ZEB2 SUMOylation in Cardiac Fibroblast Drives Post-Infarction Cardiac Remodeling through CtBP1-Associated repression of Nr4a1

Yilin Wang; Shaopeng Cheng; Hongyu Chen; Zhihao Yang; Xiaoting Wu; Qiuyan Zong; Jianing Wang; Yi Jiang; Yanrong Yu; Hao Chen; Yunxing Xue; Qianwen Zhao; Jie Yang; Dongjin Wang

doi:10.1016/j.isci.2026.117091

. 2026 Aug 6;29(8):117091. doi: 10.1016/j.isci.2026.117091

ZEB2 SUMOylation in Cardiac Fibroblast Drives Post-Infarction Cardiac Remodeling through CtBP1-Associated repression of Nr4a1

Yilin Wang ^1,^2,¹⁰, Shaopeng Cheng ^2,^3,¹⁰, Hongyu Chen ^2,^4,¹⁰, Zhihao Yang ^2,³, Xiaoting Wu ^2,⁵, Qiuyan Zong ^1,², Jianing Wang ⁴, Yi Jiang ^2,⁴, Yanrong Yu ^1,², Hao Chen ^7,⁸, Yunxing Xue ^2,^3,^∗, Qianwen Zhao ^6,^∗∗, Jie Yang ^2,^3,^9,^∗∗∗, Dongjin Wang ^1,^2,^11,^∗∗∗∗

PMCID: PMC13475699 PMID: 42602983

Summary

The transcription factor ZEB2 has been implicated in cardiovascular disease, but its role and post-translational regulation in cardiac fibroblast-to-myofibroblast transition (FMT) and fibrosis after myocardial infarction (MI) are not fully understood. ZEB2 protein expression was significantly increased in infarcted murine hearts, activated cardiac fibroblasts, and human fibrotic myocardium. Myofibroblast-specific ZEB2 knockout alleviates cardiac dysfunction and adverse remodeling post-MI. Mechanistically, SENP1 regulated SUMO-1 modification of ZEB2 at K462, K479, K611, and K774. Reconstituting ZEB2-deficient systems with wild-type ZEB2 (ZEB2-WT) restored pathological fibrosis and FMT, whereas the SUMOylation-deficient mutant (ZEB2-4KR) conferred cardioprotective effects. ZEB2 SUMOylation activated the pro-fibrotic PI3K/AKT-mTORC1 signaling pathway. ZEB2-4KR weakens CtBP1 binding and enhances ZEB2 occupancy at the Nr4a1 promoter. Importantly, Nr4a1 knockdown abolished the protective effects of ZEB2-4KR. These findings identify SUMO-1 modification of ZEB2 as a central driver and molecular switch of pathological cardiac remodeling after MI and suggest that targeting ZEB2 SUMOylation may represent a promising therapeutic strategy.

Keywords: ZEB2, myocardial infarction, SUMOylation, Nr4a1, myofibroblasts, cardiac remodeling

Graphical abstract

Highlights

•
ZEB2 upregulated post MI drives FMT, cardiac fibrosis, and adverse remodeling
•
Myofibroblast-specific ZEB2 deletion attenuates post-MI cardiac dysfunction and fibrosis
•
SENP1-mediated SUMO-1 stabilizes ZEB2, activating pro-fibrotic PI3K/AKT-mTORC1 signaling
•
SUMOylation-deficient ZEB2-4KR induces Nr4a1, protecting against post-MI remodeling

Molecular biology; Epigenetics; Cell biology

Introduction

Myocardial infarction (MI) remains a leading cause of mortality worldwide, imposing a substantial burden on healthcare systems.¹ A critical determinant of long-term prognosis is pathological cardiac remodeling, characterized by ventricular dilation, wall thinning, and excessive myocardial fibrosis.²^,³ This maladaptive fibrosis, driven by dysregulated extracellular matrix (ECM) deposition, impairs myocardial compliance and accelerates the progression to heart failure (HF).⁴ Despite advancements in reperfusion strategies and pharmacological therapies, no therapies specifically target post-MI cardiac fibrosis.⁵^,⁶ This unmet medical need underscores the urgent imperative to identify novel molecular targets and elucidate the underlying mechanisms driving fibrotic remodeling.

Cardiac fibroblasts (CFs), the predominant stromal cells in the heart, are central orchestrators of both physiological repair and pathological fibrosis following injury.⁷ Following MI, CFs undergo a fibroblast-to-myofibroblast transition (FMT), acquiring a contractile phenotype and secreting excessive ECM components.⁸ While essential for acute wound healing, sustained FMT activation is a major driver of adverse remodeling and cardiac dysfunction.⁹ However, the molecular mechanisms controlling FMT, particularly the regulation of posttranslational modifications (PTMs) of key regulatory proteins, remain incompletely defined.

Nuclear transcriptional regulators play pivotal roles as determinants of CF activation and fate commitment during FMT.¹⁰^,¹¹^,¹² Among these, zinc finger binding E-box 2 (ZEB2), has been linked to coronary artery disease (CAD) risk.¹³ ZEB2 is a recognized regulator of epithelial-to-mesenchymal transition (EMT), and has been implicated in various cardiovascular contexts, including endothelial-to-mesenchymal transition (EndoMT), cardiac development, and ischemia-induced injury.¹³^,¹⁴^,¹⁵ It exerts context-dependent effects on inflammation, survival, and remodeling across diverse cell types such as cardiomyocytes, endothelial cells, and macrophages.¹⁶^,¹⁷^,¹⁸ Although ZEB2 has been linked to FMT regulation, its specific mechanistic role within CFs post-MI and the regulatory pathways involved are poorly understood.¹⁹

Crucially, we observed pronounced upregulation of ZEB2 protein, but not mRNA levels, in both infarcted myocardium and activated CFs. Mass spectrometry (MS) analysis identified SUMO1 (small ubiquitin-like modifier 1) as a key modifier of ZEB2. SUMOylation, a reversible PTM, involves the covalent attachment of SUMO proteins to lysine residues. This process is catalyzed by an E1-activating enzyme (SAE1/SAE2), E2-conjugating enzyme (UBC9), and E3 ligase, while deSUMOylation is mediated by SENP (sentrin/SUMO-specific proteases) family members.²⁰ SUMOylation critically modulates protein stability, localization, and activity, influencing diverse processes including inflammation and fibrosis.²¹ Notably, SUMOylation fine-tunes several transcription factors pivotal in cardiovascular disease, such as TEAD1, HEY1, and SRF, regulating their control over gene networks involved in vascularization and hypertrophy.²²^,²³^,²⁴ However, the functional impact of SUMOylation on transcription factors, particularly within CFs, remains poorly defined. Building on the identification of ZEB2 SUMO1 association and its observed PTM upregulation, we hypothesized that ZEB2 is upregulated in CFs post-MI to coordinate fibrotic responses via SUMO1-mediated modification, serving as a key regulator of FMT and maladaptive remodeling.

In this study, we provide the first evidence demonstrating ZEB2 modification by SUMO1 at specific lysine residues (K462, K479, K611 and K774). This SUMOylation is dynamically reversed by the SUMO-specific protease SENP1 in vitro. Critically, we demonstrate that ZEB2 SUMOylation deficiency attenuates fibrosis and preserves cardiac function following MI. While our co-IP assay detected SUMO-1 modification conjugation in TGF-β1-stimulated CFs or infarcted myocardium, it revealed no significant changes in SUMO-2/3 modification. These findings demonstrate that SUMO-1-mediated modification of ZEB2 represents a potential mechanism and suggest that targeting ZEB2 SUMOylation may provide a basis for novel therapeutic strategies to adverse cardiac remodeling.

Results

ZEB2 is upregulated in ischemic fibrotic hearts and cardiac fibroblasts post-MI

To investigate whether ZEB2 expression is associated with myocardial infarction, we first measured ZEB2 expression levels in ventricular aneurysms from patients with chronic myocardial infarction undergoing surgical aneurysmectomy, where it co-localized with vimentin-positive cells (Figure 1A). In addition, no significant changes in ZEB2 mRNA expression were detected in the infarcted myocardium (Figure 1B). To investigate ZEB2 dynamics following MI, we assessed its expression in murine hearts. The immunoblotting analysis demonstrated that ZEB2 was upregulated at day 7 and day 14 post-MI (Figures S1A and S1B). Critically, ZEB2 protein expression was significantly increased within the infarct zone, coinciding with increased α-SMA expression; no significant changes were detected in the non-infarct area (Figures 1C and D, F-G). Immunofluorescence staining confirmed ZEB2 upregulation in the infarcted region, revealing strong co-localization with vimentin-positive cells, in contrast to the non-infarcted zone where co-localization was weaker (Figure 1I). ZEB2 expression is predominantly localized to interstitial, vimentin-positive cells within the infarct area. These vimentin⁺ cells exhibit a fibroblast-like morphology and occupy the interstitial compartment, consistent with activated CFs. To determine the cellular source of increased ZEB2, we isolated cardiomyocytes and fibroblasts from sham and MI hearts at day 14. Immediate analysis upon isolation showed a marked increase in ZEB2 protein specifically in cardiac fibroblasts from MI hearts, while cardiomyocyte ZEB2 expression remained unaltered (Figures 1J–1L). Consistent with the in vivo findings, stimulation with TGF-β1 robustly induced ZEB2 protein expression in cultured cardiac fibroblasts (Figures 1M and 1N). Notably, however, ZEB2 mRNA levels remained unchanged both in vivo (post-MI tissues and isolated cells) and in vitro (TGF-β1-stimulated fibroblasts) (Figures 1E–1H, and 1O). Immunofluorescence staining showed that TGF-β1 stimulation increased ZEB2 expression in the nucleus (Figure 1P). Collectively, these results demonstrate that ZEB2 protein expression is significantly upregulated in cardiac fibroblasts within the infarct zone following MI, independently of mRNA level changes, suggesting post-transcriptional regulation.

ZEB2 is increased in ischemic myocardium and cardiac fibroblasts post-MI

(A) Immunofluorescence staining for ZEB2 (red), Vimentin (green), and DAPI (blue) in the human infarcted myocardium (MI) ventricular aneurysm tissue undergoing surgical aneurysmectomy (chronic phase after MI) and healthy control myocardium. Scale bars, 20 μm, n = 3.

(B) qPCR analysis and quantification of ZEB2 in healthy control and infarcted myocardium with ventricular aneurysms from patients with prior myocardial infarction. Data were normalized to the internal control and expressed relative to healthy control myocardium.

(C–H) C57BL/6J mice were subjected to sham or MI surgery for 2 weeks. Western blotting and qRT-PCR were performed on the heart lysates from the infarct zones and non-infarct zones at 2 weeks post-MI to analyze the expression of ZEB2, n = 6 mice per group. Data were normalized and expressed relative to the corresponding sham group.

(I) Immunofluorescence co-staining for vimentin (green) with ZEB2 (red) and DAPI (blue) in the heart post-MI. Scale bars, 20 μm, n = 6 mice per group.

(J–L) ZEB2 protein expression was measured in cardiomyocytes and fibroblasts isolated from the sham or MI heart on 14 days after the surgery. n = 6 mice per group.

(M–O) Cardiac fibroblasts were isolated and cultured and then treated with TGF-β1(10 ng/mL) for 24 h. Western blotting, qPCR analysis, and quantification of ZEB2 are shown. n = 3 individual experiments. Data are normalized and expressed relative to untreated cardiac fibroblasts.

(P) Immunofluorescence co-staining of ZEB2 (green) and DAPI (blue) in cardiac fibroblasts. Scale bars, 20 μm. n = 3 individual experiments. Data are expressed as mean ± SD. Two-tailed Student’s t test were used for analysis. ∗∗∗p < 0.001, ns means no significant statistical analysis.

Myofibroblast-specific knockout of ZEB2 attenuates adverse remodeling and cardiac dysfunction post-MI

To investigate the temporal role of ZEB2 in established myofibroblast-mediated fibrosis, we generated inducible myofibroblast-specific ZEB2 knockout mice (ZEB2^MF−KO) by crossing ZEB2^f/f mice with Postn-MerCreMer mice (Figures S2A and S2B). Genotyping PCR confirmed the ZEB2 floxed allele and successful Cre-mediated recombination in ZEB2^MF−KO mice (Figure S2C). To verify that loxP site insertion does not impair basal ZEB2 expression, we examined mRNA in hearts from ZEB2^f/f and wild-type (WT) mice, which revealed no significant difference between groups (Figure S2D). ZEB2^f/f and ZEB2^MF−KO mice were subjected to MI surgery. To allow initial myofibroblast transition prior to gene deletion, tamoxifen injections were administered for 5 consecutive days starting on day 8 post-MI. The immunofluorescence of ZEB2 and postn in the infarcted myocardium from ZEB2^f/f and ZEB2^MF−KO mice confirmed myofibroblast-specific ZEB2 knockout (Figure S2E). Finally, immunoblot analysis of fibroblasts isolated from MI hearts confirmed efficient ZEB2 knockout in ZEB2^MF−KO mice (Figures S2F–S2G). TTC staining at day 7 post-MI (pre-tamoxifen) verified comparable initial infarct size between ZEB2^f/f and ZEB2^MF−KO mice (Figures S2H and S2I). Assessment of cardiac function demonstrated that ZEB2^MF−KO mice exhibited significantly preserved EF and FS compared to ZEB2^f/f mice post-MI (Figures 2A–2C; Table S2). ZEB2^MF−KO mice also displayed a reduced heart weight-to-body weight (HW/BW) ratio (Figure 2D). Histological staining revealed that myofibroblast-specific ZEB2 knockout reduced scar size and suppressed adverse left ventricular remodeling post-MI (Figures 2E–2J). Furthermore, ZEB2 knockout in myofibroblasts markedly attenuated cardiac fibrosis, as evidenced by reduced expression of profibrotic genes (Figures 2K–2M). Collectively, these data demonstrate that myofibroblast-specific ZEB2 knockout alleviates adverse remodeling and cardiac dysfunction following MI.

Myofibroblast-specific ZEB2 knockout attenuates adverse remodeling and cardiac dysfunction after MI

(A–C) ZEB2^f/f and ZEB2^MF−KO mice were injected with tamoxifen for 5 continuous days after being subjected to MI surgery for 7 days. At 14 days post-MI, cardiac function indices were measured by echocardiography. EF (ejection fraction) and FS (fractional shortening) were quantified. n = 8 mice per group.

(D) Quantification of HW/BW (heart weight/body weight) between ZEB2^f/f and ZEB2^MF−KO mice. n = 8 mice per group.

(E–H) Heart sections were stained with Masson trichrome at 2 weeks after MI. The scar size was measured and quantified as a percentage of the total circumference of the left ventricle. n = 8 mice per group. Scale bars, 500 μm and 50 μm.

(I and J) HE staining of cardiomyocyte cell size in left ventricular non-infarct zone. Scale bars, 500 μm and 50 μm.

(K and L) Western blot and analysis of α-SMA and Collagen I in the infarct zone at 2 weeks post-MI. n = 6 mice per group.

(M) mRNA expression levels of *Acta2, Col1α1, Col3α1, Ctgf* were detected by real-time quantification PCR. n = 6 mice per group. Data are expressed as mean ± SD; two-way ANOVA followed by Tukey’s multiple comparisons test was used for analysis. ∗p < 0.05, ∗∗p < 0.01, ∗∗∗p < 0.001.

SUMO1 modification enhances ZEB2 stability via conserved lysine residues

To elucidate the molecular mechanism underlying ZEB2’s regulation of cardiac remodeling, we performed co-immunoprecipitation (Co-IP) coupled with mass spectrometry (MS) on CFs transfected with flag-tagged ZEB2 (Adv-Flag-ZEB2) or Adv-EGFP control (Figure 3A). MS analysis of ZEB2 immunocomplexes identified peptides corresponding to SUMO1 (Figure 3B). Given our observation of post-transcriptional upregulation of ZEB2 protein (Figure 1) and the well-established role of SUMOylation in modulating transcription factor (TF) function and target gene expression, we hypothesized that ZEB2 undergoes SUMOylation. Consistent with this, Co-IP of endogenous proteins demonstrated SUMO1 conjugation of ZEB2 in both TGF-β1-treated CFs and infarcted myocardium (Figures 3C and 3D). We next assessed ZEB2 SUMO1 modification in CMs and cardiac microvascular endothelial cells (CMECs) exposed to hypoxic conditions. Our results revealed no-significant changes in ZEB2 SUMOylation levels in CMs and CMECs compared to baseline, while CFs exhibited a robust and pronounced increase in ZEB2 SUMOylation specifically in response to TGF-β1 stimulation (Figure S3A). In addition, we examined potential modification by SUMO2/3. However, Co-IP assay detected no changes in SUMO2 or SUMO3 conjugation of ZEB2 in either MI heart tissue or TGF-β1-treated CFs compared with control (Figures S3B and S3C). These findings indicate that both TGF-β1 stimulation and MI specifically and significantly enhance SUMO1-mediated modification of ZEB2. To identify potential SUMOylation sites, we co-infected CFs with lentivirus expression His-tagged SUMO1 (Lenti-His-SUMO1) and flag-tagged ZEB2 containing a TEV cleavage site and biotinylated peptide (pEBB-TB-ZEB2). Subsequent LC-MS/MS analysis identified nine lysine residues (K462, K475, K479, K532, K537, K611, K689, K696, K774) within ZEB2-bound SUMO1 peptide. We then employed the group-based prediction system-SUMO (GPS-SUMO) and joint-analyzer of SUMOylation sites (JASSA) to predict SUMOylation sites of ZEB2. Intersection of the MS-identified sites with the bioinformatically predicted sites revealed four high-confidence candidate SUMOylation residues: K462, K479, K611 and K774 (Figures 3E and S4A). Notably, these lysine residues were highly conserved across vertebrates, underscoring their potential functional significance. We constructed ZEB2 point mutants at each of these lysine (K462R, K479R, K611R and K774R) and a quadruple mutant (4KR) (Figure S4B). Immunoprecipitation assay demonstrated that individual lysine-to-arginine mutations (K462R, K479R, K611R, K774R) modestly reduced ZEB2 SUMOylation, while the 4KR mutation almost completely abolished SUMO1 conjugation (Figure 3F). To evaluate the contribution of these sites, we produced lentiviruses expressing every single mutant or the 4KR mutant and infected ZEB2^KO CFs. Expression of ZEB2-4KR markedly reduced the TGF-β1-induced α-SMA expression in ZEB2^KO CFs (Figures 3G and 3H). SUMOylation can regulate target proteins by influencing stability, subcellular localization, or protein-DNA interactions.²⁵ We therefore assessed whether SUMOylation impacts these properties of ZEB2. Immunofluorescence analysis indicated that SUMOylation-deficient ZEB2 (ZEB2-4KR) remained predominantly to the nucleus, similar to ZEB2-WT (Figure 3I), indicating that SUMOylation does not regulate ZEB2 nuclear localization. By contrast, cycloheximide (CHX) chase assays revealed that ZEB2-4KR had a significantly shorter half-life than ZEB2-WT (Figures S4C and S4D). Furthermore, the proteasome inhibitor MG132 prevented degradation of both ZEB2-WT and ZEB2-4KR, confirming that both forms are degraded via the proteasome pathway (Figures 3J and 3K). In conclusion, our findings demonstrate that ZEB2 undergoes covalent SUMO1 modification in CFs, both in vivo and in vitro, that this modification occurs primarily at conserved lysine residues K462, K479, K611, and K774, and that SUMOylation is critical for maintaining ZEB2 protein stability.

SENP1 is the primary DeSUMOylase regulating ZEB2 deSUMOylation

SUMOylation is a dynamic, reversible post-translational modification regulated by the opposing activities of SUMO ligases (E3 ligases) and deSUMOylases (SENPs).²⁰ The SENP family includes SENP1-3 and SENP5-7, with SENP1 and SENP2 exhibiting broad specificity for all SUMO isoforms, while SENP3 and SENP5-7 preferentially target SUMO2/3-modified proteins and SUMO chains.²⁶ Given our findings that ZEB2 is modified by SUMO1, we investigated whether SENP1 or SENP2 regulates ZEB2 deSUMOylation. To identify the specific deSUMOylase for ZEB2, we co-transfected HEK293T cells with plasmids expressing flag-tagged wild-type ZEB2 (Flag-ZEB2-WT) or its SUMOylation-deficient mutant (Flag-ZEB2-4KR), His-SUMO1, and Myc-tagged SENP1 (Myc-SENP1) or Myc-tagged SENP2 (Myc-SENP2). Co-IP assays revealed that SENP1 overexpression significantly attenuated ZEB2 SUMOylation. In contrast, SENP2 overexpression only partially reduced this modification (Figures 4A and 4B). To confirm the specific role of SENP1, we performed loss-of-function experiments. siRNA-mediated knockdown of SENP1, but not SENP2, significantly enhanced ZEB2 SUMOylation in HEK293T cells (Figures 4C and 4D). Collectively, these findings demonstrate that SENP1 serves as the primary deSUMOylase responsible for dynamically regulating ZEB2 SUMOylation.

SENP1 is the key deSUMOylation that mediates ZEB2 deSUMOylation

(A) HEK293T cells were transfected with Flag-tagged ZEB2-WT, Flag-ZEB2-4KR, Myc-SENP1, His-SUMO1 plasmids. Lysates from cells were immunoprecipitated with anti-Flag antibody, and immunoblot assays were performed using SUMO1 antibody.

(B) HEK293T cells were transfected with Flag-tagged ZEB2-WT, Flag-ZEB2-4KR, Myc-SENP2, His-SUMO1 plasmids. Lysates from cells were immunoprecipitated with anti-Flag antibody, and immunoblot assays were performed using SUMO1 antibody.

(C and D) HEK293T cells were transfected with Flag-tagged ZEB2-WT, Flag-ZEB2-4KR and His-SUMO1 plasmids and si-NC, si-SENP1 or si-SENP2. Lysates from cells were immunoprecipitated with anti-Flag antibody, and immunoblot assays were performed using SUMO1 antibody. n = 3 individual experiments.

ZEB2 SUMOylation is essential for TGF-β1-induced myofibroblast transformation in vitro

To definitively establish the functional significance of ZEB2 SUMOylation and eliminate confounding effects from endogenous ZEB2, we utilized CFs isolated from ZEB2^KO mice (generation detailed in Figure S5). These ZEB2^KO CFs were infected with adenovirus expressing either flag-tagged wild-type ZEB2 (Adv-Flag-ZEB2-WT) or its SUMOylation-deficient mutant (Adv-Flag-ZEB2-4KR) to reconstitute ZEB2 expression (Figures S6A and S6B). Following TGF-β1 stimulation for 24 h, collagen gel contraction assays revealed that reconstitution with ZEB2-WT further enhanced contractility compared to GFP controls in ZEB2^KO mice CFs. Crucially, expression of ZEB2-4KR significantly attenuated TGF-β1-induced contractile functions (Figures 5A and 5B). These in vitro gain-of-function experiments in a ZEB2-null background provide direct evidence that ZEB2 SUMOylation is required for acquisition of myofibroblast contractile properties. We then examined the impact of ZEB2 SUMOylation on FMT and proliferation. Co-staining for α-SMA (green) and EdU (red) in ZEB2^KO CFs revealed that EdU+ nuclei were predominantly localized with α-SMA+ cells, indicating that proliferating cells are largely myofibroblasts undergoing FMT (Figures 5C and 5D). Consistent with the functional data, western blot analysis demonstrated that ZEB2-4 KR expression in ZEB2^KO CFs significantly reduced TGF-β1-induced α-SMA protein levels compared with ZEB2-WT (Figures 5E and 5F). Furthermore, qPCR analysis showed significantly reduced mRNA expression of myofibroblast marker genes (Acta2, Col1α1, Col3α1, and Ctgf) in TGF-β1-stimulated ZEB2^KO CFs expressing ZEB2-4KR compared to those expressing ZEB2-WT (Figure 5G). Collectively, these results, obtained in a ZEB2-null background, demonstrate that SUMOylation is essential for ZEB2 to promote TGF-β1-induced FMT in CFs. Reconstitution of ZEB2^KO CFs with the SUMOylation-deficient ZEB2-4KR mutant robustly suppressed TGF-β1-driven FMT and its associated functional, phenotypic, and transcriptional changes, unequivocally highlighting the requirement of ZEB2 SUMOylation for this process.

SUMOylation-deficient ZEB2 attenuates adverse remodeling and cardiac dysfunction post-MI

To investigate the functional significance of ZEB2 SUMOylation in adverse remodeling in vivo, we performed cardiac-specific reconstitution in ZEB2^KO mice. ZEB2^KO mice were intravenously injected with recombinant adeno-associated virus serotype 9 (rAAV9) expressing either flag-tagged WT-ZEB2 (rAAV9-ZEB2-WT-3×Flag) or the SUMOylation-deficient mutant (rAAV9-ZEB2-4KR-3×Flag) (Figure S7A). Two weeks after injection, robust and comparable ZEB2 protein expression was confirmed in the left ventricular myocardium cardiac fibroblasts of both rAAV9-ZEB2-WT and rAAV9-ZEB2-4KR groups, with no significant difference (Figures S7B–S7D). ZEB2 levels achieved in AAV-treated knockout hearts are in the same order of magnitude as the endogenous, about 2- to 2.3-fold increase observed in wild-type hearts at day 14 post-MI. All mice then underwent either sham or MI surgery. Echocardiographic assessment revealed that the impact of ZEB2 reconstitution on post-MI cardiac function depended critically on its SUMOylation status. Reconstitution with ZEB2-WT in ZEB2^KO mice exacerbated MI-induced cardiac dysfunction, as reflected by significantly reduced ejection fraction (EF) and fractional shortening (FS) compared to GFP MI mice. Conversely, reconstitution with the SUMOylation-deficient ZEB2-4KR mutant significantly attenuated cardiac dysfunction, preserving higher EF and FS compared to both ZEB2-WT-reconstituted and GFP-reconstituted ZEB2^KO mice post-MI (Figures 6A–6C; Table S3). HW/BW ratio was significantly reduced in ZEB2-4KR-reconstituted ZEB2^KO mice compared to ZEB2-WT-reconstituted and GFP-reconstituted controls post-MI (Figure 6D). Histological analysis demonstrated that ZEB2-4KR-reconstituted ZEB2^KO mice exhibited smaller infarct sizes and attenuated cardiomyocyte hypertrophy in the non-ischemic regions compared to ZEB2-WT-reconstituted and GFP-reconstituted controls post-MI (Figures 6E–6J). The protein expression of α-SMA and collagen I was significantly decreased in ZEB2-4KR mice remodeling myocardium (Figures 6K–6L). The qPCR analysis further revealed significantly decreased mRNA expression of Acta2, Col1α1, Col3α1 and Ctgf in the myocardium of ZEB2-4KR-reconstituted ZEB2^KO mice compared to ZEB2-WT-reconstituted and GFP-reconstituted controls post-MI (Figure 6M).

To validate these findings in an alternative model of pathological remodeling, we subjected mice to ISO-induced cardiac stress (10 mg/kg/d for 14 days). Echocardiography demonstrated that ZEB2-4KR-reconstituted ZEB2^KO mice were protected from ISO-induced cardiac dysfunction compared to ZEB2-WT-reconstituted and GFP-reconstituted controls (Figures S8A–S8C; Table S4). ZEB2-4KR-reconstitution significantly attenuated the ISO-induced increase in HW/BW ratio, ISO-triggered cardiomyocyte enlargement, and interstitial fibrosis (Figures S8D–S8I). Consistently, protein levels of α-SMA and collagen I were markedly downregulated in the hearts of ISO-challenged ZEB2-4KR-reconstituted ZEB2^KO mice compared with the ZEB2-WT-reconstituted group (Figures S8J–S8K), and mRNA expression of fibrosis-related genes was similarly reduced (Figure S8L). Collectively, these results from two distinct models of cardiac fibrosis demonstrated that genetic ablation of ZEB2 SUMOylation (via the 4KR mutant) potently attenuates adverse remodeling and dysfunction in vivo. Reconstitution of ZEB2^KO mice with the SUMOylation-deficient ZEB2-4KR mutant confers significant cardiac protection against both MI and ISO-induced pathological changes.

SUMOylation-deficient ZEB2 impaired CtBP1 recruitment, thereby increasing Nr4a1 expression and PI3K/AKT-mTOR pathway inhibition

To elucidate the molecular mechanism by which ZEB2 SUMOylation regulates cardiac remodeling, we employed an integrated transcriptomic approach. In the first arm of the study, ZEB2^KO CFs were transfected with Adv-Flag-ZEB2-WT or Adv-Flag-ZEB2-4KR followed by TGF-β1 stimulation and RNA-seq analysis. Both Flag-ZEB2-WT and flag-ZEB2-4KR altered the cell transcriptome, including differential expression of 113 genes, with a majority being downregulated and upregulated (Figure 7A). Among these DEGs, we identified the top 9 genes that were upregulated and 11 downregulated by ZEB2 SUMOylation deficiency in ZEB2^KO CFs followed by TGF-β1 stimulation (Figure 7B). Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway analysis further indicated their enrichment in several profibrogenic pathways, particularly the PI3K-AKT and mTOR signaling pathways (Figure 7C). To understand the basis of these transcriptional changes, we performed genome-wide profiling of chromatin binding for overexpressed flag-ZEB2-WT and flag-ZEB2-4KR in ZEB2^KO CFs using CUT&Tag-seq. Both ZEB2-WT and ZEB2-4KR exhibited discrete binding peaks across the genome (Figure 7D). Notably, although both proteins bound promoter regions, ZEB2-4KR exhibited a significant increase in promoter-binding capability overall compared to ZEB2-WT (Figure 7G). Integration of the RNA-seq and CUT&Tag-seq datasets identified five genes whose expression was upregulated by ZEB2-4KR and whose promoters were bound by ZEB2-4KR. Among the top five candidate target genes (Nr4a1, Rorb, Hmox1, Macro and Dht1) (Figures 7E and 7F), Nr4a1 ranked highest. Nr4a1 has been documented as a well-established suppressor of cardiac fibrosis.²⁷^,²⁸^,²⁹ We focused subsequent analyses on its regulation by ZEB2. Consistent with the CUT&Tag-seq data, the chromatin immunoprecipitation (ChIP) assay confirmed that both ZEB2-WT and ZEB2-4KR bound to the Nr4a1 promoter region, but crucially, with ZEB2-4KR displaying significantly stronger promoter occupancy (Figures 7G and 7H).

SUMOylation-Deficient ZEB2 attenuates Fibrosis via Enhanced Nr4a1 repression and PI3K/AKT-mTOR pathway inhibition

ZEB2^KO CFs were infected with Adv-ZEB2-WT and Adv-ZEB2-4KR, followed by exposure to vehicle or TGF-β1 (10 ng/mL) for 24 h. RNA-Seq analysis was performed to reveal differentially expressed genes (DEGs).

(A) Volcano plot.

(B) Heatmap of differentially expressed genes.

(C) Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway Analysis.

(D) CUT&Tag seq was performed as described in STAR Methods. Heatmap of peaks.

(E) Location of ZEB2-WT and ZEB2-4KR binding events respective to various genomic features.

(F) Integration of the RNA-seq and CUT&Tag-seq datasets identified five genes whose expression was upregulated by ZEB2-4KR and whose promoters were bound by ZEB2-4KR.

(G) CUT&Tag gene plot tracks for the read coverage of Nr4a1.

(H) The independent CH-IP assay was performed on Adv-GFP, Adv-ZEB2-4KR and Adv-ZEB2-WT infected CFs to verify ZEB2 binding to the promoter regions of Nr4a1. n = 3 independent biological cultures.

(I) Cell lysates were immunoprecipitated with Flag antibody and immunoblot assays were performed using CtBP1 and Smad2/3 antibodies.

(J and K) ZEB2^KO mice were injected with rAAV9-GFP, rAAV9-ZEB2-WT-3×Flag and rAAV9-ZEB2-4KR-3×Flag and then subjected to MI surgery for 14 days. Western blotting analysis for Nr4a1.

(L and M) ZEB2^KO CFs were infected with Adv-ZEB2-WT and Adv-ZEB2-4KR and then treated for TGF-β1 (10 ng/mL) for 24 h. Western blotting analysis for Nr4a1.

(N) Schematic diagram of the construction of wild type and mutant luciferase reporter plasmids of Nr4a1 promoter.

(O) Luciferase activation driven by the wild type or mutant promoter of Nr4a1 after normalization to Renilla luciferase in HEK293T cells. n = 3. Data are expressed as mean ± SD. One-way of two-way ANOVA followed by Tukey’s multiple comparisons test was used for analysis. ∗∗p < 0.01, ∗∗∗p < 0.001, ns means no significant statistical analysis.

Previous studies have established that ZEB2 binds E-box motifs (CACCTG) and represses target gene expression, at least in part by recruiting co-regulators such as Smad2/3 and the co-repressor CtBP1.³⁰ We hypothesized that SUMOylation modulates this repressor function by altering ZEB2’s protein-protein interactions. Co-IP assays in TGF-β1-treated CFs showed that ZEB2-WT and ZEB2-4KR associated with Smad2/3 to a similar extent, indicating that SUMOylation does not affect ZEB2-Smad2/3 complex formation. In contrast, the SUMOylation-deficient ZEB2-4KR mutant displayed a significantly reduced interaction with CtBP1 compared with ZEB2-WT (Figure 7I). These findings suggest that, despite its stronger binding to the Nr4a1 promoter, ZEB2-4KR is less able to recruit the CtBP1 co-repressor, thereby shifting the role of ZEB2 at this locus from a transcriptional repressor toward a state of relieved repression and enhanced Nr4a1 expression.

Furthermore, western blotting demonstrated that under MI or TGF-β1 stimulation, overexpression of ZEB2-WT downregulated Nr4a1 expression, whereas ZEB2-4KR robustly increased Nr4a1 protein levels (Figures 7J–7M). To functionally test whether SUMOylation-dependent differences in repressor activity are mediated through the canonical E-box motif, we performed dual-luciferase reporter assays in HEK293T cells transfected with either a WT Nr4a1 promoter-reporter construct or a construct containing a mutated E-box motif (CACCTG----AAAAAA). As expected, ZEB2-WT significantly repressed the activity of the WT-Nr4a1 promoter, whereas ZBE2-4KR failed to exert comparable repression, resulting in higher promoter activity than ZEBE-WT. Notably, mutation of the canonical E-box motif abolished ZEB2-WT-mediated-repression and eliminated the difference between ZEB2-WT and ZEB2-4KR, demonstrating that the SUMOylation-dependent repressive function of ZEB2 on the Nr4a1 promoter requires an intact E-box motif (Figures 7N–7O).

KEGG pathway analysis implicated ZEB2 SUMOylation in regulating both the PI3K/AKT and mTOR signaling pathways. To validate this in vivo, we intravenously injected ZEB2^KO mice with rAAV9-ZEB2-WT-3×Flag or rAAV9-ZEB2-4KR-3×Flag, followed by MI surgery for 2 weeks. As shown in Figures 8A and 8B, MI significantly increased phosphorylation of AKT (P-AKT, Ser473), phosphorylation of S6K (P-S6K, Thr389), and phosphorylation of 4E-BP1 (P-4E-BP1, Ser65). While ZEB2-WT overexpression further enhanced this activation, the SUMOylation-deficient mutant ZEB2-4KR markedly suppressed PI3K-AKT and mTORC1 complex activity. Consistently, in TGF-β1-stimulated ZEB2^KO CFs, ZEB2-WT activated PI3K-AKT and mTOR signaling, whereas the ZEB2-4KR mutant inhibited these pathways (Figures 8C and 8D). To investigate whether Nr4a1 mediates ZEB2 SUMOylation-dependent regulation of fibrotic signaling, we infected ZEB2^KO CFs with Adv-Flag-ZEB2-4KR in the presence or absence of Adv-sh-Nr4a1 prior to TGF-β1 stimulation. Knockdown of Nr4a1 by shRNA abolished the suppression effects of ZEB2-4KR on fibrotic activation and signaling, as evidenced by restored levels of P-AKT, P-S6K, and P-4E-BP1 (Figures 8E and 8F). Collectively, these findings demonstrate that ZEB2 SUMOylation critically regulates activation of the PI3K/AKT-mTORC1 signaling axis during cardiac remodeling, at least in part, by de-repression and upregulation of the antifibrotic transcription factor Nr4a1 consequent to impaired CtBP1 recruitment.

ZEB2 SUMOylation critically regulates activation of PI3K/AKT-mTORC1 signaling axis

ZEB2^KO mice were injected with rAAV9-EGFP, rAAV9-ZEB2-WT-3×Flag and rAAV9-ZEB2-4KR-3×Flag and then subjected to MI surgery for 14 days. Heart tissues were collected and lysed for western blotting analysis.

(A) Western blotting analysis for P-AKT (Ser473), P-S6K (Thr389), P-4EBP1(Ser65).

(B) Quantification of the protein. n = 6 mice per group.

(C and D) ZEB2^KO CFs were infected with Adv-ZEB2-WT and Adv-ZEB2-4KR and then treated for TGF-β1 (10 ng/mL) for 24 h. Western blotting analysis for P-AKT (Ser473), P-S6K (Thr389), P-4EBP1 (Ser65).

(E and F) ZEB2^KO CFs were infected with Adv-ZEB2-4KR-3×Flag in the presence or absence of Adv-Sh-Nr4a1 and then stimulated with TGF-β1 for 24 h. Western blot and analysis and quantification of P-AKT (Ser473), P-S6K (Thr389), P-4EBP1 (Ser65) expression. n = 3 individual experiments. Data are expressed as mean ± SD. two-way ANOVA followed by Tukey’s multiple comparisons test was used for analysis. ∗p < 0.05, ∗∗p < 0.01, ∗∗∗p < 0.001.

Nr4a1 knockdown abolishes the cardioprotective effects of SUMOylation-deficient ZEB2-4KR in post-MI hearts

Finally, to determine whether SUMOylation-deficient ZEB2 regulates MI-induced cardiac dysfunction via Nr4a1, we intravenously injected ZEB2^KO mice rAAV9-ZEB2-4KR-3×Flag in the presence or absence of rAAV9-Col1a2-Sh-Nr4a1 for 2 weeks. Robust and comparable flag overexpression and Nr4a1 knockdown were confirmed (Figures S9A and S9B). All mice then underwent MI surgery followed by echocardiographic assessment of cardiac function (Figure S9C). Reconstitution of ZEB2^KO mice with ZEB2-4KR significantly attenuated MI-induced cardiac dysfunction. Specifically, ZEB2-4KR-reconstituted mice exhibited a significant increase in ejection fraction (EF) and fractional shortening (FS) compared to GFP-reconstituted control mice subjected to MI. Conversely, Nr4a1 knockdown further aggravated cardiac dysfunction post-MI. Notably, Nr4a1 knockdown completely abolished the cardioprotective effect conferred by the SUMOylation-deficient ZEB2-4KR mutant, as evidenced by the loss of improvement in EF and FS (Figures 9A–9C, Table S5). HW/BW ratio showed that the protective effects of ZEB2 SUMOylation deficiency were abrogated by Nr4a1 knockdown (Figure 9D). Consistent with the functional data, histological analysis using Masson and HE staining revealed that ZEB2-4KR reconstitution reduced MI-induced adverse remodeling and fibrosis, whereas Nr4a1 knockdown exacerbated these pathological changes and negated the protective effects of ZEB2-4KR (Figures 9E–9J). At the molecular level, western blotting and qPCR analysis demonstrated that ZEB2-4KR reconstitution significantly decreased the protein expression of α-SMA and collagen I in the remodeling myocardium. Nr4a1 knockdown abolished the protective effects of ZEB2-4KR on fibrotic protein expression (Figures 9K–9L). Furthermore, qPCR analysis confirmed that mRNA levels of fibrosis genes, including Acta2, Col1α1, Col3α1 and Ctgf, were markedly reduced in the myocardium of ZEB2-4KR-reconstituted mice compared to GFP controls post-MI. In contrast, Nr4a1 knockdown significantly increased the expression of these fibrotic genes and counteracted the inhibitory effect of ZEB2-4KR (Figure 9M). Collectively, these findings demonstrate that the cardioprotective effects of SUMOylation-deficient ZEB2-4KR against MI are mediated via Nr4a1 and are completely abolished by Nr4a1 knockdown.

Discussion

The present study established ZEB2 as a critical driver of pathological cardiac remodeling post-MI. We found that ZEB2 protein, but not mRNA, was upregulated in infarcted myocardium and activated CFs, implicating post-translational regulation. Critically, myofibroblast-specific knockout of ZEB2 markedly attenuated post-MI cardiac dysfunction and adverse remodeling, demonstrating an indispensable, cell-autonomous role in fibrosis. Mechanistically, we identified SENP1-mediated deSUMOylation as a key regulator of ZEB2 stability, targeting specific lysine residues (K462, K479, K611, K774). Reconstitution of ZEB2-deficient systems with a SUMOylation-deficient mutant (ZEB2-4KR) more effectively attenuated cardiac fibrosis and FMT than ZEB2-WT both in vivo and in vitro. Crucially, ZEB2 SUMOylation was essential for activating the pro-fibrotic PI3K/AKT-mTORC1 signaling axis. Notably, the SUMOylation-deficient ZEB2-4KR mutant displayed markedly reduced interaction with the co-repressor CtBP1 compared with ZEB2-WT. This diminished CtBP1 binding enhanced the transcriptional activation of the antifibrotic gene Nr4a1 by ZEB2-4KR. Importantly, Nr4a1 knockdown abolished the protective effects of ZEB2-4KR, exacerbating fibrosis and cardiac dysfunction. These findings establish ZEB2 SUMOylation as a pivotal mechanism promoting pathological remodeling through repression of Nr4a1.

ZEB2 has been implicated in the CM responses to myocardial infarction (MI). Single-cell analyses have shown that ZEB2 is upregulated in stressed CMs shortly after MI, where it plays a critical protective role by promoting paracrine signaling that enhances endothelial cell migration, angiogenesis, and infarct healing.¹⁸ Consistently, Gladka et al. reported that ZEB2 in CMs regulates Ca²⁺ handling and hypertrophic signaling pathways.¹⁶ In contrast, our data reveal a potent profibrotic role for ZEB2 in CFs during the remodeling phase. Immunoblotting of whole-heart lysates demonstrated increased ZEB2 protein levels at days 7 and 14 post-MI, while cell type-specific analysis at day 14 revealed that this sustained upregulation originates predominantly from CFs, with no detectable increase in CMs. Given that fibroblast activation peaks around day 7, it is likely that ZEB2 elevation in whole-heart lysates at this time mainly reflects CF activation, although a transient early induction in CMs cannot be completely excluded. These results support a temporal shift in ZEB2 function from an early CM-centered stress response to a later fibroblast-driven profibrotic program.

miR-208a provides an additional layer of cell-type specificity to this model. As an intronic miRNA encoded with Myh6, miR-208a is primarily transcribed in CMs under physiological and pathological conditions and has been consistently validated as CM-enriched across multiple models of cardiac remodeling.³¹^,³² Detection of miR-208a in fibroblasts is therefore best explained by exosome-mediated transfer from CMs in the fibrotic microenvironment, rather than by endogenous transcription in CFs.³³^,³⁴^,³⁵ Our data revealed a miR-208a-independent mechanism regulating ZEB2 in CFs. SUMO1-mediated SUMOylation of ZEB2 enhances its interaction with the co-repressor CtBP1 and represses Nr4a1, thereby driving profibrotic gene expression and fibroblast activation. Thus, whereas ZEB2 in CMs contributes to cardioprotection early after MI, ZEB2 in CFs sustains maladaptive remodeling through a distinct, post-translationally controlled pathway. This cell-type and time-dependent duality underscores the need to consider both the cellular origin and the temporal context when targeting the ZEB2 axis in cardiac disease.

In cardiac fibroblasts, our data indicate that the profibrotic actions of ZEB2 are governed predominantly by PTMs, with SUMOylation emerging as a central regulatory mechanism. Although SUMO-1 and SUMO-2/3 are known to have both overlapping and distinct targets and subcellular distributions, our affinity-MS and follow-up co-IP analyses in MI hearts and TGF-β1-stimulated CFs detected robust SUMO-1, but not SUMO-2/3, conjugation of ZEB2 (Figures S3B and S3C). Within this SUMO-1-restricted context, IP-MS combined with GPS-SUMO and JASSA predictions identified four confidence SUMO-1 sites (K462, K479, K611, K774), two of which (K479, K611) overlap with sites annotated for ZEB2 in public databases. These residues differ from K391 and K866, previously reported in tumor cells as SUMO sites.³⁶ Supporting this tissue specificity, neither K391R nor K866R mutations affected α-SMA expression or ZEB2 SUMOylation in TGF-β1-stimulated CFs (Data not shown), underscoring the context-dependence of SUMO-site usage. Beyond SUMOylation, other ZEB2 PTMs likely converge on its function in CFs. Acetylation at K377 and ubiquitylation at K362, previously described in other systems, probably cooperate with SUMO-1 to fine-tune ZEB2 stability, nuclear retention, and co-factor engagement. Acetylation at K377 may modulate chromatin binding and co-regulator recruitment, whereas ubiquitylation at K632 is linked to protein turnover and may counteract stabilizing PTMs. Notably, one of our SUMO-1 sites, K479, lies within the originally defined Smad-binding domain (aa437-487). The functional impact of K479 SUMOylation in CF activation therefore is likely mediated by changes in the surrounding structural context that influence Smad complex assembly and/or chromatin association, rather than by direct disruption of the minimal Smad binding motif.

Functionally, SUMO-1 modification of ZEB2 at K462, K479, K611 and K774 is required for its full profibrotic activity in CFs. In ZEB2-deficient CFs, re-expression of ZEB2-WT restores and enhances TGF-β1-induced α-SMA expression, whereas the SUMO-deficient ZEB2-4KR mutant markedly attenuates myofibroblast activation. Mechanistically, SUMO-1 conjugation at these sites facilitates efficient formation of ZEB2-CtBP1 complexes; TGF-β1 stimulation enhances ZEB2-WT binding to CtBP1, while this interaction is substantially reduced in ZEB2-4KR (Figure 7I). By contrast, ZEB2-4KR retains normal association with Smad2/3, consistent with K479 lying outside the minimal Smad-binding core. These findings indicate that SUMOylation primarily modulates the composition and repressive potency of ZEB2 co-repressor complexes Nr4a1, rather than Smad recruitment.

This SUMO-dependent profibrotic program is counterbalanced by SENP1. SENP1 expression was decreased post-MI, and its deSUMOylation of targets like HSP90ab1 mitigates fibrosis.³⁷ Our data demonstrate that SENP1-mediated deSUMOylation of ZEB2 exerts protective effects (Figure 4). Notably, while ZEB2 knockout attenuates fibrosis, reconstitution of ZEB2-KO mice with ZEB2-4KR confers even greater functional benefit, suggesting that loss of SUMO-1modifciation not only abolishes the profibrotic function of ZEB2 but may also partially convert it into a dominant-negative or otherwise protective form. Together, these observations define a fibroblast-specific, SUMO-1-dependent ZEB2 axis in pathological remodeling and, within the limits of our current model, support the idea that selectively targeting SUMO-1-mediated regulation of ZEB2 could form the basis for novel antifibrotic therapies. These disparities likely reflect cell-context-dependent engagement of E3 ligases and methodological variations. A key paradox emerges: while ZEB2 knockout attenuates fibrosis, reconstituting ZEB2^KO mice with the SUMOylation-deficient ZEB2-4KR mutant confers superior cardioprotection. This suggests that abolishing SUMOylation does more than nullify ZEB2’s profibrotic function; it likely induces a functional conversion. The 4KR mutant loses repressive activity against antifibrotic genes while potentially gaining dominant-negative properties, thereby actively engaging protective mechanisms.

Structurally, SUMOylation dynamically modulates the interactome of ZEB2, reducing its affinity for Nr4a1. The proximity of the identified SUMO sites to the Nr4a1 interaction domain suggests a plausible structural mechanism for this regulation. Collectively, our work establishes ZEB2 SUMOylation as a pivotal, therapeutically targetable mechanism driving pathological remodeling through PTM-dependent repression of Nr4a1. Our findings establish transcriptional repression of Nr4a1 as the central mechanism by which ZEB2 SUMOylation promotes cardiac fibrosis. This aligns with the emerging role of Nr4a1 as an anti-fibrotic effector in the heart. While Goran et al. demonstrated that macrophage Nr4a1 attenuates post-MI fibrosis, Zhang et al. reported that fibroblast-specific Nr4a1 deletion reduced fibrosis in pressure overload, suggesting context-dependent outcomes.³⁸^,³⁹ Critically, both findings, alongside other evidence, support the fundamental role of Nr4a1 as an anti-fibrotic factor.⁴⁰^,⁴¹^,⁴² Our data further solidify this concept by demonstrating that ZEB2 SUMOylation promotes fibrosis primarily through transcriptional repression of this protective Nr4a1 gene (Figure 7). In fibroblasts, ZEB2 SUMOylation dominantly represses Nr4a1 (Figure 7), thereby enabling unfettered PI3K/AKT-mTORC1 activation. Critically, Nr4a1 knockdown abolished the cardioprotection conferred by SUMOylation-deficient ZEB2-4KR (Figures 8A–8I), confirming its non-redundant role downstream of ZEB2 SUMOylation. The PI3K/AKT-mTORC1 axis is a well-established driver of pathological remodeling, with mTORC1 hyperactivation promoting fibrosis and dysfunction post-MI.⁴³ We unveil a novel regulatory paradigm in which ZEB2 SUMOylation, but not its basal expression, is essential for pathway activation specifically in fibroblasts (Figures 8A–8F). This contrasts starkly with canonical growth factor-driven PI3K activation, highlighting cell- and stimulus-specific control. Crucially, the ZEB2-4KR mutant suppressed PI3K/AKT-mTORC1 signaling even under profibrotic stimulation (Figures 8B–8D), demonstrating that SUMOylation licenses ZEB2 to amplify fibrotic signaling independent of upstream triggers.

Our combined approach, which uses ZEB2^KO mice with AAV9-mediated reconstitution of ZEB2-WT or ZEB2-4KR, provides rigorous in vivo evidence that ZEB2 SUMOylation drives pathological remodeling. Although AAV9-CMV efficiently targets cardiac, including CMs, our data show that it also transduces CFs, restoring ZEB2 or flag-ZEB2 expression in vimentin-positive CFs. MI-induced ZEB2-SUMOylation is restricted to CFs and not detectable in CMs or CMECs under our experimental conditions. We did not assess extracellular vesicle (EV) V release or EV-mediated ZEB2 transfer in this study; therefore, our conclusions do not involve an EV-dependent mechanism. Thus, the divergent outcomes of ZEB2-4KR not ZEB2-WT protected against adverse remodeling by CF-autonomous, SUMO-dependent effects rather than by differences in CMs. Importantly, AAV9-ZEB2 was used solely in a ZEB2-deficient background, and the restored ZEB2 levels were comparable to, or only slightly above, the 2- to 2.3-fold endogenous increase seen in WT hearts post-MI.

The identification of ZEB2 SUMOylation as a pivotal, therapeutically targetable driver of pathological cardiac remodeling, achieved by repressing the cardioprotective nuclear receptor Nr4a1, establishes its inhibition as a highly promising strategy. This approach offers specificity compared to broadly targeting downstream effectors such as AKT, which has context-dependent protective roles, or mTORC1 with rapamycin; by acting upstream, targeting ZEB2-SUMO simultaneously restores Nr4a1-mediated protection and normalizes hyperactivated PI3K/AKT/mTORC1 signaling.⁴⁴^,⁴⁵^,⁴⁶ In contrast to contexts where Nr4a1 inhibition may be beneficial, such as atherosclerosis, targeting Nr4a1 in cardiac fibrosis appears detrimental. Studies demonstrate that Nr4a1 activation exerts anti-fibrotic effects. Genetic deletion of Nr4a1 exacerbated cardiac fibrosis and dysfunction in a model of diabetic cardiomyopathy, while pharmacological activation of Nr4a1 attenuated cardiac fibrosis and improved function in pressure-overload induced heart failure.⁴⁷^,⁴⁸^,⁴⁹ Consequently, a gene therapy approach utilizing AAV9 to deliver a SUMOylation-resistant ZEB2 mutant (e.g., ZEB2-4KR) presents a potentially durable and cardiac-specific therapeutic avenue, leveraging the protective knockout phenotype by derepressing Nr4a1 and inhibiting pathological signaling. However, this strategy faces significant challenges inherent to gene therapy, including potential long-term safety concerns, immunogenicity, complex manufacturing, and high costs, necessitating further preclinical validation.

In summary, our findings establish reversible ZEB2 SUMOylation as a critical molecular switch coordinating FMT and PI3K/AKT/mTORC1 signaling in CFs. SUMOylation and deSUMOylation dynamically regulate the assembly of the ZEB2 transcriptional complex and its DNA-binding affinity in response to profibrotic stimuli, thereby fine-tuning its transcriptional output. These results provide fundamental mechanistic insights into how context-specific PTMs orchestrate transcriptional programs governing cardiac fibrosis, revealing ZEB2 SUMOylation as a promising therapeutic target for modulating pathological remodeling.

Limitations of the study

Although we have established the functional significance of ZEB2 SUMOylation in CFs and our IP-MS results suggest RNF216 as a potential E3 ligase, the specific E3 ligase responsible for mediating this modification in CFs remains unidentified. This gap limits our understanding of the upstream regulatory mechanisms controlling ZEB2 SUMOylation. Furthermore, using AAV9 with a CMV promoter in the manuscript lacks cell-type and temporal specificity, which drives ZEB2 expression beyond CFs and complicates attribution of the observed phenotypes to CF-autonomous SUMOylation. Moreover, the potential impact of ZEB2 SUMOylation within CFs and the neighboring cells and the reciprocal signaling pathways mediating pathological intercellular crosstalk remains unexplored. Finally, validating the levels of ZEB2 SUMOylation and their correlation with Nr4a1 expression and disease severity in human failing heart specimens is essential to confirm the translational relevance of these findings.

Resource availability

Lead contact

Further information and requests for resources and reagents should be directed to and will be fulfilled by the lead contact, Dongjin Wang (wangdongjin@njglyy.com).

Materials availability

This study did not generate new unique reagents.

Data and code availability

•
Data: Transcriptomic: Raw RNA-sequencing are deposited in the genome sequencing archive (GSA: CRA046490) at https://ngdc.cncb.ac.cn/gsa/browse/CRA046490 and CUT&Tag-sequencing data are deposited in the genome sequencing archive (GSA:CRA046488) at https://ngdc.cncb.ac.cn/gsa/browse/CRA046488.
•
Code: This paper does not report original code.
•
Additional information: Any additional information required to reanalyze the data reported in this paper is available from the lead contact upon request.

Acknowledgments

Thanks to all the authors and funders who contributed to the article. This study was supported by grants from the National Natural Science Foundation of China (grant nos. 82300312, 82241212, 82202377, 82500322) and China Postdoctoral Science Foundation (grant nos. 2023M741651 and 2023M731626) and the Natural Science Foundation of Jiangsu Province (grant no. BK20220176).

Author contributions

D.J.W., J.Y., and Q.W.Z were involved in the conceptualization of the study. J.Y. and Y.X.X designed and implemented the methodology. Y.L.W., S.P.C., and H.Y.C. contributed to the investigations and performed the visualization of the data. Z.H.Y, X.T.W., Q.Y.Z., J.N. W., Y.J., and H. C. performed the visualization of the data. D.J.W., J.Y., and X.T.W. acquired funding for the studies. D.J.W., J.Y., and Q.W.Z. administered the project. Y.R.Y. and Y.X.X. provided supervision for the study. All authors were involved in the review and editing of the manuscript.

Declaration of interests

The authors have declared that no competing interest exists.

STAR★Methods

Key resources table

REAGENT or RESOURCE	SOURCE	IDENTIFIER
Antibodies

Rabbit monoclonal anti-ZEB2	Protein tech	Cat NO.67514-1, RRID: AB_2882735
Rabbit polyclonal anti-CtBP1	Protein tech	Cat NO.10972-1, RRID: AB_2086628
Rabbit polyclonal anti-Nr4a1	Protein tech	Cat NO.12235-1, RRID: AB_10644125
Rabbit polyclonal anti-SENP2	Protein tech	Cat NO.29772-1, RRID: AB_3086158
Rabbit polyclonal anti-SUMO1	Protein tech	Cat NO.10329-1, RRID: AB_2286872
Rabbit monoclonal anti-SUMO2/3	Protein tech	Cat NO.67154-1, RRID: AB_2882451
Mouse monoclonal anti-α-smooth muscle actin	Sigma-Aldrich	Cat #: A2574
Mouse monoclonal anti-Flag	Sigma-Aldrich	Cat #: F9291
Rabbit polyclonal anti-beta Tubulin	Abcam	Cat #: ab6046
Mouse monoclonal anti-bata Actin	Abcam	Cat #: ab6276
Rabbit polyclonal anti-Myc tag	Abcam	Cat #: ab9106
Rabbit polyclonal anti-SENP1	Abcam	Cat #: ab236094
Rabbit multiclonal anti-Collagen I	Abcam	Cat #: ab316222
Mouse monoclonal anti-Vimentin	Abcam	Cat #: ab8978
Rabbit polyclonal anti-Vimentin	Abcam	Cat #: ab137321
Rabbit multiclonal anti-periostin	Abcam	Cat #: ab315604
Rabbit monoclonal anti-HA-tag	Cell Signaling Technology	Cat: #3724
Rabbit monoclonal anti-His-tag	Cell Signaling Technology	Cat: #2365
Rabbit monoclonal anti-P-AKT(Ser473)	Cell Signaling Technology	Cat: #4060
Mouse monoclonal anti-AKT (pan)	Cell Signaling Technology	Cat: #2920
Rabbit polyclonal anti-Phospho-P70 S6K (Thr389)	Cell Signaling Technology	Cat: #9205
Rabbit polyclonal anti-P70 S6K	Cell Signaling Technology	Cat: #2708
Rabbit polyclonal anti-Phospho-4E-BP1(Ser65)	Cell Signaling Technology	Cat: #9451
Rabbit polyclonal anti-4E-BP1	Cell Signaling Technology	Cat: #9452
Rabbit polyclonal anti-SMAD2/3	Cell Signaling Technology	Cat: #3102
Goat Anti-Rabbit IgG H&L (Alexa Flour ®555)	Abcam	Cat #: ab150078
Goat Anti-mouse IgG H&L (Alexa Flour ®555)	Abcam	Cat #: ab150118
Goat Anti-Rabbit IgG H&L (Alexa Flour ®647)	Abcam	Cat #: ab150079
Goat Anti-mouse IgG H&L (Alexa Flour ®647)	Abcam	Cat #: ab150115
DAPI	Abcam	Cat #: ab104139
Cycloheximide	Selleck	Cat #: S7418
MG132	Selleck	Cat #: S2619
Isoproterenol	Selleck	Cat #: S2566

Bacterial and virus strains

rAAV9-ZEB2-WT-3×Flag	This paper	N/A
rAAV9-ZEB2-4 KR-3×Flag	This paper	N/A
rAAV9-Col1a2-Sh-Nr4a1	This paper	N/A
rAAV9-Col1a2-GFP	This paper	N/A
Adv-GFP	This paper	N/A
Adv-ZEB2-WT	This paper	N/A
Adv-ZEB2-4 KR	This paper	N/A
Adv-Sh-Nr4a1	This paper	N/A
Lentivirus-Flag-ZEB2-K774R	This paper	N/A
Lentivirus-Flag-ZEB2-K611R	This paper	N/A
Lentivirus-Flag-ZEB2-K479R	This paper	N/A
Lentivirus-Flag-ZEB2-K462R	This paper	N/A
Lentivirus-Flag-ZEB2-4 KR	This paper	N/A
Lentivirus-Flag-ZEB2-WT	This paper	N/A
Lentivirus-His-SUMO1	This paper	N/A

Biological samples

Human chronic post-MI left ventricular aneurysm myocardium	This paper	N/A
Healthy adult myocardium from organ donors	This paper	N/A

Chemicals, peptides, and recombinant proteins

Recombinant Human TGF-β1	PEPROTECH	Cat #: 100-21

Critical commercial assays

RNA sequencing	Illumina	N/A
CUT&Tag sequencing	Illumina	N/A

Deposited data

Raw RNA-sequencing data	GSA: CRA046490	https://ngdc.cncb.ac.cn/gsa/browse/CRA046490
Raw CUT&Tag sequencing	GSA: CRA046488	https://ngdc.cncb.ac.cn/gsa/browse/CRA046488

Experimental models: cell lines

HEK293T cells	ATCC	CRL-3216

Experimental models: organisms/strains

Mouse: ZEB2^f/f	Cyagen Biosciences Inc	N/A
Mouse ZEB2^KO	Cyagen Biosciences Inc	N/A
Mouse: Postn-MerCreMer	Cyagen Biosciences Inc	N/A
Mouse C57BL/6J	GemPharmatech	N/A

Oligonucleotides

siRNA targeting sequence: SENP1: 5′-TACTGAACTAAGACATCG-3′	This paper	N/A
siRNA targeting sequence: SENP2: 5′-GGAGCCTGACCTATCAGAA-3′	This paper	N/A
Primers for mZEB2 Flox-F2(5′- CTCACCTTTTGTTCTCTGACACCT-3′)	This paper	N/A
Primers for mZEB2 Flox-R2 (5′- TACTTGAGCTTATTTGTGGGGAGG-3′)	This paper	N/A
Primers for mZEB2 AIB4828-E6-F: (5′- ACATCAAGTACCGCCACGAG -3′)	This paper	N/A
Primers for mZEB2 AIB4828-E6-R (5′- TGAACTTGCGGTTACCTGCT -3′)	This paper	N/A
Primers for Postn-Cre: Forward (5′- GGTGGGACATTTGAGTTGCT -3′)	This paper	N/A
Primers for Postn-Cre: Reverse (5′- CCTTGCAATAAGTAAAACAGCTC -3′)	This paper	N/A
Primers for mZEB2^KO Forward (5′- CATATTGGCAGAACGAAAACG -3′)	This paper	N/A
Primers for mZEB2^KO Reverse (5′- CATTCAACTCACAAGTCCAATCAG -3′)	This paper	N/A
Primers for mNr4a1 in ChIP-PCR: forward: 5′- TTGGGGCTAGTGTAATGGGC-3′	This paper	N/A
Primers for mNr4a1 in ChIP-PCR: forward: 5′- GTGATACCCCAGTGCAGGTT -3′	This paper	N/A

Recombinant DNA

Flag-ZEB2-WT plasmid	This paper	N/A
Flag-ZEB2-K462R plasmid	This paper	N/A
Flag-ZEB2-K4479R plasmid	This paper	N/A
Flag-ZEB2-K611R plasmid	This paper	N/A
Flag-ZEB2-K774R plasmid	This paper	N/A
Flag-ZEB2-4 KR plasmid	This paper	N/A
HA-SUMO1 plasmid	This paper	N/A
His-SUMO1 plasmid	This paper	N/A
Myc-SENP1 plasmid	This paper	N/A
Myc-SENP2 plasmid	This paper	N/A
Vector plasmid	This paper	N/A
WT-Nr4a1-promoter pGL6 luciferase reporter plasmid	This paper	N/A
Nr4a1-mutant-promoter pGL6 luciferase reporter plasmid	This paper	N/A

Software and algorithms

GraphPad Prism 10 software	GraphPad software	https://www.graphpad.com/
ImageJ	ImageJ software	https://imagej.net/ij/
ChemiDoc™ Imaging System	Bio-Rad	https://www.bio-rad.com/
Biorender	Biorender	https://Biorender.com
R	R core team	https://www.r-project.org/
Adobe Photoshop	Adobe	https://www.adobe.com/products/ilustrator/free-trial-download.html
Dual-Luciferase® Reporter Assay System	Promega	#E1910
VisualSonics VeVo 2100 imaging system	Toronto	N/A

Other

Pierce BCA Protein Assay Kit	Thermo Fisher Scientific	Cat: #23227
protein A/G Agarose beads	Thermo Fisher Scientific	Cat: #80106G
Lipofectamine® RNAiMAX Reagent	Invitrogen	Cat #: 13778075
Cell contraction Assay Kit	Corning	Cat #: 354236
TRIzol Regent	Takara Bio	Cat #: TKR-9109
HiFScript	Cowin Biotech	Cat #: CW3371
SYBR Green-based master mix	Cowin Biotech	Cat #: CW3371

Open in a new tab

Experimental model and study participant details

Human samples

All procedures involving human myocardium tissues were conducted in accordance with institutional guidelines and approved by the Medical Ethics Committee of Affiliated Nanjing Drum Tower Hospital, Nanjing University Medical College (approval number: 2023–197-02). Written informed consent was obtained from all participants. Clinical specimens used in this study were collected from the Department of Cardiac Surgery of Nanjing Drum Tower Hospital. The human ventricular samples used were obtained from three patients with chronic post-MI left ventricular aneurysm undergoing surgery. In these patients, the index MI has occurred at least six months before surgery. The control myocardium tissues were harvested from 3 organ donors. The tissues were divided into several segments, which were used for immunofluorescence or protein extraction.

Animals

All animal experiments were reviewed and approved by the guidelines for the Care and Use of Laboratory Animals published by the US National Institutes of Health (NIH Publication, 8th Edition, 2011) and approved by the Nanjing University Drum Tower Hospital Committee on Animal Care (Permit Number: IACUC: 20230315-12). Mice were housed in a specific pathogen-free (SPF) environment with 12-h light/dark cycles and provided ad libitum access to food and water. This study utilized male C57BL/6JGpt mice (Strain NO. N000013) which were purchased from GemPharmatech (Nanjing, China) for next animal experiments. Male C57BL/6JGpt mice, aged 6–8 weeks, were used for the myocardial infarction (MI) model. Male mice aged 5–6 weeks were used for cardiac fibroblasts isolation and culture. Conditional transgenic mice targeting ZEB2 gene (ZEB2^f/f) were generated by Cyagen Biosciences Inc (Suzhou, Jiangsu, China) in the C57BL/6J background. Systemic knockout of ZEB2 mice were generated by Cyagen Biosciences Inc (Suzhou, Jiangsu, China). ZEB2^f/f mice were bred with Postn-MerCreMer (purchased from Cyagen Biosciences Inc) mice to generate myofibroblast-specific knockout of ZEB2 in the adult mice. Mice were intraperitoneally treated with 30 mg/kg of tamoxifen dissolved in corn oil for 5 consecutive days. Detailed information was described in Supplementary materials.

Primary cells

Murine cardiac fibroblasts were isolated from healthy mice hearts and digested in the Type II collagenase solution. To minimize dedifferentiation, only CFs between passages 2 were utilized in subsequent experiments.

Cell lines

Method details

Construction of recombinant adeno-associated virus and injection

Recombinant adeno-associated virus serotype 9 (rAAV9) purchased from Zebrafish Biotech Co., Ltd. (Nanjing, China), was used to over express ZEB2-WT or ZEB2-4 KR in cardiac tissues of ZEB2^KO mice. A cDNA encoding the ZEB2-WT sequence and ZEB2-4 KR mutant (K462, K479, K611 and K774) were inserted into the rAAV9 packaging vector to generate rAAV9-ZEB2-WT-3×Flag and rAAV9-ZEB2-4 KR-3×Flag. Male ZEB2^KO mice, six weeks old, were intravenously injected with rAAV9-ZEB2-WT-3×Flag or rAAV9-ZEB2-4KR-3×Flag or rAAV9-vector (3×10¹¹ vg/mouse) through the tail vein for 2 weeks prior to MI surgery or ISO injection. The efficiency of viral transfection was assessed by measuring ZEB2 expression through western blotting and immunofluorescence staining. Another group is ZEB2^KO mice with injection of rAAV9-Sh-Nr4a1 in the presence or absence of rAAV9-ZEB2-4 KR-3×Flag for two weeks prior to MI surgery. The sequence of Sh-Nr4a1 is CATGTGCCTTTAAGCCTATAG. The sequence of Scramble as follows: TTCTCCGAACCGTGTCACGT.

Generation of knockout of myofibroblast mice

Conditional transgenic mice targeting ZEB2 gene (ZEB2^f/f) were generated by Cyagen Biosciences Inc (Suzhou, China) in the C57BL/6J background. The conditional ZEB2-flox (Exon 6) allele was generated by CRISPR/Cas-9-mediated homologous recombination. Targeted allele design was based on Ensemble-202, transcript ENSMUSP00000069685.5. Briefly, ten exons were identified with the ATG start codon in exon 2 and the TAA stop codon in exon 10. The schematic diagram illustrated the targeted strategy, including the homologous arms and loxP sites flanking Exon6 (215bp coding sequence). One loxP site was placed in the intron upstream of Exon6 and second loxP site was placed in the intron downstream of Exon6. No other coding sequence was altered apart from the insertion of the loxP sites. CRISPR/Cas-9 and a ZEB2-specific sgRNA were used to create a double-strand break near Exon6. The Postn-MerCreMer mice were purchased from Cyagen Biosciences Inc. Postn-MerCreMer mice express a tamoxifen-activated recombinase under the control of Postn promoter, thereby enabling myofibroblasts-specific gene knockout.⁵⁰^,⁵¹^,⁵² ZEB2^f/f mice were bred with the Postn-MerCreMer mice to generate myofibroblast-specific ZEB2 knockout in the adult mice. Mice were intraperitoneally treated with 30 mg/kg of tamoxifen dissolved in corn oil for 5 consecutive days at 8 days post-MI. ZEB2^f/f mice tamoxifen littermates were represented as controls. Mice of different strains were genotyped with PCR. Primers were as follows: 5′- CTCACCTTTTGTTCTCTGACACCT-3′(sense), 5′- TACTTGAGCTTATTTGTGGGGAGG-3′(antisense), which yielded 233bp products for ZEB2^f/f. Postn-Cre was genotyped by the primers as follows: 5′- GGTGGGACATTTGAGTTGCT-3′(sense) and 5′- CCTTGCAATAAGTAAAACAGCTC -3′(antisense), which yielded 268 bp products.

Generation of ZEB2 knockout mice

ZEB2 systemic knockout mice (ZEB2^KO) mice were constructed by crossbreeding ZEB2^f/f mice and Ddx-4-Cre mice. Ddx4-Cre is a germline Cre driver in which Cre recombinase is expressed in male and female embryonic germ cells, leading to germline deletion of Loxp-flanked alleles. ZEB2^f/f mice, in which Exon6 is flanked by loxP sites were crossed with Ddx4-Cre mice to obtain ZEB2^+/− mutants carrying a germline-deleted ZEB2 null allele, and ZEB2^+/− mice were then intercrossed to generated ZEB2^−/− and ZEB2^+/+ wild type littermates. Mice of different strains were genotyped with PCR. Primers were as follows: 5′-CATATTGGCAGAACGAAAACG-3′(sense), 5′- CATTCAACTCACAAGTCCAATCAG -3′(antisense), which yielded 556 bp products for ZEB2^KO.

Myocardial infarction model

Myocardial infarction (MI) was induced in 8-10-week-old male mice by permanent ligation of the left anterior descending (LAD) coronary artery. Briefly, mice were anesthetized with 3% isoflurane and subsequently ventilated with 1.5–2% isoflurane using a rodent ventilator. Following anesthesia, a left thoracotomy was performed between the fourth and fifth intercostal spaces, and the underlying muscles were dissected to expose the heart. The LAD was then permanently ligated with a 6-0 prolene suture. Sham-operated mice underwent an identical surgical procedure, excluding the LAD ligation. After ligation (or sham procedure), any accumulated fluid within the chest cavity was evacuated, and the incision was meticulously closed in layers. Post-operatively, mice recovered on a heating pad. All mice were sacrificed, and heart samples were harvested for analysis 14 days post-surgery.

ISO injection models

To establish heart failure models, male ZEB2^KO mice aged 6 weeks were injected with rAAV9-ZEB2-WT-3×Flag and rAAV9-ZEB2-4 KR-3×Flag virus through tail vein for 2 weeks. Then the mice were subjected to subcutaneous administration of isoproterenol (ISO: 10 mg/kg/day; S2566, Selleck Chemicals) for a period of 2 weeks. At the end of the treatment period, the animals were euthanized, and heart tissues were harvested for subsequent analyses.

Echocardiography analysis

Cardiac function was assessed by echocardiography at baseline (pre-surgery) and day 14. Imaging was performed using a high-frequency small-animal ultrasound system (VisualSonics VeVo 2100 Imaging System, Toronto, Canada) equipped with a 30-MHz linear transducer. Following anesthesia with isoflurane (1.5% in O₂ at 1 L/min), mice were positioned on a heated pad to maintain body temperature between 36.9°C and 37.3°C and heart rate at approximately 500 bpm. From the left ventricular short-axis view, M-mode tracings were acquired to measure functional parameters, including left ventricular end-diastolic diameter (LVEDd), end-systolic diameter (LVESd), ejection fraction (LVEF), and fractional shortening (LVFS). All analyses were performed by an investigator blinded to treatment groups and genotypes.

Masson trichrome staining

Two weeks after myocardial infarction (MI), mice heart was harvested and fixed in 4% paraformaldehyde for 24 h. Subsequently, samples were dehydrated through a graded ethanol series and embedded in paraffin. Tissue sections of 5 μm thickness were prepared and subjected to Masson’s trichrome staining using a commercially available kit (Sigma-Aldrich, Shanghai, China) in accordance with the manufacturer’s protocol. The extent of myocardial fibrosis was quantified using NIH ImageJ software. For each mouse, fibrosis quantification was based on the average of three randomly selected fields per section, across three non-consecutive heart sections. Each experimental group consisted of six mice.

Hematoxylin and Eosin (HE) staining

The hearts of the mice were perfused with cold PBS and excised. Then the hearts were fixed in 4% paraformaldehyde for 48h and embedded in paraffin. HE staining was performed according to the manufacturer’s protocols of the H&E buffer (Service bio, G1005).

Isolation and culture of adult mouse cardiac fibroblasts

Cardiac fibroblast isolation was performed as follows. Mice were sterilized with 75% ethanol prior to dissection. Under sterile conditions within a laminar flow hood, the heart was rapidly excised using aseptic instruments, rinsed extensively in PBS to remove residual blood, and transferred to a sterile surface. After switching instruments, the atria and great vessels were removed, and the ventricles were minced into about 1 mm³ fragments. These fragments were digested in 8 mL of Type II collagenase solution (100 U/ml, Worthington, LS004176) with constant shaking at 37°C for 20 min. The resulting supernatant was harvested and combined with 10 mL of DMEM containing 50% FBS to neutralize collagenase activity. This digestion cycle (digestion/harvest/neutralization) was repeated 5–7 times until complete tissue dissociation. The pooled digests were filtered through a 100 μm mesh, centrifuged (5 min, 1500 rpm), and the supernatant discarded. The cell pellet was resuspended in 3 mL red blood cell lysis buffer for 3 min. Lysis was terminated by adding an equal volume of DMEM, followed by centrifugation (5 min, 1500 rpm) and supernatant removal. Cells were finally resuspended in DMEM supplemented with 10% FBS. The suspension was plated onto a 10 cm culture dish. After 2 h of incubation, the non-adherent fraction was removed and replaced with fresh medium. The adherent cells were identified as cardiac fibroblasts (CFs). To minimize dedifferentiation, only CFs between passages 2 were utilized in subsequent experiments.

Cell culture

HEK293T cells were obtained from the American Type Culture Collection (CRL-3216, ATCC) and cultured in Dulbecco’s modified Eagle’s medium/high-glucose (DMEM) supplemented with 10% fetal bovine serum, streptomycin and penicillin (all from Gibco, Carlsbad, USA). The cells were confirmed to be free of microbial contamination (including mycoplasma) prior to use. HEK293T cells were used for the plasmid’s transient transfection and si-RNA transfection. Adult mice cardiac fibroblasts were isolated from 5 to 6 weeks old mice.

Plasmids transient transfection

The Flag-tagged full-length ZEB2 (Flag-ZEB2-WT), Flag-ZEB2-K391R, Flag-ZEB2-K462R, Flag-ZEB2-K479R, Flag-ZEB2-K611R, Flag-ZEB2-K774R, Flag-ZEB2-K866R, Flag-ZEB2-4KR, HA-SUMO1, His-SUMO1, Myc-SENP1 and Myc-SENP2 Plasmids were constructed by Zebrafish Biotech Co., Ltd. (Nanjing, China). For transient transfection, HEK293T (CRL-3216, ATCC) cells were transfected with the indicated plasmids using Lipofectamine 3000 (#1742691, Life) according to manufacturer’s recommendations.

Adenovirus infection

Flag-tagged WT full-length ZEB2 gene and Flag-ZEB2-4KR were constructed into adenovirus vector, as Adv-Flag-ZEB2-WT, and Adv-Flag-ZEB2-4KR. ZEB2^KO CFs were infected with Adv-Flag-ZEB2-WT, and Adv-Flag-ZEB2-4KR in DMEM medium (MOI = 100). After 6 h, fresh completed DMEM were added for 24h following the TGF-β1 treatment.

Lentivirus transfection

Flag-tagged WT full-length ZEB2 gene (Flag-ZEB2-WT), Flag-ZEB2-K462R, Flag-ZEB2-479R, Flag-ZEB2-K611R, Flag-ZEB2-K774R, Flag-ZEB2-4KR were packed as lentivirus and then transfected into CFs. After 24 h plating, cardiac fibroblasts were infected with lentiviruses diluted in DMEM medium (MOI = 100). 5 μg/mL polybrene was added to enhance the transfection efficacy. The medium was replaced with virus-free medium 24 h following infection, and then cells were cultured for an additional 48 h prior to experiments.

siRNA transfection Interference

HEK293T cells were transfected with negative control (NC) and small interfering RNA (siRNA) using Lipofectamine 3000, following the manufacturer’s instructions. The sequences of siRNAs used in this study are as follows: SENP1 siRNA-1 sense: 5′-TACTGAACTAAGACATCG-3′, SENP2 siRNA-1 sense: 5′-GGAGCCTGACCTATCAGAA-3′. The siRNA was transfected into cells using Lipofectamine RNAiMAX Reagent (Invitrogen, 13778075) according to the manufacturer’s instructions. At the culmination of a 48-h period following siRNA transfection, cells were harvested and used for further experiments.

Collagen gel contraction assay

Fibroblast-mediated collagen contraction was evaluated using the Cell Contraction Assay Kit (Corning, 354236, USA) in accordance with the manufacturer’s instructions. The assay was performed using a final collagen concentration of 5 mg/mL and a cell density of 1×10⁵ cells/mL embedded within the collagen matrix. Glacial acetic acid (100%) was diluted to a 0.2% solution, sterilized by filtration through a 0.2 μm membrane filter, and subsequently cooled to 4 °C. Under sterile conditions, type I collagen was mixed with the 0.2% acetic acid solution to prepare a 3 mg/mL collagen stock. An optimized volume of 1 M NaOH (5 μL) was added to the collagen/media mixture, with the same volume used consistently across all gels. Each experiment was independently performed at least three times, with triplicate samples included for each condition.

Immunofluorescence staining

For immunofluorescence staining of myocardium sections, the sections were first dewaxed to water, followed by antigen retrieval using microwave heating. Tissues were then blocked with 3% bovine serum albumin (BSA) at room temperature for 1 h and incubated overnight at 4 °C with appropriately diluted primary antibodies. For immunofluorescence staining of cells, CFs were cultured on cell slides placed in confocal plates. After cell attachment, experimental treatments were applied. Post-treatment, cells were fixed with 4% paraformaldehyde for 15 min, permeabilized with 0.5% Triton X-100 for 10 min and blocked with 5% BSA for 1 h at room temperature. Cells were then incubated overnight at 4 °C with the corresponding primary antibodies. Subsequently, fluorescent secondary antibodies were applied for 1 h at room temperature. A fluorescent mounting medium containing DAPI (ab104139, Abcam, Cambridge, UK) was used for nuclear staining. Finally, a laser scanning confocal microscope (Olympus, Japan) was used to acquire immunofluorescence staining images and the fluorescence intensity was quantified by Image-Pro Plus software (version 6.0) and was analyzed by a person blinded to treatment.

Real-time quantification PCR

Total RNA was isolated from murine cardiac tissues or CFs using TRIzol reagent (Takara Bio, #TKR-9109, Japan). The extracted RNA was then reverse transcribed to cDNA using an All-in-One RT Master Mix (HiFScript, Cowin Biotech, #CW3371, China). Gene expression analysis was performed by quantitative PCR (qPCR) using an SYBR Green-based master mix (SuperStar Universal, Cowin Biotech, #CW3360, China) on a CFX96 real-time PCR system (Bio-Rad, USA). Relative mRNA levels were quantified by the comparative Ct (2^−ΔΔCt) method. Hypoxanthine-guanine phosphoribosyl transferase (HPRT) served as the endogenous reference gene for normalization, unless otherwise specified. Data are expressed as mean fold change ±SD relative to control. Primer sequences are provided in the appendix (Table S1).

Western blotting

Protein extraction was performed on both myocardium and cultured cells respectively. Cells and heart tissues were collected and lysed on ice using RIPA lysis buffer (#P0013B, Beyotime Biotechnology) supplemented with 1% PMSF (Solarbio) and 1% protease inhibitor cocktail (#78438, Thermo Fisher Scientific). Lysates were centrifuged at 12,000 rpm for 15 min at 4 °C to collect the supernatant containing total protein. Protein concentrations were determined using the Pierce BCA Protein Assay Kit (#23227, Thermo Fisher Scientific). Samples were mixed with 5× loading buffer, denatured at 99 °C for 10 min in a metal bath, and subsequently subjected to SDS-PAGE. Protein samples (20–40) were resolved by SDS-PAGE and transferred onto PVDF membranes using a wet transfer method. Membranes were blocked with 5% non-fat milk or BSA for 1–2 h at room temperature and then incubated overnight at 4 °C with the appropriate primary antibodies. After washing, membranes were incubated with HRP-conjugated secondary antibodies for 1–2 h at room temperature. The signals were detected using the Electrochemiluminescence (ECL) substrate (Thermo, #32132) and images were obtained by ChemiDoc Imaging System (Bio-Rad, CA, US).

Co-immunoprecipitation mass spectrometry analysis

Cell and tissue lysates (800 μg) were prepared and incubated with antibodies against IgG, ZEB2, FLAG tag overnight at 4°C followed by the addition of 30 μL of protein A/G Agarose beads (Cat: 80106G, Thermo Fisher Scientific) for 4h at 4°C. After being washed three times with cold wash buffer, the immunoprecipitated complexes were immediately identified by SDS-PAGE and Western blotting analysis.

RNA sequencing (RNA-seq) and data analysis

RNA sequencing was performed by Sinotech Genomics Co., Ltd. (Shanghai, China). Total RNA was isolated from cells using TRIzol reagent (Invitrogen), followed by quality assessment with a NanoDrop One spectrophotometer (Thermo Scientific) and Bioanalyzer (Agilent) to confirm an RNA integrity number (RIN) > 7. Poly (A)+ mRNA was enriched using the NEBNext Poly (A) mRNA Magnetic Isolation Module (#E7490, NEB), and strand-specific cDNA libraries were constructed with the NEBNext Ultra II Directional RNA Library Prep Kit (#E7760, NEB). Sequencing was performed on an Illumina NovaSeq 6000 platform to generate 150-bp paired-end reads, yielding approximately 8 Gb (53.3 million reads, 26.7 million pairs) per sample.

CUT&Tag-sequencing and data processing

Chromatin profiling was performed using the Hyperactive Universal CUT&Tag Assay Kit for Illumina (Vazyme, China) according to the manufacturer’s protocol. Primary cardiac fibroblasts (CFs), isolated as described previously, were rested in complete medium at 37°C for 12 h prior to the assay. ZEB2-knockout CFs (ZEB2^KO CFs) were transduced with either Adv-Flag-ZEB2-WT or Adv-Flag-ZEB2-4 KR adenovirus, stimulated with TGF-β1 (24 h), and then subjected to CUT&Tag analysis. For the assay, approximately 1×10⁵ cells per sample were bound to concanavalin A-coated magnetic beads. Primary antibody incubation was performed overnight at 4°C. Samples were subsequently incubated for 1 h with a secondary antibody pre-complexed with Hyperactive pA/G-Tn5 Transposase. Tagmented DNA fragments were released by TTBL digestion (37°C, 1 h), purified, and amplified to generate sequencing libraries. Libraries were sequenced on an Illumina platform to generate paired end reads. Bioinformatic analysis proceeded as follows: Raw reads were aligned to the mm10 reference genome using Bowtie2 (v2.4.4). Peak calling was performed using MACS2 (v2.2.7.1) with default parameters. Differential peak analysis between specified sample groups was conducted using the DiffBind R package (v3.4.11), defining significant differences as |fold change| > 1.5 and adjusted p-value <0.05. Genomic annotations of differential peaks were assigned using ChIPseeker (v1.30.3). De novo motif discovery and enrichment analysis within peak regions were performed using HOMER (v4.11) with the findMotifsGenome.pl command.

Chromatin immunoprecipitation (ChIP) assays and sequencing

ChIP was performed using CHIP Assay Kit (#P2078, Beyotime Biotechnology). Briefly, cells were cross-linked with 1% formaldehyde for 8 min at room temperature before being washed with ice-cold phosphate-buffered saline (PBS), Solution I (10 mM of HEPES, pH 7.5, 10 mM of EDTA, 0.5 mM of EGTA, 0.75% Triton X-100) and Solution II (10 mM of HEPES, pH 7.5, 200 mM of NaCl, 1 mM of EDTA, 0.5 mM of EGTA). Then cells were incubated in lysis buffer (150 mM of NaCl, 25 mM of Tris, pH 7.5, 1% Triton X-100, 0.1% SDS, 0.5% deoxycholate) supplemented with protease inhibitor tablets. DNA was fragmented into 200 bp pieces using a Branson 250 sonicator. Aliquots of lysates containing 100 μg of protein were used for each immunoprecipitation reaction with indicated antibodies followed by adsorption to Protein A/G PLUS-Agarose beads (Thermo, USA). Precipitated DNA–protein complexes were washed sequentially with RIPA buffer (50 mM of Tris, pH 8.0, 150 mM of NaCl, 0.1% SDS, 0.5% deoxycholate, 1% Nonidet P-40, 1 mM of EDTA), high salt buffer (50 mM of Tris, pH 8.0, 500 mM of NaCl, 0.1% SDS, 0.5% deoxycholate, 1% Nonidet P-40, 1 mM EDTA), LiCl buffer (50 mM of Tris, pH 8.0, 250 mM of LiCl, 0.1% SDS, 0.5% deoxycholate, 1% Nonidet P-40, 1 mM of EDTA) and TE buffer (10 mM of Tris, 1 mM of EDTA, pH 8.0). DNA–protein cross-link was reversed by heating the samples to 65°C overnight. Proteins were digested with proteinase K (Sigma), and DNA was phenol/chloroform extracted and precipitated by 100% ethanol. Precipitated genomic DNA was amplified by PCR and the Primers of Nr4a1 were used for PCR assay. The following primers were used for ChIP-assay of targeting genes: Mice Nr4a1, forward: 5′- TTGGGGCTAGTGTAATGGGC-3′

Mice Nr4a1, reverse: 5′- GTGATACCCCAGTGCAGGTT-3′.

Luciferase assays

Luciferase reporter assays were conducted to assess promoter activity using the Dual-Luciferase Reporter Assay System (Promega, #E1910) per the manufacturer’s protocol. HEK293T cells were plated in 24-well plates (1 ×10⁵ cells/well) and cultured for 24 h prior to transfection. Wild-type or mutant fragments of the Nr4a1 promoter were cloned into the pGL6-Basic luciferase reporter vector. Each reporter construct was co-transfected into HEK293T cells with either a ZEB2-WT expression plasmid, a ZEB2-4KR expression plasmid, or an empty vector control. Cells were harvested 24–48 h post-transfection. Firefly luciferase activity, driven by the Nr4a1 promoter variants, was measured and normalized to co-transfected Renilla luciferase activity (phRL vector) to account for variations in transfection efficiency.

Quantification and statistical analysis

Data are presented as mean ± standard deviation (SD) derived from independent biological replicates. The value of n represents the sample size for human myocardium and animal studies, and the times of independent biological replicates for in vitro cell experiments. Normality (Shapiro-Wilk or Kolmogorov-Smirnov test) and homogeneity of variance (Brown-Forsythe or Bartlett’s test) were assessed for all datasets. For comparisons between two groups meeting both assumptions, unpaired two-tailed Student’s t-tests were applied. When assumptions were violated, the Mann-Whitney U test was used. Multi-group comparisons employed either one-way or two-way analysis of variance (ANOVA). Where ANOVA indicated significance (p < 0.05), post hoc pairwise comparisons were performed using the Bonferroni method. Statistical significance was defined as p < 0.05, with non-significant results denoted as ns. Full statistical details (including specific tests, n values, p-values, and degrees of freedom where applicable) are reported in the relevant Figure Legends. All analyses were conducted using GraphPad Prism (version 9.0.0).

Footnotes

Supplemental information can be found online at https://doi.org/10.1016/j.isci.2026.117091.

Contributor Information

Yunxing Xue, Email: albert_xue@163.com.

Qianwen Zhao, Email: cola@njmu.edu.cn.

Jie Yang, Email: yangjie1130@njglyy.com.

Dongjin Wang, Email: wangdongjin@njglyy.com.

Supplemental information

Document S1. Figures S1–S10 and Tables S1–S5 and Data S1

mmc1.pdf^{(6.7MB, pdf)}

References

1.Tsao C.W., Aday A.W., Almarzooq Z.I., Anderson C.A.M., Arora P., Avery C.L., Baker-Smith C.M., Beaton A.Z., Boehme A.K., Buxton A.E., et al. Heart Disease and Stroke Statistics-2023 Update: A Report From the American Heart Association. Circulation. 2023;147:e93–e621. doi: 10.1161/CIR.0000000000001123. [DOI] [PMC free article] [PubMed] [Google Scholar]
2.Frangogiannis N.G. The Extracellular Matrix in Ischemic and Nonischemic Heart Failure. Circ. Res. 2019;125:117–146. doi: 10.1161/CIRCRESAHA.119.311148. [DOI] [PMC free article] [PubMed] [Google Scholar]
3.Talman V., Ruskoaho H. Cardiac fibrosis in myocardial infarction-from repair and remodeling to regeneration. Cell Tissue Res. 2016;365:563–581. doi: 10.1007/s00441-016-2431-9. [DOI] [PMC free article] [PubMed] [Google Scholar]
4.Travers J.G., Kamal F.A., Robbins J., Yutzey K.E., Blaxall B.C. Cardiac Fibrosis: The Fibroblast Awakens. Circ. Res. 2016;118:1021–1040. doi: 10.1161/CIRCRESAHA.115.306565. [DOI] [PMC free article] [PubMed] [Google Scholar]
5.Pesce M., Duda G.N., Forte G., Girao H., Raya A., Roca-Cusachs P., Sluijter J.P.G., Tschöpe C., Van Linthout S. Cardiac fibroblasts and mechanosensation in heart development, health and disease. Nat. Rev. Cardiol. 2023;20:309–324. doi: 10.1038/s41569-022-00799-2. [DOI] [PubMed] [Google Scholar]
6.Zhao X., Williamson T., Gong Y., Epstein J.A., Fan Y. Immunomodulatory Therapy for Ischemic Heart Disease. Circulation. 2024;150:1050–1058. doi: 10.1161/CIRCULATIONAHA.124.070368. [DOI] [PMC free article] [PubMed] [Google Scholar]
7.Plikus M.V., Wang X., Sinha S., Forte E., Thompson S.M., Herzog E.L., Driskell R.R., Rosenthal N., Biernaskie J., Horsley V. Fibroblasts: Origins, definitions, and functions in health and disease. Cell. 2021;184:3852–3872. doi: 10.1016/j.cell.2021.06.024. [DOI] [PMC free article] [PubMed] [Google Scholar]
8.Moore-Morris T., Guimarães-Camboa N., Banerjee I., Zambon A.C., Kisseleva T., Velayoudon A., Stallcup W.B., Gu Y., Dalton N.D., Cedenilla M., et al. Resident fibroblast lineages mediate pressure overload-induced cardiac fibrosis. J. Clin. Investig. 2014;124:2921–2934. doi: 10.1172/JCI74783. [DOI] [PMC free article] [PubMed] [Google Scholar]
9.Zhao M., Wang L., Wang M., Zhou S., Lu Y., Cui H., Racanelli A.C., Zhang L., Ye T., Ding B., et al. Targeting fibrosis, mechanisms and cilinical trials. Signal Transduct. Target. Ther. 2022;7:206. doi: 10.1038/s41392-022-01070-3. [DOI] [PMC free article] [PubMed] [Google Scholar]
10.Chen L., Li H., Liu X., Zhang N., Wang K., Shi A., Gao H., Akdis D., Saguner A.M., Xu X., et al. PBX/Knotted 1 homeobox-2 (PKNOX2) is a novel regulator of myocardial fibrosis. Signal Transduct. Target. Ther. 2024;9:94. doi: 10.1038/s41392-024-01804-5. [DOI] [PMC free article] [PubMed] [Google Scholar]
11.Johansen A.K.Z., Kasam R.K., Vagnozzi R.J., Lin S.C.J., Gomez-Arroyo J., Shittu A., Bowers S.L.K., Kuwabara Y., Grimes K.M., Warrick K., et al. Transcription Factor 21 Regulates Cardiac Myofibroblast Formation and Fibrosis. Circ. Res. 2025;136:44–58. doi: 10.1161/CIRCRESAHA.124.325527. [DOI] [PMC free article] [PubMed] [Google Scholar]
12.Tani H., Sadahiro T., Yamada Y., Isomi M., Yamakawa H., Fujita R., Abe Y., Akiyama T., Nakano K., Kuze Y., et al. Direct Reprogramming Improves Cardiac Function and Reverses Fibrosis in Chronic Myocardial Infarction. Circulation. 2023;147:223–238. doi: 10.1161/CIRCULATIONAHA.121.058655. [DOI] [PubMed] [Google Scholar]
13.Cheng P., Wirka R.C., Shoa Clarke L., Zhao Q., Kundu R., Nguyen T., Nair S., Sharma D., Kim H.J., Shi H., et al. ZEB2 Shapes the Epigenetic Landscape of Atherosclerosis. Circulation. 2022;145:469–485. doi: 10.1161/CIRCULATIONAHA.121.057789. [DOI] [PMC free article] [PubMed] [Google Scholar]
14.Comijn J., Berx G., Vermassen P., Verschueren K., van Grunsven L., Bruyneel E., Mareel M., Huylebroeck D., van Roy F. The two-handed E box binding zinc finger protein SIP1 downregulates E-cadherin and induces invasion. Mol. Cell. 2001;7:1267–1278. doi: 10.1016/s1097-2765(01)00260-x. [DOI] [PubMed] [Google Scholar]
15.Thompson E.W., Redfern A.D., Brabletz S., Berx G., Agarwal V., Ganesh K., Huang R.Y., Ishay-Ronen D., Savagner P., Sheng G., et al. EMT and cancer: what clinicians should know. Nat. Rev. Clin. Oncol. 2025;22:711–733. doi: 10.1038/s41571-025-01058-2. [DOI] [PubMed] [Google Scholar]
16.Gladka M.M., Kohela A., de Leeuw A.E., Molenaar B., Versteeg D., Kooijman L., van Geldorp M., van Ham W.B., Caliandro R., Haigh J.J., et al. Hypoxia-responsive zinc finger E-box-binding homeobox 2 (ZEB2) regulates a network of calcium-handling genes in the injured heart. Cardiovasc. Res. 2024;120:1869–1883. doi: 10.1093/cvr/cvae163. [DOI] [PMC free article] [PubMed] [Google Scholar]
17.Zhang M.Y., Zhang B., Tao Q.Q., Wang X.Y., Guan X.X., Dong Z., Guo X.F., Qi W.Y., Cheng W.Z., Zhang X.Y., et al. Single-cell RNA sequencing identifies a unique ZEB2-Regulated Clec4d(+) macrophage state contributing to early inflammatory injury in cardiac hypertrophy. Eur. J. Pharmacol. 2025;1002 doi: 10.1016/j.ejphar.2025.177867. [DOI] [PubMed] [Google Scholar]
18.Gladka M.M., Kohela A., Molenaar B., Versteeg D., Kooijman L., Monshouwer-Kloots J., Kremer V., Vos H.R., Huibers M.M.H., Haigh J.J., et al. Cardiomyocytes stimulate angiogenesis after ischemic injury in a ZEB2-dependent manner. Nat. Commun. 2021;12:84. doi: 10.1038/s41467-020-20361-3. [DOI] [PMC free article] [PubMed] [Google Scholar]
19.Jahan F., Landry N.M., Rattan S.G., Dixon I.M.C., Wigle J.T. The Functional Role of Zinc Finger E Box-Binding Homeobox 2 (Zeb2) in Promoting Cardiac Fibroblast Activation. Int. J. Mol. Sci. 2018;19 doi: 10.3390/ijms19103207. [DOI] [PMC free article] [PubMed] [Google Scholar]
20.Lamoliatte F., McManus F.P., Maarifi G., Chelbi-Alix M.K., Thibault P. Uncovering the SUMOylation and ubiquitylation crosstalk in human cells using sequential peptide immunopurification. Nat. Commun. 2017;8 doi: 10.1038/ncomms14109. [DOI] [PMC free article] [PubMed] [Google Scholar]
21.Haerinck J., Goossens S., Berx G. The epithelial-mesenchymal plasticity landscape: principles of design and mechanisms of regulation. Nat. Rev. Genet. 2023;24:590–609. doi: 10.1038/s41576-023-00601-0. [DOI] [PubMed] [Google Scholar]
22.Shi X., Dang X., Huang Z., Lu Y., Tong H., Liang F., Zhuang F., Li Y., Cai Z., Huo H., et al. SUMOylation of TEAD1 Modulates the Mechanism of Pathological Cardiac Hypertrophy. Adv. Sci. 2024;11 doi: 10.1002/advs.202305677. [DOI] [PMC free article] [PubMed] [Google Scholar]
23.Ren R., Ding S., Ma K., Jiang Y., Wang Y., Chen J., Wang Y., Kou Y., Fan X., Zhu X., et al. SUMOylation Fine-Tunes Endothelial HEY1 in the Regulation of Angiogenesis. Circ. Res. 2024;134:203–222. doi: 10.1161/CIRCRESAHA.123.323398. [DOI] [PMC free article] [PubMed] [Google Scholar]
24.Xu Y., Zhang H., Chen Y., Pober J.S., Zhou M., Zhou J.H., Min W. SRF SUMOylation modulates smooth muscle phenotypic switch and vascular remodeling. Nat. Commun. 2024;15:6919. doi: 10.1038/s41467-024-51350-5. [DOI] [PMC free article] [PubMed] [Google Scholar]
25.Li X.C., Zeng Y., Sun R.R., Liu M., Chen S., Zhang P.Y. SUMOylation in cardiac disorders - a review. Eur. Rev. Med. Pharmacol. Sci. 2017;21:1583–1587. [PubMed] [Google Scholar]
26.Hickey C.M., Wilson N.R., Hochstrasser M. Function and regulation of SUMO proteases. Nat. Rev. Mol. Cell Biol. 2012;13:755–766. doi: 10.1038/nrm3478. [DOI] [PMC free article] [PubMed] [Google Scholar]
27.Liu W., Li G., Shi J., Gao Y., Fang P., Zhao Y., Zhong F., Guo X., Lyu Y., Da X., et al. NR4A1 Acts as a Novel Regulator of Platelet Activation and Thrombus Formation. Circ. Res. 2025;136:809–826. doi: 10.1161/CIRCRESAHA.124.325645. [DOI] [PMC free article] [PubMed] [Google Scholar]
28.Ji J.J., Qian L.L., Zhu Y., Jiang Y., Guo J.Q., Wu Y., Yang Z.W., Yao Y.Y., Ma G.S. Kallistatin/Serpina3c inhibits cardiac fibrosis after myocardial infarction by regulating glycolysis via Nr4a1 activation. Biochim. Biophys. Acta. Mol. Basis Dis. 2022;1868 doi: 10.1016/j.bbadis.2022.166441. [DOI] [PubMed] [Google Scholar]
29.Medzikovic L., van Roomen C., Baartscheer A., van Loenen P.B., de Vos J., Bakker E.N.T.P., Koenis D.S., Damanafshan A., Creemers E.E., Arkenbout E.K., et al. Nur77 protects against adverse cardiac remodelling by limiting neuropeptide Y signalling in the sympathoadrenal-cardiac axis. Cardiovasc. Res. 2018;114:1617–1628. doi: 10.1093/cvr/cvy125. [DOI] [PubMed] [Google Scholar]
30.Kumar P.A., Kotlyarevska K., Dejkhmaron P., Reddy G.R., Lu C., Bhojani M.S., Menon R.K. Growth hormone (GH)-dependent expression of a natural antisense transcript induces zinc finger E-box-binding homeobox 2 (ZEB2) in the glomerular podocyte: a novel action of gh with implications for the pathogenesis of diabetic nephropathy. J. Biol. Chem. 2010;285:31148–31156. doi: 10.1074/jbc.M110.132332. [DOI] [PMC free article] [PubMed] [Google Scholar]
31.van Rooij E., Sutherland L.B., Qi X., Richardson J.A., Hill J., Olson E.N. Control of stress-dependent cardiac growth and gene expression by a microRNA. Science. 2007;316:575–579. doi: 10.1126/science.1139089. [DOI] [PubMed] [Google Scholar]
32.Montgomery R.L., Hullinger T.G., Semus H.M., Dickinson B.A., Seto A.G., Lynch J.M., Stack C., Latimer P.A., Olson E.N., van Rooij E. Therapeutic inhibition of miR-208a improves cardiac function and survival during heart failure. Circulation. 2011;124:1537–1547. doi: 10.1161/CIRCULATIONAHA.111.030932. [DOI] [PMC free article] [PubMed] [Google Scholar]
33.Guo Y., Bie Z.D., Li X. Hypoxic cardiomyocyte-derived exosomes regulate cardiac fibroblast activation, apoptosis, migration and ferroptosis through miR-208a/b. Gen. Physiol. Biophys. 2023;42:149–158. doi: 10.4149/gpb_2022061. [DOI] [PubMed] [Google Scholar]
34.Baksh S.S., Hodgkinson C.P. Conservation of miR combo based direct cardiac reprogramming. Biochem. Biophys. Rep. 2022;31 doi: 10.1016/j.bbrep.2022.101310. [DOI] [PMC free article] [PubMed] [Google Scholar]
35.Yang J., Yu X., Xue F., Li Y., Liu W., Zhang S. Exosomes derived from cardiomyocytes promote cardiac fibrosis via myocyte-fibroblast cross-talk. Am. J. Transl. Res. 2018;10:4350–4366. [PMC free article] [PubMed] [Google Scholar]
36.Long J., Zuo D., Park M. Pc2-mediated sumoylation of Smad-interacting protein 1 attenuates transcriptional repression of E-cadherin. J. Biol. Chem. 2005;280:35477–35489. doi: 10.1074/jbc.M504477200. [DOI] [PubMed] [Google Scholar]
37.Liu Z., Bian X., Li L., Liu L., Feng C., Wang Y., Ni J., Li S., Lu D., Li Y., et al. SENP1-Mediated HSP90ab1 DeSUMOylation in Cardiomyocytes Prevents Myocardial Fibrosis by Paracrine Signaling. Adv. Sci. 2024;11 doi: 10.1002/advs.202400741. [DOI] [PMC free article] [PubMed] [Google Scholar]
38.Zhang X., Chang X., Chai R., Zhang X., Li J., Guo Z., Qiu Z., Song Y., Shi S., Hu Y., Du B. Xin-Fu-Kang oral liquid mitigates chronic heart failure through NR4A1-Dependent regulation of endoplasmic reticulum-mitochondrial crosstalk in Cardiomyocytes. Phytomedicine. 2025;140 doi: 10.1016/j.phymed.2025.156467. [DOI] [PubMed] [Google Scholar]
39.Marinković G., Koenis D.S., de Camp L., Jablonowski R., Graber N., de Waard V., de Vries C.J., Goncalves I., Nilsson J., Jovinge S., Schiopu A. S100A9 Links Inflammation and Repair in Myocardial Infarction. Circ. Res. 2020;127:664–676. doi: 10.1161/CIRCRESAHA.120.315865. [DOI] [PubMed] [Google Scholar]
40.Fuchs C.D., Claudel T., Mlitz V., Riva A., Menz M., Brusilovskaya K., Haller F., Baumgartner M., Königshofer P., Unger L.W., et al. GLP-2 Improves Hepatic Inflammation and Fibrosis in Mdr2(-/-) Mice Via Activation of NR4a1/Nur77 in Hepatic Stellate Cells and Intestinal FXR Signaling. Cell. Mol. Gastroenterol. Hepatol. 2023;16:847–856. doi: 10.1016/j.jcmgh.2023.08.003. [DOI] [PMC free article] [PubMed] [Google Scholar]
41.Yang R., Zong T., Wang N., Wang F., Su Y. NR4A1 Alleviates Subretinal Fibrosis by Inhibiting Macrophage to Myofibroblast Transition. Investig. Ophthalmol. Vis. Sci. 2025;66:47. doi: 10.1167/iovs.66.6.47. [DOI] [PMC free article] [PubMed] [Google Scholar]
42.Chen J., Jia J., Ma L., Li B., Qin Q., Qian J., Ge J. Nur77 deficiency exacerbates cardiac fibrosis after myocardial infarction by promoting endothelial-to-mesenchymal transition. J. Cell. Physiol. 2021;236:495–506. doi: 10.1002/jcp.29877. [DOI] [PubMed] [Google Scholar]
43.Liu D., Wang H., Li J., Sheng S., Wang S., Tian Y. Non-lethal sonodynamic therapy mitigates hypertensive renal fibrosis through the PI3K/AKT/mTORC1-autophagy pathway. Sci. Rep. 2025;15:4534. doi: 10.1038/s41598-025-86973-1. [DOI] [PMC free article] [PubMed] [Google Scholar]
44.Yan L., Wei X., Tang Q.Z., Feng J., Zhang Y., Liu C., Bian Z.Y., Zhang L.F., Chen M., Bai X., et al. Cardiac-specific mindin overexpression attenuates cardiac hypertrophy via blocking AKT/GSK3beta and TGF-beta1-Smad signalling. Cardiovasc. Res. 2011;92:85–94. doi: 10.1093/cvr/cvr159. [DOI] [PMC free article] [PubMed] [Google Scholar]
45.Paoletti E. mTOR Inhibition and Cardiovascular Diseases: Cardiac Hypertrophy. Transplantation. 2018;102:S41–S43. doi: 10.1097/TP.0000000000001691. [DOI] [PubMed] [Google Scholar]
46.Huang H.Y., Chang H.F., Tsai M.J., Chen J.S., Wang M.J. 6-Mercaptopurine attenuates tumor necrosis factor-alpha production in microglia through Nur77-mediated transrepression and PI3K/Akt/mTOR signaling-mediated translational regulation. J. Neuroinflammation. 2016;13:78. doi: 10.1186/s12974-016-0543-5. [DOI] [PMC free article] [PubMed] [Google Scholar]
47.Wang D., Yin Y., Wang S., Zhao T., Gong F., Zhao Y., Wang B., Huang Y., Cheng Z., Zhu G., et al. FGF1(DeltaHBS) prevents diabetic cardiomyopathy by maintaining mitochondrial homeostasis and reducing oxidative stress via AMPK/Nur77 suppression. Signal Transduct. Target. Ther. 2021;6:133. doi: 10.1038/s41392-021-00542-2. [DOI] [PMC free article] [PubMed] [Google Scholar]
48.Zheng Y., Xu Y., Ji L., San W., Shen D., Zhou Q., Meng G., Shi J., Chen Y. Roles of distinct nuclear receptors in diabetic cardiomyopathy. Front. Pharmacol. 2024;15 doi: 10.3389/fphar.2024.1423124. [DOI] [PMC free article] [PubMed] [Google Scholar]
49.Medzikovic L., de Vries C.J.M., de Waard V. NR4A nuclear receptors in cardiac remodeling and neurohormonal regulation. Trends Cardiovasc. Med. 2019;29:429–437. doi: 10.1016/j.tcm.2018.11.015. [DOI] [PubMed] [Google Scholar]
50.Song S., Zhang X., Huang Z., Zeng L., Cai F., Wang T., Song Y., Guo J., Li M., Liu C., et al. Cardiac fibroblast Foxm1 deficiency prevents pressure overload-induced cardiac remodeling via the Usp10/MKK6-p38gamma MAPK axis. Cell Death Differ. 2026;33:765–778. doi: 10.1038/s41418-025-01609-1. [DOI] [PMC free article] [PubMed] [Google Scholar]
51.Song S., Zhang X., Huang Z., Zhao Y., Lu S., Zeng L., Cai F., Wang T., Pei Z., Weng X., et al. TEA domain transcription factor 1(TEAD1) induces cardiac fibroblasts cells remodeling through BRD4/Wnt4 pathway. Signal Transduct. Target. Ther. 2024;9:45. doi: 10.1038/s41392-023-01732-w. [DOI] [PMC free article] [PubMed] [Google Scholar]
52.Zhao Q., Shao T., Zhu Y., Zong G., Zhang J., Tang S., Lin Y., Ma H., Jiang Z., Xu Y., et al. An MRTF-A-ZEB1-IRF9 axis contributes to fibroblast-myofibroblast transition and renal fibrosis. Exp. Mol. Med. 2023;55:987–998. doi: 10.1038/s12276-023-00990-6. [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Document S1. Figures S1–S10 and Tables S1–S5 and Data S1

mmc1.pdf^{(6.7MB, pdf)}

Data Availability Statement

•
Data: Transcriptomic: Raw RNA-sequencing are deposited in the genome sequencing archive (GSA: CRA046490) at https://ngdc.cncb.ac.cn/gsa/browse/CRA046490 and CUT&Tag-sequencing data are deposited in the genome sequencing archive (GSA:CRA046488) at https://ngdc.cncb.ac.cn/gsa/browse/CRA046488.
•
Code: This paper does not report original code.
•
Additional information: Any additional information required to reanalyze the data reported in this paper is available from the lead contact upon request.

[bib1] 1.Tsao C.W., Aday A.W., Almarzooq Z.I., Anderson C.A.M., Arora P., Avery C.L., Baker-Smith C.M., Beaton A.Z., Boehme A.K., Buxton A.E., et al. Heart Disease and Stroke Statistics-2023 Update: A Report From the American Heart Association. Circulation. 2023;147:e93–e621. doi: 10.1161/CIR.0000000000001123. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib2] 2.Frangogiannis N.G. The Extracellular Matrix in Ischemic and Nonischemic Heart Failure. Circ. Res. 2019;125:117–146. doi: 10.1161/CIRCRESAHA.119.311148. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib3] 3.Talman V., Ruskoaho H. Cardiac fibrosis in myocardial infarction-from repair and remodeling to regeneration. Cell Tissue Res. 2016;365:563–581. doi: 10.1007/s00441-016-2431-9. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib4] 4.Travers J.G., Kamal F.A., Robbins J., Yutzey K.E., Blaxall B.C. Cardiac Fibrosis: The Fibroblast Awakens. Circ. Res. 2016;118:1021–1040. doi: 10.1161/CIRCRESAHA.115.306565. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib5] 5.Pesce M., Duda G.N., Forte G., Girao H., Raya A., Roca-Cusachs P., Sluijter J.P.G., Tschöpe C., Van Linthout S. Cardiac fibroblasts and mechanosensation in heart development, health and disease. Nat. Rev. Cardiol. 2023;20:309–324. doi: 10.1038/s41569-022-00799-2. [DOI] [PubMed] [Google Scholar]

[bib6] 6.Zhao X., Williamson T., Gong Y., Epstein J.A., Fan Y. Immunomodulatory Therapy for Ischemic Heart Disease. Circulation. 2024;150:1050–1058. doi: 10.1161/CIRCULATIONAHA.124.070368. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib7] 7.Plikus M.V., Wang X., Sinha S., Forte E., Thompson S.M., Herzog E.L., Driskell R.R., Rosenthal N., Biernaskie J., Horsley V. Fibroblasts: Origins, definitions, and functions in health and disease. Cell. 2021;184:3852–3872. doi: 10.1016/j.cell.2021.06.024. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib8] 8.Moore-Morris T., Guimarães-Camboa N., Banerjee I., Zambon A.C., Kisseleva T., Velayoudon A., Stallcup W.B., Gu Y., Dalton N.D., Cedenilla M., et al. Resident fibroblast lineages mediate pressure overload-induced cardiac fibrosis. J. Clin. Investig. 2014;124:2921–2934. doi: 10.1172/JCI74783. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib9] 9.Zhao M., Wang L., Wang M., Zhou S., Lu Y., Cui H., Racanelli A.C., Zhang L., Ye T., Ding B., et al. Targeting fibrosis, mechanisms and cilinical trials. Signal Transduct. Target. Ther. 2022;7:206. doi: 10.1038/s41392-022-01070-3. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib10] 10.Chen L., Li H., Liu X., Zhang N., Wang K., Shi A., Gao H., Akdis D., Saguner A.M., Xu X., et al. PBX/Knotted 1 homeobox-2 (PKNOX2) is a novel regulator of myocardial fibrosis. Signal Transduct. Target. Ther. 2024;9:94. doi: 10.1038/s41392-024-01804-5. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib11] 11.Johansen A.K.Z., Kasam R.K., Vagnozzi R.J., Lin S.C.J., Gomez-Arroyo J., Shittu A., Bowers S.L.K., Kuwabara Y., Grimes K.M., Warrick K., et al. Transcription Factor 21 Regulates Cardiac Myofibroblast Formation and Fibrosis. Circ. Res. 2025;136:44–58. doi: 10.1161/CIRCRESAHA.124.325527. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib12] 12.Tani H., Sadahiro T., Yamada Y., Isomi M., Yamakawa H., Fujita R., Abe Y., Akiyama T., Nakano K., Kuze Y., et al. Direct Reprogramming Improves Cardiac Function and Reverses Fibrosis in Chronic Myocardial Infarction. Circulation. 2023;147:223–238. doi: 10.1161/CIRCULATIONAHA.121.058655. [DOI] [PubMed] [Google Scholar]

[bib13] 13.Cheng P., Wirka R.C., Shoa Clarke L., Zhao Q., Kundu R., Nguyen T., Nair S., Sharma D., Kim H.J., Shi H., et al. ZEB2 Shapes the Epigenetic Landscape of Atherosclerosis. Circulation. 2022;145:469–485. doi: 10.1161/CIRCULATIONAHA.121.057789. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib14] 14.Comijn J., Berx G., Vermassen P., Verschueren K., van Grunsven L., Bruyneel E., Mareel M., Huylebroeck D., van Roy F. The two-handed E box binding zinc finger protein SIP1 downregulates E-cadherin and induces invasion. Mol. Cell. 2001;7:1267–1278. doi: 10.1016/s1097-2765(01)00260-x. [DOI] [PubMed] [Google Scholar]

[bib15] 15.Thompson E.W., Redfern A.D., Brabletz S., Berx G., Agarwal V., Ganesh K., Huang R.Y., Ishay-Ronen D., Savagner P., Sheng G., et al. EMT and cancer: what clinicians should know. Nat. Rev. Clin. Oncol. 2025;22:711–733. doi: 10.1038/s41571-025-01058-2. [DOI] [PubMed] [Google Scholar]

[bib16] 16.Gladka M.M., Kohela A., de Leeuw A.E., Molenaar B., Versteeg D., Kooijman L., van Geldorp M., van Ham W.B., Caliandro R., Haigh J.J., et al. Hypoxia-responsive zinc finger E-box-binding homeobox 2 (ZEB2) regulates a network of calcium-handling genes in the injured heart. Cardiovasc. Res. 2024;120:1869–1883. doi: 10.1093/cvr/cvae163. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib17] 17.Zhang M.Y., Zhang B., Tao Q.Q., Wang X.Y., Guan X.X., Dong Z., Guo X.F., Qi W.Y., Cheng W.Z., Zhang X.Y., et al. Single-cell RNA sequencing identifies a unique ZEB2-Regulated Clec4d(+) macrophage state contributing to early inflammatory injury in cardiac hypertrophy. Eur. J. Pharmacol. 2025;1002 doi: 10.1016/j.ejphar.2025.177867. [DOI] [PubMed] [Google Scholar]

[bib18] 18.Gladka M.M., Kohela A., Molenaar B., Versteeg D., Kooijman L., Monshouwer-Kloots J., Kremer V., Vos H.R., Huibers M.M.H., Haigh J.J., et al. Cardiomyocytes stimulate angiogenesis after ischemic injury in a ZEB2-dependent manner. Nat. Commun. 2021;12:84. doi: 10.1038/s41467-020-20361-3. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib19] 19.Jahan F., Landry N.M., Rattan S.G., Dixon I.M.C., Wigle J.T. The Functional Role of Zinc Finger E Box-Binding Homeobox 2 (Zeb2) in Promoting Cardiac Fibroblast Activation. Int. J. Mol. Sci. 2018;19 doi: 10.3390/ijms19103207. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib20] 20.Lamoliatte F., McManus F.P., Maarifi G., Chelbi-Alix M.K., Thibault P. Uncovering the SUMOylation and ubiquitylation crosstalk in human cells using sequential peptide immunopurification. Nat. Commun. 2017;8 doi: 10.1038/ncomms14109. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib21] 21.Haerinck J., Goossens S., Berx G. The epithelial-mesenchymal plasticity landscape: principles of design and mechanisms of regulation. Nat. Rev. Genet. 2023;24:590–609. doi: 10.1038/s41576-023-00601-0. [DOI] [PubMed] [Google Scholar]

[bib22] 22.Shi X., Dang X., Huang Z., Lu Y., Tong H., Liang F., Zhuang F., Li Y., Cai Z., Huo H., et al. SUMOylation of TEAD1 Modulates the Mechanism of Pathological Cardiac Hypertrophy. Adv. Sci. 2024;11 doi: 10.1002/advs.202305677. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib23] 23.Ren R., Ding S., Ma K., Jiang Y., Wang Y., Chen J., Wang Y., Kou Y., Fan X., Zhu X., et al. SUMOylation Fine-Tunes Endothelial HEY1 in the Regulation of Angiogenesis. Circ. Res. 2024;134:203–222. doi: 10.1161/CIRCRESAHA.123.323398. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib24] 24.Xu Y., Zhang H., Chen Y., Pober J.S., Zhou M., Zhou J.H., Min W. SRF SUMOylation modulates smooth muscle phenotypic switch and vascular remodeling. Nat. Commun. 2024;15:6919. doi: 10.1038/s41467-024-51350-5. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib25] 25.Li X.C., Zeng Y., Sun R.R., Liu M., Chen S., Zhang P.Y. SUMOylation in cardiac disorders - a review. Eur. Rev. Med. Pharmacol. Sci. 2017;21:1583–1587. [PubMed] [Google Scholar]

[bib26] 26.Hickey C.M., Wilson N.R., Hochstrasser M. Function and regulation of SUMO proteases. Nat. Rev. Mol. Cell Biol. 2012;13:755–766. doi: 10.1038/nrm3478. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib27] 27.Liu W., Li G., Shi J., Gao Y., Fang P., Zhao Y., Zhong F., Guo X., Lyu Y., Da X., et al. NR4A1 Acts as a Novel Regulator of Platelet Activation and Thrombus Formation. Circ. Res. 2025;136:809–826. doi: 10.1161/CIRCRESAHA.124.325645. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib28] 28.Ji J.J., Qian L.L., Zhu Y., Jiang Y., Guo J.Q., Wu Y., Yang Z.W., Yao Y.Y., Ma G.S. Kallistatin/Serpina3c inhibits cardiac fibrosis after myocardial infarction by regulating glycolysis via Nr4a1 activation. Biochim. Biophys. Acta. Mol. Basis Dis. 2022;1868 doi: 10.1016/j.bbadis.2022.166441. [DOI] [PubMed] [Google Scholar]

[bib29] 29.Medzikovic L., van Roomen C., Baartscheer A., van Loenen P.B., de Vos J., Bakker E.N.T.P., Koenis D.S., Damanafshan A., Creemers E.E., Arkenbout E.K., et al. Nur77 protects against adverse cardiac remodelling by limiting neuropeptide Y signalling in the sympathoadrenal-cardiac axis. Cardiovasc. Res. 2018;114:1617–1628. doi: 10.1093/cvr/cvy125. [DOI] [PubMed] [Google Scholar]

[bib30] 30.Kumar P.A., Kotlyarevska K., Dejkhmaron P., Reddy G.R., Lu C., Bhojani M.S., Menon R.K. Growth hormone (GH)-dependent expression of a natural antisense transcript induces zinc finger E-box-binding homeobox 2 (ZEB2) in the glomerular podocyte: a novel action of gh with implications for the pathogenesis of diabetic nephropathy. J. Biol. Chem. 2010;285:31148–31156. doi: 10.1074/jbc.M110.132332. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib31] 31.van Rooij E., Sutherland L.B., Qi X., Richardson J.A., Hill J., Olson E.N. Control of stress-dependent cardiac growth and gene expression by a microRNA. Science. 2007;316:575–579. doi: 10.1126/science.1139089. [DOI] [PubMed] [Google Scholar]

[bib32] 32.Montgomery R.L., Hullinger T.G., Semus H.M., Dickinson B.A., Seto A.G., Lynch J.M., Stack C., Latimer P.A., Olson E.N., van Rooij E. Therapeutic inhibition of miR-208a improves cardiac function and survival during heart failure. Circulation. 2011;124:1537–1547. doi: 10.1161/CIRCULATIONAHA.111.030932. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib33] 33.Guo Y., Bie Z.D., Li X. Hypoxic cardiomyocyte-derived exosomes regulate cardiac fibroblast activation, apoptosis, migration and ferroptosis through miR-208a/b. Gen. Physiol. Biophys. 2023;42:149–158. doi: 10.4149/gpb_2022061. [DOI] [PubMed] [Google Scholar]

[bib34] 34.Baksh S.S., Hodgkinson C.P. Conservation of miR combo based direct cardiac reprogramming. Biochem. Biophys. Rep. 2022;31 doi: 10.1016/j.bbrep.2022.101310. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib35] 35.Yang J., Yu X., Xue F., Li Y., Liu W., Zhang S. Exosomes derived from cardiomyocytes promote cardiac fibrosis via myocyte-fibroblast cross-talk. Am. J. Transl. Res. 2018;10:4350–4366. [PMC free article] [PubMed] [Google Scholar]

[bib36] 36.Long J., Zuo D., Park M. Pc2-mediated sumoylation of Smad-interacting protein 1 attenuates transcriptional repression of E-cadherin. J. Biol. Chem. 2005;280:35477–35489. doi: 10.1074/jbc.M504477200. [DOI] [PubMed] [Google Scholar]

[bib37] 37.Liu Z., Bian X., Li L., Liu L., Feng C., Wang Y., Ni J., Li S., Lu D., Li Y., et al. SENP1-Mediated HSP90ab1 DeSUMOylation in Cardiomyocytes Prevents Myocardial Fibrosis by Paracrine Signaling. Adv. Sci. 2024;11 doi: 10.1002/advs.202400741. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib38] 38.Zhang X., Chang X., Chai R., Zhang X., Li J., Guo Z., Qiu Z., Song Y., Shi S., Hu Y., Du B. Xin-Fu-Kang oral liquid mitigates chronic heart failure through NR4A1-Dependent regulation of endoplasmic reticulum-mitochondrial crosstalk in Cardiomyocytes. Phytomedicine. 2025;140 doi: 10.1016/j.phymed.2025.156467. [DOI] [PubMed] [Google Scholar]

[bib39] 39.Marinković G., Koenis D.S., de Camp L., Jablonowski R., Graber N., de Waard V., de Vries C.J., Goncalves I., Nilsson J., Jovinge S., Schiopu A. S100A9 Links Inflammation and Repair in Myocardial Infarction. Circ. Res. 2020;127:664–676. doi: 10.1161/CIRCRESAHA.120.315865. [DOI] [PubMed] [Google Scholar]

[bib40] 40.Fuchs C.D., Claudel T., Mlitz V., Riva A., Menz M., Brusilovskaya K., Haller F., Baumgartner M., Königshofer P., Unger L.W., et al. GLP-2 Improves Hepatic Inflammation and Fibrosis in Mdr2(-/-) Mice Via Activation of NR4a1/Nur77 in Hepatic Stellate Cells and Intestinal FXR Signaling. Cell. Mol. Gastroenterol. Hepatol. 2023;16:847–856. doi: 10.1016/j.jcmgh.2023.08.003. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib41] 41.Yang R., Zong T., Wang N., Wang F., Su Y. NR4A1 Alleviates Subretinal Fibrosis by Inhibiting Macrophage to Myofibroblast Transition. Investig. Ophthalmol. Vis. Sci. 2025;66:47. doi: 10.1167/iovs.66.6.47. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib42] 42.Chen J., Jia J., Ma L., Li B., Qin Q., Qian J., Ge J. Nur77 deficiency exacerbates cardiac fibrosis after myocardial infarction by promoting endothelial-to-mesenchymal transition. J. Cell. Physiol. 2021;236:495–506. doi: 10.1002/jcp.29877. [DOI] [PubMed] [Google Scholar]

[bib43] 43.Liu D., Wang H., Li J., Sheng S., Wang S., Tian Y. Non-lethal sonodynamic therapy mitigates hypertensive renal fibrosis through the PI3K/AKT/mTORC1-autophagy pathway. Sci. Rep. 2025;15:4534. doi: 10.1038/s41598-025-86973-1. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib44] 44.Yan L., Wei X., Tang Q.Z., Feng J., Zhang Y., Liu C., Bian Z.Y., Zhang L.F., Chen M., Bai X., et al. Cardiac-specific mindin overexpression attenuates cardiac hypertrophy via blocking AKT/GSK3beta and TGF-beta1-Smad signalling. Cardiovasc. Res. 2011;92:85–94. doi: 10.1093/cvr/cvr159. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib45] 45.Paoletti E. mTOR Inhibition and Cardiovascular Diseases: Cardiac Hypertrophy. Transplantation. 2018;102:S41–S43. doi: 10.1097/TP.0000000000001691. [DOI] [PubMed] [Google Scholar]

[bib46] 46.Huang H.Y., Chang H.F., Tsai M.J., Chen J.S., Wang M.J. 6-Mercaptopurine attenuates tumor necrosis factor-alpha production in microglia through Nur77-mediated transrepression and PI3K/Akt/mTOR signaling-mediated translational regulation. J. Neuroinflammation. 2016;13:78. doi: 10.1186/s12974-016-0543-5. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib47] 47.Wang D., Yin Y., Wang S., Zhao T., Gong F., Zhao Y., Wang B., Huang Y., Cheng Z., Zhu G., et al. FGF1(DeltaHBS) prevents diabetic cardiomyopathy by maintaining mitochondrial homeostasis and reducing oxidative stress via AMPK/Nur77 suppression. Signal Transduct. Target. Ther. 2021;6:133. doi: 10.1038/s41392-021-00542-2. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib48] 48.Zheng Y., Xu Y., Ji L., San W., Shen D., Zhou Q., Meng G., Shi J., Chen Y. Roles of distinct nuclear receptors in diabetic cardiomyopathy. Front. Pharmacol. 2024;15 doi: 10.3389/fphar.2024.1423124. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib49] 49.Medzikovic L., de Vries C.J.M., de Waard V. NR4A nuclear receptors in cardiac remodeling and neurohormonal regulation. Trends Cardiovasc. Med. 2019;29:429–437. doi: 10.1016/j.tcm.2018.11.015. [DOI] [PubMed] [Google Scholar]

[bib50] 50.Song S., Zhang X., Huang Z., Zeng L., Cai F., Wang T., Song Y., Guo J., Li M., Liu C., et al. Cardiac fibroblast Foxm1 deficiency prevents pressure overload-induced cardiac remodeling via the Usp10/MKK6-p38gamma MAPK axis. Cell Death Differ. 2026;33:765–778. doi: 10.1038/s41418-025-01609-1. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib51] 51.Song S., Zhang X., Huang Z., Zhao Y., Lu S., Zeng L., Cai F., Wang T., Pei Z., Weng X., et al. TEA domain transcription factor 1(TEAD1) induces cardiac fibroblasts cells remodeling through BRD4/Wnt4 pathway. Signal Transduct. Target. Ther. 2024;9:45. doi: 10.1038/s41392-023-01732-w. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib52] 52.Zhao Q., Shao T., Zhu Y., Zong G., Zhang J., Tang S., Lin Y., Ma H., Jiang Z., Xu Y., et al. An MRTF-A-ZEB1-IRF9 axis contributes to fibroblast-myofibroblast transition and renal fibrosis. Exp. Mol. Med. 2023;55:987–998. doi: 10.1038/s12276-023-00990-6. [DOI] [PMC free article] [PubMed] [Google Scholar]

PERMALINK

ZEB2 SUMOylation in Cardiac Fibroblast Drives Post-Infarction Cardiac Remodeling through CtBP1-Associated repression of Nr4a1

Yilin Wang

Shaopeng Cheng

Hongyu Chen

Zhihao Yang

Xiaoting Wu

Qiuyan Zong

Jianing Wang

Yi Jiang

Yanrong Yu

Hao Chen

Yunxing Xue

Qianwen Zhao

Jie Yang

Dongjin Wang

Summary

Graphical abstract

Highlights

Introduction

Results

ZEB2 is upregulated in ischemic fibrotic hearts and cardiac fibroblasts post-MI

Figure 1.

Myofibroblast-specific knockout of ZEB2 attenuates adverse remodeling and cardiac dysfunction post-MI

Figure 2.

SUMO1 modification enhances ZEB2 stability via conserved lysine residues

Figure 3.

SENP1 is the primary DeSUMOylase regulating ZEB2 deSUMOylation

Figure 4.

ZEB2 SUMOylation is essential for TGF-β1-induced myofibroblast transformation in vitro

Figure 5.

SUMOylation-deficient ZEB2 attenuates adverse remodeling and cardiac dysfunction post-MI

Figure 6.

SUMOylation-deficient ZEB2 impaired CtBP1 recruitment, thereby increasing Nr4a1 expression and PI3K/AKT-mTOR pathway inhibition

Figure 7.

Figure 8.

Nr4a1 knockdown abolishes the cardioprotective effects of SUMOylation-deficient ZEB2-4KR in post-MI hearts

Figure 9.

Discussion

Limitations of the study

Resource availability

Lead contact

Materials availability

Data and code availability

Acknowledgments

Author contributions

Declaration of interests

STAR★Methods

Key resources table

Experimental model and study participant details

Human samples

Animals

Primary cells

Cell lines

Method details

Construction of recombinant adeno-associated virus and injection

Generation of knockout of myofibroblast mice

Generation of ZEB2 knockout mice

Myocardial infarction model

ISO injection models

Echocardiography analysis

Masson trichrome staining

Hematoxylin and Eosin (HE) staining

Isolation and culture of adult mouse cardiac fibroblasts

Cell culture

Plasmids transient transfection

Adenovirus infection

Lentivirus transfection

siRNA transfection Interference

Collagen gel contraction assay

Immunofluorescence staining

Real-time quantification PCR

Western blotting

Co-immunoprecipitation mass spectrometry analysis

RNA sequencing (RNA-seq) and data analysis

CUT&Tag-sequencing and data processing

Chromatin immunoprecipitation (ChIP) assays and sequencing

Luciferase assays

Quantification and statistical analysis

Footnotes