Mycophenolic acid exerts dichotomous regulation of hepatic lipogenesis in a metabolic context-dependent manner

Abstract

With the increasing global burden of fatty liver disease (FLD), liver transplantation (LT) recipients face heightened risks of steatosis recurrence in graft livers, necessitating urgent reappraisal of immunosuppressant-associated metabolic effects. The effects of mycophenolic acid (MPA), a commonly used immunosuppressant in LT, on hepatic lipid metabolism following LT remain unclear. This study employed a multifaceted approach encompassing cellular, animal, and molecular biology techniques to systematically evaluate the role of MPA in hepatic lipid metabolism. Our results demonstrated that MPA promoted lipid synthesis in healthy hepatic models by upregulating the expression of genes associated with lipogenesis, including SREBP-1c, FASN, and PPARγ, while paradoxically inhibiting lipid accumulation in steatotic hepatic models. Mechanistically, we identified inosine monophosphate dehydrogenase 2 (IMPDH2) as the molecular switch governing this dichotomy. Notably, significantly increased interaction between IMPDH2 and PPARγ was observed following MPA treatment. Our findings highlight the intricate role of MPA in hepatic lipid homeostasis and underscore the importance of further elucidating the effects of MPA across a spectrum of liver conditions to inform tailored treatment strategies following LT.

Supplementary Information

The online version contains supplementary material available at 10.1038/s41598-026-42743-1.

Introduction

Fatty liver disease (FLD) has rapidly evolved into a global health challenge; it currently affects 25–30% of the worldwide population and is emerging as a leading cause of chronic liver disease^1,2. Projections indicate that FLD-related complications will surpass viral hepatitis as the primary indication for liver transplantation (LT), particularly for patients with end-stage cirrhosis and hepatocellular carcinoma³.

Globally, FLD patients who undergo LT face the risk of recurrent hepatic steatosis in their new grafts, primarily due to the persistence of underlying metabolic risk factors such as obesity, insulin resistance, diabetes and dyslipidaemia. Moreover, given that 15–53% of potential living donors exhibit biopsy-proven FLD, disqualifying 3–21% of donor livers⁴, it is anticipated that a greater proportion of donors may exhibit steatosis or steatohepatitis. These epidemiological realities highlight the urgent need to develop strategies to mitigate FLD recurrence in both graft recipients and marginal donors.

Following LT, immunosuppressants are administered to reduce the risk of graft rejection. The immunosuppressive regimen includes drugs from multiple pharmacological classes: mammalian target of rapamycin (mTOR) inhibitors (e.g., rapamycin and everolimus), corticosteroids, calcineurin inhibitors and Mycophenolic Mofetil (MMF)⁵. While these agents effectively prevent graft rejection, their metabolic impacts remain poorly characterized. Emerging evidence suggests that mTOR inhibitors may suppress hepatocyte lipogenesis through activation of AMPK⁶. However, a consensus regarding FLD-specific immunosuppressive therapy remains elusive in clinical practice. This knowledge gap is clinically significant given the metabolic heterogeneity of LT recipients and the potential for pharmacological exacerbation of steatosis.

MMF, an ester prodrug of mycophenolic acid (MPA), is rapidly hydrolyzed to MPA in the gastrointestinal tract. MMF/MPA has gained prominence in LT protocols because of its favourable renal safety profile⁷. MPA exerts immunosuppression by inhibiting inosine monophosphate dehydrogenase (IMPDH), specifically targeting isoform 2 (IMPDH2). Particularly in lymphocytes, this depletion affects the pool of guanosine nucleotides, compounds that are essential for DNA and RNA synthesis⁸(Fig. 1A). However, key questions regarding MPA’s metabolic effects remain unanswered. While altered MPA pharmacokinetics have been reported in metabolic dysfunction-associated steatotic liver disease (MASLD) models⁸, the functional impact on hepatic lipid metabolism is unclear. Clinical studies reported that 41% of cardiac transplant recipients develop MMF-associated hyperlipidaemia⁹, whereas renal transplant cohorts show neutral lipid effects¹⁰.

Fig. 1.

MPA treatment promotes lipid accumulation in hepatocytes. (A) Schematic overview of the mechanism of action of MPA. MPA targets IMPDH, a rate-limiting enzyme in the de novo purine biosynthesis pathway, thereby suppressing guanosine nucleotide synthesis. (B) Assessment of the cytotoxicity of MPA in HepG2 and Huh7 cells using CCK8 assays. (C) Nile Red/Hoechst 33,258 co-staining demonstrating MPA-induced lipid droplet accumulation (red: neutral lipids; blue: nuclei). (D) Measurement of Nile Red fluorescence intensity by flow cytometry. (E‒F) MPA increases the cellular triglyceride (TG) and total cholesterol (TC) content in both cell lines. ^*P < 0.05, ^**P < 0.01, ^***P < 0.001; MPA treatment group vs. control group. The data are presented as the mean ± SEM (n = 6). Representative results from at least 3 replicates are shown.

In this study, we investigated the effect of MPA on hepatic lipogenesis in hepatic cells and in model mice maintained on a high-fat diet (HFD). Our results demonstrate that the dichotomous regulation of hepatic lipogenesis by MPA is contingent upon baseline metabolic status and is mediated by IMPDH2. Our findings reveal a previously unrecognized duality in the hepatic effects of MPA and provide a molecular rationale for the development of personalized immunosuppression strategies that can be used in the treatment of FLD-susceptible transplant populations.

Materials and methods

Cell culture and treatment

The human hepatocytes, along with the human embryonic kidney epithelial cell line HEK293T, were purchased from Meisen CTCC (Hangzhou, China). All cells were cultured in DMEM (Gibco, USA) supplemented with 10% FBS (HyClone) and 50 units/ml penicillin-streptomycin (Gibco, USA) at 37 °C under 5% CO₂ in a humidified atmosphere. A lipid-accumulating cell model was established following published protocols^11,12 by treating cells with free fatty acids (FFAs), prepared as a 1:2 molar mixture of palmitic acid and oleic acid (Sigma-Aldrich). This mixture was supplemented with 0.5% (w/v) bovine serum albumin (BSA). The cells were seeded at 1.5 × 10⁶ cells per well in 6-well plates, incubated overnight, and then exposed to FFAs at 1000 µM for 24 h. The medium containing the FFA mixture was then aspirated, and the cells were cultured for an additional 24 h either with or without MPA. Cells treated with an equivalent volume of PBS served as controls.

Cell viability was measured using a Cell Counting Kit-8 (CCK8) (Dojindo Laboratories) according to the manufacturer’s instructions. The cells were plated in 96-well plates at 7000 cells per well (6 replicates; n = 6) and incubated overnight to permit adherence. Various concentrations of MPA (0.5, 1, and 3 µM) were then added; the control cells received an equivalent volume of PBS containing the same amount of DMSO. After 48 h of incubation, 10 µL of CCK8 solution was added to each well, and the absorbance of the wells at a wavelength of 450 nm was determined using a microplate reader (Thermo Fisher). The absorbance of the untreated controls was taken to indicate 100% survival. Three independent experiments were performed. MPA concentrations were selected to span the clinically relevant plasma trough range and maintain cell viability (> 80%), in accordance with our previous study¹³.

Fluorescence detection of intracellular lipid deposition

Intracellular lipid droplets were visualized with Nile Red staining. The cells were rinsed with cold PBS, incubated in PBS containing 0.75 µg/mL Nile Red staining solution (Sigma-Aldrich) for 15 min, washed with distilled water, and counterstained with Hoechst 33,258 for 5 min. Fluorescence images of the intracellular lipid droplets were captured using an inverted fluorescence microscope (IX71-22FL/PH; Olympus). The fluorescence intensity of Nile Red was measured via flow cytometry on a BD Accuri C6.

Gene knockdown by lentiviral vector-mediated delivery of short hairpin RNA

To achieve stable gene knockdown, HepG2 cells were transduced with lentiviral shRNA vectors. HEK293T cells were transfected with viral packaging plasmids containing pVSVG, pPMD and pREV to generate lentiviral particles. Stably transfected cells were selected following preliminary screening in the presence of puromycin for two weeks. After a pilot investigation, shRNA vectors demonstrating the most effective gene knockdown were identified through RT-qPCR and Western blot analyses.

Animals, diets and treatment

The care and handling of the mice used in this study adhered strictly to the “Guide for the Care and Use of Laboratory Animals” and to the ARRIVE guidelines. The study was also conducted in accordance with the guidelines set by the Laboratory Animal Ethics Committee of Zhejiang Sci-Tech University. Male C57BL/6J mice, 4–6 weeks of age, were procured from Shanghai SLAC Laboratory Animal Co., Ltd. (Shanghai, China). The animals were group-housed with 2–4 mice per cage in a pathogen-free environment under controlled conditions of 18–23 °C and ~ 50% humidity.

After one week of adaptation, the mice were randomly divided into two groups. The mice in the normal diet (ND) group (n = 18) were given a normal diet, whereas the mice in the HFD group (Vehicle, n = 18) were given a high-fat diet and high-sugar drinking water. The ND diet consisted of 10% fat, 20% protein, 70% carbohydrate, and purified drinking water; the HFD contained 60% fat, 20% protein, and 20% carbohydrate, and 45% fructose (v/v) was included in the drinking water. The food was purchased from Double Lion Experimental Animal Feed Technology Co., Ltd. Mice in the ND group and vehicle groups were further subdivided into subgroups (n = 6): control, low-dose MPA (0.94 mg/kg), and high-dose MPA subgroup (2.82 mg/kg). To maintain equal drug concentrations across all animals, the dosing volumes were adjusted according individual body weight. MPA was administered intraperitoneally twice weekly, and the animals in the control group were given the same volume of saline (Figs. 4A and 5A). All mice had free access to food and drinking water during the entire experiment and were weighed weekly.

Fig. 4.

MPA promotes hepatic steatosis in mice fed a normal diet (ND mice). (A) Experimental timeline for the ND groups. (B) Body weight and liver index (liver weight/body weight ×100%) during the experiment (n = 6). (C) HE staining revealing MPA-induced microvesicular steatosis in hepatocytes. (D) Hepatic TG/TC quantification showing lipid accumulation mediated by treatment with MPA. ^*P < 0.05, MPA treatment group vs. control group. (E) Alterations in serum lipid profiles (TG, TC, LDL-C); ^*P < 0.05, ^**P < 0.01, MPA treatment group vs. control group. (F) Detection of liver function markers (ALT, AST) across groups. Representative results from at least 3 replicates are shown.

Fig. 5.

MPA ameliorates hepatic steatosis in high-fat diet (HFD)-induced mice. (A) Timeline for the HFD groups. (B) Body weight dynamics and liver indexes during treatment (n = 6). (C) HE staining showing histopathological improvement in hepatic steatosis after MPA treatment. (D) MPA-mediated reduction in hepatic TG/TC content. ^**P < 0.01, MPA treatment group vs. vehicle. (E) Serum lipid profile normalization (TC and LDL-C); ^*P < 0.05, MPA treatment group vs. vehicle. (F) Hepatoprotective effects of MPA evidenced by decreased ALT/AST levels. The data are presented as the mean ± SEM (n = 6). Representative results from at least 3 replicates are shown.

After 12 weeks of treatment, mice were deeply anesthetized and euthanized. The mice were anesthetized with 4% isoflurane (United States Pharmacopeia, USP) in 100% oxygen at a flow rate of 1 L/min for induction. Anesthesia was maintained with 1.75–2.5% isoflurane at a 0.5 L/min oxygen flow. Following the surgical procedure, euthanasia was humanely performed by rapid cervical dislocation while the animals were under deep anesthesia. Blood samples were obtained from the orbital veins of the mice. The livers of the animals were then immediately isolated and weighed, a portion of each liver was preserved in 4% paraformaldehyde for histological investigation, and the remaining portion was snap-frozen in liquid nitrogen.

Biochemical examination

Serum biochemical profiles, including triglyceride (TG), total cholesterol (TC), alanine aminotransferase (ALT), aspartate aminotransferase (AST), high-density lipoprotein cholesterol (HDL-C), low-density lipoprotein cholesterol (LDL-C), the hepatic enzyme alanine aminotransferase (ALT) and aspartate aminotransferase (AST) levels, were quantified using a commercial detection system. Hepatic lipid extraction was performed by homogenizing 100 mg of liver in ethanol at a 1:9 (w/v) ratio. The resulting supernatant was analysed using commercial TG/TC detection kits. For cellular lipid quantification, cultured cells were lysed with 2% Triton X-100, followed by measurement with TG/TC kits (Jiancheng, Nanjing).

Histological examination and staining

The livers were fixed in 4% paraformaldehyde. After a series of steps, including dehydration, clearing, and embedding in paraffin, 4–6 μm sections were generated. The sections were stained with haematoxylin for 3 min and with eosin for another 5 min. The images were captured on a Nikon Eclipse Ti microscope.

RNA isolation and RT-qPCR analysis

Total RNA was isolated from cells and liver tissues using an RNA Extraction Kit (Sigma-Aldrich) following the manufacturer’s instructions. The first strand of cDNA was synthesized from 1 µg of the extracted total RNA using Ready-To-Go™ first-strand beads (GE Healthcare). QPCR was conducted with GoTaq Real-Time qPCR mix (Promega). The conditions were as follows: 95 °C for 5 min (initial denaturation); 40 cycles of 95 °C for 10 s and 60 °C for 30 s. GAPDH served as the reference gene for normalization of target gene expression. Fold change was calculated by applying the 2^−ΔΔCq formula and normalizing to GAPDH. The logarithmic values were then converted to an exponential scale (2^ΔΔCq). The primers used in this process are listed in Supplementary Table 1.

Immunoblot analysis

Protein samples were isolated, quantified using a Bicinchoninic Acid Assay (BCA) kit (Vazyme, Nanjing, China), separated on 12% SDS-PAGE gels (20 µg per lane) and transferred to polyvinylidene fluoride (PVDF) membranes (Millipore, Bedford, MA, USA). The PVDF membranes to which the proteins had been transferred were incubated with 5% skim milk at room temperature for one hour, incubated with the relevant target protein primary antibody overnight at 4 °C, and washed three times with Tris-buffered saline containing Tween 20 (TBST). The membranes were then incubated with horseradish peroxidase-labeled goat anti-rabbit and goat anti-mouse secondary antibodies at room temperature for 1 h. Visualization was performed through the use of an ECL detection kit (PerkinElmer Inc., MA) and an advanced imaging system (Clinx 600EXP, Shanghai, China). GAPDH was used as an internal reference. Band intensities were measured using ImageJ software and normalized to loading controls.

For the Co-Immunoprecipitation (Co-IP) assay, cell lysates were treated with an antibody against PPARγ (Cell Signaling Technology). This was followed by the introduction of protein A/G-conjugated agarose beads (Beyotime Biotechnology). IgG was used as a negative control. The precipitated immune complexes were washed with PBS and boiled in loading buffer before being subjected to Western blot analysis. Supplementary Table 2 provides a list of the antibodies used in this study.

Statistical analysis

The data presented in this study are expressed as the mean ± SEM. Comparisons between two groups were performed using Student’s t test. For comparisons involving three or more groups, one-way ANOVA with Tukey’s post hoc test was used for multiple comparisons. All statistical analyses were performed using GraphPad Prism 9. P values less than 0.05 were considered to indicate statistical significance.

Results

MPA-promoted lipid accumulation in hepatocytes is associated with upregulated lipogenic transcriptional regulators

The cytotoxicity of MPA in hepatocytes was assessed using CCK-8 assays. The experiments covered concentrations spanning the clinical plasma trough range, all of which maintained cell viability > 80%. Our findings revealed negligible cytotoxicity of MPA in HepG2 and Huh7 cells when it was applied at concentrations less than 3 µM (Fig. 1B). Notably, Nile Red staining demonstrated that MPA induced the accumulation of lipid droplets (Fig. 1C), this was confirmed by flow cytometry, which revealed increased fluorescence intensity of the MPA-treated cells (Fig. 1D). Compared with those levels in control cells, quantitative analyses revealed increases in TG and TC levels in the MPA-treated cells (Fig. 1E and F), consistent with previous reports linking MMF derivatives to dyslipidaemia^9,14. Mechanistically, MPA upregulated key lipogenic regulators, including peroxisome proliferator-activated receptor gamma (PPARγ), fatty acid synthase (FASN), and acetyl-CoA carboxylase 1 (ACC1), in HepG2 cells (Fig. 2A, C and D) and in Huh7 cells (Fig. 2B, E and F). These findings position MPA as a modulator of hepatic lipogenesis.

Fig. 2.

MPA treatment led to the upregulations of lipogenic transcriptional regulators in hepatocytes. (A-B) RT-qPCR analysis of the mRNA expression of key lipogenic genes (SREBP-1c, ChREBP, PPARγ, FASN, and ACC1) in MPA-treated cells; ^*P < 0.05 (MPA treatment group vs. control group); (C-D) Western blot showing increased protein levels of FASN, SREBP-1c and PPARγ following MPA treatment in HepG2 cells; (E-F) Western blot showing increased protein levels of FASN, SREBP-1c and PPARγ following MPA treatment in Huh7 cells. The data are presented as the mean ± SEM (n = 6). Representative results from at least 3 replicates are shown.

MPA inhibits lipid synthesis in steatotic hepatocytes

Given the clinical significance of post-transplant hepatic steatosis, we established an in vitro model of hepatic steatosis using FFA-treated cells. Strikingly, MPA significantly attenuated FFA-induced lipid deposition, as demonstrated by both qualitative Nile Red fluorescence imaging (Fig. 3A) and quantitative reductions in TG and TC levels (Fig. 3B). Mechanistic investigations revealed that MPA downregulated lipogenic regulators, including sterol regulatory element-binding protein 1 (SREBP1), carbohydrate-responsive element-binding protein (ChREBP), PPARγ, FASN and ACC1, at both the transcriptional (Fig. 3C and D) and translational (Fig. 3E-H) levels in steatotic cells. These findings suggest that MPA exerts anti-steatotic effects.

Fig. 3.

MPA attenuates lipid accumulation in steatotic hepatocytes. (A) Nile red staining showing the MPA-mediated reduction in free fatty acid (FFA)-induced lipid droplets. (B) Quantitative reversal of FFA-elevated TG/TC levels after MPA treatment; ^*P < 0.05, ^**P < 0.01; MPA treatment group vs. control group. (C-D) RT-qPCR analysis of the expression of lipogenesis-associated genes (SREBP-1c, FASN, ACC1, ChREBP and PPARγ) under FFA challenge; ^*P < 0.05, ^**P < 0.01, ^***P < 0.001 for the MPA treatment group vs. the control group. (E-F) Western blot demonstrating MPA-induced suppression of lipogenic proteins (FASN, SREBP-1c and PPARγ) in FFA-treated HepG2 cells; (G-H) Western blot demonstrating MPA-induced suppression of lipogenic proteins (FASN, SREBP-1c and PPARγ) in FFA-treated Huh7 cells. Representative results from at least 3 replicates are shown.

Dichotomous effects of MPA in mouse models

To explore the effects of MPA on hepatic lipid metabolism, we conducted parallel investigations in two distinct mouse models. In mice fed a normal diet (ND), intraperitoneal administration of MPA for 12 weeks (Fig. 4A) slightly increased the animals’ body weights and liver indexes (Fig. 4B). The promotion of hepatic steatosis by MPA was confirmed by HE staining, which revealed the presence of lipid vacuoles in hepatocytes (Fig. 4C), and the hepatic TG content increased significantly from 2.5 ± 0.3 mmol/g (control) to 3.1 ± 0.2 mmol/g in the mice treated with MPA, and the TC content increased from 0.5 ± 0.01 mmol/g to 1.1 ± 0.05 mmol/g (Fig. 4D). Evaluation of the animals’ serum biochemistry indicated that the circulating levels of TG, TC and LDL-C were markedly elevated in the MPA-treated mice (Fig. 4E), although liver function markers (ALT and AST) remained unaffected (Fig. 4F). These results suggest that MPA promotes hepatic steatosis in mice maintained on a normal diet.

To further investigate the effects of MPA on FLD, we employed a mouse model of high-fat diet (HFD)-induced hepatic steatosis. The grouping of the animals and the treatment protocols used are outlined in Fig. 5A. Comparative analysis revealed significant hepatic remodeling in the HFD-fed mice (vehicles) compared with the ND controls, as manifested by elevated liver indexes (4.6% ± 0.4 vs. 3.5% ± 0.3, Figs. 4B and 5B) and histopathological findings of marked microvesicular steatosis (Fig. 5C vs. Figure 4C), validating the successful establishment of the diet-induced hepatic steatosis mouse model. Unlike in the ND controls, MPA treatment had no discernible effect on the body weight or liver index of the vehicles (Fig. 5B). Histological analysis revealed decreased accumulation of lipid droplets (Fig. 5C), and this was corroborated by the observed reductions in hepatic TG and TC levels (Fig. 5D). Additionally, serum examination indicated that MPA treatment improved serum parameters, including reductions in TC, LDL-C (Fig. 5E), AST and ALT (Fig. 5F). These findings demonstrate the anti-steatotic effects of MPA in hepatic steatosis mouse model.

Collectively, these findings suggest that MPA administration has dichotomous effects in mice: it promotes lipogenesis in healthy livers while inhibiting lipid synthesis in fatty livers. The bidirectional effects were not strictly dose dependent (compare Figs. 4D and 5D), suggesting threshold-based modulation of MPA in lipid homeostasis pathways.

The regulatory effect of MPA is associated with the IMPDH2/PPARγ axis

MPA acts by depleting de novo guanine nucleotide synthesis through the inhibition of IMPDH, particularly IMPDH2¹³. This prompted us to investigate the response of IMPDH2 to diet and MPA treatment. Initially, we compared the expression of IMPDH2 in mice fed a normal diet with that in mice fed a high-fat diet. Our results revealed that hepatic IMPDH2 expression was higher in HFD mice than in ND mice, indicating diet-dependent regulation (Fig. 6A). We subsequently found that MPA paradoxically upregulated IMPDH2 in normal hepatocytes while suppressing IMPDH2 in steatotic hepatocytes (Fig. 6B). We then performed lentiviral IMPDH2 knockdown in normal HepG2 cells (Fig. 6C) and found that IMPDH2 knockdown reduced the formation of lipid droplets (Fig. 6D) and abrogated FFA-induced TG accumulation (Fig. 6E). Our findings establish IMPDH2 as a critical mediator of the bidirectional regulatory effects of MPA on hepatic lipogenesis in normal hepatocytes. Importantly, the metabolic effects of MPA appear to be dependent on the baseline hepatic lipid profile, and potentially mediated through context-dependent modulation of IMPDH2.

Fig. 6.

The bidirectional regulatory effect of MPA on lipogenesis is mediated via the IMPDH2/PPARγ axis. (A) Western blot showing HFD-induced hepatic IMPDH2 upregulation. (B) Context-dependent IMPDH2 modulation: MPA upregulates IMPDH2 in normal hepatocytes while suppressing IMPDH2 in steatotic cells. (C) Determination of the efficiency of lentivirus-mediated IMPDH2 knockdown. (D) IMPDH2 knockdown inhibits the formation of lipid droplets induced by FFAs. (E) IMPDH2 silencing abrogates FFA-induced TG accumulation. (F) Co-IP demonstrating an obvious interaction of IMPDH2-PPARγ. (G) Western blot analysis demonstrates that IMPDH2 knockdown reduces PPARγ levels. The data are presented as the mean ± SEM (n = 6). ^**P < 0.01. Representative results from at least 3 replicates are shown.

To investigate the molecular mechanism underlying IMPDH2-mediated regulation of hepatic lipogenesis, particularly its potential interaction with key lipogenic regulators, we performed Co-IP assays. The results suggested an interaction between IMPDH2 and PPARγ (Fig. 6F). Further experiments confirmed that IMPDH2 knockdown reduced PPARγ levels, as shown by Western blot analysis (Fig. 6G). Collectively, these findings imply that IMPDH2/PPARγ signalling may represent a pathway contributing to the effects of MPA on hepatic lipid biosynthesis.

Discussion

In the past two decades, the incidence of FLD has increased significantly, indicating that this condition is poised to emerge as a predominant contributor to the growing burden of LT¹⁵. However, LT recipients remain at substantial risk of developing recurrent fatty liver post-transplantation^16,17. MMF is often used together with other agents to reduce the risk of graft rejection while also balancing the side effects of long-term immunosuppression¹⁸. However, there is scant direct evidence suggesting that MMF affects lipoprotein metabolism or leads to hyperlipidaemia, thus limiting its application in preventing FLD recurrence. Our study reveals a pharmacological paradox: MPA, the metabolite of MMF, has dual effects on hepatic lipid metabolism, promoting lipogenesis in healthy livers and suppressing it in fatty livers. This dichotomy may explain the conflicting clinical reports on MMF-associated dyslipidaemia, and they indicate its potential utility in post-transplant FLD management^9,10. The steatosis-alleviating effects of MPA in the hepatic steatosis mouse model position MPA as a potential adjunct therapy for transplant recipients, although stratification of patients on the basis of metabolic status is essential.

MPA exerts its therapeutic effects primarily through inhibition of IMPDH2-mediated guanosine nucleotide biosynthesis, and this results in depletion of intracellular guanosine nucleotide pools. Intriguingly, our study revealed paradoxical upregulation of IMPDH2 expression in MPA-treated hepatocytes. This counterintuitive phenomenon could be attributed to the existence of a compensatory feedback mechanism wherein cells initiate transcriptional upregulation of IMPDH2 to counteract the suppression of enzymatic activity. Conversely, reduced expression of IMPDH2 was observed in FFA and MPA co-treated hepatocytes. Knockdown of IMPDH2 counteracted the FFA-induced lipogenesis in HepG2 cells. A similar reduction in IMPDH2 was also noted in the livers of mice on a normal diet. Notably, our findings contrast with Bremer’s report of minimal IMPDH2 transcriptional regulation in CD4 + cells¹⁹ and thus highlight the existence of lipid metabolism-specific drug effects.

During MPA treatment, there was a notable increase in the expression of PPARγ and FASN in hepatocytes. However, in hepatocytes with steatosis, MPA suppressed the expression of these genes. These findings suggest that the dual effects of MPA on lipid synthesis are associated with regulation of the expression of lipogenesis-related genes, specifically PPARγ, FASN, and SREBP-1c²⁰.

Given the association between IMPDH2 and PPARγ revealed in this study, we infer that the dichotomous effects of MPA on lipid synthesis are related to its regulation of the IMPDH2-PPARγ axis, which reveals fundamental crosstalk between nucleotide biosynthesis and lipid metabolism. We propose that this crosstalk functions as a metabolic sensor, where IMPDH2 acts as a GTP-level sensor, allosterically modulating PPARγ’s transcriptional output, while PPARγ binding may in turn influence IMPDH2 enzymatic activity. The interaction could impair PPARγ’s activity or alter its stability, thereby coupling cellular nucleotide availability to lipid metabolic transcription–revealing direct crosstalk between purine biosynthesis and lipogenesis (Fig. 7). Beyond this, MPA may broadly reshape the hepatic transcriptional landscape by modulating other key regulators of lipogenesis, such as liver X receptor (LXR) and activating transcription factor 4 (ATF4), through nucleotide depletion following IMPDH2 inhibition. Future studies delineating these transcriptional networks will further clarify the systemic metabolic effects of IMPDH2 inhibition.

Fig. 7.

Schematic of IMPDH2/PPARγ axis in MPA’s bidirectional control of hepatic lipid metabolism.

These mechanistic insights elucidate MPA’s bidirectional regulation of hepatic lipid metabolism, informing optimized post-liver transplantation (LT) therapeutic strategies. Critically, our findings position MMF/MPA as a promising candidate for repurposing in FLD recurrence prevention within metabolically defined subgroups. We propose the following clinical framework: (1) Metabolic stratification: pre-transplant, quantify hepatic steatosis via MRI-PDFF or histology, Post-transplant, monitor metabolic risk monthly (such as TG, HDL-C, and HbA1c); (2) Therapeutic application: For steatotic/high-risk patients, initiate MMF early post-LT to leverage its lipid-lowering effects; for non-steatotic patients, avoid MMF monotherapy, augment with statins if dyslipidaemia emerges. Collectively, these findings provide a precision medicine roadmap for MMF in transplant care, where prospective validation of this stratification protocol could reposition MMF from conventional immunosuppression to targeted metabolic therapy.

This study has several limitations that should be acknowledged. First, our findings need clinical validation. Second, extra-hepatic effects of MPA (e.g., on adipose tissue) were not assessed, which is critical for understanding MPA’s systemic metabolic impact. Third, the functional role of IMPDH2 was examined only in normal hepatocytes, its involvement in steatotic hepatocytes remains to be validated. Finally, immune cell interactions relevant to transplantation were not explored. Future studies should integrate clinical data, metabolic profiling (e.g., GC-MS), and immune analysis to advance translational applications.

Supplementary Information

Below is the link to the electronic supplementary material.

Author contributions

Z. Xu performed the experiments and the data analysis and drafted the manuscript. H. L. and Y. N. were responsible for the acquisition, analysis and interpretation of the data. Y. F. and X. J. were involved in the data analysis. C. J and H. L were responsible for funding and study supervision. H. L and K. C. were responsible for funding, study design, data interpretation, study supervision, and critical revision of the manuscript.

Funding

This work was supported by the National Natural Science Foundation of China (No. 81972281) and the Medical Science and Technology Project of Zhejiang Province (No. 2025KY454).

Data availability

The analysed data sets generated during the study are available from the corresponding author upon reasonable request.

Declarations

Competing interests

The authors declare no competing interests.

Ethical approval and consent to participate

The present study was approved by the ethics committee of Zhejiang Sci-Tech University (Hangzhou, China; Approval No. IACUC-20220307).