ALYREF stabilizes MZF1 via m5C modification to exacerbate cardiac remodeling and atrial fibrillation in heart failure

Abstract

Background

Patients with heart failure (HF) frequently develop atrial fibrillation, but the precise molecular mechanisms are unclear, leading to limited therapeutic strategies.

Methods

Isoproterenol hydrochloride (ISO)-induced mouse HF model and ISO-treated HL-1 cardiomyocyte model were employed. A combination of qRT-PCR, Western blot, ChIP, dual-luciferase reporter assay, MeRIP-qPCR, RIP-qPCR, actinomycin D assay, and mitochondrial function assessments were used to systematically investigate the regulatory mechanism of the ALYREF/MZF1/HTRA1 axis and its role in cardiac remodeling and atrial fibrillation susceptibility.

Results

HTRA1 expression was elevated in HF mice and human atrial fibrillation data. In cellular models, HTRA1 overexpression induced mitochondrial damage, cellular hypertrophy, and inflammation via the HIF-1α pathway. The transcription factor MZF1 directly bound to and activated the HTRA1 promoter. NSUN2-mediated m5C modification of MZF1 mRNA was required for its recognition and stabilization by the m5C reader protein ALYREF, thereby positively regulating the MZF1/HTRA1 axis. In the ISO-induced HF mouse model, knockdown of ALYREF effectively reduced MZF1 and HTRA1 expression, alleviated myocardial hypertrophy, fibrosis, cardiac dysfunction, and atrial fibrillation inducibility.

Conclusion

This study elucidates a novel signaling axis wherein NSUN2-mediated m5C modification of MZF1 mRNA enables its stabilization by ALYREF, leading to transcriptional upregulation of HTRA1. This axis drives cardiomyocyte mitochondrial dysfunction, hypertrophy, and inflammation in the context of HF, ultimately increasing susceptibility to atrial fibrillation.

Supplementary Information

The online version contains supplementary material available at 10.1007/s00018-026-06273-3.

Introduction

Atrial fibrillation and heart failure (HF) form a clinically intertwined entity, sharing common risk factors and promoting each other’s progression pathologically [1, 2]. Atrial fibrillation represents the most prevalent arrhythmia in patients with HF, while HF constitutes the terminal stage of various cardiovascular diseases, characterized by impaired systolic and/or diastolic function leading to systemic and pulmonary congestion [3, 4]. Their coexistence portends a poor clinical prognosis: HF exacerbates the arrhythmic substrate in atrial fibrillation, posing substantial therapeutic challenges, and conversely, atrial fibrillation accelerates the progression of HF [5]. Despite therapeutic advances, morbidity and mortality remain high [6, 7], underscoring an urgent need to elucidate the fundamental molecular pathways driving their co‑occurrence.

Pathological cardiac remodeling is a hallmark of HF progression, involving cardiomyocyte hypertrophy, interstitial fibrosis, chronic inflammation, and metabolic dysfunction [8, 9]. These processes collectively compromise cardiac output and electrical stability, thereby establishing the pathological basis of atrial fibrillation [6, 10]. Mitochondrial dysfunction, characterized by oxidative stress and energy metabolism dysfunction, is increasingly recognized as a central mediator of cardiac remodeling, contributing to both contractile impairment and pro‑arrhythmic signaling [11, 12]. Identifying upstream regulators that integrate these pathological features may reveal novel therapeutic targets for HF‑associated atrial fibrillation.

The high temperature requirement factor A1 (HTRA1) has emerged as a potential key player in tissue homeostasis and disease [13, 14]. Beyond its role in protein quality control, HTRA1 modulates critical signaling pathways involved in inflammation, fibrosis, and cellular stress responses [15–17]. Notably, recent epigenomic studies have demonstrated that HTRA1 exhibits hypomethylation and concomitant upregulation in human failing myocardium, suggesting its involvement in HF pathogenesis [18]. Our preliminary analysis of human atrial fibrillation dataset GSE282504 further indicated elevated HTRA1 expression, implicating its role in arrhythmogenesis. Evidence from other disease contexts, such as dilated cardiomyopathy and age‑related macular degeneration, shows that HTRA1 overexpression can drive fibrosis, mitochondrial impairment, and activation of hypoxia‑inducible factor‑1α (HIF‑1α) signaling [17, 19]. However, the precise role and regulatory mechanism of HTRA1 in cardiomyocyte hypertrophy remain unclear.

Transcriptional regulation represents a critical layer controlling gene expression in disease. Myeloid zinc finger 1 (MZF1) is a dual regulatory‑function transcription factor involved in cell proliferation, metabolism, and inflammatory responses [20–23]. Studies have shown that MZF1 enhances the post‑transcriptional accumulation of HIF‑1α by promoting c‑Myc transcriptional activity, thereby facilitating glioma progression [21]. Our bioinformatic analysis identified a predicted MZF1‑binding site within the HTRA1 promoter region, suggesting that MZF1 may act as a transcriptional regulator of HTRA1.

Recent advances highlight the importance of epitranscriptomic modifications, such as 5‑methylcytosine (m5C), in fine‑tuning gene expression and RNA stability, with profound implications in cancer, neurodegenerative diseases, and metabolic disorders [24]. The methyltransferase NOP2/Sun RNA methyltransferase 2 (NSUN2) mediates m5C deposition on target RNAs [25], while the reader protein Aly/REF export factor (ALYREF) specifically recognizes m5C‑modified mRNAs and enhances their stability [26–28]. Previous studies have reported significant upregulation of ALYREF in HF samples [29]. Preliminary bioinformatic screening (e.g., via RMBase3 and StarBase) in this study predicted the presence of m5C modification sites on MZF1 mRNA and identified ALYREF as a potential binding protein. This leads to a compelling hypothesis: ALYREF may stabilize MZF1 mRNA through recognition of its m5C modification, thereby enhancing its impact on HTRA1 transcriptional activity and promoting HF progression and atrial fibrillation susceptibility.

This study aimed to validate the upregulation of HTRA1 in experimental HF; define the functional consequences of HTRA1 knockdown and overexpression in cardiomyocytes; elucidate MZF1 as a transcriptional activator of HTRA1; identify NSUN2 as the m5C writer responsible for MZF1 modification and ALYREF as the reader that stabilizes MZF1 mRNA; and confirm the ameliorative effects of ALYREF knockdown on pathological cardiac remodeling and atrial fibrillation susceptibility in HF mice. Our findings delineate a previously unrecognized NSUN2-ALYREF-MZF1-HTRA1 signaling axis crucial to the pathogenesis of HF and atrial fibrillation, offering new potential targets for therapeutic intervention.

Materials and methods

Bioinformatics analysis

The gene expression dataset GSE282504 was downloaded from the Gene Expression Omnibus (GEO) database. This dataset contains left atrial tissue samples from 20 patients and 10 controls. Raw count data were processed using R software (version 4.2.0). Gene annotation was performed using the human genome reference GRCh38.112 (gff3 format) to map Ensembl IDs to Symbol IDs. Normalization was carried out using the voom function with quantile normalization from the limma package. Differential expression analysis was performed using the limma package, and differentially expressed genes were identified with thresholds of |logFC| > 0.32 and p < 0.05, without multiple testing correction due to the exploratory nature of the analysis and the relatively modest effect sizes.

Animals

Male C57BL/6J mice (6–8 weeks old) were obtained from SiPeiFu Biotechnology (Beijing, China). All animals were housed under specific pathogen‑free conditions at a temperature of 21–26 °C and relative humidity of 40–70%, with free access to food and water. Mice were acclimatized for one week before the start of experiments. All procedures involving animals were approved by the Animal Ethics Committee of The First Affiliated Hospital of Kunming Medical University (No. 2025DF052).

HF model establishment and treatment

The HF model was induced by intraperitoneal injection of isoproterenol hydrochloride (ISO; #I129810-5 g, Aladdin, Shanghai, China) at a dose of 10 mg/kg daily for two consecutive weeks, as previously described [30]. Control mice received an equal volume of normal saline via intraperitoneal injection on the same schedule. For gene intervention studies, mice in HF + sh-ALYREF group received a single tail-vein injection of 50 µL adeno-associated virus (AAV9 serotype, 5.0 × 10¹² vg/mL, carrying shRNA targeting ALYREF under the control of the aMHC promoter) two weeks prior to ISO injections. Mice in HF + sh-NC group received an equal volume of AAV carrying a non-targeting shRNA via the same route. The shRNA sequences are listed in Table 1. All mice were euthanized at 4 weeks after the initiation of ISO/saline injections for subsequent analyses.

Table 1.

shRNA sequences used in this study

Gene	Forward sequence (5’-3’)	Reverse sequence (5’-3’)
sh-MZF1-1 Mouse	CCGGCATCACGGCAGAGCCAGATATCTCGAGATATCTGGCTCTGCCGTGATGTTTTTG	AATTCAAAAACATCACGGCAGAGCCAGATATCTCGAGATATCTGGCTCTGCCGTGATG
sh-MZF1-2 Mouse	CCGGATCACGGCAGAGCCAGATATGCTCGAGCATATCTGGCTCTGCCGTGATTTTTTG	AATTCAAAAAATCACGGCAGAGCCAGATATGCTCGAGCATATCTGGCTCTGCCGTGAT
sh-ALYREF-1 Mouse	CCGGCGAAACAACTTCCCGACAAATCTCGAGATTTGTCGGGAAGTTGTTTCGTTTTTG	AATTCAAAAACGAAACAACTTCCCGACAAATCTCGAGATTTGTCGGGAAGTTGTTTCG
sh-ALYREF-2 Mouse	CCGGCCAGCTTGTCACATCACAGATCTCGAGATCTGTGATGTGACAAGCTGGTTTTTG	AATTCAAAAACCAGCTTGTCACATCACAGATCTCGAGATCTGTGATGTGACAAGCTGG
sh-HTRA1-1 Mouse	CCGGGCGTCATAAGTACAACTTTATCTCGAGATAAAGTTGTACTTATGACGCTTTTTG	AATTCAAAAAGCGTCATAAGTACAACTTTATCTCGAGATAAAGTTGTACTTATGACGC
sh-HTRA1-2 Mouse	CCGGCCTTCGCAATTCCATCCGATACTCGAGTATCGGATGGAATTGCGAAGGTTTTTG	AATTCAAAAACCTTCGCAATTCCATCCGATACTCGAGTATCGGATGGAATTGCGAAGG
sh-HIF-1α-1 Mouse	CCGGCCCATTCCTCATCCGTCAAATCTCGAGATTTGACGGATGAGGAATGGGTTTTTG	AATTCAAAAACCCATTCCTCATCCGTCAAATCTCGAGATTTGACGGATGAGGAATGGG
sh-HIF-1α-2 Mouse	CCGGAGTCGACACAGCCTCGATATGCTCGAGCATATCGAGGCTGTGTCGACTTTTTTG	AATTCAAAAAAGTCGACACAGCCTCGATATGCTCGAGCATATCGAGGCTGTGTCGACT
sh-NSUN2-1 Mouse	CCGGCCTGAAGATGATCCTTTATTTCTCGAGAAATAAAGGATCATCTTCAGGTTTTTG	AATTCAAAAACCTGAAGATGATCCTTTATTTCTCGAGAAATAAAGGATCATCTTCAGG
sh-NSUN2-2 Mouse	CCGGCCTGAAGATGATCCTTTATTTCTCGAGAAATAAAGGATCATCTTCAGGTTTTTG	AATTCAAAAACCTGAAGATGATCCTTTATTTCTCGAGAAATAAAGGATCATCTTCAGG
sh-NC Mouse	CCGGCGCCAGATTTGTTCGCCTTATCTCGAGATAAGGCGAACAAATCTGGCGTTTTTG	AATTCAAAAACGCCAGATTTGTTCGCCTTATCTCGAGATAAGGCGAACAAATCTGGCG

Following euthanasia, body weight was recorded, and the heart was rapidly excised, rinsed in cold saline, blotted dry, and weighed. The heart-to-body weight ratio (HW/BW) was calculated as an index of cardiac hypertrophy. The atria were carefully dissected from the heart, weighed, and the atrial weight-to-body weight ratio (LAW/BW) was calculated as an index of atrial remodeling.

Transthoracic echocardiography

Transthoracic echocardiography was performed to assess cardiac function. Mice were anesthetized with isoflurane (#R510-22-10, RWD Life Science, Shenzhen, Guangdong, China) via inhalation and placed in a supine position. After removal of chest hair, two‑dimensional M‑mode images of the left ventricle were obtained from the parasternal long‑axis view using a small‑animal ultrasound system (#Vevo2100, FujiFilm, Tokyo, Japan). Measurements including fractional shortening (FS), left ventricular internal diameter at end‑diastole (LVIDd), and left ventricular internal diameter at end‑systole (LVIDs) were recorded.

Atrial fibrillation electrocardiogram (ECG) monitoring

Atrial fibrillation ECG monitoring was conducted under continuous anesthesia maintained with 1–2% isoflurane (700 mL/min) via a mask (induction at 5%). Rectal temperature was monitored and kept at 37–38 °C using a heating pad. Needle electrodes were placed subcutaneously in the right forelimb and both hindlimbs, and signals were acquired for 5–10 min using a small‑animal ECG system (#DE03-VET, Dawei Medical, Xuzhou, Jiangsu, China) and recorded with LabChart 4.2.3 software. After recording, mice were allowed to recover on a warm pad with supplemental oxygen. ECG traces were analyzed for stable segments (≥ 1 min) to identify P‑waves, QRS complexes, and rhythm irregularities. Atrial fibrillation was defined as abnormal electrocardiographic traces characterized by irregular rhythm, loss of P waves, and irregular R-R intervals lasting for at least 1 s, as previously described [31]. For each mouse, the following quantitative atrial fibrillation endpoints were assessed: atrial fibrillation incidence, number of atrial fibrillation episodes, total atrial fibrillation burden, and ventricular rate during atrial fibrillation episodes.

Histological morphology analysis

Atrial tissues were fixed in 4% paraformaldehyde (#G1101, Servicebio, Wuhan, Hubei, China), paraffin‑embedded, and sectioned at 4 μm. For hematoxylin and eosin (H&E) staining, sections were deparaffinized, rehydrated, stained with hematoxylin solution (#G1004, Servicebio) for 5 min and eosin solution (#G1001, Servicebio) for 2 min, dehydrated, cleared, and mounted. For semi‑quantitative assessment of myocardial disarray and inflammatory infiltration, H&E‑stained sections were evaluated by a pathologist blinded to the experimental groups. Three random fields per section were examined at 200× magnification. The percentage of lesional area (inflammatory infiltration and necrosis) per field was scored as follows: 0, no lesion; 1, lesion area < 25%; 2, 25%–49%; 3, 50%–75%; 4, > 75%.

To assess fibrosis, Masson’s trichrome staining was performed using a commercial kit (#G1346, Solarbio, Beijing, China) following the manufacturer’s protocol, which included sequential staining with hematoxylin, ponceau‑acid fuchsin, phosphomolybdic acid, and aniline blue; collagen fibers were stained blue. Cardiomyocyte cross‑sectional area was measured after staining with fluorescein‑labeled wheat germ agglutinin (WGA, #MP6329, 1:100; Maokang Biotechnology, Shanghai, China) overnight at 4 °C. All sections were examined and imaged using a light or fluorescence microscope.

Immunohistochemistry (IHC)

Paraffin‑embedded cardiac sections were deparaffinized, rehydrated, and subjected to antigen retrieval in Tris‑EDTA buffer (#G1206, Servicebio) using microwave heating. After endogenous peroxidase activity was blocked, sections were incubated overnight at 4 °C with a primary antibody against HTRA1 (#55011-1-AP, 1:200; Proteintech, Wuhan, Hubei, China). After washing, slides were incubated with a horseradish peroxidase‑conjugated secondary antibody (#PV‑6000, Zhongshan Golden Bridge, Beijing, China) at 37 °C for 30 min. Signal was developed using a DAB substrate kit, and nuclei were counterstained with hematoxylin. Sections were dehydrated, cleared, and mounted with neutral balsam. Images were acquired using a light microscope.

Cell culture and treatment

The murine cardiomyocyte cell line HL‑1 (Immocell Biotechnology, Xiamen, Fujian, China) was cultured in DMEM medium (#E600003-0500, Sangon Biotech, Shanghai, China) at 37 °C in a humidified incubator with 5% CO₂. To induce cellular hypertrophy, cells were treated with 30 µM ISO for 48 h. For functional studies, cells were assigned to different experimental sets, including control (untreated), ISO‑treated (Model), and ISO‑treated cells transfected with non‑targeting control (sh‑NC), overexpression negative control (OE‑NC), or gene‑specific shRNA/overexpression constructs targeting HTRA1, MZF1, or ALYREF, or shRNA targeting HIF-1α, either alone or in combination. All shRNA transfections were performed using Lipofectamine™ 2000 transfection reagent (#11668019, Thermo Fisher Scientific, Waltham, USA). Specific rescue experiments were performed by co‑treating ALYREF‑knockdown cells with MZF1 overexpression vectors, and HTRA1‑overexpressing cells with the HIF‑1α inhibitor KC7F2 (#HY-18777, MCE, New Jersey, USA) at a concentration of 20 µM for 24 h. All transfections were carried out prior to ISO stimulation.

Cell viability assay

Cell viability was measured using a Cell Counting Kit‑8 (CCK‑8; #C0038, Beyotime, Shanghai, China) according to the manufacturer’s protocol. Briefly, HL‑1 cells from different groups were seeded in 96‑well plates at a density of 1 × 10⁶ cells/mL. After 24 h of culture, 10 µL of CCK‑8 reagent was added to each well, followed by incubation at 37 °C for 1 h. Absorbance was measured at 450 nm using a microplate reader (#Multiskan FC, Thermo Fisher Scientific).

Immunofluorescence staining for α‑actinin

Cells grown on coverslips were fixed with 4% paraformaldehyde for 15 min, permeabilized with 0.5% Triton X‑100 for 20 min, and blocked with 3% BSA. Subsequently, cells were incubated overnight at 4 °C with a rabbit anti‑α‑actinin primary antibody (#11313-2-AP, Proteintech) diluted in antibody dilution buffer (#P0103, Beyotime). After washing, a Cy3‑conjugated goat anti‑rabbit secondary antibody (#A0516, Beyotime) was applied for 1 h at room temperature. Nuclei were counterstained with DAPI. Images were captured using a fluorescence microscope (#BZ‑X800, Keyence, Osaka, Japan).

MitoTracker staining

Cells seeded in 48‑well plates were incubated with MitoTracker Red CMXRos (#C1049B, Beyotime) at a working concentration of 100 nM in pre‑warmed culture medium for 15 min at 37 °C. After washing with fresh medium, cells were immediately imaged using a fluorescence microscope (#BZ‑X800, Keyence). Fluorescence intensity was quantified using ImageJ software to assess mitochondrial mass.

Mitochondrial membrane potential assay

Mitochondrial membrane potential was assessed using the JC‑1 dye (#C2006, Beyotime). Following the indicated treatments, cells were collected and incubated with 1 mL JC‑1 working solution at 37 °C for 20 min. After washing with JC‑1 staining buffer, cells were resuspended in PBS and immediately analyzed by flow cytometry (#RMNNC-3000, Agilent, Santa Clara, USA).

Adenosine triphosphate (ATP) content and mitochondrial reactive oxygen species (ROS) measurement

Intracellular ATP level was determined using an ATP Assay Kit (#BC0300, Solarbio). Cells were lysed, and the supernatant was collected after centrifugation. ATP content was measured following the manufacturer’s protocol by recording absorbance at 340 nm using a microplate reader (#Multiskan FC, Thermo Fisher Scientific).

Mitochondrial ROS production was assessed using the MitoSOX™ Red reagent (#S0061S, Beyotime). Cells were incubated with 5 µM MitoSOX Red working solution at 37 °C for 15 min, washed with PBS, and immediately imaged under a fluorescence microscope (#BZ‑X800, Keyence).

Seahorse XF assay

Mitochondrial respiration and glycolytic function were evaluated using a Seahorse XF Analyzer (#Seahorse XF Pro, Agilent). HL‑1 cells were seeded into Seahorse XF24 cell culture plates at a density of 1 × 10⁴-3 × 10⁴ cells/well and incubated overnight. Prior to the assay, cells were washed and incubated in pre‑warmed XF assay medium (pH 7.4) for 1 h at 37 °C in a non‑CO₂ incubator. The Seahorse XF Cell Mito Stress Test Kit (#103015-100, Agilent) was used according to the manufacturer’s instructions. Oxygen consumption rate (OCR) and extracellular acidification rate (ECAR) were measured in real‑time.

Enzyme-linked immunosorbent assay (ELISA)

Plasma samples were collected in EDTA‑coated tubes and centrifuged at 1000×g for 15 min at 4 °C to obtain supernatant. For atrial tissue lysates, tissues were homogenized in PBS and centrifuged at 5000×g for 10 min at 4 °C to collect the supernatant. Cell culture supernatants from HL‑1 cells or cell lysates prepared in RIPA buffer were collected after centrifugation. Levels of interleukin‑1β (IL‑1β; #E‑EL‑M0037, Elabscience, Wuhan, Hubei, China), interleukin‑6 (IL‑6; #E‑EL‑M0044, Elabscience), and tumor necrosis factor‑α (TNF‑α; #E‑EL‑M3063, Elabscience) were measured using commercial mouse ELISA kits according to the manufacturer’s instructions. Briefly, samples or standards were added to pre‑coated 96‑well plates and incubated at 37 °C for 90 min. After washing, biotinylated detection antibody was added, followed by incubation with HRP‑conjugated streptavidin. Color was developed with TMB substrate and the reaction was stopped with sulfuric acid. Absorbance at 450 nm was measured using a microplate reader (#Multiskan FC, Thermo Fisher Scientific).

Quantitative real-time PCR (qRT‑PCR)

Total RNA was extracted from atrial tissues or HL‑1 cells using TRIzol reagent (#G3013, Servicebio) following the manufacturer’s protocol. RNA concentration and purity were measured using a micro‑volume spectrophotometer (#FMUV‑2000, BigFly, Hangzhou, Zhejiang, China). cDNA was synthesized from total RNA using the PrimeScript™ RT reagent Kit (#RR037Q, Takara, Kusatsu, Shiga, Japan). qRT‑PCR was performed with TB Green^® Premix Ex Taq™ II (#RR820Q, Takara) on a real‑time PCR system (#SLAN‑96 S, Hongshi, Foshan, Guangdong, China). The reaction mixture (20 µL) contained 10 µL TB Green Premix, 0.8 µL each of forward and reverse primers (10 µM), 1 µL cDNA template, and RNase‑free water. The cycling conditions were: 95 °C for 3 min, followed by 40 cycles of 95 °C for 5 s and 60 °C for 20 s. Melt‑curve analysis was performed to confirm amplification specificity. GAPDH was used as the internal control. Relative mRNA expression of target genes (HTRA1, MZF1, ALYREF, ANP, BNP, β‑MHC) was calculated using the 2^−ΔΔCt method. Primer sequences are listed in Table 2.

Table 2.

Primers of qRT-PCR

Gene	Forward sequence (5’-3’)	Reverse sequence (5’-3’)
HTRA1 Mouse	GCTGATGTGGTGGAGAAGATCG	CTCCGATACGATGAATCCTGACC
MZF1 Mouse	AACACTGGACCAGACCTCTCCT	TCTGCCGTGATGCTGTCCATCT
ALYREF Mouse	TGTCCAACCTGGACTTCGGAGT	CCTAAACTTCGTCCAGAGCGATC
ANP Mouse	ACGTGGAGTTGAGTTCCTTGGC	TTCTCTGCCAGGACTTCCAAGC
BNP Mouse	TCCTAGCCAGTCTCCAGAGCAA	GGTCCTTCAAGAGCTGTCTCTG
β-MHC Mouse	GCTGGAAGATGAGTGCTCAGAG	TCCAAACCAGCCATCTCCTCTG
NSUN2 Mouse	GATGTGTCTGCTGAGTTGCCAG	GGTCGGATTTGTGTATGCCTGC
GAPDH Mouse	CATCACTGCCACCCAGAAGACTG	ATGCCAGTGAGCTTCCCGTTCAG

Western blot (WB) analysis

Protein was extracted from atrial tissues or HL‑1 cells using RIPA lysis buffer (#BL504A, Biosharp, Hefei, Anhui, China) supplemented with protease inhibitors (#A32955, Thermo Fisher Scientific). Protein concentration was determined with a BCA assay kit (#BL521A, Biosharp). Samples were separated by SDS‑PAGE gels and transferred onto PVDF membranes (#IPVH00010, Millipore, Darmstadt, Germany). After blocking with 5% skim milk, membranes were incubated overnight at 4 °C with primary antibodies: anti‑HTRA1 (#55011‑1‑AP, 1:1000; Proteintech), anti‑HIF‑1α (#ab308433, 1:1000; Abcam, Cambridge, MA, USA), anti‑MZF1 (#A10356, 1:1000; ABclonal, Wuhan, Hubei, China), anti‑ALYREF (#ab202894, 1:1000; Abcam), anti‑Cx40 (#EPR28370-62, 1:1000; Abcam), anti‑Cx43 (#26980-1-AP, 1:1000; Proteintech), anti‑SERCA2a (#13985-1-AP, 1:1000; Proteintech), anti‑NCX1 (#55075-1-AP, 1:1000; Proteintech), or anti‑GAPDH (#ab8245, 1:1000; Abcam). Membranes were washed and incubated with HRP‑conjugated secondary antibodies (goat anti‑rabbit, #ab6721 or goat anti‑mouse, #ab6789; 1:5000, Abcam) for 2 h at room temperature. Bands were visualized using Immobilon Western Chemiluminescent HRP Substrate (#WBKLS0100, Millipore) and imaged with a chemiluminescence detection system (#ChampChemi 910, SINSAGE, Beijing, China).

Chromatin immunoprecipitation (ChIP) assay

The binding of MZF1 to the HTRA1 promoter was examined using a ChIP assay kit (#P2078, Beyotime). Atrial tissues or HL‑1 cells were cross‑linked with 1% formaldehyde and lysed in SDS Lysis Buffer. Chromatin was fragmented by sonication to an average size of 200‑1000 bp. After centrifugation, the supernatant was diluted and pre‑cleared with Protein A/G agarose. Immunoprecipitation was performed overnight at 4 °C with an anti‑MZF1 antibody or control IgG. Protein‑DNA complexes were collected with Protein A/G agarose, washed, and eluted. Cross‑links were reversed by incubation with NaCl at 65 °C, followed by proteinase K digestion. DNA was purified by phenol‑chloroform extraction and ethanol precipitation. Enrichment of the HTRA1 promoter region was quantified by qRT‑PCR as described above, with primers listed in Table 2.

Dual‑luciferase reporter assay

To assess the transcriptional regulation of HTRA1 by MZF1, wild‑type (WT) and mutant (MUT) HTRA1 promoter fragments containing the predicted MZF1‑binding site were cloned into a pGL3‑basic vector. HL‑1 cells were co‑transfected with the reporter plasmids (HTRA1‑WT or HTRA1‑MUT) and either sh‑MZF1 or sh‑NC using Lipo8000™ transfection reagent (#C0533, Beyotime). Cells were also transfected with a Renilla luciferase plasmid as an internal control. After 48 h, luciferase activity was measured using the Dual‑Luciferase Reporter Assay Kit (#RG027, Beyotime) on a TECAN Spark microplate reader. Firefly luciferase signals were normalized to Renilla luciferase activity for each sample.

Methylated RNA immunoprecipitation-qPCR (MeRIP‑qPCR)

The m5C modification level of MZF1 mRNA was assessed using the GenSeq^® m5C MeRIP Kit (#GS-ET-003, Cloud-Seq Biotechnology, Shanghai, China) with an anti‑m5C antibody. Total RNA was extracted from atrial tissues or HL‑1 cells. RNA samples were fragmented and immunoprecipitated with antibody‑coated magnetic beads. After washing, bound RNA was eluted, purified, and reverse‑transcribed. Enrichment of m5C‑modified MZF1 mRNA was quantified by qRT‑PCR using primers specific for the target region. IgG‑immunoprecipitated samples and input RNA served as controls. Relative enrichment was calculated using the 2^−ΔΔCt method.

RNA immunoprecipitation-qPCR (RIP‑qPCR)

The interaction between ALYREF protein and MZF1 mRNA was assessed using an RIP assay kit (#JKR23003, Jinkairui, Wuhan, Hubei, China). HL‑1 cells were lysed in RIP lysis buffer. Cell lysates were incubated with Protein A/G magnetic beads pre‑coated with anti‑ALYREF antibody or control IgG at 4 °C overnight. After washing, the immunoprecipitated RNA‑protein complexes were eluted and digested with proteinase K. RNA was extracted and reverse‑transcribed into cDNA. The enrichment of MZF1 mRNA in the immunoprecipitated samples was quantified by qRT‑PCR.

Actinomycin D assay

To assess MZF1 mRNA stability, HL‑1 cells were treated with 100 µg/mL actinomycin D (#HY‑17559, MCE) to inhibit transcription. Cells were harvested at 0, 3, 6, and 9 h after actinomycin D addition. Total RNA was extracted using TRIzol reagent and reverse‑transcribed with the PrimeScript RT Kit. MZF1 mRNA level was quantified by qRT‑PCR.

Statistical analysis

All in vivo experiments used 10 mice per group. For atrial fibrillation-related analyses, only mice that exhibited atrial fibrillation episodes (6 per group) were included. For molecular analyses, 3 randomly selected mice per group were used. All in vitro experiments were performed with three independent biological replicates (n = 3). Animals were randomly assigned to experimental groups using the Excel RAND() function. A double-blind design was implemented: group allocation was performed by an independent third party, and both the experimenter and the pathologist responsible for histological assessment were blinded to group assignments until data analysis was completed. Exclusion criteria were predefined as mice that died during modeling or failed to exhibit cardiac functional impairment on echocardiography. Data were analyzed by GraphPad Prism 8.0 (La Jolla, CA, USA) and expressed as mean ± standard deviation. p < 0.05 was considered as a significant difference. Normality and homogeneity of variance were assessed before parametric analysis. Comparisons between two groups were performed using two-tailed unpaired Student’s t-test. Comparisons among multiple groups were performed using one-way ANOVA followed by Tukey’s post-hoc test.

Results

HF mice demonstrated HTRA1 upregulation and atrial fibrillation susceptibility

Analysis of the publicly available dataset GSE282504 revealed a significant upregulation of HTRA1 in atrial fibrillation group compared with control group (Fig. 1A). To investigate the role of HTRA1 in a HF context, a mouse model was established via daily ISO injections for two weeks. Consistent with successful HF induction, echocardiography confirmed impaired cardiac function in HF mice, as evidenced by reduced FS, as well as increased LVIDd and LVIDs (Fig. 1B, S1A). HF mice exhibited significant cardiac remodeling, as indicated by an increased HW/BW ratio (Fig. 1C), along with an elevated LAW/BW ratio (Fig. 1D). Given that atrial pathology is critical for arrhythmogenesis, atrial tissue remodeling was next assessed. Atrial tissue analysis revealed myocardial disarray and elevated inflammatory infiltration (Fig. 1E), cardiac fibrosis (Fig. 1F), and cardiomyocyte hypertrophy as evidenced by enlarged cardiomyocyte cross-sectional area (Fig. 1G). Moreover, ELISA revealed higher levels of IL-1β, IL-6, and TNF-α in both plasma and atrial tissue lysates from HF mice compared to control group (Figure S1B-C). WB analysis showed decreased expression of gap junction proteins Cx40 and Cx43, along with altered calcium-handling protein expression (decreased SERCA2a and increased NCX1) in atrial tissues of HF mice, indicating significant atrial electrical remodeling (Fig. 1H). To determine whether these structural and electrical alterations translated into functional arrhythmic consequences, ECG analysis was performed. HF mice exhibited atrial arrhythmia episodes characterized by irregular rhythm, loss of P waves, and irregular R-R intervals (Fig. 1I). Quantitative analysis showed that HF mice had significantly higher atrial fibrillation incidence, more frequent atrial fibrillation episodes, increased total atrial fibrillation burden, and elevated ventricular rate during atrial fibrillation episodes compared with control mice, in which no atrial fibrillation events were observed (Fig. 1J-M). These findings indicate that HF mice have increased susceptibility to atrial fibrillation. Critically, HTRA1 expression was markedly increased at the mRNA and protein levels in atrial tissue from these HF mice (Fig. 1N-P, S1D). Therefore, HTRA1 upregulation was observed in both human atrial fibrillation samples and the ISO-induced HF mouse model, prompting further investigation into its functional role in HF-associated atrial fibrillation.

Fig. 1.

HTRA1 was upregulated in HF mice with atrial fibrillation. Analysis of human atrial fibrillation dataset. A Violin plot showing HTRA1 expression in control and atrial fibrillation samples from the GSE282504 dataset. C57BL/6J mice were subjected to HF by daily intraperitoneal injection of ISO for two weeks. B Transthoracic echocardiography images from CON and HF mice; quantification of left ventricular FS. C HW/BW ratio of mice was assessed to evaluate ventricular hypertrophy. D LAW/BW ratio of mice showing atrial hypertrophy. E H&E staining showing myocardial disarray and inflammatory infiltration in mouse atrial tissue; quantitative scoring of myocardial disarray and inflammatory infiltration. F Masson’s trichrome staining demonstrating cardiac fibrosis. G WGA staining was employed to evaluate cardiomyocyte cross-sectional area in mouse atrial tissue. H WB analysis of gap junction proteins Cx40 and Cx43 and calcium-handling proteins SERCA2a and NCX1 in atrial tissues. I ECG tracings showing spontaneous atrial arrhythmia episodes in HF mice, characterized by irregular rhythm, loss of P waves, and irregular R-R intervals. J-M Quantitative atrial fibrillation endpoints including incidence, number of episodes, total atrial fibrillation burden, and ventricular rate. N-P The mRNA and protein levels of HTRA1 in atrial tissues were determined by qRT-PCR, WB and IHC assays. Data are presented as mean ± SD (n = 10 per group; n = 6 for atrial fibrillation analyses; n = 3 for molecular analyses). Exact p-values are indicated in the figures

HTRA1 knockdown attenuated ISO-induced cardiomyocyte hypertrophy, inflammation, and mitochondrial dysfunction in vitro

To investigate the functional role of HTRA1, loss-of-function experiments was performed in HL-1 cardiomyocytes subjected to ISO treatment. Efficient knockdown of HTRA1 was first validated (Figure S2A-B). Consistent with a potential pathogenic role, ISO treatment significantly reduced cell viability and upregulated HTRA1 expression in HL-1 cells, which was compromised by HTRA1 knockdown (Fig. 2A-C). Next, whether HTRA1 contributes to cardiomyocyte hypertrophy was examined. HTRA1 knockdown attenuated the ISO-induced increase in cardiomyocyte surface area and upregulation of hypertrophic marker genes (ANP, BNP, β-MHC) (Fig. 2D-G). Furthermore, HTRA1 knockdown markedly reduced the levels of inflammatory cytokines (IL-1β, IL-6, TNF-α) in HL-1 cells that was elevated by ISO treatment (Fig. 2H). Since mitochondrial dysfunction is a critical driver of cardiomyocyte pathology, mitochondrial parameters were further evaluated. At the mitochondrial level, knockdown of HTRA1 rescued the ISO-induced reduction in mitochondrial mass, as shown by MitoTracker staining (Fig. 2I). It also reversed the ISO-mediated decline in ATP production (Fig. 2J) and prevented the loss of mitochondrial membrane potential (Fig. 2K). Notably, HTRA1 knockdown significantly diminished the excessive mitochondrial ROS generation triggered by ISO (Fig. 2L). Functional assays using the Seahorse analyzer demonstrated that HTRA1 knockdown reversed the impaired OCR and the enhanced ECAR caused by ISO treatment (Fig. 2M-O). Collectively, these experiments demonstrated that knockdown of HTRA1 effectively counteracted ISO-induced cardiomyocyte hypertrophy, inflammation, and mitochondrial dysfunction, highlighting its critical role in mediating these pathological processes.

Fig. 2.

Knockdown of HTRA1 influenced ISO-induced cardiomyocyte hypertrophy, inflammation, and mitochondrial dysfunction in vitro. HL-1 cardiomyocytes were treated with ISO or vehicle and co-transfected with sh-HTRA1 or sh-NC. A Cell viability was assessed by CCK-8 assay. B-C qRT-PCR and WB were performed to analyze HTRA1 mRNA and protein expression levels. D-E Immunofluorescence staining and quantification of α-actinin (red) was conducted to visualize cell size. Nuclei were stained with DAPI (blue). F-G The mRNA levels of hypertrophic markers ANP, BNP, and β-MHC were assessed by qRT-PCR. H ELISA demonstrating IL-1β, IL-6, and TNF-α levels in the culture supernatant. I MitoTracker staining was utilized to assess mitochondrial mass. J Intracellular ATP content was quantified by kit assay. K Mitochondrial membrane potential was measured by JC-1 staining using flow cytometry. L MitoSOX Red fluorescence was performed to evaluate mitochondrial ROS level. M-O OCR and ECAR were measured by Seahorse analyzer. Data are presented as mean ± SD from three independent biological replicates (n = 3). Exact p-values are indicated in the figures

HTRA1 overexpression exacerbated cardiomyocyte mitochondrial dysfunction via activating HIF-1α signaling

For exploring whether the effect of HTRA1 is mediated through HIF-1α signaling, ISO-treated HL-1 cells were subjected to HTRA1 overexpression (as validated in Figure S2C-D) and then co-treated with the HIF-1α translation inhibitor KC7F2. Overexpression of HTRA1 in ISO-treated HL-1 cells further enhanced the expression of HIF-1α protein, indicating a potential link between HTRA1 and HIF-1α (Fig. 3A). HTRA1 overexpression also aggravated mitochondrial damage in ISO-treated HL-1 cells, including reduced mitochondrial mass (Fig. 3B), increased mitochondrial ROS (Fig. 3C), loss of membrane potential (Fig. 3D), decreased ATP production (Fig. 3E), decreased OCR (Fig. 3F), and enhanced ECAR (Fig. 3G). To test whether these detrimental effects depend on HIF-1α activation, the effect of KC7F2 co-treatment on HTRA1 overexpression was assessed. Treatment with KC7F2 effectively mitigated the mitochondrial dysfunction exacerbated by HTRA1 overexpression (Fig. 3B-E), suggesting that HTRA1 disrupted mitochondrial function in ISO-treated HL-1 cells at least in part through activation of the HIF-1 signaling pathway.

Fig. 3.

HTRA1 overexpression affected mitochondrial dysfunction via activation of the HIF-1α pathway. ISO-treated HL-1 cells were transfected with OE-HTRA1 and then co-treated with the HIF-1α inhibitor KC7F2 or DMSO vehicle. A WB analysis of HIF-1α protein expression. B-E Mitochondrial mass, mitochondrial ROS level, mitochondrial membrane potential and ATP content were respectively quantified by MitoTracker staining, MitoSOX detection, flow cytometry and kit assay. F-G Seahorse analyzer was used to assess OCR and ECAR profile. Data are presented as mean ± SD from three independent biological replicates (n = 3). Exact p-values are indicated in the figure

To further validate the role of HIF-1α in mediating cardiomyocyte pathology, loss-of-function experiments were performed using shRNA targeting HIF-1α in ISO-treated HL-1 cells. Efficient knockdown of HIF-1α expression was confirmed by qRT-PCR and WB (Figure S3A-B). HIF-1α knockdown recapitulated the protective effects observed with KC7F2. Specifically, knockdown of HIF-1α significantly ameliorated ISO-induced reduction in cell viability (Figure S3C) and attenuated cardiomyocyte hypertrophy, as evidenced by decreased cell size and reduced mRNA expression of hypertrophic markers ANP, BNP, and β-MHC (Figure S3D-G). In addition, HIF-1α knockdown markedly suppressed ISO-induced production of inflammatory cytokines IL-1β, IL-6, and TNF-α (Figure S3H). At the mitochondrial level, HIF-1α knockdown rescued ISO-induced mitochondrial mass reduction (Figure S3I), restored ATP production (Figure S3J), improved OCR (Figure S3K), and normalized ECAR (Figure S3L). Collectively, these genetic loss-of-function data corroborated the pharmacological findings and further confirmed that HIF-1α was a critical downstream mediator of HTRA1-induced cardiomyocyte pathology.

MZF1 bound to the HTRA1 promoter and mediated cardiomyocyte hypertrophy and inflammation

To identify potential upstream regulators of HTRA1, MZF1 expression was examined in atrial tissues HF mice. Increased MZF1 expression was observed in atrial tissue (Fig. 4A-B). Given that MZF1 functions as a transcription factor, whether it directly regulates HTRA1 transcription was next investigated. Bioinformatic analysis using JASPAR predicted a potential binding site for MZF1 within the HTRA1 promoter region (Fig. 4C). ChIP assay confirmed this interaction in vivo, as evidenced by enriched binding of MZF1 to the HTRA1 promoter in atrial tissue from HF mice (Fig. 4D). In vitro, MZF1 knockdown was performed in ISO-treated HL-1 cells and its knockdown efficiency was verified (Figure S2E-F). Knockdown of MZF1 downregulated both MZF1 and HTRA1 expression in ISO-treated HL-1 cells (Fig. 4E). ChIP assay revealed that MZF1 binding of the HTRA1 promoter was markedly reduced following MZF1 knockdown in ISO-treated HL-1 cells (Fig. 4F). Dual-luciferase reporter assays demonstrated that knockdown of MZF1 specifically reduced the activity of the HTRA1 promoter in HL-1 cells transfected with the wild-type HTRA1 promoter construct (HTRA1 WT), but not in cells transfected with the construct harboring mutations in the MZF1 binding site (HTRA1 MUT) (Fig. 4G). ChIP assay confirmed specific binding of MZF1 to the WT promoter in HL-1 cells transfected with the HTRA1 WT construct, whereas no such binding was observed in cells transfected with the HTRA1 MUT construct (Fig. 4H). After establishing that MZF1 regulated HTRA1 transcription, its functional role in cardiomyocyte pathology was next examined. MZF1 knockdown decreased MZF1 expression, attenuated cardiomyocyte hypertrophy (reduced cell size and ANP/BNP/β-MHC expression), and mitigated inflammation (decreased IL-1β, IL-6, TNF-α levels) in ISO-treated HL-1 cells (Fig. 4I-M). Collectively, these results demonstrated that MZF1 knockdown reduced HTRA1 promoter activity and attenuated cardiomyocyte hypertrophy and inflammation in ISO-treated HL-1 cells.

Fig. 4.

MZF1 bound to the HTRA1 promoter and mediated cardiomyocyte hypertrophy. Mice were subjected to daily intraperitoneal injection of ISO or saline for two weeks. A-B MZF1 mRNA and protein levels in atrial tissue from CON and HF mice were analyzed by qRT-PCR and WB. C Schematic diagram of the predicted MZF1 binding site on the HTRA1 promoter via JASPAR. D ChIP assay showing the interaction between MZF1 and HTRA1 promoter in atrial tissue. In vitro, HL-1 cells were treated with ISO or vehicle and transfected with sh-MZF1 or sh-NC. E qRT-PCR analysis of MZF1 and HTRA1 mRNA expression. F ChIP assay showing MZF1 binding to the HTRA1 promoter in HL-1 cells. HL-1 cells were co-transfected with a wild-type (WT) or mutant (MUT) HTRA1 promoter reporter plasmid, along with sh-MZF1 or sh-NC, and treated with ISO. G Dual-luciferase reporter assay was employed to confirm the effect of sh-MZF1 on the HTRA1 promoter activity. H ChIP assay using chromatin from cells transfected with the HTRA1 WT or MUT reporter plasmids, confirming specific binding of MZF1 and HTRA1. I-J Efficiency of MZF1 knockdown in ISO-treated HL-1 cells in was verified by qRT-PCR and WB. K qRT-PCR analysis of hypertrophic markers ANP, BNP, and β-MHC mRNA expression in ISO-treated HL-1 cells transfected with sh-MZF1 or sh-NC. L-M The effects of sh-MZF1 on cell size and inflammatory cytokine levels in ISO-treated cardiomyocytes were assessed via α-actinin staining and ELISA, respectively. Data are presented as mean ± SD. For in vivo experiments, n = 10 mice per group (n = 6 for atrial fibrillation analyses); for molecular analyses, n = 3 mice per group. For in vitro experiments, n = 3 independent biological replicates. Exact p-values are indicated in the figures

NSUN2-mediated m5C modification of MZF1 was required for MZF1 stabilization by ALYREF

Bioinformatic analysis using the RMBase3 database predicted the presence of m5C modification sites on MZF1 mRNA, while the StarBase database identified ALYREF as a potential RNA-binding protein of MZF1. These findings prompted us to further investigate whether MZF1 is subject to m5C modification and recognized by ALYREF. qRT-PCR and WB further confirmed that ALYREF was upregulated in HF mouse atrial tissues (Fig. 5A-B). MeRIP-qPCR assay revealed that the m5C modification level on MZF1 mRNA was increased in HF model (Fig. 5C). RIP-qPCR assay confirmed that the binding of ALYREF protein to MZF1 mRNA was enhanced in atrial tissues of HF mice (Fig. 5D). ChIP assay further showed that MZF1 binding to the HTRA1 promoter was enriched in atrial tissues of HF mice (Fig. 5E). To functionally validate the role of ALYREF in MZF1 regulation, in vitro experiments were performed in HL-1 cells with ALYREF knockdown (as verified in Figure S2I-J) followed by ISO treatment. ALYREF knockdown reduced both ALYREF and MZF1 expression in ISO-treated HL-1 cells (Fig. 5F-G). RIP assay confirmed that the binding of ALYREF protein to MZF1 mRNA was enhanced in ISO-treated HL-1 cells, which was inhibited by ALYREF knockdown (Fig. 5H). Furthermore, knockdown of ALYREF decreased the m5C modification level on MZF1 mRNA in ISO-treated HL-1 cells (Fig. 5I). Importantly, ALYREF knockdown reduced the stability of MZF1 mRNA in ISO-treated HL-1 cells (Fig. 5J). Subsequently, the functional impact of ALYREF knockdown on cardiomyocyte pathology was also evaluated. ALYREF knockdown attenuated ISO-induced cardiomyocyte hypertrophy and inflammation (Fig. 5K-N). In addition, ALYREF knockdown ameliorated ISO-induced mitochondrial dysfunction, as evidenced by reduced mitochondrial ROS, restored membrane potential, increased ATP production, improved OCR, and normalized ECAR (Fig. 5O-U). Taken together, in ISO-treated HL-1 cells, ALYREF stabilized MZF1 mRNA by recognizing its m5C modification, thereby promoting cardiomyocyte hypertrophy, inflammation, and mitochondrial dysfunction.

Fig. 5.

ALYREF influenced the m5C modification of MZF1 mRNA and regulated cardiomyocyte pathology. Mice were subjected to daily intraperitoneal injection of ISO or saline for two weeks. A-B qRT-PCR and WB analysis of ALYREF mRNA and protein expression in atrial tissue from CON and HF mice. C The m5C modification level on MZF1 mRNA in mouse atrial tissues was measured by MeRIP-qPCR. D RIP-qPCR analysis of ALYREF binding to MZF1 mRNA. E ChIP analysis of MZF1 binding to the HTRA1 promoter. In vitro validation of ALYREF function. HL-1 cells were treated with ISO or vehicle and transfected with sh-ALYREF or sh-NC. F-G The mRNA and protein levels of ALYREF and MZF1 were evaluated by qRT-PCR and WB. (H) RIP assay showing the binding of ALYREF protein to MZF1 mRNA. I The m5C modification level on MZF1 mRNA in HL-1 cells was assessed by MeRIP-qPCR. (J) Actinomycin D chase assay was conducted to analyze MZF1 mRNA stability. K-L Cell size of HL-1 cells was detected by α-actinin staining. M The mRNA expression of hypertrophic markers (ANP, BNP and β-MHC) was quantified by qRT-PCR. N ELISA assay demonstrating the inflammatory cytokine levels (IL-1β, IL-6 and TNF-α). O MitoSOX Red fluorescence showing mitochondrial ROS level. P ATP content was assessed by kit assay. Q-R Flow cytometry of JC-1 staining was performed to measure mitochondrial membrane potential. S-U OCR and ECAR were evaluated by Seahorse analyzer. Data are presented as mean ± SD. For in vivo experiments, n = 10 mice per group (n = 6 for atrial fibrillation analyses); for molecular analyses, n = 3 mice per group. For in vitro experiments, n = 3 independent biological replicates. Exact p-values are indicated in the figures

To identify the potential m5C writer responsible for MZF1 modification, the expression of NSUN2 and NSUN6 was examined in atrial tissues from HF mice. WB analysis revealed that both NSUN2 and NSUN6 were upregulated in HF mouse atrial tissues, with NSUN2 showing a more pronounced increase (Figure S4A). To investigate whether NSUN2 mediated m5C modification of MZF1, loss-of-function experiment was performed in ISO-treated HL-1 cells. Efficient knockdown of NSUN2 was first validated by qRT-PCR and WB (Figure S4B-C), and ALYREF overexpression efficiency was verified (Figure S4D-E). Knockdown of NSUN2 significantly reduced MZF1 mRNA and protein expression, MZF1 mRNA stability and m5C enrichment, and the interaction of NSUN2 with MZF1 mRNA in ISO-treated HL-1 cells, which were reversed by MZF1 overexpression but not by ALYREF overexpression (Figure S4F-J). Moreover, whether NSUN2 knockdown phenocopies the protective effects of ALYREF knockdown on cardiomyocyte pathology was examined. NSUN2 knockdown significantly increased cell viability, attenuated cardiomyocyte hypertrophy, and suppressed inflammatory cytokine production, which was counteracted by overexpression of ALYREF or MZF1 (Figure S4K-N). These results collectively indicated that NSUN2-mediated m5C modification of MZF1 was required for ALYREF-dependent stabilization of MZF1.

ALYREF promoted cardiomyocyte pathology via upregulating the MZF1/HTRA1 axis

Having established that ALYREF stabilized MZF1 mRNA and that MZF1 transcriptionally activated HTRA1, we next sought to determine whether ALYREF promoted cardiomyocyte pathology through the MZF1/HTRA1 axis. To address this, ISO-treated HL-1 cells were subjected to ALYREF knockdown in combination with MZF1 overexpression (as validated in Figure S2G-J). Knockdown of ALYREF downregulated the expression of ALYREF, MZF1, and HTRA1 in ISO-treated HL-1 cells, whereas overexpression of MZF1 specifically reversed the effects of ALYREF knockdown on MZF1 and HTRA1 expression (Fig. 6A-B). While ALYREF knockdown attenuated cardiomyocyte hypertrophy and inflammation, concurrent overexpression of MZF1 reversed these protective effects, leading to increased cell size, elevated expression of hypertrophic markers, and enhanced levels of inflammatory cytokines (Fig. 6C-F). These results demonstrated that ALYREF acted upstream of MZF1 to regulate the HTRA1-mediated cardiomyocyte pathology.

Fig. 6.

ALYREF affected cardiomyocyte hypertrophy via regulating MZF1/HTRA1 axis. ISO-treated HL-1 cells were co-transfected with sh-ALYREF, with or without the MZF1 overexpression. A-B The mRNA and protein expression levels of ALYREF, MZF1, and HTRA1 were analyzed by qRT-PCR and WB. C-D The α-actinin staining showing cell size of HL-1 cells. (E-F) The levels of myocardial hypertrophy markers (ANP, BNP and β-MHC) and inflammatory cytokines (IL-1β, IL-6 and TNF-α) were respectively quantified by qRT-PCR and ELISA. Data are presented as mean ± SD from three independent biological replicates (n = 3). Exact p-values are indicated in the figures

ALYREF knockdown ameliorated cardiac hypertrophy, fibrosis, cardiac dysfunction, and atrial fibrillation susceptibility in HF mice

To evaluate the therapeutic potential of targeting ALYREF in vivo, ALYREF knockdown was performed in the ISO-induced HF mouse model using AAV-mediated delivery, with knockdown efficiency validated in atrial tissues (Figure S2K-L). Consistent with our in vitro findings, knockdown of ALYREF significantly reduced the expression of ALYREF, MZF1, and HTRA1 in atrial tissues from HF mice (Fig. 7A-B). IHC analysis further revealed that ALYREF knockdown significantly reduced HIF-1α protein expression in atrial tissues of HF mice (Fig. 7C). Next, the impact of ALYREF knockdown on the pathological features of HF mice was assessed. ALYREF knockdown also attenuated inflammation (reduced IL-1β, IL-6 and TNF-α plasma levels) (Fig. 7D-E), decreased cardiac hypertrophy (lower HW/BW ratio and cardiomyocyte size) (Fig. 7F-H), reduced cardiac fibrosis (Fig. 7I), and improved cardiac histopathology (alleviated myocardial disarray and inflammatory infiltration) in HF mouse atrial tissues (Fig. 7J). Echocardiography revealed that ALYREF knockdown improved cardiac function (increased FS) without significant changes in ventricular dimensions (Fig. 7K-L, 8A-B). Most importantly, ALYREF knockdown markedly reduced the severity of atrial fibrillation in HF mice. ECG recordings showed that ALYREF knockdown improved rhythm regularity and partially restored P-waves (Fig. 7M). Quantitative analysis further revealed that ALYREF knockdown significantly reduced atrial fibrillation incidence, the number of atrial fibrillation episodes, total atrial fibrillation burden, and ventricular rate during atrial fibrillation episodes compared with the HF + sh-NC group (Fig. 7N-Q). Taken together, these results delineate a novel signaling pathway wherein ALYREF, by enhancing m5C modification and stability of MZF1 mRNA, promotes MZF1-mediated transcriptional upregulation of HTRA1. This HTRA1 upregulation leads to HIF-1α activation, mitochondrial dysfunction, cardiomyocyte hypertrophy, and inflammation, ultimately contributing to HF development and increased susceptibility to atrial fibrillation.

Fig. 7.

ALYREF knockdown improved cardiac remodeling, dysfunction, and atrial fibrillation susceptibility in HF mice. C57BL/6J mice received tail vein injection of AAV carrying sh-ALYREF or sh-NC two weeks prior to ISO injection to induce HF. (A-B) qRT-PCR and WB analysis of ALYREF, MZF1, and HTRA1 mRNA and protein expression in atrial tissue. (C) IHC staining and quantification of HIF-1α in atrial tissues. (D-E) Plasma levels of inflammatory cytokines were quantified by ELISA. (F) HW/BW ratio assessing ventricular hypertrophy. (G-I) WGA staining and Masson’s trichrome staining were respectively performed to evaluate cardiomyocyte cross-sectional area and cardiac fibrosis. (J) H&E staining of atrial tissue sections demonstrating the changes in histopathological conditions. (K-L) Cardiac function status in mice was assessed through echocardiographic images, as well as quantification of FS. (M) Representative ECG tracings from each group revealing mouse heart rate condition. (N-Q) Atrial fibrillation incidence, number of atrial fibrillation episodes, total atrial fibrillation burden, and ventricular rate during atrial fibrillation episodes were assessed. Data are presented as mean ± SD (n = 10 per group; n = 6 for atrial fibrillation analyses; n = 3 for molecular analyses). Exact p-values are indicated in the figures

Discussion

The clinical management of coexisting atrial fibrillation and HF remains a major challenge [7]. Current therapeutic strategies, such as rate control, anticoagulation, and catheter ablation, can alleviate symptoms but often fail to reverse progressive cardiac remodeling and show limited efficacy in preventing new-onset atrial fibrillation in HF patients [32, 33]. This therapeutic gap stems from an incomplete understanding of the specific molecular mechanisms driving atrial fibrillation in the context of HF [2]. This study identified a novel signaling axis in which NSUN2-mediated m5C modification of MZF1 enabled its stabilization by ALYREF, which in turn transcriptionally activated HTRA1. The upregulation of HTRA1 led to cardiomyocyte mitochondrial dysfunction, hypertrophy, and inflammation through activation of the HIF‑1α signaling pathway, collectively driving pathological cardiac remodeling and arrhythmogenesis. This finding not only reveals a new molecular link between HF and atrial fibrillation but also provides potential therapeutic targets for intervention.

A complex bidirectional relationship exists between HF and atrial fibrillation, with shared pathological underpinnings rooted in sustained myocardial stress‑induced cardiac remodeling [6, 34]. In this context, the concept of atrial cardiomyopathy has been proposed as a mechanistic framework to bridge atrial fibrillation and heart failure, as atrial cardiomyopathy is increasingly recognized as a progressive, multifaceted disease of the atrial myocardium involving structural, electrical, mechanical, and molecular remodeling that can both precede and sustain arrhythmia and hemodynamic deterioration [35]. In this study, we successfully recapitulated the phenotypes of systolic dysfunction, cardiac hypertrophy, fibrosis, and increased atrial fibrillation susceptibility using an ISO‑induced mouse HF model, consistent with clinical observations and previous animal studies [36, 37]. Notably, atrial-specific analysis in this work revealed pronounced fibrosis, elevated inflammatory cytokine levels, and remodeling of gap junction (Cx40 and Cx43) and calcium-handling proteins (SERCA2a and NCX1), collectively indicating significant atrial remodeling that may contribute to the increased atrial fibrillation susceptibility. Furthermore, this study revealed significant upregulation of HTRA1 both in this model and in public atrial fibrillation datasets, directly linking HTRA1 expression to post‑HF atrial fibrillation susceptibility for the first time. As a serine protease, HTRA1 has been implicated in extracellular matrix remodeling and inflammatory responses in other disease contexts [13, 38]. Our work extended HTRA1 function to cardiomyocytes, demonstrating that HTRA1 overexpression was sufficient to induce mitochondrial dysfunction, inflammation, and a hypertrophic phenotype in vitro. This aligns with reports of HTRA1 impairing mitochondrial function and promoting inflammation in retinal pigment epithelial cells [17, 39], suggesting HTRA1 may serve as a common node linking cellular stress to metabolic dysregulation.

Mitochondrial integrity is essential for maintaining cardiomyocyte energy supply, calcium cycling, and redox balance [40, 41]. Mitochondrial dysfunction is a shared pathological feature of HF and atrial fibrillation [42, 43]. Furthermore, HIF‑1α activation has been shown to induce mitochondrial dysfunction in cardiomyocytes [44, 45]. Our data demonstrated that HTRA1 overexpression exacerbated ISO‑induced reduction in mitochondrial mass, loss of membrane potential, decreased ATP production, excessive ROS generation, and impaired cellular oxygen consumption, which were significantly reversed by the HIF‑1α inhibitor KC7F2. Studies indicate that aberrant HIF‑1α activation can drive metabolic reprogramming in the heart, promoting glycolysis while suppressing oxidative phosphorylation, thereby contributing to pathological hypertrophy [46, 47]. Thus, HIF‑1α activation is likely a core mechanism through which HTRA1 induces mitochondrial dysfunction and cardiomyocyte hypertrophy. Future studies should clarify whether HTRA1 regulates HIF‑1α directly or indirectly via its protease activity.

Research indicates that nuclear receptor RXR heterodimers, together with HDAC enzymes, regulate HTRA1 transcription in NSCLC cells [48]. Although bioinformatics analyses have linked HTRA1 to the HF immune microenvironment and disease progression [49, 50], the mechanisms controlling its transcriptional expression in cardiomyocytes remained unexplored. Here, this study provided the first evidence that the transcription factor MZF1 was a direct upstream regulator of HTRA1. MZF1 expression was increased in atrial tissues of HF mice, and it promoted HTRA1 transcription by binding to a specific region of HTRA1 promoter. MZF1, a zinc‑finger transcription factor, has been reported to regulate cell proliferation and inflammatory responses [51]. Our study newly positioned MZF1 in the context of pathological cardiac remodeling, showing that MZF1 knockdown effectively attenuated cardiomyocyte hypertrophy and inflammation, closely associated with its suppression of HTRA1 expression. The identification of this MZF1‑HTRA1 regulatory axis offers a fresh perspective for understanding the mechanisms of myocardial gene expression regulation in HF models.

Another key advance of this study lies in uncovering the upstream epitranscriptomic regulation of the MZF1‑HTRA1 pathway. The m5C modification has recently emerged as a critical player in regulating mRNA stability and translation efficiency [24]. ALYREF, an m5C “reader,” has been shown to influence neurological disorders and cancer metastasis by modulating m5C modifications on target RNAs [52, 53]. A recent study further demonstrates that ALYREF affects cardiac remodeling by regulating extracellular matrix protein synthesis in cardiac fibroblasts [54]. In this study, we identified NSUN2 as the m5C writer responsible for MZF1 modification, and demonstrated that ALYREF bound to and stabilized MZF1 mRNA in an m5C-dependent manner. Through bioinformatic prediction and experimental validation, we found that ALYREF bound to and stabilized MZF1 mRNA. Knockdown of ALYREF reduced the m5C modification level and stability of MZF1 mRNA, thereby downregulating MZF1 and its target gene HTRA1. This finding tightly linked RNA modification to the transcriptional network governing cardiac remodeling. While upregulation of ALYREF in HF samples has been reported [29, 54], our study expanded its functional significance by showing that NSUN2-mediated m5C deposition on MZF1 was required for ALYREF-dependent stabilization, driving downstream pro‑hypertrophic and pro‑inflammatory signaling. In vivo data confirmed that ALYREF knockdown significantly ameliorated ISO‑induced cardiac hypertrophy, fibrosis, systolic dysfunction, and atrial fibrillation susceptibility in HF mice, strongly supporting the pathophysiological importance of the NSUN2/ALYREF/MZF1/HTRA1 axis in the progression from HF to atrial fibrillation.

Several limitations of this study should be noted. First, the mechanistic investigations were primarily conducted in HL‑1 cells and mouse models; their generalizability to human cardiomyocytes and patient tissues requires further validation. Second, the specific downstream effectors through which HTRA1 affects mitochondrial function remain incompletely defined. Third, whether ALYREF participates in HF regulation by modifying other mRNA targets warrants exploration. Fourth, although our in vitro functional assays demonstrated that ALYREF knockdown ameliorated mitochondrial dysfunction, ultrastructural analysis of mitochondria by electron microscopy was not performed. Fifth, while MeRIP-qPCR was used to assess m5C enrichment on MZF1 mRNA, this method lacks single-base resolution. Future studies employing RNA bisulfite sequencing are warranted to identify the precise m5C sites critical for ALYREF binding and MZF1 stabilization. Finally, developing small‑molecule inhibitors that specifically target key nodes in this pathway, such as the m5C‑reading activity of ALYREF or the protease activity of HTRA1, and evaluating their therapeutic potential represent important directions for future translational research.

Conclusion

This study systematically elucidates the role of the NSUN2/ALYREF/MZF1/HTRA1 axis in HF-associated atrial fibrillation through in vitro and in vivo experiments. We found that in the HF model, NSUN2-mediated m5C modification of MZF1 enabled its stabilization by the RNA m5C reader ALYREF, promoting MZF1 expression, which in turn transcriptionally activated the downstream target gene HTRA1. HTRA1 upregulation led to cardiomyocyte mitochondrial dysfunction, hypertrophy, and inflammation via the HIF‑1α activation, ultimately driving cardiac remodeling and increasing susceptibility to atrial fibrillation. These findings not only reveal a novel molecular link between heart failure and atrial fibrillation but also provide a theoretical basis for developing new therapeutic strategies targeting this pathway.

Supplementary Information

Below is the link to the electronic supplementary material.

Abbreviations

ALYREF

Aly/REF export factor

ATP

Adenosine triphosphate

CCK‑8

Cell Counting Kit‑8

ChIP

Chromatin immunoprecipitation

ECAR

Extracellular acidification rate

ECG

Electrocardiogram

ELISA

Enzyme‑linked immunosorbent assay

FS

Fractional shortening

H&E

Hematoxylin and eosin

HF

Heart failure

HIF‑1α

Hypoxia‑inducible factor‑1α

HTRA1

High temperature requirement factor A1

HW/BW

Heart weight‑to‑body weight

IHC

Immunohistochemistry

IL‑1β

Interleukin‑1β

IL‑6

Interleukin‑6

ISO

Isoproterenol hydrochloride

LVIDd

Left ventricular internal diameter at end‑diastolic

LVIDs

Left ventricular internal diameter at end‑systolic

m5C

5‑Methylcytosine

MeRIP

Methylated RNA Immunoprecipitation

MZF1

Myeloid zinc finger 1

NSUN2

NOP2/Sun RNA methyltransferase 2

OCR

Mitochondrial oxygen consumption rate

qRT‑PCR

Quantitative Real‑Time PCR

RIP

RNA immunoprecipitation

ROS

Reactive oxygen species

TNF‑α

Tumor necrosis factor‑α

WB

Western blot

WGA

Wheat germ agglutinin

Author contributions

Siqi Nian: Conceptualization, Investigation, Data curation, Funding, Formal analysis, Writing-original draft, Writing-review & editing. Lulu Zhao: Conceptualization, Investigation, Data curation, Formal analysis, Writing-original draft, Writing-review & editing. Xiaokang He: Conceptualization, Investigation, Data curation, Formal analysis, Writing-original draft, Writing-review & editing. Ping Xia: Formal analysis, Software, Methodology, Writing-review & editing. Yanqing Liu: Software, Methodology, Writing-review & editing. Baotong Hua: Conceptualization, Investigation, Project administration, Resources, Funding, Supervision, Writing-review & editing. Jun Li: Conceptualization, Project administration, Supervision, Writing-review & editing. Hongxing Zhang: Conceptualization, Investigation, Project administration, Resources, Supervision, Writing-review & editing.

Funding

This work was supported by the Joint Special Project for Basic Research of the Department of Science and Technology of Yunnan Province and Kunming Medical University (Grant No. 202401AY070001-341).

Data availability

All data generated and/or analyzed during the current study are available from the corresponding author on reasonable request.

Declarations

Ethics approval

All procedures involving animals were approved by the Animal Ethics Committee of The First Affiliated Hospital of Kunming Medical University (No. 2025DF052).

Consent for publication

The data used in this study has never been published before.

Conflict of interest

The authors declare that they have no competing interests.