MMF inhibits poxvirus infection by disrupting IMPDH2 interaction with USP5 and inducing its rod-and-ring assemblies

Abstract

Monkeypox virus (MPXV), a pathogenic orthopoxvirus, has caused major outbreaks and emerged as a global public health threat. Although antivirals approved for smallpox are used therapeutically against monkeypox, their clinical utility is limited by drug availability and emerging resistance. The conserved strategy by which viruses remodel host nucleotide metabolism to secure biosynthetic precursors for replication and spread has emerged as a pivotal target for the development of broad-spectrum antiviral therapeutics. In this study, leveraging the high genetic and biological similarity between vaccinia virus (VACV) and MPXV, we employed VACV as a surrogate model to screen 10 FDA-approved inhibitors targeting nucleotide metabolism enzymes, aiming to identify potential novel inhibitors against MPXV. Mycophenolate mofetil (MMF), an inosine 5′-monophosphate dehydrogenase type II (IMPDH2) inhibitor, displayed potent inhibition effects against both VACV and MPXV. Subsequent downstream time-course studies revealed that MMF targets a post-entry stage of the viral replication cycle. Mechanistic studies suggest that MMF inhibits IMPDH2 activity by suppressing ubiquitin-specific protease 5 (USP5)-mediated deubiquitination of IMPDH2 and inducing rod-and-ring (R&R) assembly, leading to reducing dNTP pools and enhancing antiviral effects. In conclusion, our findings demonstrate that MMF is an effective antiviral drug against VACV and MPXV infection and establish a host-directed therapeutic strategy to combat future orthopoxvirus outbreaks.

Highlights

• •MMF inhibits poxvirus infection by targeting IMPDH2.
• •MMF inhibits IMPDH2 enzyme activity by suppressing USP5-mediated deubiquitination of IMPDH2.
• •MMF inhibits IMPDH2 enzyme activity by inducing its ‘Rod and Ring’ assembly.

Introduction

Mpox is a zoonotic disease caused by infection with monkeypox virus (MPXV), a member of the genus Orthopoxvirus within the family Poxviridae (Alakunle et al., 2024; Li et al., 2023; Lim et al., 2023). The genus Orthopoxvirus comprises multiple species, including Vaccinia virus (VACV), Variola virus (VARV) and monkeypox virus (MPXV) (Mitjà et al., 2023; Shchelkunova and Shchelkunov, 2022). Since 2022, a substantial global increase in MPXV infections has been reported, prompting the World Health Organization (WHO) to declare mpox a Public Health Emergency of International Concern (PHEIC) (MacIntyre et al., 2024; Nuzzo et al., 2022). FDA has approved two drugs for strategic stockpiling against smallpox: tecovirimat and brincidofovir (BCV). However, tecovirimat exhibits a relatively low genetic barrier to viral resistance, and resistant mutants are expected following widespread use (Smith et al., 2023). Moreover, the clinical efficacy of BCV against mpox remains limited due to severe adverse effects and drug resistance (Andrei and Snoeck, 2023; Rizk et al., 2025; Smith et al., 2023). Therefore, developing novel drugs against orthopoxviruses is critically important to provide valuable leads for rapid and effective countermeasures against MPXV.

Nucleotides are essential components of living cells, serving as the fundamental building blocks of DNA and RNA. Viruses, as obligate intracellular parasites lacking their own metabolic systems, rely on the metabolic resources of host cells to support their replication (Dsouza and Yang, 2025; Hisam et al., 2026). In virus-infected cells, the activities of nucleotide metabolic enzymes are often altered. Increasing evidence suggests that viruses hijack host nucleotide metabolic enzymes to enhance nucleotide biosynthesis, thereby promoting efficient viral replication (Chen et al., 2025; Wan et al., 2024). For example, inosine monophosphate dehydrogenase (IMPDH), the rate-limiting enzyme in guanine nucleotide biosynthesis, has been identified as a critical host factor for porcine epidemic diarrhea virus (PEDV) replication (Zhou et al., 2026). In addition, methylenetetrahydrofolate dehydrogenase 2 (MTHFD2) has been reported to regulate nucleotide synthesis required for Newcastle disease virus (NDV) replication (Tang et al., 2023). Therefore, inhibitors targeting nucleotide metabolic enzymes have long been considered promising candidates for antiviral drug development.

Vaccinia virus (VACV) is the prototype poxvirus and serves as a closely related surrogate for studying highly pathogenic poxviruses (e.g., mpox and variola viruses) due to their > 95% genomic identity (Alakunle et al., 2020; Dsouza et al., 2023). In this study, we used VACV as a model virus to screen FDA-approved drugs targeting nucleotide metabolic enzymes, aiming to identify effective agents against MPXV. We found that MMF significantly inhibited VACV and MPXV replication in HeLa and A549 cells without discernible cytotoxicity. Mechanistically, MMF targets inosine 5′-monophosphate dehydrogenase type II (IMPDH2) and inhibits its activity by suppressing ubiquitin-specific protease 5 (USP5)-mediated deubiquitination and inducing inactive rod and ring (R&R) assemblies. Consequently, MMF suppresses viral replication by decreasing intracellular 2′-deoxynucleoside-5′-triphosphate (dNTP) pools required for viral DNA synthesis. Importantly, MMF reduces viral loads and improves survival in VACV-infected mice. Collectively, our findings demonstrate that MMF is a promising antiviral agent against multiple poxviruses and highlight IMPDH as a promising target for host-directed antiviral therapy.

Results

Screening of nucleotide metabolism enzyme inhibitors against vaccinia virus

Viral infections often reprogram host cell nucleotide metabolism to support viral replication (Dsouza and Yang, 2025). Studies have shown that inhibitors of nucleotide metabolic enzymes, including ribavirin, mycophenolic acid, and 6-thioguanine, suppress viral replication by limiting the supply of essential raw materials. (Ariav et al., 2021; Dsouza and Yang, 2025) (Fig. 1A). To evaluate the effects of inhibitors targeting nucleotide metabolic enzymes on VACV replication, we infected HeLa cells with the reporter virus GFP-A5, in which GFP is fused to the viral core protein A5, and monitored viral infection by fluorescence. (Fig. 1B). Cytarabine (Ara-C) was used as a positive control in the screen because it inhibits viral replication through a mechanism distinct from that of the compounds tested, namely direct termination of DNA chain elongation (Shen et al., 2024). The results showed that the IMPDH-targeting inhibitors MPA, MMF, and AVN-944 efficiently reduced viral infection, leading to a robust and reproducible reduction in GFP signal (Fig. 1C). Fluorescence quantification in the right panel further supports these screening results (Fig. 1D). Next, a VACV infection assay was performed to evaluate the antiviral effects of the candidate compounds. In HeLa cells, these compounds inhibited VACV replication in a concentration-dependent manner. Notably, MMF suppressed VACV replication at a very low concentration (EC₅₀ = 0.019 μM) while maintaining low cytotoxicity (CC₅₀ > 10 μM), resulting in an SI value > 9357 (Fig. 1F). In comparison, MPA and AVN-944 exhibited weaker antiviral activity than MMF (EC₅₀ = 0.125 μM and 0.061 μM, respectively) and showed no detectable cytotoxicity in HeLa cells at concentrations of 39.81 μM and 10 μM, yielding SI values > 318 and >163, respectively (Fig. 1E and G). Collectively, these data indicate that the superior efficacy and safety profile of MMF highlights its potential as a host-directed antiviral candidate.

Fig. 1.

Screening of nucleotide metabolism inhibitors and evaluation of antiviral efficacy. A Schematic diagram of the nucleotide biosynthesis pathways and key enzymes targeted by inhibitors. B Overview of the compound screening workflow. HeLa cells were cultured overnight in 96-well plates, then treated with 10 μM candidate compounds, and incubated for 24 h under CO₂ conditions. After incubation, fluorescence intensity was recorded using an automated scanning system to assess antiviral activity. C HeLa cells were treated with 10 μM drugs from panel A and infected with A5-GFP at an MOI of 0.1 for 24 h. Cytarabine (Ara-C) was used as the positive control compound. GFP fluorescence was visualized using a fluorescence microscope. Scale bar: 100 μm. D Quantification of GFP fluorescence intensity from panel C. Relative fluorescence values were normalized to cell number. E–G HeLa cells were treated with 10 μM drugs and infected with A5-GFP at an MOI of 0.1 for 24 h. Dose–response curves showing antiviral efficacy (EC₅₀) and cytotoxicity (CC₅₀) of (E) Mycophenolic acid (MPA), (F) Mycophenolate mofetil (MMF), and (G) AVN-944. The upper green curves represent antiviral activity, while the lower red curves indicate cytotoxicity. (mean ± SD of n = 3 biological replicates in panels C–G). ns, not significant, ∗P < 0.05, ∗∗P < 0.01, ∗∗∗P < 0.001.

MMF inhibits VACV and MPXV by targeting their replication stage

To validate the findings from the screening, we assessed the antiviral activity of MMF in several human cell lines, including HeLa and A549 cells. Cells were infected with VACV and MPXV at an MOI of 0.1. MMF was added simultaneously, while dimethyl sulfoxide (DMSO) served as the negative control (Fig. 2A and B). To determine the temporal window of MMF-mediated antiviral activity, a time-of-addition assay was performed in HeLa cells infected with VACV and MPXV (Fig. 2C). Viral replication was significantly inhibited when MMF was added at 1 h post-infection (Fig. 2D and E). Taken together, these findings demonstrate that MMF targets the replication stage of poxvirus infection and exerts broad-spectrum antiviral activity against members of the poxviridae family, highlighting its potential as a host-directed antiviral agent.

Fig. 2.

MMF inhibits poxvirus replication. A-B Cells were infected with VACV and MPXV at an MOI of 0.1 for 24 h, and simultaneously treated with 10 μM MMF or DMSO. Viral titers in the HeLa cells (A) or A549 cells (B) were measured by plaque assay. C Schematic representation of the experimental design used to evaluate the effects of MMF on different stages of the viral life cycle. D–E HeLa cells were treated with 10 μM MMF or DMSO and subsequently infected with VACV (D) or MPXV (E). Viral binding, entry, and replication were quantified by qPCR, whereas viral release was assessed by plaque assay. (mean ± SD of n = 3 biological replicates in panels A, B, D and E). ns, not significant, ∗∗∗P < 0.001.

MMF suppresses VACV replication by targeting IMPDH2

MMF is known to effectively inhibit IMPDH activity in mammalian cells (Wu et al., 2025). IMPDH serves as the rate-limiting enzyme in de novo purine biosynthesis and has two isoforms, IMPDH1 and IMPDH2 (Ogawa-Iio et al., 2025). Among them, IMPDH2 is the predominant isoform upregulated in rapidly proliferating cells and virus-infected environments (Zhou et al., 2026). To determine whether MMF inhibits VACV replication through IMPDH2, IMPDH2 expression was silenced using shRNA, and knockdown efficiency was confirmed by qRT-PCR (Fig. 3A). Control and IMPDH2-knockdown cells were then treated with MMF or DMSO as a vehicle control, followed by infection with VACV-WR at an MOI of 0.1 for 24 h. Plaque assay results for viral titration showed that IMPDH2 knockdown significantly suppressed VACV replication, and MMF did not further enhance the antiviral effect in IMPDH2-deficient cells (Fig. 3B). In contrast, ectopic expression of IMPDH2 partially abolished the antiviral effect of MMF (Fig. 3C). The incomplete rescue likely reflects continued binding of MMF to a fraction of the overexpressed IMPDH2, thereby maintaining partial inhibition of IMPDH2 activity and antiviral efficacy. Next, we examined the effect of MMF on IMPDH2 enzymatic activity. Cells were infected with VACV at an MOI of 0.1 for 24 h and treated with either MMF or DMSO as a control. We found that MMF significantly suppressed IMPDH2 enzymatic activity, supporting the notion that MMF inhibits viral replication by impairing IMPDH2 catalytic function (Fig. 3D). These results support a model in which MMF inhibits VACV replication primarily through suppression of IMPDH2 activity. Following that, the interaction between MMF and IMPDH2 was examined using a cellular thermal shift assay (CETSA). The results showed that the IMPDH2 protein level gradually decreased with increasing temperature. Notably, the IMPDH2 protein level in the MMF-treated group was higher than that in the untreated group, indicating that MMF binds to IMPDH2 (Fig. 3E).

Fig. 3.

MMF targets IMPDH2 to inhibit viral replication. A Relative IMPDH2 mRNA levels quantified by qRT-PCR in control and IMPDH2-knockdown HeLa cells. B–C Wild-type and IMPDH2-knockdown HeLa cells (B) and control and IMPDH2-overexpressing HeLa cells (C), were infected with VACV at an MOI of 0.1 for 24 h and treated with 10 μM MMF or DMSO. Viral titers were measured by plaque assay. D HeLa cells were mock-infected or infected with VACV at an MOI of 0.1 and treated with 10 μM MMF or DMSO for 24 h, after which cell lysates were harvested to quantify IMPDH2 levels using an ELISA kit. E HeLa cells were treated with 10 μM MMF or DMSO, heated at indicated temperatures, and analyzed by Western blot using an antibody against IMPDH2, with GAPDH serving as a control. The thermal shift curve shows normalized IMPDH2 protein levels. Quantification of IMPDH2 band intensity at 50 °C from three independent CETSA experiments. F HeLa cells were treated with increasing concentrations of guanosine (0–1000 μM) for 24 h. Cell viability was measured by CCK-8 assay. G HeLa cells were infected with VACV at an MOI of 0.1 and treated with 10 μM MMF, and then treated increasing concentrations of guanosine for 24 h. Viral titers were measured by plaque assay. H Wild-type and IMPDH2 knockdown of HeLa cells were infected with VACV at an MOI of 0.1 and treated 1000 μM guanosine or DMSO for 24 h. Viral titers were measured by plaque assay. I MMF is converted to its active form, mycophenolic acid (MPA), via esterase-mediated hydrolysis. J IMPDH2 (pink) structure in complex with MPA (orange). Magnified views show hydrogen bonds (yellow dashes, distances in Å) between MPA and residues Ser276, Gly326, Thr333, and Gln441. Right panel shows 90° rotated view of the binding site. (mean ± SD of n = 3 biological replicates in panels A–F). ns, not significant, ∗P < 0.05, ∗∗P < 0.01, ∗∗∗P < 0.001.

IMPDH2, the rate-limiting enzyme for de novo GTP synthesis, is targeted by MMF to reduce GTP availability and thereby inhibit viral replication (Shand et al., 2024). Therefore, we determined the role of guanosine on viral replication. First, the cytotoxicity of guanosine was assessed, and we confirmed that treatment with 1000 μM guanosine did not induce cytotoxicity in HeLa cells (Fig. 3F). Then, guanosine dose-dependently reversed the antiviral effect of MMF, and 1000 μM guanosine completely restored viral replication in the presence of MMF (Fig. 3G). Additionally, we evaluated the effect of guanosine supplementation in IMPDH2-knockdown and control cells. Guanosine supplementation effectively restored viral replication in IMPDH2-knockdown cells (Fig. 3H). MMF is hydrolyzed to its active metabolite MPA (Allison, 2005), and molecular docking analysis showed that MPA occupied the catalytic pocket of IMPDH2 while forming stable interactions with conserved residues including S276, G326, T333, and Q441 (Fig. 3I and J). These structural features provide a mechanistic basis for MMF-mediated inhibition of IMPDH2 activity. These results provide evidence that MMF suppresses viral replication through direct inhibition of IMPDH2, leading to depletion of guanine nucleotide pools.

USP5 mediates the deubiquitination of IMPDH2 and enhances its enzymatic activity

It has been reported that enzyme activity is often closely associated with its expression level (Yuan et al., 2017). To determine whether VACV enhances IMPDH2 enzyme activity by increasing its expression, we performed transcriptomic analysis on infected and mock-infected HeLa cells. Volcano plot analysis identified differentially expressed genes based on fold change and statistical significance. Genes with a P-value < 0.05 and a fold change ≥ 1.2 were considered significant. The results revealed that the transcription level of IMPDH2 (Gene ID: NM000884) showed no significant change in infected and mock-infected cells (Fig. 4A; Supplementary Table S1), which was further confirmed by qPCR (Fig. 4B). Next, we conducted proteomic analysis to examine whether VACV modulates IMPDH2 expression at the post-transcriptional level. The quantitative proteomic analysis identified IMPDH2 as one of the upregulated proteins in VACV-infected cells (Fig. 4C). Western blot analysis further confirmed a time-dependent accumulation of IMPDH2 protein following VACV infection at 12 and 24 hpi, which paralleled the expression kinetics of the viral protein B5 (Fig. 4D). These findings suggest that IMPDH2 is stabilized at the post-translational level following VACV infection, thereby facilitating viral replication.

Fig. 4.

USP5-mediated deubiquitination of IMPDH2 facilitates viral replication. A Volcano plots showing differentially expressed genes comparing mock-infected or infected. B HeLa cells were infected with VACV at an MOI of 0.1, and cell samples were collected at 12 and 24 h, respectively. Total RNA was extracted and subjected to RT-qPCR analysis for IMPDH2 mRNA expression. C Volcano plot of quantitative proteomic analysis depicting differentially expressed proteins upon VACV mock-infection or infection. D HeLa cells seeded in 12-well plates were infected at an MOI of 0.1. At 24 h post-transfection (hpt), cells were lysed for immunoblotting with antibodies against IMPDH2, B5 and GAPDH. E Relative USP5 mRNA levels in control or USP5 knockdown HeLa cells. F Wild-type and USP5 knockdown of HeLa cells were either mock-infected or infected with VACV at an MOI of 0.1. At 24 hpt, cells were lysed for immunoblotting with antibodies against IMPDH2,USP5,B5 and GAPDH. G HeLa cells were either mock-infected or infected with VACV at an MOI of 0.1 for 24 h, and proximity ligation assay (PLA) was performed with anti-USP5 and anti-IMPDH2 antibodies to detect USP5-IMPDH2 interaction., Scale bar: 10 μm. H HeLa cells seeded in 6-cm dishes were either mock-infected or infected with VACV. At 24 hpi, cell lysates were immunoprecipitated with a rabbit anti-IMPDH2 antibody or control IgG, followed by Western blot analysis with against USP5, IMPDH2, B5 and GAPDH. I Wild-type and USP5 knockdown of HeLa cells seeded in 6-cm dishes were either mock-infected or infected with VACV and treated with the proteasome inhibitor MG132 (10 μM). At 24 hpi, cell lysates were immunoprecipitated with a rabbit anti-IMPDH2 antibody, followed by Western blot analysis with against UB, USP5, IMPDH2, B5 and GAPDH. J Wild-type and USP5 knockdown of HeLa cells were either mock-infected or infected with VACV at an MOI of 0.1 for 24 h, after which cell lysates were harvested to quantify IMPDH2 levels using an ELISA kit. K Wild-type or USP5-knockdown of HeLa cells were infected with VACV at an MOI of 0.1 for 24 h. Viral titers were measured by plaque assay. L–O HeLa cells were mock-infected or infected with VACV at an MOI of 0.1 and treated with 10 μM MMF or DMSO for 24 h. L Proximity ligation assay (PLA) was performed with anti-USP5 and anti-IMPDH2 antibodies to detect USP5-IMPDH2 interaction., Scale bar: 10 μm. M Cell lysates were subjected to Western blot analysis using antibodies against USP5, IMPDH2, B5, GAPDH. N Cells were subjected to immunoprecipitation with a rabbit anti-IMPDH2 antibody or control IgG. Whole-cell lysates (WCLs) and immunoprecipitated proteins were analyzed by immunoblotting with antibodies against IMPDH2, USP5, B5 and GAPDH. O HeLa cells were treated with the proteasome inhibitor MG132 (10 μM) and then cells were subjected to immunoprecipitation with a rabbit anti-IMPDH2 antibody. WCLs and immunoprecipitated proteins were analyzed by immunoblotting with antibodies against UB, IMPDH2, USP5, B5 and GAPDH. (Data are representative of at least three independent experiments. mean ± SD of n = 3 biological replicates in panels A, C, F and I–K). ns, not significant, ∗∗∗P < 0.001.

It has been reported that USP5 can deubiquitinate and stabilize IMPDH2 during cancer progression (Jiang et al., 2025). However, the role of USP5 in regulating IMPDH2 during viral infection remains unclear. To determine whether VACV-induced USP5 regulates IMPDH2 protein stability, USP5 was depleted using shRNA, and the efficiency of gene silencing was verified by qRT-PCR (Fig. 4E). Next, we examined the correlation between USP5 and IMPDH2 expression in control and USP5-knockdown HeLa cells subjected to mock or VACV infection and observed a positive correlation. Finally, we found that USP5 knockdown decreased basal IMPDH2 protein levels and reversed its VACV-induced upregulation. (Fig. 4F). To validate the interaction between IMPDH2 and USP5, we used proximity ligation assays (PLA) and coimmunoprecipitation assays. PLA revealed abundant red puncta indicative of an interaction between IMPDH2 and USP5, which was further confirmed by co-immunoprecipitation showing that IMPDH2 specifically associated with USP5 in VACV-infected cells (Fig. 4G and H). To explore whether USP5 stabilizes IMPDH2 through the ubiquitin-proteasome pathway, we treated HeLa cells with MG132, a proteasome inhibitor, in the context of mock or VACV infection, combined with manipulation of USP5 expression. USP5 knockdown increased IMPDH2 ubiquitination and decreased its protein level (Fig. 4I). Functional validation using an ELISA-based assay to quantify IMPDH2 protein levels confirmed that VACV infection significantly increased IMPDH2 enzymatic activity, and USP5 knockdown substantially reduced IMPDH2 protein levels in infected cells (Fig. 4J). Finally, viral titer assays revealed that USP5 knockdown significantly reduced the production of viral progeny (Fig. 4K). Collectively, these findings demonstrate that USP5 deubiquitinates IMPDH2 and protects it from proteasomal degradation during VACV infection.

Next, we investigated whether MMF treatment affects the USP5-IMPDH2 interaction. PLA analysis revealed that MMF treatment significantly decreased the proximity between USP5 and IMPDH2 (Fig. 4L). Western blot analysis showed that MMF treatment reduced total IMPDH2 protein levels regardless of VACV infection status (Fig. 4M). Co-IP experiments further confirmed that MMF weakened the interaction between IMPDH2 and USP5 (Fig. 4N). Moreover, ubiquitination analysis showed that MMF treatment promoted IMPDH2 ubiquitination, thereby counteracting its stabilization induced by VACV infection (Fig. 4O). Taken together, our findings indicate that MMF interferes the interaction between IMPDH2 and USP5, thereby inhibiting IMPDH2 deubiquitination and promoting its ubiquitination to suppress VACV infection.

MMF induces IMPDH2 rods and rings (R&R) formation and inhibits its enzymatic activity

Previous studies have shown that IMPDH exists as a dimer and can further oligomerize into catalytically active tetramers. Under conditions of IMPDH inhibition or guanine nucleotide depletion, IMPDH assembles into characteristic “rods and rings” (R&R) structures (Calise and Chan, 2020; Gedeon et al., 2023). This process is thought to maintain a dynamic equilibrium between catalytically active tetramers and higher-order R&R assemblies (Fig. 5A), thereby enabling rapid adaptation to changes in nucleotide demand and biosynthesis (Nass et al., 2020). To investigate how VACV infection and MMF treatment affect the relationship between IMPDH2 enzymatic activity and oligomeric state, we analyzed cell lysates by Native-PAGE. Native PAGE analysis showed that VACV infection reduced high-molecular-weight IMPDH2 oligomers, whereas MMF treatment promoted the accumulation of high-molecular-weight IMPDH2 oligomers consistent with R&R structures (Fig. 5B). To visualize MMF-induced IMPDH2 reorganization at the cellular level, we performed immunofluorescence microscopy. The dispersed distribution of IMPDH2 in VACV-infected HeLa cells is consistent with it being in its catalytically active form. In striking contrast, MMF treatment induced the formation of prominent filamentous R&R structures in both mock-infected and VACV-infected cells (Fig. 5C). Quantification revealed that more than 80% of MMF-treated cells displayed filamentous IMPDH2, compared with fewer than 20% of DMSO-treated controls (Fig. 5D).

Fig. 5.

MMF inhibits viral replication by inducing IMPDH2 Rod and Ring assembly, thereby inhibiting its enzymatic activity. A Schematic model illustrating stepwise assembly of IMPDH2 from monomers into higher-order oligomers and Rod and Ring (R&R) filaments. B–D HeLa cells were either mock-infected or infected with VACV at an MOI of 0.1 and treated with 10 μM MMF or DMSO. B At 24 hpt, cell lysates were resolved by semi-denaturing detergent agarose gel electrophoresis (SDD-AGE) or standard SDS-PAGE, followed by immunoblotting using antibodies against IMPDH2, B5 and GAPDH. C Cells were fixed and stained for IMPDH2 (red), VACV (green) and nuclei (DAPI, blue), and imaged using confocal microscopy. Scale bars: 10 μm. D The quantity of IMPDH2 filaments in each cell was measured as in panel C, with at least 100 cells analyzed per group. E IMPDH2 knockdown cells overexpress the drug-binding site mutant and wild-type of IMPDH2 were treated 10 μM MMF, and imaged using confocal microscopy. Scale bars: 10 μm. F The quantity of IMPDH2 filaments in each cell was measured (panel D and panel F), with at least 100 cells analyzed per group. G IMPDH2 knockdown cells overexpress the drug-binding site mutant and wild-type of IMPDH2 for 24 h, after which cell lysates were harvested to quantify IMPDH2 levels using an ELISA kit. H IMPDH2 knockdown cells overexpress the drug-binding site mutant and wild-type of IMPDH2. At 12hpt, cells were treated with 10 μM MMF or DMSO, after which cell lysates were harvested to quantify IMPDH2 levels using an ELISA kit. I Cells used the same treatment conditions as in panel H. Viral titers were measured by plaque assay. (mean ± SD of n = 3 biological replicates in panels C–I). ns, not significant, ∗P < 0.05, ∗∗P < 0.01, ∗∗∗P < 0.001.

Building on these results, we sought to determine the role of MMF in promoting the assembly of IMPDH2 into R&R structures. We generated HeLa cell lines stably expressing wild-type (WT) or mutants of IMPDH2 (S276A, G326A, T333I, and Q441A), corresponding to the four hydrogen-bonding residues identified in our molecular docking analysis (Fig. 3J). Immunofluorescence analysis showed that MMF can strongly induce the formation of R&R structures in cells expressing WT IMPDH2, whereas four mutants showed varying degrees of reduction in R&R structures upon MMF treatment (Fig. 5E). Quantitative analysis indicated that a high proportion of WT-IMPDH2 overexpressing cells formed R&R structures, while R&R formation was significantly reduced in the binding site mutants to varying extents (Fig. 5F), suggesting that these residues are essential for MMF-mediated IMPDH2 polymerization. We next investigated the contribution of drug-binding-site mutants to IMPDH2 enzymatic activity. Mutation of residues 333 and 441 reduced basal IMPDH2 activity, whereas mutations at the other two sites had no significant effect on enzyme activity (Fig. 5G). Upon MMF treatment, the drug-binding mutants exhibited differential tolerance to MMF, as reflected by a reduced sensitivity of IMPDH2 enzymatic activity to MMF inhibition compared with wild-type IMPDH2 (Fig. 5H). To further determine whether the antiviral activity of MMF depends on its binding to IMPDH2, we assessed viral titers in parallel. Consistent with the enzymatic activity results, MMF-mediated suppression of viral replication was significantly attenuated in the drug-binding-site mutants, and viral titers positively correlated with residual IMPDH2 activity (Fig. 5I). Collectively, these results establish that MMF exerts its antiviral effects by targeting IMPDH2 to induce catalytically inactive R&R assemblies, leading to reduced IMPDH2 enzymatic activity.

MMF inhibits VACV infection in mice

To evaluate the antiviral efficacy of MMF in vivo, forty-eight 6-week-old female BALB/c mice were randomly divided into four groups and intranasally infected with VACV (1 × 10⁴ PFU) or mock-infected, followed by daily intraperitoneal administration of MMF (50 mg/kg), with DMSO serving as the negative control (Fig. 6A). At six days postinfection, four mice per group were euthanized to assess viral loads and pathological damage in the lung and brain tissues. The remaining mice were monitored until weight loss exceeding 30%, at which point they were euthanized and recorded as deceased. All DMSO-treated mice succumbed within eight days of inoculation, exhibiting substantial weight loss from the first day of infection. In contrast, MMF treatment significantly attenuated weight loss and extended survival time following VACV infection (Fig. 6B and C). Quantitative analysis demonstrated that MMF treatment significantly reduced VACV-E9L DNA levels in the lung and brain tissues (Fig. 6D and E). Compared with the VACV-infected group, MMF treatment attenuated peribronchial inflammation and pulmonary interstitial infiltration while preserving alveolar structural integrity (Fig. 6F). Additionally, MMF treatment alleviated pathological changes in the brain by reducing inflammatory cell infiltration and minimizing tissue damage (Fig. 6G). Together, these results demonstrate that MMF effectively suppresses VACV infection in vivo and protects mice from VACV-induced disease.

Fig. 6.

MMF treatment prolonged the survival of VACV-infected mice and reduced viral loads in multiple organs. A Schematic of the experimental design. Six-week-old female BALB/c mice (n = 12 per group) were divided into four groups. Mice were either mock-infected or infected with 1 × 10⁴ PFU of VACV. B Survival analysis. Mouse survival was monitored for 14 days according to the design in panel A. C Body weight changes. The body weight of mice was monitored daily for 14 days and is presented as the mean percentage of initial weight. D,E Viral load quantification. Viral loads in the lungs (D) and brains (E) were measured by qPCR at 6 days post-infection (dpi). F,G Histopathological analysis. Tissues for analysis were collected from mice that were euthanized at 6 days post-infection (dpi). Representative hematoxylin and eosin (H&E)-stained sections of lung (F) and brain (G) tissues were collected at the experimental endpoint. Scale bars: 100 μm. Data are representative of at least three independent experiments with similar results (mean ± SD, n = 8 biological replicates in panels B and C, n = 4 biological replicates in panels D and E). ∗∗∗P < 0.001.

Discussion

Mpox currently poses a threat to global public health, yet no specific antivirals are available for its treatment. Several existing antiviral agents, including tecovirimat, brincidofovir, and cidofovir, have been explored as potential options (Poland et al., 2022). Tecovirimat was approved by the U.S. Food and Drug Administration (FDA) in 2018 for smallpox and subsequently by the European Medicines Agency in January 2022 for smallpox and cowpox (Frenois-Veyrat et al., 2022). However, evidence of drug-resistant viral strains has emerged (Vernuccio et al., 2025). Cidofovir was approved by the FDA in 1996 for the treatment of cytomegalovirus-induced retinitis in patients with AIDS (Harapan et al., 2022). Its efficacy has been demonstrated in vitro and in MPXV-infected animal models; however, clinical data on its effectiveness against human mpox are not available. Brincidofovir, a cidofovir derivative with reduced toxicity, was approved by the FDA for smallpox treatment in 2021 (Ortiz-Saavedra et al., 2022). Nevertheless, clinical trials have yet to provide evidence regarding its efficacy and safety in treating human mpox (Siegrist and Sassine, 2023). There is an urgent demand for novel broad-spectrum antipoxvirus therapies.

De novo pyrimidine and purine biosynthesis pathway plays an important role in rapidly proliferating cells, such as virus-infected host cells, to fulfill the increased nucleotide requirements. In this context, host-directed strategies offer a promising avenue for identifying antiviral drugs, as they carry a lower risk of driving drug resistance. Notably, some host cell processes targeted by this approach are already modulated by existing drugs, facilitating drug repurposing-the reuse of established agents for new therapeutic indications-which can greatly accelerate the development of antipoxvirus therapies. In this study, we screened FDA-approved nucleotide metabolic enzyme inhibitors for antivirals against VACV. As a well-characterized and widely used agent, MMF showed no obvious cytotoxicity and exhibited potent efficacy against multiple poxviruses, highlighting its promise as an effective antiviral therapy.

In humans, there are two isoforms of IMPDH that share 84% sequence identity: IMPDH2 is up-regulated in proliferating cells while IMPDH1 plays a housekeeping role and is expressed in most tissues (Burrell and Kollman, 2022). Given that IMPDH2 exhibited a more pronounced phenotype in the context of viral infection, we next sought to characterize its functional role during VACV replication. In our study, we demonstrated that MMF effectively reduces viral DNA levels, and suppression of its cellular target, IMPDH2, also reduces poxvirus replication. Similar to our conclusion, several studies have showed that MMF significantly enhances antiviral efficacy against RSV, SARS-CoV-2, and influenza (Cho et al., 2017; Hu et al., 2026; Wu et al., 2025). We found that knockdown of IMPDH2 recapitulated the antiviral effect observed with MMF treatment and this phenotype was rescued by exogenous guanosine, a downstream metabolite of the IMPDH2-catalyzed reaction. Furthermore, CETSA and molecular docking indicated a binding interaction between MMF and IMPDH2.

Jiang et al. have found that USP5 removed Lys48-linked ubiquitin chains from IMPDH2 through its deubiquitinase activity, preventing its ubiquitin-mediated degradation and stabilizing IMPDH2 (Jiang et al., 2025). Qiao et al. found that USP5 inhibits IRF3-triggered antiviral immune responses through its deubiquitinase activity (Qiao et al., 2025). Nevertheless, the regulatory function of USP5 in IMPDH2 activity in the context of VACV infection remains to be elucidated. In this study, we found that VACV infection promoted USP5-mediated deubiquitination of IMPDH2. Conversely, MMF binds to the active pocket of IMPDH2 and interferes with its interaction with USP5, thereby inhibiting IMPDH2 deubiquitination and depleting guanine nucleotides to suppress VACV infection. However, the underlying mechanisms require further exploration.

IMPDH catalyzes the rate-limiting step in de novo GMP synthesis and is an important target for immunosuppressive and antiviral drugs. In cultured cells, IMPDH inhibitors promote the rapid polymerization of IMPDH protein into R&R structures (Carcamo et al., 2011; Keppeke et al., 2019). Immunofluorescence assays showed that MMF, the prodrug of MPA, promotes IMPDH2 polymerization into catalytically inactive R&R assemblies, thereby functionally sequestering the enzyme and suppressing VACV replication. The antiviral effect of MMF was partially attenuated in cells reconstituted with the IMPDH2 G326A or T333I mutants, likely due to impaired assembly of IMPDH2 R&R structures. These findings indicate that residues G326 and T333 play critical roles in MMF binding to IMPDH2. Notably, both mutations also reduced the basal enzymatic activity of IMPDH2, preventing full restoration of enzyme activity upon reconstitution. Consistent with this observation, VACV replication was only partially rescued in cells expressing these mutants. Further studies are needed to elucidate the underlying mechanisms.

While MMF is well-established as an immunosuppressive agent that inhibits lymphocyte proliferation via IMPDH blockade, accumulating evidence supports its direct, broad-spectrum antiviral activity across multiple virus families (Castañeda Cataña et al., 2025; Hart et al., 2014; Manchala et al., 2019). Our findings demonstrate that MMF exerts significant antiviral activity against poxvirus both in vitro and in vivo.

Conclusions

In summary, our work demonstrates that MMF effectively suppresses viral replication both in vitro and in vivo. Mechanistic studies revealed that MMF inhibits IMPDH2 activity by suppressing USP5-mediated deubiquitination of IMPDH2 and promoting the assembly of “R&R” structures, an inactive conformation, thereby reducing the synthesis of nucleotide precursors required for VACV replication. Collectively, our findings provide a potential antiviral strategy against orthopoxviruses, particularly MPXV, and support MMF as a promising candidate for the treatment of MPXV infection.

Materials and methods

Cells, viruses and plasmids

The HeLa, A549, BHK-21, and HEK-293T cell lines were obtained from the American Type Culture Collection (ATCC). HeLa, BHK-21, and HEK293T cells were cultured in Dulbecco’s Modified Eagle Medium (DMEM; Gibco, #11995065), and A549 cells were maintained in F12K medium (Gibco, #11330500), both supplemented with 10% fetal bovine serum (FBS; Gibco, #10099141). For overexpression of the mutant IMPDH2 (S276A, G326A, T333I, Q441A) and knockdown of IMPDH2, cells were transfected with the human VR-Flag-IMPDH2-mutation plasmid and the pLKO.1-sh-IMPDH2 plasmid. Transfection was performed using Lipofectamine 3000 Transfection Reagent (Thermo Fisher, China) according to the manufacturer’s instructions, followed by selection with puromycin at 2 μg/mL (Selleck, #S9631). All cell cultures were maintained with 1% antibiotics (penicillin-streptomycin) and incubated at 37 °C with 5% CO₂ in a humidified incubator.

Vaccinia virus (VACV) Western Reserve (WR) strain was propagated as described elsewhere (Zhu et al., 2023). The recombinant pGFP-A5 fluorescent virus was constructed through homologous recombination. Briefly, the GFP fragment, along with the A5 homology arms lacking the downstream terminator, was inserted into the multiple cloning site of the pUC19 vector using the primers listed in Supplementary Table S2. This generated the recombinant plasmid pUC19-GFP. BHK-21 cells were infected with VACV-WR at an MOI of 0.1 for 2 h and subsequently transfected with the pUC19-GFP plasmid for 36 h. The recombinant virus was then purified by fluorescence screening following homologous recombination. The recombinant locus was confirmed by sequencing. The resulted recombinant virus was named VACV(A5-GFP). Additionally, the Mpox virus clade IIb (MPXV-B.1-China-C-Tan-CQ01) was generously provided by Zheng Zhang at Shenzhen Third People’s Hospital (Ju et al., 2025). All mpox experiments were conducted in a Biosafety Level 3 laboratory, adhering strictly to standard operating procedures. Virus stocks were prepared, and infectivity titrations were performed in BHK-21 cells for MPXV and VACV amplification.

The shIMPDH2 and shUSP5 primers were sourced from the Broad Institute’s genetic platform (https://portals.broadinstitute.org/gpp/public/gene/search). The IMPDH2 mutant plasmid was constructed using the cDNA of human IMPDH2, which was inserted into the VR vector through homologous recombination at the EcoRI restriction site. The primer sequences are provided in Supplementary Table S2.

Antibodies and chemical reagents

The following antibodies were used in this study: anti-IMPDH2 (mouse-specific; Proteintech, #67663-1-Ig), anti-USP5 (rabbit-specific; Proteintech, #15158-1-AP), anti-UB (rabbit-specific; Proteintech, #10201-2-AP), anti-GAPDH (rabbit-specific, 14C10; Cell Signaling Technology, #2118), and anti-VACV-B5 (rabbit-specific, made in our lab). Secondary antibodies included goat anti-rabbit IgG-HRP (Santa Cruz Biotechnology, #sc-2357) and goat anti-mouse IgG-HRP (Santa Cruz Biotechnology, #sc-2005). The chemical reagents used include the nucleotide metabolism enzyme inhibitors from MedChemExpress provided in Supplementary Table S3, MG-132 (Selleck, #S2619) and guanosine (MedChemExpress, #HY-N0097).

Drug screening

Ten nucleotide synthesis-related enzyme inhibitors were collected, dissolved in DMSO to a final concentration of 10 mM and stored at −20 °C until use. HeLa cells were seeded at a density of 1 × 10⁴ cells per well in a 96-well plate and allowed to incubate overnight. The cells were then infected with VACV(A5-GFP) (MOI = 0.1) in the presence of the compounds (10 μM) dissolved in DMEM. Two hours post-infection, the supernatant was removed, and the cells were treated with fresh medium containing the compounds (10 μM) for an additional 24 h. After fixation with 4% paraformaldehyde (PFA) for 1 h, cells were washed three times with PBS. High-content analysis was performed using a scanning system to measure the fluorescence intensity.

Antiviral activity assessment and cell viability assay

HeLa cells were infected VACV (A5-GFP) at an MOI 0.1 for 24 h and treated with increasing concentrations of candidate drugs (0–100 μM). Fluorescence intensity was recorded using an automated scanning system to assess antiviral activity, and the EC₅₀ was calculated. Cell viability was assessed using Cell Counting Kit-8 (CCK-8) assay (Abcam, #ab228554). HeLa cells were treated with candidate drugs at different concentrations for 24 h. Afterward, cell viability was detected at 450 nm, and the CC₅₀ was calculated using GraphPad Prism 10. Finally, the selectivity index (SI) was determined using the formula SI = EC₅₀/CC₅₀.

Viral infection and titration

The cell culture medium from cells at 90% confluency was removed and replaced with VACV (A5-GFP), MPXV, or VACV (wild-type) diluted in DMEM. Cells were infected for 2 h at 37 °C. The inoculum was then discarded, and the cells were washed twice with PBS and maintained in fresh medium with 2% FBS. For viral titer determination, samples were collected at 24 hpi. Viruses from each treatment were collected by freezing and thawing three times, and the virus titers of VACV (A5-GFP), MPXV, and VACV (wild-type) were determined by plaque assay in BHK-21 cells, as previously described.

Time of addition experiments

HeLa cells were seeded in 12-well plates. Binding group: Cells were pre-treated with MMF for 1 h, washed three times with PBS at 4 °C, and then infected with VACV (MOI = 25) for 1 h at 4 °C in the presence of MMF (10 μM). After infection, cells were harvested for DNA extraction with the DNeasy Blood / Tissue DNA mini kit (Qiagen), and viral DNA- E9L was quantified by quantitative PCR. Entry group: Cells were infected with VACV (MOI = 25) for 1 h at 4 °C, washed three times with PBS at 4 °C, and then treated with MMF (10 μM) and cultured in DMEM containing 2% FBS for 1 h at 37 °C. Cells were subsequently harvested for DNA extraction and absolute quantification of viral DNA. Replication group: Cells were infected with VACV (MOI = 0.1) for 1 h, washed, and then cultured in DMEM containing 2% FBS with MMF (10 μM) for 24 h at 37 °C. Cells were harvested for RNA extraction with the Applied Biosystems SYBR Select Master Mix (Thermo Fisher), and virus E9L and host GAPDH as control was analyzed by quantitative real-time PCR. Release group: Cells were infected with VACV (MOI = 0.1) for 1 h, washed, and then cultured in DMEM containing 2% FBS with MMF (10 μM) for 16 h at 37 °C. Viruses were harvested, and viral progeny were released by three freeze–thaw cycles. Viral titers were determined by plaque assay.

Enzyme activity assay

HeLa cells were seeded at a density of 2 × 10⁶ cells per well in a 6-well plate and infected with VACV for 24 h, with simultaneous treatment with MMF. The cells were then collected and subjected to three freeze-thaw cycles. Enzyme activity was measured according to the manufacturer’s protocol (CUSABIO, # CSB-E17855h), and the results were analyzed accordingly.

Cellular thermal shift assay (CETSA)-Western blot (WB)

CETSA were performed to detect the interaction of MMF with IMPDH2, as previously described (Luo et al., 2022; Tu et al., 2023). The soluble protein lysate of HeLa cells was aliquoted into PCR tubes and treated with MMF (10 μM) or DMSO for 1 h at room temperature (RT) prior to CETSA heat pulse. The solutions were heated at the indicated temperatures (50–85 °C) for 3 min, followed by cooling at 4 °C for 3 min in a thermocycler (Applied biosystems, USA). After centrifugation for 20 min (12,000×g, 4 °C), the soluble supernatant was subject to Western blotting.

Western blot assay

Cells were washed once with ice-cold PBS and lysed in wells with 1 × cell lysis buffer and 1 × PMSF (Beyotime Biotechnology) for Western blot analysis on ice. Proteins were resolved on 10% Tris SwePAGE precast gels (Servicebio; G2177-50T) and transferred to nitrocellulose membranes (Vazyme; E803-01). Membranes were blocked with 5% nonfat milk dissolved in 1 × phosphate-buffered saline (PBS; Beyotime, C0221A) with 0.1% Tween 20 for 2 h at RT, then incubated with primary antibodies diluted in 1 × PBST containing 3% (w/v) BSA (Beyotime, ST2254) overnight at 4 °C. The membranes were washed three times with 1 × PBST and incubated with a secondary antibody conjugated with horseradish peroxidase diluted 1:5000 in 1 × PBST containing 5% nonfat milk for 1 h at RT. Signals were detected using Chemistar High-Sig ECL Western blot substrate (Tanon, 4600) and imaged on a Tanon 5200 system (Tanon).

Molecular docking technology

The receptor protein-IMPDH2 was searched in the Uniprot database (https://www.uniprot.org/). The 3D structure of the protein was downloaded from the RCSB PDB database (https://www.rcsb.org/). The 2D structure of the molecule ligands was downloaded from the PubChem database (https://pubchem.ncbi.nlm.nih.gov/). ChemBio 3D software was used to calculate and export the 3D structure by minimizing energy. PyMOL 2.4.0 software was performed the dehydration of the receptor protein and Autodock software was used to carry out hydrogenation and charge calculation of proteins. The parameters for the receptor protein docking site were set to include the active pocket sites where small-molecule ligands bind. Finally, AutoDock Vina was used to perform molecular docking between the receptor protein and the small-molecule ligand MPA, an active compound of MMF.

Bioinformatics and statistical analyses

Volcano plot was generated to present the differentially expressed mRNA or proteins based on the screening criteria. Quantitative variables were presented as mean ± standard deviation (SD) or as the median and interquartile range. Comparisons of continuous variables were performed using one-way analysis of variance (ANOVA) or nonparametric tests. A P-value <0.05 was considered significant.

Confocal microscopy

Cells seeded onto confocal dishes were transfected with the indicated plasmids. At 24 hpt, medium was removed, and cells were fixed with 4% paraformaldehyde (Beyotime, #P0099) at RT for 10 min, followed by three washes with PBST. The cells were then permeabilized by adding 1% (vol/vol) Triton X-100 in PBS for 10 min and blocked with 3% BSA in PBS for 1 h. Cells were incubated overnight at 4 °C with primary antibodies. After washing three times with PBS, the cells were incubated with a fluorescein-conjugated secondary antibody. The cells were washed with PBS and stained with 4’,6-diamidino-2-phenylindole (DAPI; Sigma-Aldrich, #D9542). Images were captured using an inverted fluorescence microscope with a 60 × oil immersion objective.

Proximity ligation assay (PLA)

HeLa cells were infected with VACV for 24 h. PLA was performed using the Duolink In Situ Detection Reagents Green (Sigma-Aldrich, #DUO92014). According to the manufacturer’s instructions, cells were sequentially fixed, recovered, and permeabilized. They were then blocked with Duolink blocking buffer at 37 °C for 1 h, followed by incubation with mouse anti-IMPDH2 antibody and rabbit anti-USP5 antibody at 37 °C for 2 h. Next, cells were treated with pre-diluted anti-rabbit and anti-mouse minus probes at 37 °C for 1 h. Finally, cells were incubated with 1 × ligase for 30 min and 1 × polymerase for 100 min, then mounted on slides with Duolink In Situ Mounting Medium containing DAPI.

SDD-AGE assay

HeLa cells were lysed on ice for 20 min in a non-denaturing buffer containing 0.5% Triton X-100, 50 mM Tris–HCl (pH 7.5), 150 mM NaCl, 1% glycerol, and a protease inhibitor cocktail. Cell lysates were centrifuged at 12,000×g at 4 °C for 15 min, then mixed with 5 × sample loading buffer containing 2.5 × TBE, 2.5% SDS, 25% glycerol, and 0.25% bromophenol blue. Proteins were electrophoresed on a vertical 1% agarose gel in running buffer containing 0.5 × TBE and 0.1% SDS. Electrophoresis was performed at 80 V for 120 min, and then transferred to a NC membrane at 4 °C. Immunoblot analysis of the membranes was conducted as mentioned above.

Mice

The 6-week-old female BALB/c mice were purchased from GemPharmatech. The mice were housed in specific pathogen-free barrier facilities at the Suzhou Institute of Systems Medicine.

VACV infection in mice and tissue collection

6-week-old mice were divided into four groups (n = 12 per group): (1) DMSO, (2) MMF-treated, (3) VACV-infected and DMSO-treated, (4) VACV-infected and MMF-treated. Consider a 30% weight loss as the threshold for death. Four mice per group were euthanized at six days post-infection, and lung and brain tissues were collected.

DNA loads in mouse tissues

Brain and lung tissues from four mice per group were dissected at 6 days post-infection (dpi). Each 100 mg tissue sample was homogenized in 1 mL PBS, and the homogenates were briefly centrifuged for 10 min (12,000×g, 4 °C) to collect the supernatant. Viral DNA was extracted from the supernatants and quantified by q-PCR. Absolute quantification of the VACV E9L gene was performed with the FastPure Viral DNA/RNA Mini Kit Pro (Vazyme, #RC323-01). Primer sequences are listed in Supplementary Table S2.

Hematoxylin-eosin staining

Tissues were fixed in 4% paraformaldehyde for more than 24 h, dehydrated through graded ethanol solutions, and incubated in a 1:1 xylene / ethanol solution for 5 min, followed by two sequential 10 min incubations in pure xylene. Specimens were embedded in paraffin, sectioned, and stained with hematoxylin and eosin (H&E) according to standard protocols. Slides were examined by light microscopy.

Statistical analysis

Results are expressed as mean ± standard deviation (SD). Statistical analysis was performed using GraphPad Prism 10.0 software (GraphPad Software Inc., USA). Statistical significance was determined using unpaired two-tailed Student’s t-test for two-group comparisons, or one-way ANOVA followed by Tukey’s or Holm-Sidak’s multiple-comparisons test where appropriate, as specified in the figure legends. Unless otherwise indicated in the figure legends, P values were calculated from three independent biological replicates. All experiments were repeated independently, as noted in the figure legends. A P-value ≥ 0.05 was considered non-significant, P < 0.05 was considered statistically significant (∗), P < 0.01 was considered highly significant (∗∗), and P < 0.001 was considered extremely significant (∗∗∗).

Data availability

All data generated or analyzed during this study are included in this published article (and its supplemental files). Original data and materials will be available upon requests. The sequences used in this study were obtained from the National Center for Biotechnology Information (NCBI, https://www.ncbi.nlm.nih.gov/).

Ethics statement

All experimental animal procedures were carried out in accordance with the Animal Welfare Act and Guide for the Care and Use of Laboratory Animals, as approved by the Suzhou Institute of Systems Medicine (approval number: ISM-IACUC-20250113).

Author contributions

Qian Sun: investigation, writing-original draft, methodology, visualization, formal analysis, and data curation. Kesen Liu: investigation, writing-original draft, methodology, visualization, formal analysis, and data curation. Wandi Cao, Chengyue Wu, Hanhua Zhang, Xingya Wang: validation, and writing-review and editing. Chen Peng, Jie Sun, Anbing Zhang, Zhuo Zhou: project administration, writing-review & editing. Xing Liu: conceptualization, data curation, investigation, writing-original draft, methodology, visualization, formal analysis, funding acquisition, project administration, resources, supervision, writing-review & editing.

Conflict of interest

All authors declare no competing interests.

Appendix A. Supplementary data

The following are the Supplementary data to this article: