MLL4 protects cardiomyocytes against ischemia-reperfusion injury through STAT3-mediated mitochondrial function

Abstract

Myocardial ischemia-reperfusion injury (MIRI) is an inevitable pathophysiological response during the revascularization process following myocardial ischemia. Despite its clinical significance, effective targeted therapies for MIRI remain an unmet medical need. Mixed-lineage leukemia 4 (MLL4), a member of the SET family of histone methyltransferases, exhibits particular methyltransferase action toward histone H3 lysine 4 (H3K4). This study establishes a protective role for MLL4 in MIRI pathogenesis. Utilizing cardiomyocyte-specific Mll4 knockout mice and an in vivo ischemia-reperfusion (I/R) model induced by left anterior descending coronary artery ligation, we observed significant upregulation of MLL4 expression in cardiac tissue following I/R. Genetic ablation of Mll4 in cardiomyocytes markedly exacerbated both acute and chronic phases of MIRI. In vitro, Mll4 knockdown in neonatal rat cardiomyocytes (NRCMs) amplified mitochondrial dysfunction and apoptosis under hypoxia/reoxygenation (H/R) conditions. Integrated analysis of Cleavage Under Targets and Tagmentation sequencing (CUT&Tag-seq) and RNA sequencing (RNA-seq) revealed that Mll4 deficiency induces a pronounced reduction in H3K4 monomethylation (H3K4me1) and histone H3 lysine 27 acetylation (H3K27ac) enrichment at the Stat3 genomic locus. Mechanistically, MLL4 functions as a transcriptional activator of Stat3 by depositing H3K4me1 and H3K27ac, thereby facilitating STAT3 transcription. This regulatory cascade ultimately governs STAT3-dependent mitochondrial homeostasis. Collectively, these findings identify MLL4 as a critical epigenetic regulator of MIRI and suggest its therapeutic targeting may offer a promising strategy for mitigating reperfusion injury.

Graphical abstract

Highlights

• •MLL4 is upregulated during MIRI and NRCMs treated with H/R.
• •Genetic ablation of Mll4 in cardiomyocytes markedly exacerbated both acute and chronic phases of MIRI by amplifying mitochondrial dysfunction and apoptosis.
• •Stat3 inhibition abrogates MLL4 overexpression induced protective effect.
• •MLL4 functions as a transcriptional activator of Stat3 by depositing H3K4me1 and H3K27ac, thereby facilitating STAT3 transcription.
• •MLL4, as a critical epigenetic regulator of MIRI, establishes a protective role in MIRI pathogenesis. and its therapeutic targeting may offer a promising strategy for mitigating reperfusion injury.

1. Introduction

Coronary artery disease is one of the leading causes of morbidity and mortality worldwide [1,2]. Timely reperfusion of the influenced tissue can decrease myocardial infarction size; conversely, the reperfusion process can also cause irreversible damage to the myocardium [3]. MIRI is an inevitable pathophysiological phenomenon during revascularization following myocardial infarction, significantly impacting patient prognosis [4]. Comprehension of the mechanisms and developing treatment strategies for MIRI, its complex pathophysiological processes remain incompletely elucidated, and effective clinical interventions are still lacking. To enhance cardiac function and prognosis, a detailed understanding of the MIRI mechanisms is required.

Epigenetics refers to mechanisms influencing gene expression through DNA methylation, histone modifications, and non-coding RNA regulation, without altering the DNA sequence itself [5]. In recent years, epigenetics has been increasingly applied to cardiovascular disease research, highlighting its significant roles in cardiovascular development, disease progression, and therapeutic interventions [6,7]. MIRI arises from a series of complex factors, involving mechanisms such as oxidative stress [8], excessive reactive oxygen species (ROS) generation [9], calcium overload [10], inflammation, and mitochondrial dysfunction [11]. Mitochondrial dysfunction represents a central event in MIRI. Recent studies have revealed that cardiomyocyte specific Piezo1 deficiency alleviates ischemia-reperfusion injury by maintaining mitochondrial homeostasis [12], while NRF3 exacerbates cardiac dysfunction by increasing mitochondrial reactive oxygen species generation and promoting cardiomyocyte apoptosis [13]. Therefore, therapeutic strategies targeting signaling axes involved in mitochondrial function may offer novel approaches for treating MIRI. Reperfusion induces multiple forms of cardiomyocyte death in MIRI. Beyond the classical paradigms of apoptosis, autophagy, and necrosis, emerging evidence has highlighted the critical role of necroptosis in the pathogenesis of MIRI [14]. Zhang et al. demonstrated that plasma RIPK3 levels are positively correlated with the risk of major adverse cardiovascular events in patients with acute myocardial infarction undergoing percutaneous coronary intervention. Extracellular RIPK3 functions as a damage-associated molecular pattern (DAMP) that aggravates MIRI by mediating necroptotic signaling [15]. In a separate study, Zhao et al. confirmed that LGR6 exerts protective effects against MIRI by suppressing necroptosis [16]. Critically, these underlying mechanisms are closely linked to dynamic alterations in gene expression. In eukaryotic cells, gene transcription is influenced by both regulatory elements, including promoters and enhancers, and the chromatin's functional state that encompasses these genetic elements [17,18]. To date, research on the epigenome in myocardial ischemia-reperfusion injury has been limited, primarily focusing on DNA methylation, histone acetylation, and methylation [7].

Histone modification, a type of post-translational modification, is significant in epigenetic regulation and serves as a core mechanism for regulating gene expression. Among these modifiers, MLL4 has a significant function, particularly in the methylation of the histone H3K4 site [19,20]. Previous research has shown that MLL4 is crucial for regulating gene expression associated with cardiac development and is closely related to heart development [21,22]. MLL4 not only acts as an essential regulator of gene expression during development but also likely contributes significantly to the maintenance of cardiac function and the pathogenesis of heart diseases [23]. Current research suggests that targeting histone-modifying enzymes for drug development has become an important epigenetic intervention strategy, exhibiting remarkable therapeutic potential in major diseases such as malignant tumors and neurodegenerative disorders [24]. Nevertheless, systematic and in-depth investigations into the biological functions of the histone methyltransferase MLL4 in adult myocardial tissue, as well as its precise mechanisms in the initiation and progression of cardiovascular diseases, remain limited and warrant further research.

In the current study, we found that MLL4 expression was upregulated during MIRI. By generating myocardial-specific Mll4 knockout (Mll4-cKO) mice, we demonstrated that MLL4 exerts a defensive part in MIRI, and further clarified that this protective effect is mediated through epigenetic regulatory mechanisms. Specifically, MLL4 regulates the transcriptional activity of Stat3 by modulating the levels of histone modifications H3K4me1 and H3K27ac at an intronic enhancer region of the Stat3 gene. These changes in Stat3 transcription subsequently alter mitochondrial membrane potential, thereby modulating cardiomyocyte apoptosis and ultimately influencing the event and movement of myocardial ischemia-reperfusion damage.

2. Materials and methods

2.1. Animals

All male C57BL/6 rodents, 8-to 10-week-old, were bought from GemPharmatech Co. Ltd., (Chengdu, China). Mll4^flox/flox rodents were crossed with mice carrying the cardiomyocyte-specific TNNT2 promoter to create transgenic mice harboring both LoxP destinations and the Tnnt2-(rtTa-tetO-Cre) component through half-breed breeding. At eight weeks of age, the rodents were managed with doxycycline at a concentration of 0.2 g/100 mL in drinking water for 2 weeks to initiate cardiomyocyte-specific Mll4 knockout. All mice were kept at the Test Creature Center of West China Clinic, Sichuan College (Chengdu, China). They were kept in a controlled habitat of 22 °C, humidity of 55% ± 5%, and a 12-h cycle of light and dark. The mice had free access to nourishment and water. Each exertion was taken to decrease the torment of creatures. Upon completion of the test, the mice were given anesthesia and euthanized by breaking their necks.

2.2. Animal model of I/R

Mice were used to build the I/R animal model [25]. Animals were immobilized to one mouse plate at 37 °C. Moreover, the precordial region was sterilized through the iodophor. The total rodents underwent inhalation anesthesia with 2% isoflurane previous to surgery. The left anterior descending coronary artery (LAD) was tied to a surgical slipknot by a 7-0 silk suture to create the ischemic mice model. Sham-operated mice were processed through the identical surgery tying no LAD. Electrocardiography was implemented to assess the excellent production of the ischemia model. Following half an hour, the slipknot for cardiac reperfusion was unfolded. Further, mice died at discrepant time points. Significantly, myocardial tissues were gathered for in-depth analyses.

2.3. Cell culture and H/R treatment

H9c2 cardiomyocyte cell lines were taken from Procell Life Science Innovation Co., Ltd. (Wuhan, China) and refined in 4.5 g/L high-glucose DMEM medium in a CO₂ hatchery (37 ^°C, 5% CO₂). When the cell thickness came to around 90%, entry or plating was performed. For H/R treatment, the H9c2 cells were refined for 8 h in glucose-deprived DMEM under hypoxic conditions employing a three-gas low oxygen consistent temperature hatchery (Esco Lifesciences Group, Singapore) with 5% CO₂, and 1.1% O₂. Reoxygenation was performed by refining the H9c2 cells in DMEM containing 10% FBS for 2 h, 6 h, 12 h, and 24 h in a hatchery with 5% CO₂ and 95% air.

2.4. NRCMs preparation and culture

Cardiomyocytes from 0-1-day-old neonatal rats were isolated using enzymatic digestion. These are the specific steps: The skin surface of the neonatal rats was thoroughly disinfected by spraying with 75% ethanol. The hearts were quickly excised and immediately placed in pre-cooled PBS. After washing away the blood, a small amount of trypsin was added, and the hearts were minced into 1 mm³ tissue fragments using curved scissors. A Pasteur pipette was used to rinse the scissors with trypsin, and the tissue fragments containing trypsin were moved to a 50 mL centrifuge tube. The tissue was then digested at room temperature on a shaker at the lowest speed for 3-4 h. After digestion, the enzyme solution was collected, and an equal volume of complete medium was introduced to terminate the digestion. The tissue fragments were moved to a new 50 mL centrifuge tube, and a digestion medium was added for further digestion, followed by gentle pipetting. The tube containing the tissue mixture was rotated at the lowest speed at room temperature for half an hour. The tissue fragments were gently pipetted repeatedly to release cells into the supernatant, which was then filtered through a cell strainer. The remaining tissue fragments were further digested with a digestion medium for 15 min and filtered through the cell strainer. All broken-down cells were spun in a centrifuge at 300 rpm for 5 min. Cardiomyocytes were separated using the differential adhesion method. For the transfection, we used NRCMs and introduced Stat3-siRNA (50 nM) or Nc-siRNA with Lipo3000 (Invitrogen, L3000015). The liquid the cells were growing in was changed 8 h after they were treated, and then the cells were kept in a complete liquid for another 48 h. Next, the cells were gathered for more experiments. For Mll4 knockout and overexpression tests, NRCMs and dCas9 were mixed and either MLL4-sgRNA or Nc-sgRNA lentivirus (MOI = 20). Transfection efficiency was observed under a fluorescence microscope 48 h post-transfection. The sequence of Stat3-siRNA is listed in Supplementary Table 1.

2.5. Plasmid construction and lentivirus packaging

Single-guide RNAs (sgRNAs) targeting the murine Mll4 gene and a non-targeting control sgRNA were designed and synthesized by Shanghai Genechem Co., Ltd. Following annealing, the resulting double-stranded DNA fragments were cloned into the BsmBI-linearized LV-sgRNA-MS2-P65-HSF1-EGFP vector (GV468, Genechem). The ligated product was transformed into competent cells, and positive clones were identified by colony PCR and verified by Sanger sequencing. The dCas9-VP64-mCherry vector was also obtained from Genechem. For lentivirus production, 293T cells were co-transfected with the transfer vector along with packaging plasmids Helper 1.0 and 2.0. After 48 h, the supernatant was collected, centrifuged at 1000 ×g for 10 min, filtered through a 0.45-μm membrane, and concentrated by ultracentrifugation at 25,000 rpm for 2 h. Viral titers were determined via serial dilution. The following Mll4-targeting sgRNA sequences are listed in Supplementary Table 3.

2.6. Echocardiography

3% isoflurane was used to anesthetize the mice. Once the rodents were fully anesthetized, the concentration of isoflurane was adjusted to 1.5% to maintain a stable anesthetic state. The mice were then positioned on their backs on an operating board, and the hair on the left side of the chest was removed. An appropriate amount of ultrasound coupling gel was applied to the left chest between the 3rd and 4th intercostal spaces to ensure optimal contact between the ultrasound probe and the heart. Relevant parameters reflecting cardiac function were recorded and analyzed.

2.7. Western blotting

For total protein extraction, the heart tissues were dissolved within the RIPA buffer comprised of a complete protease and phosphatase suppressor cocktail (New Cell &Molecular Biotech). Protein density was evaluated through the bicinchoninic acid (BCA) assay kit (23225, Thermo Fisher Scientific). After that, the proteins were segregated via precast SDS–PAGE gel (4-20%, YEASEN). Also, the bands were moved onto PVDF membranes (Merck) using electricity. The PVDF membranes were treated with a special blocking solution without protein for 20 min. Then, they were mixed with the main antibody and kept in the fridge overnight at 4 °C. You can find more details about the antibody in Supplementary Table S1. After performing three washes using TBST, the PVDF membranes were left with the secondary antibody for 1 h at room temperature. We checked all the protein bands using ECL buffer (32209, Thermo Fisher Scientific). We used ImageJ software to measure the results of Western blotting. This software is from the NIH in Bethesda, MD, USA, and can be found at http://rsb. infonihgov/nih-image/.

2.8. RT-qPCR

The total RNAs were obtained from heart tissues or cells through the Cell/Tissue Total RNA Kit (19221ES50, YEASEN). After synthesis of cDNA with Hifair® one-step RT-gDNA digestion SuperMix for qPCR Kit (11142ES60, YEASEN). Real-time PCR intensification was executed through the ChamQ Universal SYBR qPCR Master Mix (11201ES08, YEASEN) on the ABI QuantStudio6 Q6 Real-time PCR framework. The comparative expression of mRNA was computed through the ΔΔCt approach in light of standard methods. Supplementary Table S2 displays the gene-specific primer sequences.

2.9. Immunofluorescent staining for NRCMs and cardiac tissues

Cryosections and NRCMs were settled with 4% paraformaldehyde, made permeable with 0 5% Triton X-100, and blocked with 2% bovine serum egg whites. At that point, they were treated overnight at 4 °C with essential antibodies: WGA counter-acting agent (1:1000, Merck, L4895), p-STAT3 counter-acting agent (1:200, CST, #9145), and MLL4 counter-acting agent (1:200, Merck, ABE1867). The areas were cleaned in three washes using PBS for 5 min each, and after that, treated with the auxiliary antibodies for 1 h at room temperature. Fluoroshield histology mounting medium with DAPI was utilized to cover the slides, and all areas were inspected with a confocal magnifying instrument (ECLIPSE Ti, Nikon).

2.10. TUNEL staining

We checked for Myocardial tissue apoptosis using the TUNEL Apoptosis Detection Kit [26]. In short, the 5-μm-thick heart samples were rinsed in three washes with PBS for 5 min each and then placed in 4% formalin for 15 min at standard room temperature. The TUNEL solution was added, and the slides were put in an incubator at 37 °C for 1 h. The slides were cleaned in three washes with PBS for 5 min apiece, and then coverslips were put on using a DAPI solution. All parts were checked and photographed with a special microscope that uses light.

2.11. CUT&Tag-seq

CUT&Tag assay was performed to profile histone modification-associated DNA fragments in NRCMs. In this study, CUT&Tag was employed instead of ChIP-seq, primarily due to its advantages of requiring fewer cells, yielding low background signals, and offering a high signal-to-noise ratio, making it more suitable for epigenomic studies using limited quantities of cardiomyocytes. Library preparation was conducted using the Hyperactive Universal CUT&Tag Assay Kit for Illumina (Vazyme, TD903) following the manufacturer's protocol, as previously described [27]. The amplified libraries were quantified and quality-assessed, and subsequent sequencing was performed by Seqhealth Technology Co., Ltd. (Wuhan, China).

2.12. Chromatin immunoprecipitation sequencing (ChIP-seq) datasets and CUT&Tag-seq analysis

ChIP-seq datasets of H3K4me1 and H3K27ac (CRA008354) were searched from the Genome Sequence Archive (GSA) (https://ngdc.cncb.ac.cn/gsa/) [28]. The ChIP-seq and CUT&Tag-seq analysis methods are consistent. CUT&Tag data were processed as follows: Raw sequencing reads were adapter-trimmed using Trim Galore (v0.6.7). Cleaned reads were aligned to the mouse reference genome (mm10) using Bowtie2 (v2.4.4) with default parameters. The resulting alignments were sorted and filtered using SAMtools (v1.9). Peak calling was performed with MACS2 (v2.2.9.1) using the parameters -cutoff 0.01 -f BAMPE. For visualization, BAM files were converted to bigWig format using deepTools (v3.5.1). Differential peak analysis was conducted using the DiffBind R package (v3.10) with an FDR threshold of <0.05, and peak annotation was performed using ChIPseeker (v1.22.1). Gene Ontology (GO) enrichment analysis for differentially bound regions was carried out using clusterProfiler with a significance threshold of q-value <0.05.

2.13. RNA sequencing

RNA-sequencing data GSE46224 were downloaded from the GEO (https://www.ncbi.nlm.nih.gov/geo) [29]. The datas completed transcriptome profiling in human failing (NF) and ischemic cardiomyopathy (ICM) tissue samples (n = 8 for each bunch) using next-gene sequencing. After FPKM normalization, expression values of histone methyltransferase and demethylase genes were Z-score scaled. Group comparisons (NF vs. ICM) were performed using the Wilcoxon test in R, and results were visualized with the pheatmap package (v1.0.2). In this study, RNA sequencing was performed by Seqhealth Technology Co., Ltd. (Wuhan, China). Total RNA was extracted from NRCMs and used for library preparation, followed by sequencing on an Illumina NovaSeq 6000 platform. Raw reads were processed with Trimmomatic (v0.36) to obtain high-quality clean data, which were then aligned to the reference genome using STAR (v2.5.3a). Gene-level read counts were generated with featureCounts (Subread v1.5.1). Gene expression levels were quantified as FPKM. Differential expression analysis was conducted using the edgeR package (v3.12.1), with significance defined as |fold change| ≥ 2 and P < 0.05. For cluster analysis, FPKM-normalized data were processed using the visCluster function from the ClusterGVis package (v0.1.0; https://github.com/junjunlab/ClusterGVis) to define three distinct expression patterns.

2.14. Statistical analysis

Data from the mouse and cell model were expressed as mean ± SD. An Independent-sample t-test (two-tailed) was used for comparison between two groups. One-way ANOVA followed by Tukey post hoc test was performed to compare multiple groups. A value of P < 0.05 was considered to be statistically different. The investigation was done utilizing GraphPad Prism 8.0 Software.

3. Results

3.1. MLL4 is upregulated during MIRI and NRCMs treated with H/R

To investigate the potential role of histone methyltransferases in MIRI, we first analyzed publicly available data from the dataset CRA008354 and GSE46224 [29] [28]. Our analysis of histone modifications in cardiac tissue from mice with MIRI revealed significantly elevated levels of H3K4me1 and H3K27ac across the genome (Fig. 1A). Furthermore, marked redistribution of these histone modifications was observed on the genome following model induction (Fig. 1B). Next, we compared the expression levels of histone methyltransferase-related genes in cardiac tissues from ICM patients and non-failing (NF) control individuals. Among these genes, KMT2D (MLL4) mRNA was notably upregulated in myocardial tissues from ICM patients (Fig. 1C). As a histone methyltransferase, MLL4 is known to regulate methylation at histone H3K4, a modification critical for chromatin structure and gene expression regulation [20]. H3K4me1, a hallmark epigenetic marker, is generally associated with active enhancers and open chromatin regions. Based on the methyltransferase activity of MLL4, we hypothesized that altered H3K4me1 levels in cardiac tissues may correlate with MIRI.

Fig. 1.

MLL4 is upregulated during MIRI and NRCMs treated with H/R. A. The heatmap showing the normal H3K4me1 and H3K27ac flag thickness relative to translation begin destinations (TSS) with ±3 kb flanking locales gotten from the CUT&Tag in the heart of MIRI mice. B. The genomic dispersion of H3K4me1 and H3K27ac alteration crests over chromosomes. C. Heatmap illustrating relative gene expression for histone methylation modifiers, comparing non-failing (NF) and ischemic cardiomyopathy (ICM) tissue samples ((n = 3 for every group). D. Representative plots of immunohistochemical recoloring of heart tissue from MIRI and pretense mice, scale bar = 100 μm, (n = 6 for each group). E. Representative immunoblotting pictures appearing MLL4 protein levels within the heart of MIRI mice (n = 6 for every group) and NRCMs treated with H/R (n = 6 for every group). Fand G. Measurement of E. H. Measurement of D. Factual contrasts were decided by One Independent-sample t-test (two-tailed), *P < 0.05 and ***P < 0.001.

To better comprehend the role of MLL4 in animals, we caused MIRI in adult C57BL/6 J mice by tying off a blood vessel called LAD. The amount of MLL4 protein was much higher in heart tissues from the MIRI model bunch than in the sham-operated group (Fig. 1E and F). Immunohistochemical staining of paraffin-embedded heart tissue sections also revealed a markedly increased proportion of MLL4-positive cells in MIRI mice compared with the sham group (Fig. 1D and H). Furthermore, in NRCMs treated with H/R, the protein expression levels of MLL4 were significantly upregulated (Fig. 1E and G), corroborating the in vivo findings. These findings collectively imply that MLL4 is likely key to MIRI.

3.2. Cardiac-specific Mll4 knockout promotes acute MIRI in mice

To examine the function of MLL4 in MIRI, we reconstructed cardiac-specific Mll4 knockout mice using the Tnnt2-rtTATetO-Cre system, as previously described (Fig. S1) [30]. This transgenic approach significantly decreased cardiac Mll4 expression at both the mRNA (Fig. S2B) and protein levels (Fig. S2A and 2C). Eminently, cardiac-specific Mll4 knockout did not initiate embryonic lethality or extreme postnatal development retardation (Fig. S3A–H), enabling us to survey the effect of MLL4 on cardiac function directly. The experimental design and grouping are illustrated in Fig. 2A. Following MIRI, Mll4-cKO mice exhibited significantly higher serum levels of phosphocreatine kinase (CK), CK-MB, and LDH compared to WT controls (Fig. 2B–D). Even though the region at risk/left ventricle ratio was comparable between WT and Mll4-cKO mice within the MIRI model, the infarct measure was particularly bigger in Mll4-cKO mice (Fig. 2E–G). Moreover, Mll4-cKO mice showed a noteworthy increment in TUNEL-positive cells taking after MIRI (Fig. 2H and I). Consistent with these discoveries, the protein levels of Cleaved-caspase3 were further elevated in Mll4-cKO mice after MIRI compared to WT controls (Fig. 2J and K). These discoveries collectively show that MLL4 plays a defensive part in MIRI. The deletion of MLL4 exacerbates myocardial injury, as evidenced by the expansion of infarct size, elevated myocardial injury markers, and increased apoptosis, which indicate that MLL4 may serve as a critical regulator in the repair and functional recovery of cardiac injury, potentially through modulating the injury response of cardiomyocytes.

Fig. 2.

Mll4 deletion in adult cardiomyocytes accelerates acute MIRI. A. The inquiry about flowchart. (1-7) B. Serum levels of phosphocreatine kinase (CK), C, CK-MB, and D, LDH in MIRI and pretense mice (n = 4 for each group). E-G. Representative images of Evans blue/TTC (triphenyl-2H-tetrazolium chloride)-staining and averaged infarct size in Mll4 transgenic (KO) mice after MIRI, (n = 4 for each group). H and I. Apoptosis of cardiomyocytes within the border zone by TUNEL recoloring in wild sorts and Mll4-cKO mice after MIRI, scale bar = 100 μm, (n = 8 for each group). J. The protein levels of Cleaved-caspase3 in Mll4-cKO mice after MIRI, (n = 6 for every group). K. Evaluation of (1-2) J. Data are shown as the mean ± SD. Statistical differences were determined by One-way ANOVA with Tukey test. Column bars represent mean, error bars indicate the SD. (B-D, I, K), One Independent-sample t-test (two-tailed) (F and G). AAR demonstrates region at hazard; and LV, left ventricle. *P < 0.05, **P < 0.01 and ***P < 0.001.

3.3. MLL4 deficiency exacerbates cardiac dysfunction and mitochondrial damage

To further determine whether MLL4 remains influential in the chronic phase of MIRI, we performed echocardiographic assessments at 14 days and 28 days post-reperfusion to evaluate cardiac function and structural remodeling in mice. At 14 days post-reperfusion, echocardiographic results revealed no statistically notable variations in ejection fraction (EF%) and fractional shortening (FS%) between the WT and Mll4-cKO mice (Fig. S4A–I). However, a downward trend in EF% and FS% was noted in the Mll4-cKO mice, suggesting that despite the absence of significant cardiac dysfunction in the early phase. Mll4-cKO mice might have already begun to affect cardiac function, potentially leading to progressive functional decline over time and exhibited significant alterations in cardiac volume, ventricular dimensions, and wall thickness-related parameters after 28 days of reperfusion (Fig. 3A). Specifically, LVEdV, LVEsV and LVIDs were markedly increased (Fig. 3D, E and 3I), while stroke volume (SV) (Fig. 3F), EF%, and FS% were reduced (Fig. 3B and C). Additionally, WT mice at 28 days post-reperfusion exhibited significant diastolic dysfunction (Fig. 3A), as indicated by an increased E/A ratio. Compared with WT mice, Mll4-cKO mice showed a further elevation in the E/A ratio, indicating worsened diastolic dysfunction (Fig. 3G and H). Notably, histological analysis established that the fibrotic area in the Mll4-cKO group was significantly larger than that in the WT mice, increasing from 7% to 15% (Fig. 3J and K). In addition to that, we overviewed mitochondrial ultrastructure utilizing transmission electron microscopy and found that MLL4 deficiency brought nearly erratic mitochondrial morphology and thickness compared with WT mice, exhibiting shortened mitochondrial length and reduced cross-sectional area (Fig. 3L and M). Immunohistochemical analysis further revealed that following MIRI induction, hearts from Mll4-cKO mice exhibited a significant reduction of approximately 20% in the expression of TOMM20, a marker associated with mitochondrial damage, compared to the WT mice.

Fig. 3.

MLL4 deficiency aggravates cardiac dysfunction, myocardial fibrosis and mitochondrial damage. A-I. Cardiac work was assessed by echocardiography Mll4-cKO mice at 28 days post-reperfusion (n = 6 for each group). J. Masson trichrome dyeing of WT and Mll4-cKO hearts 28 days post-surgery, scale bar = 1000 μm. K. Quantification of fibrotic area (%) in mice hearts treated as shown (n = 6 for each group). L. Representative transmission electron microscopy pictures of WT and Mll4-cKO hearts, scale bar = 1 μm. M. Measurement of mitochondria-related variables from L (n = 6 for each group). Mitochondrial region alludes to the proportion of mitochondrial range to picture area. N. Representative immunohistochemical staining images of TOMM20 in WT and Mll4-cKO mice, scale bar = 100 μm, (n = 6 for each group). O. Quantification of the TOMM20-positive area shown in panel N. Data are shown as the mean ± SD. Statistical differences were determined by One-way ANOVA with Tukey test, *P < 0.05, **P < 0.01 and ***P < 0.001. Column bars represent mean, error bars indicate the SD.

3.4. Inhibition of MLL4 aggravates the apoptosis of cardiomyocytes in vitro

To further validate our in vivo findings and investigate the role of MLL4 in cardiomyocytes, we conducted loss-of-function assays in NRCMs and H9c2 cells. Oxidative stress plays a critical role in MIRI. Upon reperfusion, oxygen supply is restored, leading to a significant production of ROS by the mitochondrial electron transport chain. These ROS contribute to cellular damage by inducing mitochondrial dysfunction, inflammatory responses, and apoptosis [1,31]. Under normal culture conditions, cells generate only a small amount of ROS, exhibiting a weak green fluorescence signal. However, following H/R treatment, the ROS-positive signal in H9c2 cells was significantly increased. Notably, ROS production was further elevated upon Mll4 knockout (Fig. 4A and B). Flow cytometry analysis revealed consistent findings, where the fluorescence threshold in the FITC channel was significantly higher in Mll4-KO cardiomyocytes than in control cells after H/R treatment (Fig. 4D). Moreover, Mll4-KO cells exhibited a substantial increase in malondialdehyde (MDA) release, approximately twice the level observed in control cells (Fig. 4F). Apoptosis levels were also significantly elevated upon Mll4 knockout, reaching up to 40% (Fig. 4J and M). Western blotting analysis of the anti-apoptotic protein BCL-XL revealed that under H/R conditions, BCL-XL protein expression was further downregulated upon Mll4 knockout (Fig. 4K and L). To assess mitochondrial membrane potential (MMP) changes, JC-1 staining was performed, which is a membrane potential-dependent fluorescent dye used to monitor changes in MMP. Mitochondria with low membrane potential exhibit green fluorescence, whereas those with high membrane potential display red fluorescence. A higher ratio of green to red fluorescence indicates more impaired mitochondrial function. After H/R treatment, control cells exhibited a reduction in red fluorescence and an increase in green fluorescence, indicating a decline in MMP and partial mitochondrial dysfunction. More strikingly, in Mll4-KO cardiomyocytes, red fluorescence was almost completely absent, with nearly all fluorescence shifting to green, suggesting severe mitochondrial membrane potential loss and significant functional impairment (Fig. 4C and G). In addition, immunofluorescence staining revealed a pronounced reduction in TOMM20 expression, along with dispersed and fragmented signals, in Mll4-KO cardiomyocytes following H/R treatment, suggestive of aggravated mitochondrial injury (Fig. 4H and I).

Fig. 4.

MLL4 deficiency aggravated oxidative stress and mitochondrial damage, and promoted apoptosis in cardiomyocytes induced by H/R. A. Representative pictures of ROS in cardiomyocytes watched by confocal microscopy, scale bar = 100 μm. B. Evaluation of A (n = 6 for every group). C. Representative pictures of JC-1 staining in cardiomyocytes treated with H/R, scale bar = 100 μm. D. Detection of ROS release in cardiomyocytes by flow cytometry. E. The cell viability in H/R treated cardiomyocytes (n = 6 for each group). F. The discharge level of Malondialdehyde (MDA) in H/R treated cardiomyocytes (n = 6 for every group). G. Measurement of mitochondria-related proportion of green/red fluorescence (n = 9 for each group). H. Representative immunofluorescence staining images for TOMM20 in cardiomyocytes, scale bar = 10 μm. I. Corresponding quantification for TOMM20, (n = 6 for every group). J, M. Representative pictures of cardiomyocyte apoptosis identified by stream cytometry (n = 6 for every group). K. Representative immunoblotting pictures showing the BCL-XL protein levels in cardiomyocytes with MLL4 inhibition after H/R treatment. L. Quantification of K (n = 6 for every group). Data are shown as the mean ± SD. Statistical differences were determined by One-way ANOVA with Tukey test, *P < 0.05, **P < 0.01 and ***P < 0.001. Column bars represent mean, error bars indicate the SD.

3.5. Inhibition of MLL4 reshapes the chromatin landscape and gene expression profiles in NRCMs

MLL4, affiliated with the histone-lysine N-methyltransferase family, is closely associated with the activation of gene transcription. To investigate the mechanisms through which MLL4 deletion aggravated MIRI, we performed comprehensive RNA-seq analysis and CUT&Tag-seq for H3K4me1 and H3K27ac in NRCMs. Pie chart analysis uncovered contrasts within the distribution of H3K4me1-enriched crests across genomic functional regions among different treatment groups (Fig. 5A). Comparative analysis of differential peaks between the H/R control bunch and the H/R Mll4 knockdown group identified 324 upregulated peaks and 1340 downregulated peaks, suggesting that Mll4 deficiency led to significant differences in H3K4me1 modification levels at the same chromatin regions across samples (Fig. 5C). Hierarchical clustering analysis further confirmed good reproducibility among samples (Fig. 5D). Gene Ontology (GO) enrichment examination of genes related with differential crests revealed that these genes were annotated to a few critical biological forms related to myocardial ischemia-reperfusion injury and transcriptional regulation, including apoptotic pathways, cell cycle regulation, negative regulation of phosphate metabolism, and RNA polymerase II transcription regulator complexes (Fig. 5E). These results suggest that in the H/R-induced cardiomyocyte injury model, Mll4 deficiency may lead to a reduction in H3K4me1 modification levels at specific chromatin regions, affecting transcription factor activity and the interaction between RNA polymerase II complexes and DNA, ultimately suppressing gene transcription activity. RNA-seq analysis demonstrated that Mll4 knockdown altered gene transcription levels in NRCMs (Fig. S5A–E). Additionally, KEGG pathway analysis and GO functional enrichment analysis showed that the genes that were expressed differently were significantly enriched in pathways and biological functions related to inflammation regulation, cell survival, and ion transport (Fig. S5F). In addition, we performed GSEA comparing the H/R control and H/R Mll4-KO groups. As shown in Figure S6 A-E, compared with the H/R control group, the H/R Mll4-KO group exhibited significant upregulation of pathways related to reactive oxygen species and oxidative stress, such as the regulation of superoxide anion generation and apoptotic signaling pathways. Conversely, pathways associated with the superoxide metabolic process and positive regulation of mitochondrial function were significantly downregulated. These findings are consistent with our in vivo and in vitro experimental results, indicating that Mll4 deletion perturbs oxidative stress related signaling pathways. Therefore, it is plausible that among these significantly altered genes, key downstream targets mediated by Mll4 may play a protective effect in MIRI. Analysis of the top ten motifs associated with H3K4me1-enriched peaks across sample groups identified corresponding transcription factors. Notably, STAT3 ranked second in motif analysis, is essential in MIRI (Fig. 5B). Furthermore, KEGG pathway analysis of RNA-seq data revealed enrichment of the JAK/STAT signaling pathway (Fig. S5F), suggesting that MLL4 may exert its protective effects in myocardial ischemia-reperfusion injury by regulating Stat3 transcription.

Fig. 5.

Inhibition of MLL4 reshapes the chromatin landscape and gene expression profiles in NRCMs. A. The pie chart outlining the genomic dispersion of H3K4me1 alteration crests over chromosomes (top), and a heatmap showing the normal H3K4me1 flag thickness relative to translation begin destinations (TSS) with ±3 kb flanking locales (bottom) gotten from the CUT&Tag measures for H3K4me1 in control H/R and Mll4-KO H/R treated NRCMs. B. Top translation figure themes distinguished by HOMER from regions with differential H3K4me1 adjustment within the Mll4-KO H/R group compared to the control H/R group in NRCMs. C. The number of related genes clarified in regions (crests), differentially adjusted for H3K4me1 levels utilizing the HOMER suite device, annotates Crests. D. Relationship heatmap of histone marks over all tests. Each group had three biological replicates. E. GO examination of genes with differentially H3K4me1 modified.

3.6. MLL4 regulates Stat3 transcription in cardiomyocytes by binding to its intronic enhancer

Previous studies have shown that STAT3 confers cardioprotective effects in MIRI through multiple mechanisms. Phosphorylated STAT3 is closely associated with the mechanisms governing cardiomyocyte apoptosis, oxidative stress levels, and the maintenance of normal mitochondrial function [4,32]. Given the pivotal role of STAT3 in MIRI, we hypothesized in our subsequent research that MLL4 may regulate mitochondrial function and apoptosis in cardiomyocytes during MIRI by modulating Stat3 transcription. To investigate whether MLL4 is directly localized to genomic regions involved in H3K4me1 level maintenance, we performed CUT&Tag in NRCMs using an antibody against MLL4 to define its genome-wide binding sites. As expected, MLL4 binding was detected at the Stat3 gene locus, and this binding was markedly reduced following MLL4 knockdown, indicating that Stat3 is a direct transcriptional target of MLL4 (Fig. 6A Top). We then performed an analysis of the mapped readsfrom the CUT&Tag-seq data for Stat3 using the IGV. MLL4 deficiency caused a significant decrease in H3K4me1 and H3K27ac occupancy at the Stat3 gene locus (Fig. 6A). In addition, to further validate the transcriptional regulation of Stat3 by MLL4, we performed enhancer luciferase reporter assay. The results showed that the En-2 fragment significantly enhanced luciferase reporter gene expression compared to the control, indicating that this specific intronic region within the Stat3 gene possesses enhancer activity capable of promoting transcription (Fig. S7A–D). This finding was consistent with the marked decrease in STAT3 protein levels observed in Mll4-cKO mice and (Fig. 6B and C). To avoid potential confounding effects from non-cardiomyocyte components in whole heart tissue homogenates, we analyzed a published scRNA-seq dataset (GSE227238). The results revealed that the expression levels of both Mll4 and Stat3 were significantly upregulated in cardiomyocytes following reperfusion (Fig. S8A–C). Furthermore, immunofluorescence staining of STAT3 with the cardiomyocyte marker cTnI on frozen heart sections from MIRI mice showed markedly increased STAT3 fluorescence intensity in cardiomyocytes compared to the sham group (Fig. S8D). We further examined the nuclear localization of STAT3 in heart tissue. In comparison to the sham-operated group, WT mice subjected to MIRI showed a marked increase in p-STAT3 red fluorescence signal, indicating enhanced nuclear translocation of STAT3 (Fig. 6E). In contrast, Mll4-cKO mice exhibited a significant reduction in p-STAT3 nuclear fluorescence following MIRI, suggesting impaired nuclear translocation of STAT3. Furthermore, to clarify the temporal relationship between MLL4 and Stat3 activation, we measured the protein expression levels of total STAT3 and p-STAT3 in mouse heart tissues at different time points after MIRI, both total STAT3 and p-STAT3 protein levels showed an increasing trend starting from 3 h after reperfusion; however, the ratio of p-STAT3 to total STAT3 at each time point was not statistically different (Fig. S9A–C). We also examined the mRNA levels of Mll4 and Stat3 in NRCMs at various time points after H/R treatment. As illustrated in Fig. S9D and E, Mll4 mRNA was significantly increased as early as 1.5 h after reoxygenation and peaked at 12 h, Stat3 mRNA also showed an increasing trend at 1.5 h, but the magnitude of upregulation was markedly lower than that of Mll4. These data indicate that MIRI stress rapidly induces Mll4 transcription, and that its transcriptional upregulation occurs earlier than or in parallel with that of Stat3. Collectively, the analysis implies that the methyltransferase MLL4 is crucial for the regulation of Stat3 gene expression during MIRI, primarily through its histone modifications of H3K4me1 and H3K27ac.

Fig. 6.

MLL4 acts as a transcriptional activator of Stat3 through H3K4me1 and H3K27ac. A. Representative IGV browser tracks appearing MLL4 authoritative crests at the Stat3 locus in NRCMs (top). Tracks showing H3K4me1 and H3K27ac (middle) improvement at the Stat3 locus in Mll4-knockdown and control NRCMs taken after H/R treatment, and representative IGV browser tracks of Stat3 mRNA reads in Mll4-knockdown and Control NRCMs upon H/R treatment (bottom). Different colored boxes mark the candidate intronic fragments (En1-En4). Each track speaks to a combined flag from three organic duplicates. B. Representative immunoblotting pictures of MLL4, STAT3, p-STAT3 (Try705) and STAT3(Ser727) expression levels in Mll4 transgenic (KO) mice after MIRI. C, D Evaluation of B, (n = 6 for each group). E. Mice with or without MLL4 knockdown were treated with MIRI and arranged for immunofluorescence recoloring of p-STAT3 and WGA in heart tissues, scale bar = 50 μm. F, G. RT-qPCR investigation of Mll4 and Stat3 mRNA expression in NRCMs treated with H/R (n = 6, n = 8 for each group). Data are shown as the mean ± SD. Statistical differences were determined by One-way ANOVA with Tukey test, *P < 0.05, **P < 0.01 and ***P < 0.001. Column bars represent mean, error bars indicate the SD.

Additionally, to validate whether Mll4 deficiency suppresses Stat3 transcriptional activity and affects STAT3 protein expression, we utilized CRISPR/Cas9 single-vector lentiviral infection to knockdown Mll4 in NRCMs (Fig. S10A and E). Consistent with the in vivo results, MLL4 protein levels were significantly reduced, accompanied by a marked downregulation of histone modification-related markers, including H3K4me1 and H3K27ac (Figs. S11A, E, F). Notably, both total STAT3 and phosphorylated STAT3 protein levels were decreased following Mll4 knockdown (Fig. S11B–D), and RT-qPCR analysis revealed a similar reduction in Stat3 mRNA levels (Fig. 6F and G). These findings indicate that, in an H/R-induced cellular model, Mll4 deletion suppresses Stat3 transcriptional activity, thereby impairing the translation and phosphorylation of STAT3 protein.

3.7. MLL4 overexpression enhances STAT3 expression and subsequently attenuates cardiomyocyte apoptosis

To investigate whether MLL4 exerts a protective effect on cardiomyocytes subjected to H/R treatment in vitro, we constructed a transcriptional activation system based on the CRISPR/dCas9 complex (synergistic activation mediator, SAM) [33]. Lentiviruses containing sgRNA and dCas9 were co-transfected into NRCMs, and successful co-expression of EGFP (green) and mCherry (red) was confirmed by fluorescence microscopy (Fig. 7A). Western blotting analysis demonstrated that the dual-vector lentiviral SAM system effectively overexpressed MLL4 protein in NRCMs. Notably, MLL4 overexpression led to an up-regulation in the expression levels of STAT3 (Fig. 7B–D), but a reduction in Cleaved-caspase3 in cardiomyocytes under H/R treatment (Fig. 7E and J), which was correlated with the enhanced expression of STAT3 promoted by MLL4 (Fig. 7G–I). Furthermore, flow cytometry results revealed that MLL4 overexpression may alleviate cardiomyocyte apoptosis by enhancing STAT3 expression (Fig. 7C and F). MitoTracker staining was used to visualize mitochondrial morphology. In WT NRCMs, mitochondria appeared elongated and rod-like, and MLL4 overexpression did not affect mitochondrial structure under basal conditions. Notably, under H/R conditions, mitochondria in NC-sgRNA NRCMs exacerbated significant morphological changes, leading to increased fragmentation, a significant reduction in mitochondrial length, and severe structural damage. Conversely, MLL4 overexpression markedly ameliorates mitochondrial impairment (Fig. 7K–L). These results collectively indicate that MLL4 overexpression attenuates H/R-induced mitochondrial structural damage and apoptosis in NRCMs.

Fig. 7.

MLL4 overexpression promotes STAT3 expression and further alleviates cardiomyocytes apoptosis. A. Representative pictures of lentiviral transfection effectiveness in NRCMs were assessed by surveying EGFP and mCherry expression utilizing fluorescence microscopy. Cells showing the double inspiration for EGFP and mCherry were considered effectively transfected, scale bar = 100 μm. B. Representative immunoblotting pictures showing expanded expression of MLL4 in NRCMs through co-transfecting with sgRNAs and dCas9 lentivirus. C. Representative pictures of cardiomyocyte apoptosis identified by flow cytometry. D. Measurement of B (n = 6 for every group). E. Representative immunoblotting pictures showing the overexpression of MLL4, STAT3, p-STAT3, and Cleaved-caspase3 in NRCMs through co-transfecting with sgRNAs and dCas9 lentivirus. F. Measurement of C (n = 6 for every group). G-J. Measurement of E (n = 6 for every group). K. Representative images of MitoTracker staining in NRCMs using confocal microscopy, scale bar = 10 μm. L. Quantification of K (n = 8 for each group). Data are shown as the mean ± SD. Statistical differences were determined by One-way ANOVA with Tukey test, *P < 0.05, **P < 0.01 and ***P < 0.001. Column bars represent mean, error bars indicate the SD.

To verify that the increased apoptosis observed in Mll4-knockdown NRCMs under H/R treatment was due to the downregulation of STAT3, we treated the Mll4-knockdown cells with a STAT3 agonist. After treatment with the STAT3 agonist Colivelin, the phosphorylation level of STAT3 was significantly increased, while the expression of pro-apoptotic proteins BAX and Cleaved-caspase3 was reduced, indicating that STAT3 activation alleviates H/R-induced cardiomyocyte apoptosis (Fig. S12A–E). Flow cytometry analysis of cell apoptosis yielded consistent results, showing that the proportion of apoptotic cells was markedly reduced to approximately 13% after treatment with the STAT3 agonist Colivelin (Fig. S12F and G). These experimental findings demonstrate that the STAT3 agonist Colivelin can alleviate cardiomyocyte apoptosis induced by Mll4 deficiency.

3.8. STAT3 inhibition abrogates MLL4 overexpression induced protective effect

To validate that STAT3 is a direct downstream target of MLL4 involved in the regulation of MIRI, we conducted a “functional rescue experiment.” Specifically, we silence Stat3 in the context of MLL4 overexpression. Following siRNA treatment, Stat3 silencing markedly enhances H/R-induced apoptosis and mitochondrial injury in NRCMs (Fig. S13A–F). Overexpression of MLL4 markedly increased the protein levels of STAT3 while significantly decreasing levels of apoptotic proteins caused by H/R (Fig. 8A–C, Fig. 8I–L). Notably, this protective effect was substantially reversed by transfection with Stat3-siRNA. Stat3 knockdown significantly increased ROS production and MDA release in H/R treated NRCMs (Fig. 8D–F). Flow cytometry analysis further demonstrated that STAT3 silencing following MLL4 overexpression markedly promoted apoptosis in NRCMs (Fig. 8G and H). Furthermore, to determine whether Stat3 overexpression could rescue mitochondrial dysfunction and apoptosis induced by Mll4 deficiency in cardiomyocytes, we overexpressed Stat3 by plasmid transfection in Mll4-knockdown cardiomyocytes subjected to H/R injury. As shown in Fig. S14A–K, Stat3 overexpression successfully attenuated mitochondrial dysfunction and apoptosis in Mll4-knockdown cardiomyocytes. These results suggest that STAT3 acts as a direct downstream target of MLL4, mediating its defensive part against H/R-induced cardiomyocyte oxidative stress and mitochondrial function, thereby ultimately modulating H/R-induced apoptosis.

Fig. 8.

STAT3 inhibition abrogates MLL4 overexpression induced protective effect. A. Representative immunoblotting images of MLL4, STAT3 and p-STAT3 in NRCMs via co-transfecting with sgRNAs and dCas9 lentivirus and siStat3 in NRCMs treated as shown. B–C. Quantification of A. D and E. Representative pictures of ROS in cardiomyocytes watched by flow cytometry (n = 8 for every group). F. The discharge level of MDA in H/R treated cardiomyocytes (n = 8 for every group). G and H. Representative images of cardiomyocyte apoptosis detected by flow cytometry (n = 6 for every group). I. Representative immunoblotting images of apoptosis-related proteins in NRCMs treated as shown. J-L. Quantification of I, (n = 6 for every group). Data are shown as the mean ± SD. Statistical differences were determined by One-way ANOVA with Tukey test, *P < 0.05, **P < 0.01 and ***P < 0.001. Column bars represent mean, error bars indicate the SD.

4. Discussion

MIRI is a global health concern for which there is currently no effective treatment. Therefore, elucidating the underlying pathogenic mechanisms of MIRI is critical for improving cardiac function and prognosis in affected patients [3]. In this research, we found that the expression of the histone methyltransferase MLL4 was upregulated in both the myocardial tissue of patients with ICM and in the hearts of mice subjected to MIRI. Further investigation revealed that MLL4 regulates the transcriptional activity of Stat3 by modulating the levels of H3K4me1 and H3K27ac at an intronic enhancer region of the Stat3 gene. Moreover, we confirmed that both Mll4 deficiency and Stat3 inhibition led to mitochondrial dysfunction and apoptosis in cardiomyocytes. In contrast, MLL4 overexpression effectively attenuated cardiomyocyte apoptosis, suggesting a defense mechanism. These findings expand the present understanding of epigenetic mechanisms in MIRI progression and highlight the potential of reversing apoptosis as a novel therapeutic strategy.

Epigenetics, as a novel regulatory mechanism, has gained increasing attention for its role in various cardiac diseases [34]. Drugs targeting histone post-translational Modifications have shown promising therapeutic impacts in both animal models and clinical studies of cardiomyopathies [24,[35], [36], [37]]. Thus, a deeper understanding of how epigenetic regulation contributes to MIRI may open new avenues for drug development. MLL4, a key member of the histone methyltransferase family responsible for H3K4 methylation, has been implicated in the regulation of cardiac developmental genes [19,38,39]. However, whether and how MLL4 regulates MIRI remained unknown. In this study, we identified a previously undescribed role of MLL4 in the pathogenesis of MIRI in the adult heart. MIRI is a complex pathophysiological process that can be broadly divided into acute and chronic phases. The acute phase occurs within minutes to hours after reperfusion, characterized by rapid cell injury and inflammation, while the chronic phase unfolds over days to weeks or months, marked by tissue repair, fibrosis, and cardiac remodeling [9]. Injury during the acute phase, such as necrosis, apoptosis, and inflammation, lays the foundation for chronic fibrotic and functional remodeling. Excessive tissue repair, including fibroblast activation and extracellular matrix deposition, leads to fibrosis. Moreover, acute loss of cardiomyocytes and microvascular damage manifest as ventricular remodeling and aberrant angiogenesis during the chronic phase, ultimately impairing organ function [40,41]. Appropriately, we inspected the part of MLL4 in both stages of MIRI. Cardiomyocyte-specific deletion of Mll4 in grown-up mice brought about in serious mitochondrial dysfunction, excessive cardiac fibrosis, adverse remodeling, and impaired cardiac work under pressure. In MIRI, mitochondrial morphology and function are closely interrelated, collectively determining the fate of cardiomyocytes [42]. Healthy mitochondria exhibit an elongated rod-like shape with intact cristae structures, a morphology that underlies their core functions in efficient ATP production, maintenance of calcium homeostasis, and regulation of ROS. During reperfusion injury, sustained swelling and calcium overload contribute to increased mitochondrial outer membrane permeability [43,44]. This leads to swelling of the mitochondrial matrix, disruption or even loss of inner membrane cristae, and the release of pro-apoptotic factors such as cytochrome c, thereby initiating apoptosis. Furthermore, MIRI disrupts mitochondrial dynamics, upsetting the balance between fusion and fission and resulting in excessive fragmentation of the mitochondrial network into small, punctate structures [45]. This observation was consistent with our transmission electron microscopy findings, which revealed further shortening of mitochondrial profiles in cardiac tissues of Mll4-cKO mice. On the other hand, MLL4 overexpression in NRCMs secured cardiomyocytes from H/R-induced oxidative stress and apoptosis. The findings imply that MLL4 is vital for sustaining mitochondrial homeostasis and preventing apoptosis, thereby exerting a protective effect in MIRI.

Given the pivotal role of MLL4 in MIRI, we further explored its molecular mechanism of action by performing RNA-seq and CUT&Tag, especially focusing on the regulation of STAT3 in MLL4-driven cardiac protection. As a transcription factor and signal transducer, STAT3 responds to diverse pathological stressors and transmits signals from the extracellular environment and cytoplasm to the nucleus [7,11]. Canonical STAT3 activation occurs through tyrosine 705 phosphorylation via JAK recruitment or serine 727 phosphorylation. STAT3 is also involved in the SAFE pathway, where it interacts with other cardioprotective signaling pathways to maintain mitochondrial function [46]. Prior research indicates that STAT3 plays a critical anti-apoptotic role by upregulating genes such as Bcl-2, Bcl-xl, and Mcl-1 while inhibiting pro-apoptotic Bax expression. This regulation reduces MOMP and cytochrome c release, thereby suppressing mitochondrial-dependent apoptosis [47,48]. Specific analysis of cell type confirmed that the upregulation of MLL4 and STAT3 after MIRI is selective in cardiomyocytes, supporting the role of the MLL4-STAT3 axis in the pathogenesis of myocardial ischemia-reperfusion injury through the autonomous action of cardiomyocytes. Based on this evidence, we hypothesized that MLL4 promotes mitochondrial function and suppresses apoptosis during MIRI by epigenetically activating Stat3 transcription. CUT&Tag data confirmed that Mll4 deficiency reduced H3K4me1 and H3K27ac levels at Stat3 intronic enhancer, correlating with decreased Stat3 transcription. Increasing evidence indicates that histone methyltransferases regulate distinct target genes in a tissue and stress-dependent manner. Ortega-Molina et al. showed that in lymphoma, KMT2D indirectly affects STAT3 via SOCS3 [49]. Conversely, we demonstrate that in cardiomyocytes, MLL4 directly regulates Stat3 transcription by modulating H3K4me1 and H3K27ac. Thus, the cardioprotective role of MLL4 in MIRI is not contradictory to its indirect inhibitory effect via SOCS3 in tumors. Instead, these observations underscore its versatility as an epigenetic regulator. In addition, time-course analyses across cellular and animal models consistently demonstrated that MLL4 transcriptional activation is an early event in response to MIRI stress, occurring prior to STAT3 induction, whereas the increase in STAT3 protein is primarily attributable to elevated total protein levels. These findings position MLL4 as an early transcriptional initiator that likely coordinates downstream STAT3 expression during MIRI. Collectively, these findings indicate that MLL4 directly controls Stat3 transcription via epigenetic modifications, and that excessive mitochondrial dysfunction and apoptosis drive MIRI progression.

Furthermore, we utilized an MLL4 transcriptional activation lentiviral system in NRCMs under H/R conditions and observed that MLL4 overexpression enhanced STAT3 expression and conferred robust anti-apoptotic effects in cardiomyocytes. To further delineate the regulatory mechanism, we performed a rescue experiment in which Stat3 was silenced via siRNA in MLL4-overexpressing NRCMs following H/R treatment. The protective property of MLL4 overexpression against apoptosis was significantly reversed by Stat3 silencing, further validating that Stat3 is a direct downstream target of MLL4 and that MLL4 exerts its cardioprotective effects through STAT3.

This study does have some limitations. First, epigenetic regulation is a complex and dynamic process, and STAT3 may interact with additional transcriptional complexes not captured in this study. Second, although our findings were validated in NRCMs and animal models, the role of MLL4 in human cardiac injury remains to be confirmed, necessitating further investigation.

In summary, we are the first to demonstrate that MLL4 contributes to MIRI through its epigenetic regulatory function, acting as a transcriptional activator of STAT3 via H3K4me1, thereby promoting Stat3 transcription and nuclear accumulation. This regulation maintains mitochondrial function, inhibits cardiomyocyte apoptosis, and confers protection in the setting of MIRI. As a critical epigenetic regulator, MLL4 primarily functions by activating enhancers to control gene expression and cell fate. Loss-of-function mutations in MLL4 are important drivers of various cancers and Kabuki syndrome [50,51]. Previous studies have reported that MLL4 participates in regulating the expression of genes essential for cardiac development and is closely associated with heart development and failure [19,23]. While prior research established the indispensability of MLL4 during embryonic heart development, our study extends the significance of this role to the injury repair process in the adult heart, demonstrating that MLL4 also serves as a key protective factor in adult cardiomyocytes against MIRI. In the future, monitoring the expression or activity of MLL4 in myocardial tissue may potentially serve as a prognostic biomarker for MIRI and heart failure, thereby facilitating clinical disease management. Furthermore, the development or screening of MLL4-specific small molecule agonists may represent a promising therapeutic direction with translational potential. This study provides a new understanding the function of MLL4 in MIRI, indicating it as a potential target for therapeutic intervention.

Funding

This study was funded by the National Natural Science Foundation of China (Grant No.82274647), the National Natural Science Foundation of China (Grant No.82500350), the Natural Science Foundation of Sichuan Province (2026NSFSC1725), and the Postdoctor Research Fund of West China Hospital, Sichuan University (2024HXBH105).

CRediT authorship contribution statement

Qiu-Yu Pang: Data curation, Formal analysis, Methodology, Writing – original draft. Xiang-Min Meng: Data curation, Methodology, Writing – review & editing. De-Yong Li: Data curation, Formal analysis. Lu You: Formal analysis, Methodology. Zhen-fang Zhou: Investigation, Software. Yu-Meng Li: Data curation, Methodology. Tian Deng: Data curation, Visualization. Jing-han Yuan: Investigation, Methodology. Hao Li: Methodology, Resources. Qi-Pu Feng: Data curation, Methodology. Bing-Mei Zhu: Conceptualization, Funding acquisition, Supervision, Writing – review & editing.

Declaration of competing interest

The authors declare the following financial interests/personal relationships which may be considered as potential competing interests: The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Qiu-Yu Pang, Xiang-Min Meng, De-Yong Li, Lu You, Zhen-fang Zhou, Yu-Meng Li, Tian Deng, Jing-han Yuan, Hao Li, Qi-Pu Feng, Bing-Mei Zhu.

Appendix A. Supplementary data

The following are the Supplementary data to this article:

Data availability

Data will be made available on request.