MicroRNA‐30c suppresses the pro‐fibrogenic effects of cardiac fibroblasts induced by TGF‐β1 and prevents atrial fibrosis by targeting TGFβRII

Abstract Atrial fibrosis serves as an important contributor to atrial fibrillation (AF). Recent data have suggested that microRNA‐30c (miR‐30c) is involved in fibrotic remodelling and cancer development, but the specific role of miR‐30c in atrial fibrosis remains unclear. The purpose of this study was to investigate the role of miR‐30c in atrial fibrosis and its underlying mechanisms through in vivo and in vitro experiments. Our results indicate that miR‐30c is significantly down‐regulated in the rat abdominal aortic constriction (AAC) model and in the cellular model of fibrosis induced by transforming growth factor‐β1 (TGF‐β1). Overexpression of miR‐30c in cardiac fibroblasts (CFs) markedly inhibits CF proliferation, differentiation, migration and collagen production, whereas decrease in miR‐30c leads to the opposite results. Moreover, we identified TGFβRII as a target of miR‐30c. Finally, transferring adeno‐associated virus 9 (AAV9)‐miR‐30c into the inferior vena cava of rats attenuated fibrosis in the left atrium following AAC. These data indicate that miR‐30c attenuates atrial fibrosis via inhibition of CF proliferation, differentiation, migration and collagen production by targeting TGFβRII, suggesting that miR‐30c might be a novel potential therapeutic target for preventing atrial fibrosis.

improves renal function. Taken together, these data suggest that miR-30c is an anti-fibrotic miRNA mediating organ fibrosis. However, it remains unclear whether miR-30c is involved in atrial fibrosis and through what underlying mechanisms.
Fibrosis is a complex process resulting from activation of select signalling pathways, such as TGF-b1 signalling. 11 TGF-b1 is essential for the development of atrial fibrosis. [12][13][14] The levels of TGF-b1 are increased in the plasma and atria of AF patients. 15,16 Moreover, animal models have revealed that transgenic overexpression of TGF-b1 results in atrial-specific fibrosis and increases AF susceptibility. [17][18][19] Thus, we used TGF-b1-induced CFs to establish an in vitro model of atrial fibrosis. TGF-b1 mediates its effects via interactions among transmembrane receptors, including type I (TGFbRI), type II (TGFbRII) and type III (TGFbRIII). Upon binding to TGF-b1, TGFbRII recruits and phosphorylates TGFbRI, which initiates signal transduction via Smad proteins. 20 TGFbRII can transduce the TGF-b1 signalling from cell membrane to the cytoplasm and then regulate a series of physiological or pathological processes including cell proliferation, differentiation 21,22 and collagen production. 23 We identified TGFbRII as a direct target of miR-30c using a microRNA target prediction tool (RNAhybrid). Recent studies show that TGFbRII is a target gene of several anti-fibrotic miRNAs. Liang et al 24 suggested that miR-153, an anti-fibrotic miRNA, inhibits the migration of pulmonary fibroblasts and reduces pulmonary fibrosis by targeting TGFbRII. As shown in the study by Zou et al 25 , miR-19a-3p/19b-3p inhibits autophagymediated fibrogenesis by targeting TGFbRII. Another study suggested that miR-145 prevents TGFb-dependent ECM accumulation and fibrosis in smooth muscle cells by targeting TGFbRII. 26 Additionally, a recent study showed that constitutive expression of dominant-negative TGFbRII in the posterior left atrium in a canine heart failure model could sufficiently attenuate fibrosis-induced changes in atrial conduction and restitution to decrease AF. 27 Collectively, these data suggested that miR-30c and TGFbRII were critical for TGF-b1-mediated fibrosis. Therefore, we hypothesize that miR-30c represses TGFbRII expression, subsequently decreasing CF proliferation, differentiation, migration and collagen synthesis. Our present work aimed to elucidate the effects of miR-30c on atrial fibrosis and TGF-b1-induced fibrosis performed with in vivo and in vitro experimental approaches, and this may provide a potential therapeutic target to prevent atrial fibrosis. Fifteen Sprague-Dawley rats weighing 200-250 grams were randomly divided into sham (n = 7) and AAC (n = 8) groups. The rats were anaesthetized intraperitoneally with sodium pentobarbital (40 mg/kg). The AAC procedures were as follows: After opening the abdomen, we separated the abdominal aorta from the right renal artery branch above 5 mm, and a 22-gauge needle was placed next to the abdominal aorta. A suture was securely tied around the needle and the aorta. After ligation, the needle was quickly removed, and the abdominal cavity was closed. The rats in the Sham group underwent an open-abdomen procedure without AAC. All surgical procedures were performed under sterile conditions. One of the Sham died from deep anesthesia and one of the AAC died from ruptured abdominal aorta. After constriction with 8 weeks, all rats are killed and the atriums of heart were removed for detecting the expressions of miR-30c and TGFbRII.
One part was rapidly frozen in liquid nitrogen for subsequent RNA isolation, protein isolation, and the remaining part was kept in 4% (w/v) paraformaldehyde (PFA) for histological analysis.

| Masson's trichrome staining
Atrial tissue samples were fixed with 4% (w/v) PFA for 48 hours, subjected to alcohol dehydration, embedded in paraffin and sliced into 4 lm thick sections, which underwent Masson's trichrome staining to highlight fibres. The percentage of fibrosis was measured as fibrosis areas/total given field areas 9 100%.

| Cell culture, stimulation and transfection
293T cells were purchased from the American Type Culture Collection (ATCC, Manassas, USA) and were cultured in Dulbecco's modified Eagle medium (DMEM, HyClone, USA) with 10% foetal bovine serum (FBS, Gibco, USA). Primary cultures of neonatal CFs were isolated from the atria of 1-to 3-day-old Sprague-Dawley rats as previously described. 33 Briefly, the tissues were diced into small pieces and carefully washed in ice-cold phosphate-buffered saline (PBS, HyClone, USA) to remove plasma contaminants. The pieces were digested with a 0.25% trypsin solution and 0.1% collagenase II solution, until the tissue was no longer visible. The isolated CFs were maintained in DMEM supplemented with 10% FBS and 1% penicillin/ streptomycin and incubated in 95% air/5% CO 2 at 37°C by removing unattached cells (including cardiomyocytes and endothelial cells) after 1.5 hours. The CFs were defined as passage 0 (P0). The cells were digested with 0.25% trypsin-EDTA for passaging. CFs at passage 2 or 3 (P2-P3) were used in the following studies. The cells were transfected for 72 hours with miR-30c mimic, miR-30c inhibitor and negative controls, according to the Lipofectamine â 3000 (Invitrogen, USA) instructions. The miR-30c mimic, mimic negative control (miR01201-

| Cell proliferation
Cell growth was analysed using cell counting kit-8 (CCK-8) solution (Dojindo, Japan) according to the manufacturer's instructions. Briefly, the cells were plated at a density of 5 9 10 3 cells per well in 96well plates and transfected as described above. Then, the medium in each well was substituted with 10 ll CCK-8 solution for 2 hours.
The absorbance at 450 nm was measured using a multimode reader.

| Cell migration assays
Cellular migration assays were performed in 24-transwell permeable supports with an 8.0 lm pore size (Corning, USA). After treatment, 5 9 10 4 cells were seeded in the upper chamber in serum-free DMEM. The lower chambers were filled with serum-free DMEM containing TGF-b1 (5 ng/mL) as a chemoattractant. The cells were allowed to migrate for 24 hours at 37°C in 5% CO 2 atmosphere. After incubation, the non-migratory cells on the top of the membrane were removed with a cotton swab. Membranes containing cells were fixed with methanol and stained with crystal violet (Beyotime, China) for 30-60 minutes. The number of migrated cells was counted from 5 randomly selected fields at 200 9 magnification using a microscope.

| Dual luciferase activity assay
To investigate the target gene of miR-30c, bioinformatic prediction algorithms were used to analyse and identify potential targets. In

| miR-30c is down-regulated in the rat abdominal aortic constriction model
To explore the potential role of miR-30c in atrial fibrosis, we first established a model of atrial fibrosis by AAC. Based on Masson's trichrome staining, we identified higher collagen production in the AAC group than in the Sham group ( Figure 1A,B). In addition, we confirmed that miR-30c was down-regulated in atrial samples from AAC model rats via quantitative Real-Time PCR (qRT-PCR) ( Figure 1C).
Furthermore, the level of miR-30c was also decreased in CFs stimulated with TGF-b1 ( Figure 1F). Taken together, these data supported a potential role for miR-30c for atrial fibrosis. The TGFbRII expressions were increased both in AAC models and cultured neonatal rat CFs stimulated with TGF-b1 ( Figure 1D-F).

| miR-30c reduces TGF-b1-induced CF proliferation, differentiation and migration
Next, we investigated the effect of miR-30c on CF proliferation, differentiation migration and ECM-related proteins in vitro. Based on the qRT-PCR analysis, we confirmed that miR-30c was successfully up-regulated after transfected with miR-30c mimic for 72 hours ( CF proliferation at 24, 48 and 72 hours with or without TGF-b1 via CCK-8 assay compared with that in negative control group (Figure 2B). In addition, the overexpressing miR-30c attenuated the differentiation of CFs into myofibroblasts as indicated by decreased the levels of vimentin and a-smooth muscle actin (a-SMA) (Figure 2C-E), which is a reliable and classic marker for the myofibroblast. [33][34][35] Furthermore, our result demonstrated that a miR-30c mimic inhibited CF stimulated with or without TGF-b1 migration via transwell migration assay ( Figure 2F,G). Similarly, overexpression of miR-30c inhibited the increased mRNA and protein levels of collagen I (Col I) and collagen3a1 (Col3a1) with or without TGF-b1 ( Figure 2H-J).
Then, we inhibited miR-30c expression by transfecting the cells with miR-30c inhibitors. Based on the qRT-PCR data, miR-30c expression was successfully inhibited in CFs ( Figure 3A). MiR-30c inhibitor increased CF proliferation with or without TGF-b1 compared with that in negative control group at 24, 48 and 72 hours via CCK-8 assay ( Figure 3B). In addition, the a-SMA and vimentin levels were significantly increased in the miR-30c inhibitor + TGF-b1 group at 72 hours ( Figure 3C-E). MiR-30c inhibitor promotes CF migration ( Figure 3F, G). Contrary to the effects of miR-30c overexpression, inhibition of miR-30c increased the expression levels of Col I and Col3a1 ( Figure 3H-J). Collectively, these data indicated that overexpression of miR-30c attenuates CF proliferation, differentiation and migration and collagen synthesis.

| miR-30c targets TGFbRII
According to the above results, we verified that miR-30c was involved in CF proliferation, differentiation and migration. Further, we wondered about the potential relationship between miR-30c and TGFbRII. Figure 4A shows the proposed microRNA binding site in 3 0 UTR of TGFbRII gene as well as mutant 3 0 UTR that was generated in this study. The luciferase activity was significantly decreased in pGL3-TGFbRII-WT + miR-30c group compared with that in pGL3-TGFbRII-mu + miR-30c group ( Figure 4B). For further verification, we confirmed that up-and down-regulation of miR-30c in CFs could decrease and increase the mRNA and protein levels of TGFbRII after transfection with a miR-30c mimic or a miR-30c inhibitor in the absence or presence of TGF-b1 ( Figure 4C-H). These outcomes verified that miR-30c could target TGFbRII.

| miR-30c prevents AAC-induced atrial fibrosis
To further determine whether miR-30c indeed plays a vital role in alleviating atrial fibrosis induced by AAC, we next used a heartspecific AAV9 via IVC injection to achieve overexpression of miR-30c in left atrium. We confirmed that miR-30c expression was efficiently increased in the left atrium of rats in the AAC + miR-30c group via qRT-PCR ( Figure 5A). In addition, the mRNA and protein expressions of TGFbRII were decreased in the AAC + miR-30c group ( Figure 5B-D). Up-regulation of the miR-30c could attenuate the growth of fibres induced by AAC via Masson's trichrome staining ( Figure 5E,F). Moreover, administration of miR-30c substantially inhibited the expression of Col I, Col3a1 and a-SMA, compared with those in AAC group ( Figure 5G,H). These data provide strong evidence that up-regulation of miR-30c has a cardio-protective effect against fibrosis and that miR-30c inhibits TGFbRII expression alleviating left atrial fibrosis.

| DISCUSSION
In this study, we presented several novel findings. First, the miR-30c expression was decreased with the progression of atrial fibrosis. In addition, overexpression of miR-30c suppressed TGF-b1-induced CF indicated that miR-30c is significantly reduced in a model of pathological left ventricular hypertrophy. 7,8 MiR-30c regulates the production of ECM in TGF-b-dependent liver fibrosis. 9,56 MiR-30c reduces renal fibrosis and improves renal function in diabetic nephropathy by targeting connective tissue growth factor (CTGF). 10 Additionally, we observed that atrial fibrosis was increased and the level of miR-30c was decreased in left atrium following AAC, which suggesting that the alteration of miR-30c expression may closely relate to fibrosis.
TGF-b1 is involved in cell proliferation, apoptosis, differentiation and migration as well as the production of ECM molecules, leading to cardiac fibrosis. 15,57-61 TGF-b1 is considered the major growth factor directly promoting myofibroblast development by inducing expression of a-SMA. The transformation of fibroblasts into myofibroblasts is a critical event in the genesis of cardiac fibrosis. 62,63 During cardiac fibrosis, CFs underwent phenotypic transition to myofibroblasts marked by increased a-SMA, which is a reliable and classic marker for the myofibroblast. [33][34][35] In addition to the enhancement of contractility, myofibroblasts characteristically demonstrate increased migratory activity. 64 CF proliferation, migration and differentiation are important factors in cardiac fibrosis. MiR-30c has been reported to be involved in many cellular functions that lead to fibrosis, including cell proliferation, migration and differentiation. 54,65 Our study indicated that miR-30c was regulated by TGF-b1. Therefore, in our study, we primarily focused on the effects of miR-30c on CF proliferation, migration and differentiation to determine the potential mechanism involved in atrial fibrosis.
Consistent with these reports, the present study confirmed that miR-30c inhibits the transformation of CFs into myofibroblasts induced by TGF-b1 by decreasing the expression of a-SMA and vimentin, whereas miR-30c inhibition promoted the expressions of a-SMA and vimentin. In addition, miR-30c mimic inhibited the differentiation of CFs as evidenced by a decrease in a-SMA staining (Fig-ure S1a). MiR-30c inhibitor promoted CF differentiation by increasing in a-SMA staining ( Figure S1b). To the best of our knowledge, this is the first report that miR-30c exerts its cardio-protective properties by blocking fibroblast differentiation to myofibroblast.
Our data also revealed that miR-30c overexpression decreased cell proliferation and migration. Collectively, miR-30c reduced CF migration, preventing the accumulation of myofibroblasts and protecting against the fibrotic process. Aberrant fibroblast proliferation and their transformation to myofibroblasts, as well as migration, are hallmarks of cardiac fibrosis, which is characterized by excessive ECM deposition and leads to distorted organ architecture and function. 66,67 Notably, miR-30c up-regulation was inversely correlated with the decreased expression levels of Col I and Col3a1 in cells treated with TGF-b1. ECM metabolic imbalance is involved in various cardiovascular diseases such as myocardial infarction, heart failure and AF. [68][69][70] Moreover, miR-30c overexpression decreased the ECM-regulated gene expression levels of Col I and Col3a1 in CFs, whereas MiR-30c inhibition promoted these processes.
TGFbRII is a transmembrane receptor necessary for TGF-b1 signal transduction activation. TGF-b1 is involved in the processes of pulmonary fibrosis and liver fibrosis. 24,52 A previous study demonstrated that constitutive expression of dominant-negative TGFbRII in the posterior left atrium resulted in a significant decrease in atrial fibrosis via inhibition of the TGF signalling pathway and attenuated fibrosis-induced changes in atrial conduction and restitution to decrease AF. 27 The increased level of TGFbRII in the TGF-b1 group showed that TGFbRII may participate in TGF-b1-induced fibrosis.
Luciferase reporter assays confirmed that miR-30c led to a reduction in luciferase activity in TGFbRII 3 0 UTR-WT group, but had no effect when miR-30c binding site in TGFbRII 3 0 UTR was mutated, implying TGFbRII is a target gene of miR-30c. The protein levels of TGFbRII were negatively correlated with miR-30c expression in AAC model and in the cellular model of fibrosis induced by TGF-b1. Overexpression of miR-30c decreased the level of TGFbRII, whereas miR-30c inhibition increased the level of TGFbRII. Therefore, TGFbRII, which is necessary for TGF-b1 signal transduction, may represent the point at which miR-30c inhibits the pro-fibrotic effects of TGF-b1.
The protective effects of miR-30c overexpression against atrial fibrosis were confirmed by overexpressing miR-30c via AAV9 in vivo.
In our vivo study, miR-30c overexpression markedly reduced atrial fibrosis compared with that in AAC group. We have determined for the first time that miR-30c overexpression might significantly reduce atrial fibrosis by targeting TGFbRII expression. Although several evidences here strongly support the functional role of miR-30c in regulating atrial fibrosis, more rigorous approaches are required to support this contention. This may include intra-myocardial rather than systemic delivery with a CF-specific promoter or using a miR-30c transgenic mouse model created using the CF-specific promoter.
Unfortunately, this study was performed in a model of atrial fibrosis, which does not induce AF by burst pacing ( Figure S2c). Previous study reported that P wave duration (PWD) was useful in identifying patients at a risk for paroxysmal AF and was significantly correlated with the atrial electromechanical delay. 71 PR interval is an amalgamated measure of atrial and atrioventricular nodal conduction. PR interval prolongation increases AF risk and all-cause mortality. 72,73 In F I G U R E 2 MiR-30c exhibited an anti-fibrotic effect on CFs. (A) CFs were transfected with a miR-30c mimic or miR-30c negative control (NC) for 72 hours. NC was used as a mimic negative control. The level of miR-30c was increased in CFs transfected with miR-30c mimic by qRT-PCR. # P < .05 compared with NC group. (B) After transfecting with a miR-30c mimic or NC, the CFs were stimulated with TGF-b1 (5 ng/mL) for 24, 48 or 72 hours. Overexpression of miR-30c inhibited CF proliferation performed with a CCK8-Kit assay. # P < .05, ## P < .01 compared with NC group; *P < .05 compared with NC + TGF-b1 group. (C-E) qRT-PCR, western blot and quantification analysis of levels of a-SMA and vimentin in CFs transfected with the miR-30c mimic or negative control in the presence or absence of TGF-b1 (5 ng/mL). # P < .05 compared with NC group; *P < .05, **P < .01 compared with NC + TGF-b1 group. (F) Representative photomicrographs of the migrating cells after transfection and staining by crystal violet (Scale bar, 100 lm). (G) The effect of the miR-30c mimic on the migration of the cells was quantified. # P < .05 compared with NC group; *P < .05 compared with NC + TGF-b1 group. (H-J) The mRNA and protein levels of Col I and Col3a1 in CFs transfected with the miR-30c mimic were decreased in the presence or absence of TGF-b1 (5 ng/mL) via qRT-PCR, representative western blot and quantitative analysis for 72 hours. # P < .05 compared with NC group; *P < .05 and **P < .01 compared with NC + TGF-b1 group. All experiments were repeated 3 times, and all data were presented as the mean AE SEM (n = 3, each) XU ET AL. In conclusion, we suggest that TGF-b1 produces both profibrotic and anti-fibrotic signals and exogenous miR-30c administration decreases atrial fibrosis by counteracting the pro-fibrotic stimuli elicited by TGF-b1. Our data demonstrate that overexpression of miR-30c decreases left atrial fibrosis by inhibiting CF F I G U R E 3 The miR-30c inhibitor reversed the anti-fibrotic effects on CFs. (A) CFs were transfected with a miR-30c inhibitor or miR-30c inhibitor negative control (NC) for 72 hours. The level of miR-30c was decreased in CFs transfected with miR-30c inhibitor by qRT-PCR. NC was used as an inhibitor negative control. # P < .05 compared with NC group. (B) After transfecting with a miR-30c inhibitor or NC, the CFs were stimulated with TGF-b1 (5 ng/mL) for 24, 48 or 72 hours. A miR-30c inhibitor promoted CF proliferation evaluating by a CCK8-Kit assay. # P < .05, ## P < .01, compared with NC group; *P < .05, **P < .01, compared with NC + TGF-b1 group. (C-E) The levels of a-SMA and vimentin were increased in CFs transfected with the miR-30c inhibitor in the presence or absence of TGF-b1 (5 ng/mL) via qRT-PCR, western blot and quantification analysis. # P < .05 compared with NC group; *P < .05, **P < .01, compared with NC + TGF-b1 group. (F) Representative photomicrographs of the migrating cell after transfection and staining by crystal violet (Scale bar, 100 lm). (G) The effect of the miR-30c inhibitor on the migration of the cells was quantified. ## P < .01, compared with NC group; **P < .01 compared with NC + TGF-b1 group. (H-J) qRT-PCR, representative western blot and quantitative analysis of the mRNA and protein levels of Col I and Col3a1 in CFs transfected with the miR-30c inhibitor or NC in the presence or absence of TGF-b1 (5 ng/mL) for 72 hours. A miR-30c inhibitor increased the levels of Col I and Col3a1. *P < .05, **P < .01, compared with NC + TGF-b1 group; # P < .05, ## P < .05, compared with NC group. All experiments were repeated 3 times, and all data were presented as the mean AE SEM (n = 3, each) F I G U R E 4 MiR-30c targets TGFbRII. (A) Diagram of miR-30c binding site in TGFbRII 3 0 UTR. TGFbRII 3 0 UTR mutant was constructed to evaluate miR-30c binding. (B) After 293T cell cotransfection, luciferase activity was measured using a Dual Luciferase Report Assay Kit. The luciferase activity was significantly decreased in TGFbRII-WT + miR-30c group compared with that in TGFbRII-WT + miR-30c NC and TGFbRII-mu + miR-30c. ### P < .001, compared with pGL3-TGFbRII-WT + NC group. && P < .01, compared with pGL3-TGFbRII-mu + miR-30c group. (C-E) qRT-PCR and western blot analyses of CFs transfected with the miR-30c mimic or negative control for 72 hours in the absence or presence of TGF-b1. The miR-30c mimic decreased the mRNA and protein levels of TGFbRII. # P < .05, ## P < .05 compared with NC group; **P < .01, ***P < .01, compared with NC + TGF-b1 group. (F-H) The level of TGFbRII was increased after transfection with a miR-30c inhibitor as measured by qRT-PCR and western blot. # P < .05 compared with NC group; *P < .05 compared with NC+TGF-b1 group. All experiments were repeated 3 times, and all data were presented as the mean AE SEM (n = 3, each) proliferation, differentiation, migration and collagen production by targeting TGFbRII for the first time. Our study indicates that the application of exogenous miR-30c may be an attractive treatment strategy for atrial fibrosis and requires a further investigation in the future.

ACKNOWLEDG EMENTS
This work was supported by the grants from the National Natural

S U P P O R T I N G I N F O R M A T I O N
Additional Supporting Information may be found online in the supporting information tab for this article.
How to cite this article: Xu J, Wu H, Chen S, et al.