Establishment and functional studies of a model of cardiomyopathy with cardiomyocyte-specific conditional knockout of Arhgef18

Cardiomyopathies (CMs) are the most prevalent category of myocardial diseases that occur in childhood, and include dilated cardiomyopathy (DCM), hypertrophic cardiomyopathy (HCM) and left-ventricular noncompaction (LVNC). CMs are associated with severe clinical symptoms, poor prognoses and limited therapeutic options, and the aetiology remains largely unexplained (Maron et al., 2006; Mirza et al., 2022; Eldemire et al., 2024). Many studies have reported associations between CMs and variant genes of sarcomeres (Eldemire et al., 2020; Gao et al., 2024), ion channels (Shan et al., 2008; Mann et al., 2012) and mitochondria (Tang et al., 2010). However, these genetic factors account for only 20-30% of the CM phenotype. Notably, the same variants can result in diverse myocardiophenotypes. Furthermore, the phenotype remains unstable in previously reported genetic animal models (Shan et al., 2008; Mann et al., 2012; Wu et al., 2022; Rhee et al., 2018). Consequently, establishment of an animal model with a stable phenotype to explore the pathogenesis of CM is important for advancing the prevention and treatment of such disease.

Our research group recruited a representative family with LVNC, and conducted whole-exome sequencing on the peripheral blood of the proband and their relatives. Additionally, we performed transcriptome sequencing on induced pluripotent stem cell-derived cardiomyocytes of the proband. With the integrated analysis of transcriptome and whole-exome sequencing data, we identified the potential pathogenic gene Rho/Rac guanine nucleotide exchange factor 18 (ARHGEF18) (Li et al., 2022; Wan et al., 2023). ARHGEF18 is a guanine nucleotide exchange factor that activates Rho GTPase by catalysing the conversion of GDP to GTP. It plays a crucial role in cytoskeletal rearrangement, cell proliferation and migration. Previous studies have reported the potential association of ARHGEF18 with cardiovascular system development (Li et al., 2018). However, its involvement in the pathogenesis of CMs remains unclear. In this study, we used a gene-editing technology to construct a cardiomyocyte-specific Arhgef18 conditional knockout (cKO) mouse model. Our aim was to investigate the functional role of Arhgef18 in CMs by examining the associated CM phenotype. This work provides a reliable animal model for studying the genetic underpinnings of CMs.

After two rounds of crosses between Arhgef18^flox/+ mice and Nkx2.5-Cre mice, a total of 141 offspring were obtained, including 35 Arhgef18^flox/flox; Nkx2.5-Cre (Arhgef18^fl/fl cKO) mice, 39 Arhgef18^flox/+; Nkx2.5-Cre (Arhgef18^fl/+ cKO) mice, 38 Arhgef18^flox/flox (Arhgef18^fl/fl) mice and 29 Arhgef18^flox/+ mice, with each group accounting for approximately one-quarter of the total population (Fig. 1A). There were 53 female and 47 male mice, with each sex representing approximately half the sample size (Fig. 1B). These data suggest that Arhgef18 cKO mice follow Mendelian inheritance patterns, and that mice with different genotypes are viable and fertilizable.

To investigate the potential role of Arhgef18 in CMs, we selected Arhgef18^fl/fl, Arhgef18^fl/+ cKO and Arhgef18^fl/fl cKO mice for subsequent experimentation. PCR was employed to confirm the genotype (Fig. 1C). Mice with both 432 bp and 603 bp bands were classified as Arhgef18^fl/fl cKO, whereas those with bands at 432 bp, 375 bp and 603 bp were classified as Arhgef18^fl/+ cKO. Subsequently, we assessed Arhgef18 mRNA and protein expression levels in heart, liver, lung and kidney tissues at the age of 4 weeks. Arhgef18 mRNA expression was decreased by ∼45%, and protein expression was decreased by ∼24%, in the heart of Arhgef18^fl/+ cKO mice compared to that of Arhgef18^fl/fl mice. In addition, Arhgef18 mRNA expression was decreased by ∼84%, and protein expression was decreased by ∼49%, in the heart of Arhgef18^fl/fl cKO mice compared to that of Arhgef18^fl/fl mice. Meanwhile, Arhgef18 mRNA expression in the heart of Arhgef18^fl/fl cKO mice was decreased by ∼72%, and protein expression was decreased by ∼33%, compared to that in the heart of Arhgef18^fl/+ cKO mice (Fig. 1D-F). Notably, there were no differences in Arhgef18 expression between liver, lung or kidney tissues. These data suggested that Arhgef18 was knocked down in the cardiomyocytes of Arhgef18 cKO mice.

The body weight of the mice was measured at various time points during the cardiac development period. The results revealed no significant differences in body weight changes between Arhgef18^fl/+ cKO mice and Arhgef18^fl/fl mice. However, compared with that of Arhgef18^fl/fl mice, the growth weight of Arhgef18^fl/fl cKO mice was slowed down after the age of 3 weeks (Fig. 2A). Additionally, the survival curves revealed no significant differences among the groups (Fig. S1A), indicating a potential growth restriction associated with alterations in cardiac morphology and function that does not impact survival. Consequently, cardiac systolic function was monitored over an extended period. Short-axis M-mode ultrasound images revealed that there were no significant differences in cardiac function or chamber size in Arhgef18^fl/+ cKO mice compared to Arhgef18^fl/fl mice at any time point (Fig. 2B,C). Conversely, the cardiac function of Arhgef18^fl/fl cKO mice declined progressively after 4 weeks, as evidenced by a reduction in the left-ventricular ejection fraction (LVEF) and left-ventricular fraction shortening (LVFS), accompanied by an increase in the left-ventricular inner diameter (LVID) and left-ventricular volume, as well as thinning of the left-ventricular anterior wall (LVAW) (Fig. 2B,C). These differences became increasingly pronounced with advancing age. Notably, similar differences were also observed between Arhgef18^fl/+ cKO and Arhgef18^fl/fl cKO mice, with the data presented in Table S1. We also found that Nppa and Nppb mRNA expression levels within the ventricular muscle tissue of Arhgef18^fl/fl cKO mice were significantly elevated compared with those within the ventricular muscle tissue of Arhgef18^fl/fl mice (Fig. 2D). In summary, these findings indicate that Arhgef18^fl/fl cKO mice exhibited early cardiac dysfunction at 4 weeks of age, followed by a progressive decline in cardiac function. However, overall survival rates were not negatively impacted in these mice.

The development of CM is a chronic and progressive process (Maron et al., 2006), thus necessitating the study of the early pathogenesis. In this study, we focused on the initial stages of CM phenotype development by examining 4-week-old mice. The gross examination and Haematoxylin and Eosin (H&E) staining of the whole heart revealed that Arhgef18^fl/fl cKO mice exhibited significantly larger bilateral ventricular volumes than those of Arhgef18^fl/+ cKO and Arhgef18^fl/fl mice (Fig. 2E). These findings indicate that Arhgef18 cKO mice present a CM phenotype characterized by biventricular enlargement, thinning of the ventricular walls and diminished left-heart systolic function, confirming the successful establishment of a genetic animal model for CMs.

Arhgef18 is involved in cytoskeletal rearrangement (Zaritsky et al., 2017). Here, we aimed to explore the alteration of the cytoskeleton in Arhgef18 cKO mice. We identified alterations in cytoskeletal proteins [vinculin, αβ-tubulin, cTNT (also known as TNNT2) and α-actinin] in the ventricles of 4-week-old mice with CM by quantitative real-time PCR (q-PCR) analysis of the encoding mRNAs and western blotting. In Arhgef18^fl/fl cKO mice, the mRNA expression levels of cytoskeletal proteins such as vinculin, αβ-tubulin and α-actinin were significantly reduced compared to those in Arhgef18^fl/fl mice (Fig. 3A). Similarly, the protein levels of vinculin, αβ-tubulin and cTNT were decreased compared to those in Arhgef18^fl/fl mice (Fig. 3B,C). In Arhgef18^fl/+ cKO mice, mRNA levels of vinculin and αβ-tubulin were lower than those in Arhgef18^fl/fl mice, with corresponding reductions in protein levels of vinculin, αβ-tubulin and cTNT (Fig. 3A-C). These findings suggest that Arhgef18 plays a crucial role in maintaining cytoskeletal integrity in cardiomyocytes, and its knockout leads to downregulation of key cytoskeletal components, potentially contributing to observed cardiac dysfunction and CM phenotype.

Given the diverse cell types in myocardium, including cardiomyocytes, fibroblasts, smooth muscle cells and endothelial cells, we employed immunofluorescence to directly assess the impact of Arhgef18 on the cytoskeleton of cardiomyocytes. Compared to Arhgef18^fl/fl mice, Arhgef18^fl/+ cKO and Arhgef18^fl/fl cKO mice exhibited reduced fluorescence intensities for cytoskeletal proteins, including vinculin, αβ-tubulin, α-actinin and cTNT, with the most pronounced decreases observed in Arhgef18^fl/fl cKO mice (Fig. 3D,E). These data indicate that Arhgef18 knockout results in reduced expression of cytoskeletal proteins in cardiomyocytes, suggesting the presence of cardiomyocyte skeleton rearrangements in Arhgef18^fl/fl cKO mice.

Arhgef18 is also involved in regulation of cell polarity (Zaritsky et al., 2017; Herder et al., 2013). The expression level of cell polarity proteins (PARD3, SCRIB and CRB2) was validated via q-PCR analysis of the encoding mRNAs, western blotting and immunofluorescence. Compared to Arhgef18^fl/fl mice, both Arhgef18^fl/fl cKO mice and Arhgef18^fl/+ cKO mice presented significantly lower Pard3 and Scrib mRNA expression, and significantly lower SCRIB and CRB2 protein expression (Fig. 4A-C). Immunofluorescence further demonstrated a marked decrease in PARD3 protein localization and intensity in the Arhgef18^fl/fl cKO mice compared to Arhgef18^fl/fl mice (Fig. 4D,E). These data indicate that Arhgef18 knockdown resulted in decreased expression of cardiomyocyte polarity proteins, suggesting the presence of myocardial tissue polarity disorders in Arhgef18 cKO mice.

In the present study, we successfully employed the Cre/LoxP system to construct an Arhgef18 cKO mouse model. These mice exhibited CM phenotype characterized by bilateral ventricular enlargement, thinning of the ventricular walls and diminished cardiac contractile function. Additionally, the cytoskeleton rearrangements and cell polarity disorders were observed in Arhgef18 cKO mice. These results suggest that Arhgef18 is a pathogenic gene involved in the development of CMs and is potentially linked to the cytoskeleton and cell polarity.

The protein encoded by ARHGEF18 is a guanine nucleotide exchange factor belonging to the Rho GTPase GEF family. ARHGEF18 catalyses the conversion of GDP to GTP and directly activates Rho GTPases, which play critical roles in cytoskeletal rearrangement, cell growth and migration (Xu et al., 2019; Zaritsky et al., 2017; Mitchell et al., 2011). Most studies have indicated that the ARHGEF18 plays a crucial role in regulating the polarity state of neuroepithelial cells and promoting retinal development by activating the RHOA–ROCK2 signalling pathway (Herder et al., 2013; Loosli, 2013; Arno et al., 2017). Clinical studies have reported that mutations in ARHGEF18 lead to increased susceptibility to nonidiopathic pulmonary hypertension associated with coronary artery disease (Li et al., 2018), and systematic knockdown of Arhgef18 results in embryonic death in mice (Beal et al., 2021). However, foundational research on ARHGEF18 in the field of CMs has not been reported. Previously, a typical LVNC line was obtained from clinical samples and subjected to whole-exome sequencing in our group (Guo et al., 2021). Concurrently, transcriptome sequencing was performed on human induced pluripotent stem cell-derived cardiomyocytes, and Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) analyses of the transcriptome sequencing results revealed significant enrichment of many differentially expressed genes related to the cytoskeleton and cell polarity, and associated with the Rho GTPase cycle pathway (Guo et al., 2021). Ultimately, combining the results on whole-exome sequencing and transcriptome sequencing indicated downregulated expression of the pathogenic gene, ARHGEF18 (Li et al., 2022; Wan et al., 2023). Phenotypic alterations caused by genetic abnormalities are the gold standard for validating the genetic basis of clinical disease, so the establishment and phenotypic validation of the Arhgef18 cKO animal model could facilitate investigation of the mechanism of CMs. Our previous study used Myh6-Cre to construct an Arhgef18 cKO mouse model that exhibits abnormal expression of cytoskeletal and polarity proteins in the myocardium but does not show CM-related phenotypes (Wan et al., 2023). The lack of phenotype was attributed to the late onset of recombination induced by Myh6-Cre [embryonic day (E)8.5]. To investigate the role and mechanism of Arhgef18 in CM, we reconstructed the Arhgef18 cKO mouse model using Nkx2.5-Cre, which initiates Cre recombinase activity earlier at E7.5.

In the present study, we observed that Arhgef18^fl/fl cKO mice presented a phenotype characterized by bilateral ventricular enlargement, thinning of the ventricular walls and left-ventricular hypoconstriction beginning at 4 weeks of age. This finding is consistent with the pronounced deceleration in body weight growth observed in Arhgef18^fl/fl cKO mice beginning at 3 weeks of age. Interestingly, the phenotype changed dynamically with the age of the Arhgef18 cKO mice, and significant differences were observed between the Arhgef18^fl/fl cKO and Arhgef18^fl/+ cKO mice. These phenotypic characteristics are consistent with the clinical presentations of DCM, with ∼30-40% of DCM aetiology being genetically linked (Khan et al., 2022). Numerous studies have indicated that common pathogenic genes associated with DCM – such as TTN (Vissing et al., 2021), LMNA (Le Dour et al., 2022), MYH7 (de Frutos et al., 2022), FLNC (Verdonschot et al., 2020), TNNT2 (Luedde et al., 2010) and ZASP/Cypher (also known as LBD3) (Lv et al., 2021; Vatta et al., 2003; Zheng et al., 2009) – predominantly encode cytoskeletal proteins, nuclear membrane proteins and ion channel proteins. Extensive research has established a correlation between DCM and LVNC. For example, LVNC has been identified in young carriers of lamin A/C mutations in familial cases of DCM (Hermida-Prieto et al., 2004). Other investigations have demonstrated that genetic mouse models carrying mutant genes identified in patients with LVNC, such as TNNT2 (Luedde et al., 2010) and ZASP/Cypher (Vatta et al., 2003; Zheng et al., 2009), present typical morphological characteristics associated with DCM but lack the clinical phenotype characteristic of LVNC, and these findings are consistent with the results of our study. Our findings suggest that although numerous genes have been implicated in CMs to date, the intricate genetic mechanisms of CMs require further study, and new related genes could be discovered.

The pathogenic gene in this study, ARHGEF18, was identified within the LVNC family line. However, the Arhgef18 cKO mice did not exhibit the LVNC-characteristic phenotype for several potential reasons. (1) The absence of Yap1 in the dense layer of the myocardium leads to a thinner dense layer and trabecular overgrowth. Conversely, the lack of Yap1 in the trabecular layer of the myocardium does not influence the growth of the dense layer (Tian et al., 2017). In this research, Nkx2.5 (also known as Nkx2-5) – a marker for the trabecular myocardium – through which deletion of Arhgef18 in mouse trabecular myocardium does not lead to thinning of the dense layer. This finding is consistent with those of previous studies (Tian et al., 2017) and reinforces that growth of the dense layer is largely independent of the trabecular layer. (2) Both Myh6-Cre (Huang et al., 2021) and Nkx2.5-Cre (Stanley et al., 2002) are cardiomyocyte-specific tools designed to knock out Arhgef18 exclusively in cardiomyocytes. However, trabeculae formation and compaction involve multiple signal pathways originating from endocardial, myocardial and epicardial sources (Liu et al., 2018). Previous reports utilizing various Cre lines (Myh6-Cre, Tie2-Cre, Nfatc1-Cre and Apj-CreER) to downregulate Ino80 expression in cardiomyocytes, epicardium or endothelial cells showed that only endothelial deletion during embryonic development results in an LVNC phenotype, with other cardiac cell types with Ino80 knockdown not showing an associated phenotype (Rhee et al., 2018). Consequently, we propose developing a new Arhgef18 cKO mouse model using endothelial-specific Cre mice or simultaneous cardiomyocyte- and endothelial-specific Cre mice. This approach offers promising avenues for investigating CMs caused by Arhgef18. (3) Environmental factors play an important role in the development of disease caused by genetic variants (Hazebroek et al., 2015; Thomas, 2004), and studies have reported that young Prdm16 cKO mice exhibit normal cardiac function, but metabolic stress interventions induce heart failure and cardiac hypertrophy (Cibi et al., 2020). Therefore, the incorporation of stressors, including aortic arch narrowing surgery and high-fat chow feeding, provides avenues for validation of the phenotype of Arhgef18 cKO mice. (4) Deletion of ARHGEF18 appears to cause the LVNC phenotype in humans, but not in mice, suggesting a possible requirement for a ‘specific’ genetic background. This specific genetic background has also been reported in HCM, including large variation in cardiac hypertrophy among different mouse strains and even the absence of cardiac hypertrophy in some mice (Semsarian et al., 2001). Therefore, the phenotype of the mouse model is regulated by multiple factors, and phenotypic verification of the pathogenic gene could elucidate the genetic mechanism of disease onset. The inconsistent phenotype of the Arhgef18 cKO mice in this study provides new ideas and research directions for exploring the genetics of CMs.

The cytoskeleton and cell polarity are crucial for maintaining normal cellular structure. In cardiomyocytes, the cytoskeleton plays a vital role in sustaining both contractile and diastolic functions; proper polarity is essential for coordinating electrical signalling between cardiomyocytes and ensuring synchronized contractions. Furthermore, there are interactions among these components, as well as regulatory mechanisms governing them. An imbalance in these interactions might contribute to the pathological processes underlying CMs (Grego-Bessa et al., 2023; Burbaum et al., 2021). In this study, we assessed the expression levels of skeleton and polarity proteins in Arhgef18 cKO mice, both in the heart tissue and in cardiomyocytes. The results indicated that cytoskeletal proteins (vinculin, αβ-tubulin, cTNT and α-actinin) and polarity proteins (PARD3, SCRIB and CRB2) were downregulated in Arhgef18 cKO mice compared with in controls. These findings confirm that cardiomyocyte-specific Arhgef18 cKO results in the aberrant expression of cytoskeletal and polarity proteins. Numerous studies have reported a significant association between the cytoskeleton and CMs. Investigations have identified several regulatory pathways linked to the cytoskeleton and cell polarity, including, but not limited to, the VEGF pathway, PCP pathway and Rho pathway (Tian et al., 2015; Ichida et al., 2001; Theis et al., 2006). In our previous study, we screened the Rho GTPase cycle pathway, which could be a potential signalling pathway of Arhgef18 in the cytoskeleton and polarity (Guo et al., 2021). We will follow up with RNA sequencing on the mouse heart to look for a related pathway, which will also be the next target signalling pathway to study.

Our study results are similar to those from our previous research (Wan et al., 2023), with both Arhgef18^flox/flox; Nkx2.5-Cre mice and Arhgef18^flox/flox; Myh6-Cre mice showing abnormalities in cytoskeletal and polarity proteins, confirming that the reduction in these proteins occurs later than the critical period of trabecular development. Arhgef18^flox/flox; Nkx2.5-Cre mice exhibit a CM phenotype, whereas Arhgef18^flox/flox; Myh6-Cre mice do not show any related CM phenotypes. This indicates that Arhgef18 likely plays a crucial role in myocardial development during the 7.5-8.5 days of embryonic development, further confirming that ARHGEF18 is a key gene in the occurrence and development of CMs.

In conclusion, we successfully established an Arhgef18 cKO mouse model, which exhibited the clinical phenotype characteristic of CMs. Our findings confirmed cytoskeletal and polarity abnormalities in Arhgef18 cKO mice. This study provides a new animal model for investigating the pathogenesis of CMs, and carried out in-depth studies on cytoskeletal and cell polarity, providing basic research evidence for the diagnosis, treatment and prevention of CMs.

Arhgef18^flox/+ mice were purchased from Cyagen Biosciences, Inc. As shown in Fig. S2A, a LoxP site was inserted both upstream of exon 3 and downstream of exon 4. Nkx2.5-Cre mice were purchased from Shanghai Model Organisms Center, Inc. The Nkx2.5-Cre mice, which utilize the Nkx2.5 promoter to drive Cre recombinase expression for LoxP site-specific recombination, were employed to achieve cardiomyocyte-specific knockout of Arhgef18 (Zhang et al., 2018; Boczonadi et al., 2014; Courtney et al., 2021; Moses et al., 2001; Kim et al., 2018; Stanley et al., 2002). As shown in Fig. S2B, Arhgef18^flox/flox; Nkx2.5-Cre (Arhgef18^fl/fl cKO), Arhgef18^flox/+; Nkx2.5-Cre (Arhgef18^fl/+ cKO), Arhgef18^flox/flox (Arhgef18^fl/fl) and Arhgef18^flox/+ mice were obtained after two rounds of crossbreeding with Arhgef18^flox/+ and Nkx2.5-Cre mice.

In this study, all mice were housed at The Animal Centre for the Children's Hospital of Chongqing Medical University, Chongqing, China, under the environmental regulations of a 12-h light/dark cycle with free access to food and water. The mice were specific pathogen-free C57BL/6 mice. Male and female mice were randomly divided into different experimental groups. All experiments were performed by operators unaware of genotype. No mice were excluded from the analysis. The body weights of the mice were measured at 1, 2, 3, 4, 8 and 12 weeks of age, and the times of births and natural deaths were recorded. The mice were sacrificed via cervical dislocation at 4 weeks of age to obtain heart, liver, lung and kidney tissues. The experimental protocol was approved by the Experimental Animal Management Committee, Children's Hospital of Chongqing Medical University, China (licence number: CHCMU-IACUC2021 1028001).

Genomic DNA (gDNA) was isolated from mouse toe tissue using a Mouse Direct PCR Kit (Bimake). gDNA amplification was performed with a PCR assay employing 2× M-PCR OPTI™ Mix (Bimake) and gene primers. The DNA was separated by 2% agarose gel electrophoresis, exposed and analysed on an iBright gel imager (Invitrogen). Among the Arhgef18-Loxp genotypes, Arhgef18^fl/fl was characterized by a band of 432 bp, Arhgef18^flox/+ by bands of 432 bp and 375 bp, and wild type by a band of 375 bp. The Nkx2.5-Cre genotype was characterized by a band of 603 bp, and the wild-type genotype was characterized by no band. The primer sequences are provided in Table S2.

Total RNA was isolated from heart, liver, lung and kidney tissues using a SimplyP Total RNA Extraction Kit (BioFlux). Single-strand cDNA synthesis was performed using an Evo M-MLV RT Mix Kit with gDNA Clean (Accurate Biology). cDNA amplification was performed according to a quantitative real-time PCR assay employing a SYBR Green premix Pro Taq HS qPCR Kit (Accurate Biology). 18S (also known as Rn18s) was used as the housekeeping gene to normalize the transcription levels of the genes of interest. Relative gene expression was calculated via the 2^−ΔΔCt method. The primer sequences are provided in Table S3. Four to six mice heart specimens per group were used.

Total protein was isolated from heart, liver, lung and kidney tissues via an ice-cold mixture of RIPA buffer, protease inhibitor and phosphatase inhibitor (Beyotime). The protein concentration was measured using a BCA kit (Beyotime). The proteins were separated by SDS‒PAGE and then transferred to polyvinylidene fluoride membranes (Millipore). The membranes were blocked for 1 h with 5% milk and then incubated overnight with primary antibody at 4°C, followed by 2 h with secondary antibody at room temperature. The membranes were exposed to enhanced chemiluminescence reagent and visualized with a gel imaging analyser developer (Bio-Rad). HSP90 was used for normalization. Quantitative analysis was performed with Image Lab software (Bio-Rad). Antibody information is provided in Table S4. Four to six mice heart specimens per group were used.

Echocardiography was performed with a Vevo 3100 system (FUJIFILM). B-mode and M-mode ultrasound images of paraspinal short-axis sections of the mice were acquired to measure relevant parameters. The mice were anaesthetized with isoflurane (3-4 l/min induction, 0.5-1 l/min maintenance), and their body temperature (36-37°C) and heart rate (480-520 beats/min) were maintained during echocardiographic inspection. The measurements were recorded as the average of at least three consecutive cardiac cycles. The LVEF, LVFS, LVAW, LVID and left-ventricular volume were analysed with Vevo LAB 5.8.0. Five to eight mice per group were used.

The mice were sacrificed via the cervical dislocation method, and the whole heart was removed and washed in PBS. After fixation in 4% paraformaldehyde for 48 h, gradient dehydration, paraffin embedding, 5 μm continuous heart longitudinal sectioning, deparaffinization, hydration, Haematoxylin staining, Eosin staining, dehydration, transparency, sealing and air drying were performed. The images were captured and analysed on a microscanning instrument (Shengqiang Full Slide Scanner, Shengqiang Technology in China). Three mice per group were used.

Cardiomyocytes were obtained from 4-week-old mice via the Langendorff perfusion system (Chengdu Instrument Factory). Cardiomyocytes were fixed with 4% paraformaldehyde for 15 min in the dark, permeabilized with 0.1% Triton X-100 for 10 min, and blocked with 5% bovine serum albumin goat serum blocking solution for 1 h at room temperature. The cells were incubated with primary antibodies overnight at 4°C, incubated with secondary antibodies at room temperature in the dark for 1 h, and stained with 4′,6-diamidino-2-phenylindole (DAPI) in the dark for 10 min. Cardiomyocyte images were acquired and statistically analysed under a fluorescence inverted microscope (Nikon) using a glass-bottomed cell culture dish (Biosharp). Antibody information is provided in Table S4. Three mice per group were used, with >20 cardiomyocytes per mouse.

All data were analysed using Prism 8.3.0 (GraphPad Software). Normally distributed continuous variables are presented as the mean±s.d. and were compared via one-way ANOVA followed by Tukey's, Dunnett's or Dunn's multiple comparison post hoc tests. The survival rates of the mice were analysed via the log-rank test. P-values less than 0.05 were considered statistically significant.

We appreciate the contribution and support of the Children's Hospital of Chongqing Medical University for providing the opportunity to carry out this study, including the use of a whole-slide scanner, fluorescence inverted microscope, and technical support from the research team.