumina Novaseq 6000. The follow-up
   bioinformatics analyses were performed on the Omicsmart platform
   ([93]https://www.omicsmart.com/).

Western blotting

   Tissues and cells used in this study were lysed using a denaturing
   lysis buffer containing HEPES, NaCl, EDTA, and 10% SDS supplemented
   with proteinase and phosphatase inhibitor cocktail (Apexbio USA K1011,
   K1015). Lysates were boiled for 3 min and homogenized by sonication
   using a digital sonifier FS-350 T (Sxsonic China). Protein
   concentrations were quantified employing a BCA protein assay kit
   (Thermo USA 23227). For western blotting, the protein samples were
   mixed with loading buffer and boiled for 5 min at 95 °C. The samples
   were then separated on a 4–20% Tris-Tricine precast gel (ACEBio CHN
   ET15420LGel), and subsequently transferred to a polyvinylidene
   difluoride (PVDF) membrane (Roche SUI 3010040001). Next, the membranes
   were blocked using a blocking buffer (LI-COR USA 927-70001) for 1.5 h,
   followed by immunoblotting with primary antibodies first and
   fluorescent secondary antibodies afterward (LI-COR USA 926-32213,
   926-68072). The immunoblots were finally imaged using the fluorescent
   Odyssey System (LI-COR USA). All western blotting images presented in
   the manuscript are representative of 3 independent experiments.

RT-qPCR

   Total RNA from liver tissues was extracted using a combination of
   Trizol reagent (Sigma USA T9424) and chloroform employing the TRnaZol
   RNA Kit (NCM Biotech CHN M5102). The reverse transcription of RNA to
   cDNA was performed using the Superscript III RT-PCR Kit (Vazyme CHN
   R323-01). The cDNA synthesis process was facilitated in a C1000 Touch™
   Thermal Cycler (Bio-Rad USA). Quantitative PCR was conducted using SYBR
   Green qPCR Master Mix (Selleck USA [94]B21202) according to the
   manufacturer’s protocols on the Qtower real-time machine (Analytikjena,
   Swavesey, Cambridge GBR). The housekeeping gene Hprt was utilized as a
   reference to normalize the relative gene expression levels.

Statistics

   All experiments conducted were independently repeated a minimum of 3
   times, and generated data were statistically analyzed using GraphPad
   Prism 9 software. To compare multiple groups, a one-way analysis of
   variance (ANOVA was adopted for statistical analysis. Quantitative
   results are presented as mean ± standard error of the mean (SEM).
   Statistical significance was determined with a threshold p-value of
   less than 0.05, indicated by asterisks as *p < 0.05, **p < 0.01,
   ***p < 0.001, and ****p < 0.0001.

Institutional review board statement

   The animal study protocol was approved by Xiamen University’s Animal
   Care and Use Committee approved the experiment (approval number
   XMULAC20190070). All methods were carried out in accordance with
   relevant guidelines and regulations. All methods are reported in
   accordance with ARRIVE guidelines (Supplementary Information [95]S1).

Supplementary Information

   [96]Supplementary Information.^ (142MB, zip)

Acknowledgements