MBNL1 regulates isoproterenol‐induced myocardial remodelling in vitro and in vivo

Myocardial remodelling is a common phenomenon in cardiovascular diseases, which threaten human health and the quality of life. Due to the lack of effective early diagnosis and treatment methods, the molecular mechanism of myocardial remodelling should be explored in depth. In this study, we observed the high expression of MBNL1 in cardiac tissue and peripheral blood of an isoproterenol (ISO)‐induced cardiac hypertrophy mouse model. MBNL1 promoted ISO‐induced cardiac hypertrophy and fibrosis by stabilizing Myocardin mRNA in vivo and in vitro. Meanwhile, an increase in MBNL1 may induce the apoptosis of cardiomyocytes treated with ISO via TNF‐α signalling. Interestingly, MBNL1 can be activated by p300 in cardiomyocytes treated with ISO. At last, Myocardin can reverse activate the expression of MBNL1. These results suggest that MBNL1 may be a potential target for the early diagnosis and clinical treatment of myocardial remodelling.


| INTRODUC TI ON
Myocardial remodelling is common in many cardiovascular diseases such as coronary atherosclerotic heart disease, hypertension, hypertrophic cardiomyopathy and heart failure and is due to the activation of multiple stimulating factors and intracellular signal transduction mechanisms. 1 Chronic heart failure and heart failure are important links in the progression of various cardiovascular diseases. Cardiac remodelling mainly includes myocardial hypertrophy and interstitial fibrosis. The structure, metabolism and function of cardiac tissues change according to gene expression states. Moreover, cardiac morphology changes, cardiac hypertrophy, cardiomyocyte apoptosis and extracellular matrix network remodelling occur and lead to chronic heart failure, arrhythmia and eventually cardiogenic shock.
Myocardial hypertrophy plays an important role in cardiac remodelling. 2,3 It cannot only stimulate collagen synthesis and promote interstitial fibrosis but also induce the activation of matrix metalloproteinases (MMPs), which degrade collagen, and eventually lead to ventricular dilation. Therefore, blocking cardiac hypertrophy is a key to the prevention and treatment of cardiac remodelling. At the molecular level, hypertrophic stimuli exert effects on cell membrane effector molecules and sensitize signal transduction pathways that are related to cardiac hypertrophy in the cytoplasm and activate the transcription of hypertrophy-related molecules 4 ; therefore, the discovery of molecules that can inhibit or promote myocardial hypertrophy would be helpful in finding ways to block the occurrence and development of myocardial hypertrophy.
Myocardin consists of 35 amino acids arranged in a helix-junction-helix structure and is specifically expressed in embryonic, adult myocardium and vascular smooth muscle cells. [5][6][7] Previous studies have confirmed that Myocardin can specifically bind the DNA sequence CarG box (CC (A/T) 6 G box) in the promoter region of target genes and activate the expression of the target genes by forming a complex with serum response factor (SRF), which also result in the regulation of the development, growth, differentiation, and apoptosis of heart and smooth muscle cells. [8][9][10][11][12] In particular, the high expression of Myocardin is an important inducer of the occurrence and development of myocardial hypertrophy and has attracted wide attention [13][14][15] ; therefore, elucidating the regulation mechanism of Myocardin expression is important.
RNA-binding proteins (RBPs) are a class of proteins that directly bind to RNAs. As key participants in post-transcriptional regulation, RBPs control RNA processing at multiple levels, including alternative splicing, RNA stability, RNA localization and translation efficiency. [16][17][18][19] The correlation of RBPs with cell differentiation and metabolic diseases has been confirmed. [20][21][22] MBNL1, which belongs to the family of muscle blind-like (MBNL) proteins, is closely linked to the occurrence and development of many diseases owing to their regulation of RNA metabolism in several ways, including splicing of alternative exons, selection of a polyadenylation site in pre-mR-NAs, influencing the stability and differential localization of mRNAs, and processing of miRNAs. [23][24][25][26][27][28] To date, the relationship between MBNL1 and myocardial hypertrophy remains unclear.
We used bioinformatics tools and found that Myocardin mRNA and the promoter region of MBNL1 contain multiple MBNL1 binding sites and a Myocardin-binding site, respectively. Therefore, we speculate that there may be an interaction between MBNL1 and Myocardin. This article focuses on the interaction between MBNL1 and Myocardin and their underlying molecular mechanism in regulating myocardial hypertrophy.

| Animal studies
Two-month-old wild-type C57BL/6 mice were purchased from the Animal Experimental Centre of Hubei Academy of Preventive Medicine. All mice were fed in SPF level animal laboratory under controlled environmental conditions (12 hours light/dark cycle and room temperature 20 ± 5°C). And all mice were free to obtain standard laboratory food and water. Mice with different treatments were infused with ISO for 2 weeks and saline-infused mice served as controls.
After 14 days, mice were anaesthetised with 50 mg/kg pentobarbital sodium (Sigma) via intraperitoneal injection. No toe pinch reflex confirmed adequate anaesthesia. Then, the mice were killed by cervical dislocation, and hypertrophic analysis was performed.

| Lentivirus shRNA plasmids construction
The target shRNAs against mouse MBNL1, Myocardin and p300 were inserted into pαMHC-clone26. A negative control showing no significant homology to any mouse or human gene was inserted into pαMHC-clone26. ShRNAs lentivirus were produced by cotransfecting 293T cells with the lentivirus expression plasmid and packaging plasmid. Interference efficiency was measured by realtime PCR and Western blotting.

| Lentivirus transduction of primary mouse cardiomyocyte
Lentivirus was added to primary mouse cardiomyocyte cultured in low glucose DMEM supplemented with 10% FBS at a ratio of 50 multiplicity of infection (MOI). After overnight incubation, fresh medium was replaced.

| Quantitative realtime PCR (qRT-PCR)
According to the manufacturer's protocol, the total RNA was extracted by using the Total RNA Extraction Kit (OMEGA). Then, the total RNA extracted was reverse transcribed using M-MLV Reverse Transcriptase (Promega). Finally, realtime PCR was performed in a StepOne Realtime PCR System (Thermo) with Fast SYBR Green Master Mix (Thermo). The relative expression levels of related genes were normalized to GAPDH. The primers for the realtime PCR

| Luciferase constructs and luciferase assay
The wild type and mutant of MBNL1 promoter were inserted into pGL3 and amplified with primers as follow:

| Immunofluorescence analysis
The treated cells were immobilized with 4% paraformaldehyde for 15 minutes, washed with PBS and blocked with normal goat serum for 20 minutes at room temperature. Then incubated with Rhodamine Phalloidin (R415, Thermo) for 15 minutes and washed with PBS, the samples were observed by laser scanning confocal microscope (OLYMPUS).

| RNA pulldown
In vitro, Myocardin mRNA and TNF-α mRNA were transcribed by

| RNA immunoprecipitation (RIP)
2 × 10 7 primary mouse cardiomyocytes were taken to carry out RIP experiments combined with anti-IgG or anti-MBNL1. The RNA in the coprecipitation was then extracted and detected by realtime PCR.

| Histological analysis (H&E, WGA, Masson's Trichrome staining)
The excised mouse hearts were fixed with formaldehyde at 4°C for 72 hours and embedded in paraffin. Then, the paraffin-embedded tissues were cut into 3-4 µm slices. The sections were then dewaxed with xylene for 15 minutes, rehydrated in gradient alcohols, and stained with haematoxylin-eosin (H&E) or wheat germ agglutinin (WGA) or Masson trichrome. Five random fields were selected for capture under an inverted microscope (Olympus, Tokyo, Japan).

| TUNEL assay
After different lentiviral transductions, apoptotic cells in each well (24-well plates) or in each section (3-4 µm slices) were visualized using TUNEL according to the manufacturer's instructions (Promega). The experimental results of TUNEL assay were observed under a confocal microscope (OLYMPUS F3000).

| TNFα secretion assays
Enzyme-linked immunosorbent assay (ELISA, BMS607-3, Invitrogen) was used to detect the secretion of TNF-α in cell cultures according to the manufacturer's instructions. Briefly, supernatant of primary cardiomyocytes with different treatment was collected and centrifuged (500 g) for 5 minutes. Then, the supernatant was subjected to ELISA assay.

| Chromatin immunoprecipitation (ChIP) assay
We performed chromatin immunoprecipitation assay by using a commercial chip assay kit (#56383, CST) according to the manufacturer's instructions. Briefly, each group to be tested was incubated with 1% formaldehyde to cross-link DNA-protein complexes. Cells were then collected, washed three times with ice-cold PBS and lysed in SDS lysis buffer. After centrifugation, the lysates were ultrasonically treated and the DNA was cut into 200-1000 bp fragments. We then immunoprecipitate the cross-linked protein at 4°C overnight using an anti-Myocardin antibody (DF2434, Affinity). IgG acted as the negative control. Finally, DNA from the coprecipitated material was extracted and used as a template for PCR to find the binding site of Myocardin. The primers for PCR analysis are shown below: F: 5′-CTGCATGAGTCAGTTTTCCA-3′, R: 5′-ATT AACT TGTCGGCAGAGAAG-3′.

| Trichostatin A (TSA) treatment
In this study, we chose TSA (Sigma) as an inhibitor of p300. We treated primary cardiomyocytes with 100 ng/mL TSA for 24 hours.
Then, cells were replaced the normal medium and carried out ISO induction. After ISO induction, proteins were collected from cells for Western blotting analysis.

| Measurement of RNA stability
Primary cardiomyocytes were transduced with MBNL1 or sh-MBNL1 lentivirus. Transcription was stopped using Actinomycin D (ACD,

| Statistical analysis
High throughput sequencing data were obtained from the GEO database (GSE12 9090). Quantitative data are expressed as mean ± SEM.
Statistical analysis of differences between two groups was performed by Student's t test. A one-way analysis of variance followed by Tukey test was performed to compare differences among multiple groups. Statistical analysis was performed with GraphPad Prism 8. P < .05 indicated statistically significant difference.

| MBNL1 is highly expressed in myocardial hypertrophy mouse model
By analysing the data in the GEO database (GSE12 9090), we found that MBNL1 was highly expressed myocardial hypertrophy samples ( Figure 1A,B). To verify the above results, a cardiac hypertrophy mouse model was induced using ISO ( Figure 1C-F). MBNL1 expression in mouse cardiomyocytes was detected using realtime PCR. As shown in Figure 1G, the level of MBNL1 was significantly higher in myocardial hypertrophy mice than in normal mice. Interestingly, the level of MBNL1 in the peripheral blood of mice treated with ISO was also significantly higher than that in the control group ( Figure 1H).
This suggests that MBNL1 may be a biomarker of myocardial remodelling.

| The overexpression of MBNL1 contributes to ISO-induced myocardial remodelling in vivo and in vitro
To investigate the potential effect of MBNL1 on cardiac remodelling, a lentivirus was used to produce a cardiac-specific MBNL1 overexpression C57BL/6 mouse model. The cardiac MBNL1 protein in each group of mice was confirmed by using Western blotting ( Figure 2A). On this basis, we performed intraperitoneal injection of ISO in half of the mice in each group, and the other half was injected with saline as a control. H&E and WGA staining techniques were used to analyse the heart tissue sections of mice in each group. Significantly increased cross-sectional areas of the heart of MBNL1 mice were observed ( Figure 2B-E). Moreover, as shown in Figure 2G-I, the heart weight to bodyweight ratios (HW/BW), heart weight to tibia length ratios (HW/TL) and lung weight to bodyweight ratios (LW/BW) in the MBNL1 group, after injection with ISO for two weeks, were all significantly increased compared with those in the control. Notably, increased fibrosis induced by ISO was observed in the hearts of mice with overexpressed MBNL1 compared with those of normal mice treated with ISO ( Figure 2F). These results suggest that MBNL1 promotes myocardial hypertrophy in vivo.
F I G U R E 1 MBNL1 is highly expressed in myocardial hypertrophy mouse model. A and B, Difference in MBNL1 expression in the heart of hypertrophic mice and normal mice is shown in the volcano plot and heat map. C, Gross hearts under natural light after treatment with ISO. Scale bars represent 2 mm. D, E, and F, Ratio of heart weight to bodyweight (HW/BW), heart weight to tibia length (HW/TL) and lung weight to bodyweight (LW/BW) in different groups (Each group n = 11, **, P < .01). G and H, mRNA levels of MBNL1 in the heart or peripheral blood of mice after intraperitoneal injection with ISO or saline (Each group n = 11, *, P < .05, **, P < .01, # , P > .05) To explore the function of MBNL1 in vitro, we used lentivirus to overexpress or silenced MBNL1 in primary cardiomyocytes. As a result, a significantly increased cross-sectional area of the primary myocardium after treatment with ISO, in the context of MBNL1 expression and compared with that of the control, was observed by myocardial cytoskeleton staining with phalloidin ( Figure 2J). Then, the changes in the expression of two cardiac hypertrophy-specific genes, namely ACTN2 and ANP, in the primary myocardium with overexpressed or silenced MBNL1 were detected using realtime PCR and Western blotting. We observed that ACTN2 and ANP expression increased or significantly decreased compared with the control group, whether at the mRNA or protein level in overexpressed or silenced MBNL1 ( Figure 2K-O). These data preliminarily confirmed that MBNL1 can promote ISO-induced cardiac hypertrophy in vitro.

| MBNL1 increases Myocardin expression by binding to UGCU at the 3'-UTR of Myocardin mRNA
As

| MBNL1 may regulate ISO-induced cardiomyocyte apoptosis via TNF-α
Excessive cardiac hypertrophy or myocardial fibrosis leads to cardiomyocyte apoptosis. In this part, we wanted to identify the effect F I G U R E 2 The overexpression of MBNL1 contributes to ISO-induced myocardial remodelling in vivo and in vitro. A, Detection of MBNL1 overexpression effects. B, Gross hearts under natural light. Scale bars represent 2 mm. C, Representative images of the cross-sections of ventricles stained with H&E (20× magnification). D, 400× microscopic views of H&E sections. Scale bars represent 50 µm. E and F, The myocyte areas and cardiac fibrosis areas were measured using wheat germ agglutinin (WGA) staining and Masson's trichome staining, respectively. Scale bars represent 50 µm; each group n = 6, *, P < .05, **, P < .01. G, H and I, Ratio of heart weight to bodyweight (HW/BW), heart weight to tibia length (HW/TL) and lung weight to bodyweight (LW/BW) in different groups (Each group n = 6, *, P < .05, **, P < .01). J, The surface areas of cardiac myocytes after treatment with saline or ISO were determined using immunostaining and confocal microscopy.
The cytoskeleton was stained with phalloidin, and the nuclei were counterstained with 4′,6-diamidino-2-phenylindole (DAPI). Scale bars represent 50 µm. L, Detection of the effect of shRNA against MBNL1 in primary cardiomyocytes. (n = 3, *, P < .05, **, P < .01, #, P > .05). K, M, N and O, Primary cardiomyocytes were transduced with MBNL1 or sh-MBNL1 lentivirus. The changes in ACTN2, ANP and MBNL1 were detected using realtime PCR and Western blotting. (n = 3, **, P < .01) of MBNL1 on cardiomyocyte apoptosis in ISO-induced myocardial remodelling. As shown in Figure 5A, the overexpression of MBNL1 significantly increased the apoptosis rate of cardiomyocytes treated with ISO in vivo. Furthermore, the results of the terminal deoxynucleotidyl transferase dUTP nick end labeling (TUNEL) assay confirmed the above results in ISO-induced primary cardiomyocytes ( Figure 5B). In various acute and chronic diseases, apoptosis is often caused by an increase in free radicals and inflammatory factors (TNFα, IL-6, etc). Consistently, we found that TNF-α secretion significantly increased in the culture medium of primary cardiomyocytes with overexpressed MBNL1 treated with ISO ( Figure 5C); however, the opposite result was obtained from cells with silenced MBNL1 and ISO treatment ( Figure 5D). Meanwhile, the expression of TNF-α was positively correlated with MBNL1 at both the RNA and protein levels ( Figure 5E,F). Through bioinformatics analysis and prediction, we found a binding site of MBNL1 in the 3'-UTR of TNF-α mRNA ( Figure 5G). Moreover, RNA pulldown and RIP assays confirmed that MBNL1 can bind to the 3'-UTR of TNF-α mRNA ( Figure 5H,I). These data reveal that MBNL1 may regulate ISO-induced cardiomyocyte apoptosis via TNF-α.

| The activation of MAPK and JNK signalling pathways can up-regulate the expression of MBNL1 via p300 signalling in myocardial remodelling
We have revealed the function and molecular mechanism of MBNL1 in regulating myocardial remodelling; however, it is unclear how MBNL1 is activated during myocardial remodelling. Therefore, we

| Myocardin can reverse activate the transcription of MBNL1
We

| D ISCUSS I ON
Hibernating myocardium has great effects on myocardial remodelling, which is caused by the decrease of coronary artery blood  45 Excessive fibrosis and cardiac hypertrophy may lead to pathological myocardial remodelling and even heart failure. In addition, human heart failure has been associated with reduced cardiac Nav1.5 Na + channel current and SCN5A mRNA abundance. 46 Studies have found that MBNL1 can regulate the splicing of SCN5A mRNA; such regulation lead to heart conduction defects in DM. 47 These studies suggest that MBNL1 may play a regulatory role in myocardial hypertrophy, as we also have confirmed in this study.
Moreover, myocardial remodelling can cause cardiomyocyte apoptosis through the production of TNF-α in large amounts and lead to heart failure. 48 Here, we found that cardiomyocytes with overexpressed F I G U R E 4 MBNL1 regulates myocardial hypertrophy and fibrosis via Myocardin in vivo and in vitro. A, Detection of Myocardin knock down effects in different tissues of mice. B, Gross hearts under natural light. Scale bars represent 2 mm. C, Representative images of crosssections of the ventricles stained with H&E (20 × magnification). D, 400× microscopic views of H&E sections. E and F, The myocyte areas and cardiac fibrosis areas were measured using WGA staining and Masson's trichome staining, respectively. Scale bars represent 50 µm, each group n = 6, *, P < .05, **, P < .01, # , P > .05. G, H and I, Ratio of heart weight to bodyweight (HW/BW), heart weight to tibia length (HW/TL) and lung weight to bodyweight (LW/BW) in different groups (Each group n = 6, *, P < .05, **, P < .01, # , P > .05). J, The surface areas of cardiac myocytes after treated with saline or ISO were determined using immunostaining and confocal microscopy. The cytoskeleton was stained with phalloidin, and the nuclei were counterstained with DAPI. Scale bars represent 50 µm. K-O, Primary cardiomyocytes with Myocardin knock down were transduced with MBNL1 or sh-MBNL1. The changes in ACTN2, ANP, Myocardin and MBNL1 were detected using realtime PCR and Western blotting (n = 3, *, P < .05, **, P < .01) MBNL1 can produce more TNF-α than normal after treatment with ISO. The prolonged presence of a large amount of TNF-α in the microenvironment of cardiomyocytes will damage the cells. This suggests that the sustained high concentration of MBNL1 may contribute to heart failure via Myocardin and TNF-α. Certainly, TNF-α is not the only regulator of cardiomyocyte apoptosis. In fact, the mechanism of ISO-induced cardiomyocyte apoptosis with MBNL1 overexpression is relatively complex; we will continue to study it in depth.
F I G U R E 5 MBNL1 may regulate ISO-induced cardiomyocyte apoptosis via TNF-α. A, The apoptosis of cardiomyocytes in mice with overexpressed MBNL1 compared with normal mice treated with ISO was measured using TUNEL assay. Scale bars represent 50 µm. B, The apoptosis of cardiomyocytes after different treatments was detected using TUNEL assay under phosphate-buffered saline (PBS) and ISO conditions. Scale bars represent 50 µm. C and D, The secretion of TNF-α in primary cardiomyocytes with overexpressed or silenced MBNL1 after treatment with ISO or PBS was measured using ELISA. (n = 3, **, P < .01). E and F, The expression levels of TNF-α in different groups of primary cardiomyocytes were measured using realtime PCR and Western blotting. (n = 3, **, P < .01 Meanwhile, the sustained activation of MBNL1 can induce cardiomyocytes to secrete TNF-α and promote the apoptosis of cardiomyocytes ( Figure S1). These findings not only provide molecular targets for the diagnosis and treatment of myocardial hypertrophy but also provide a theoretical basis for basic clinical research. Unfortunately, limited by experimental conditions, functional assessment such as in vivo ultrasound parameter tracking was lacking in this study.

ACK N OWLED G EM ENTS
This work was supported by the National Natural Science Foundation of China (No.31570764).

CO N FLI C T O F I NTE R E S T
The authors declare that they have no conflict of interest.