An increasing number of studies have shown that miRNAs participate in the development of myocardial fibrosis, but their functions and mechanisms remain to be fully elucidated [38]. Myocardial infarction (MI) is the leading cause of death from cardiovascular disease. The myocardium slowly enters a long-term progressive fibrotic process after the acute phase of infarction, which is an adaptive remodeling in the early stage, but the persistent fibrotic response accelerates HF after MI. MiRNAs were significantly differentially expressed in MI, with miR-1 expression levels increasing after MI. Inhibiting miR-1 expression through the use of a miR-1 antagomir significantly reduced left ventricular end-diastolic internal diameter, collagen proliferation and TGF-$\beta$ expression after MI but increased ejection fraction and improved myocardial fibrosis and cardiac functions [3]. However, miR-155 exerted no marked effect on left ventricular volume, left ventricular mass or ejection fSSraction, although the myofibroblast density was obviously lower than that in the control [4]. RUNX1, known as acute myeloid leukemia 1 (AML 1), is a transcription factor with a highly conserved protein sequence and is involved in the expression of genes related to cell differentiation and proliferation. MiR-101 targeted RUNX1, and therefore, overexpression of miR-101 or silencing of RUNX1 might decrease infarct size, attenuate myocardial fibrosis and inhibit apoptosis, thereby improving cardiac functions. Moreover, miR-101 played a protective role against cardiac remodeling via inactivation of the RUNX1-dependent transforming growth factor $\beta$1/Smad family member 2 (TGF-$\beta$1/Smad2) signaling pathway [5]. In addition, inhibition of miR-29 expression activated the PI3K/mTOR/HIF-1$\alpha$/VEGF pathway to promote microangiogenesis and reduce myocardial fibrosis after MI [6].

Post-MI myocardial interstitial ischemic edema can decrease myocardial compliance and contractility, increase left ventricular end-diastolic pressure and volume, and eventually progression to HF or even death. MiR-30d inhibited fibroblast proliferation and activation by directly targeting integrin $\alpha$5 to enhance cardiac functions in the acute phase of ischemic HF in mice [7]. Myocardial structure, cardiac functions and oxidative stress were enhanced, and fibrosis in myocardial tissue was reduced after Smad ubiquitin regulatory factor 1 (Smurf1) knockdown. MiR-129-5p targets Smurf1, represses the ubiquitination of phosphatase and tensin homolog (PTEN), and promotes PTEN expression, which attenuates the cardiac functions of chronic HF rats [8]. Hinkel et al. [9] provided the first evidence showing the feasibility and therapeutic efficacy of miR-21 inhibition in a large animal model of HF and found that silencing miR-21 reduced cardiac fibrosis and hypertrophy to improve cardiac function at 33 days after ischemia reperfusion injury.

In the early stage of cardiovascular diseases, such as chronic diseases, degenerative valve diseases and cardiomyopathies, left ventricular wall thickening and myocardial fibrosis progression result in reduced cardiac compliance and function, which leads to left ventricular enlargement and the development of left heart failure in the late stage. Studies have shown that miRNAs are also involved in left ventricular fibrosis due to these diseases [39]. For example, the miR-195 expression level was remarkedly reduced in hypertensive rats, with disorganized myocardial cells, thickened myocardial fibers and myocardial fibrosis in a hypertension group compared to controls. Overexpression of miR-195 inhibited TGF-$\beta$1/Smad3 signaling pathway activity and related molecules, further repressing myocardial fibrosis [10]. Diabetic cardiomyopathy (DCM) is initially characterized by early diastolic dysfunction, left ventricular remodeling, hypertrophy and myocardial fibrosis. Well-characterized miRNAs involved in the development of DCM including miR-20a-5p, miR-21 and miR-340-5p [11, 12]. MiR-340-5p expression has been found to be dramatically increased in the heart tissue of mice and cardiomyocytes under diabetic conditions, and its overexpression exacerbated the apoptosis of cardiomyocytes, elevated reactive oxygen species production and impaired mitochondrial function, leading to extensive cardiac fibrosis and severe dysfunction. Later studies revealed that miR-340-5p caused severe cardiac dysfunction by repressing the expression of the target gene myeloid leukemia-1 (Mcl-1), which suggested that miR-340-5p plays a crucial role in the development of DCM and can be targeted for therapeutic intervention [12]. In addition, other miRNAs, such as miR-21, miR-27a-5p and miR-29b, play vital roles in pressure-load induced cardiac fibrosis caused by transverse aortic constriction (TAC) [13, 14, 15]. García et al. [15] assessed the condition of 103 patients with aortic stenosis preoperatively and one year after aortic valve replacement surgery and found that the preoperative plasma expression of miR-29b paralleled the severity of hypertrophy and was a significant negative predictor of reverse remodeling after aortic valve replacement surgery, indicating that the miR-29b level may be a potential prognostic biomarker. Thottakara et al. [40] first reported that miR-4454 expression was significantly elevated in patients with hypertrophic cardiomyopathy and was correlated with the severity of myocardial fibrosis, as determined through comparison cardiac magnetic resonance, suggesting that miR-4454 may be a potential biomarker of fibrosis. Multiple miRNAs are involved in the process of myocardial fibrosis, indicating that miRNAs are important regulators of myocardial fibrosis and might be potential therapeutic targets.

The right ventricle is under volume or pressure overload due to left heart failure, heart valve diseases, pulmonary artery hypertension (PAH) and other diseases, which leads to the formation of right ventricular fibrosis, and scar tissue causes right ventricular dysfunction. In PAH, sustained pressure overload exerts mechanical stress on the right ventricular interstitium and CFs, which increases collagen production by releasing TGF-$\beta$ to promote proliferative activation of CFs as well as increasing mRNA expression and upregulating $\alpha$-smooth muscle actin activity ($\alpha$-SMA). In the early stage of the disease, increasing collagen can support the myocardium in withstanding high pressure to maintain the right ventricular structure. However, with the progression of disease, adaptive myocardial collagen accumulation follows maladaptive changes in collagen network structure and loss of ECM integrity, eventually leading to decreased cardiac compliance, cardiac dysfunction and arrhythmic events [41], such as increasing right ventricular wall stiffness and filling pressure indirectly induces supraventricular arrhythmias though transmission to the right atrium. In addition, arrhythmogenic right ventricular cardiomyopathy (ARVC) is an inherited cardiomyopathy characterized by progressive fibro-fatty replacement of right ventricular myocardial, arrhythmias and risk of sudden death. Currently, the pathophysiological role of miRNAs in ARVC has not been fully elucidated. For example, the expression of miR-21-5p, miR-135b and miR-185-5p was found to be distinctly elevated and to regulate myocardial fatty fibrosis via the Wnt and Hippo pathways in patients with ARVC [16, 42].

Hsu et al. [17] found that the levels of circulating miR-29a-3p and thrombospondin-2 (THBS2) decreased and increased, respectively, in mice and patients with PAH. MiR-29a-3p directly targets and regulates THBS2 expression to inhibit the proliferation of CFs, which exerts a direct antifibrotic effect on PAH-induced cardiac fibrosis. MiR-325-3p targets and regulates human epididymis protein 4 to activate the PI3K/AKT signaling pathway, and the action of this miRNA was found to inhibit CF transformation and attenuate right ventricular fibrosis in PAH rats [18]. A mitochondrial metabolism study of PAH-induced right ventricular fibrosis revealed that increased pyruvate dehydrogenase activity inhibited mitochondrial superoxide dismutase 2 (SOD2) and H${}_2$O${}_2$ production; activated hypoxia-inducible factor-1$\alpha$ (HIF-1$\alpha$); increased pyruvate dehydrogenase kinase isoforms 1 and 3, TGF-$\beta$1 and CTGF expression; increased CF proliferation and collagen production; promoted fibrosis formation and reduced right ventricular function. Among these outcomes, HIF-1$\alpha$ activation, in particular, reflects increased DNA methyltransferase 1 expression, which has been associated with a decrease in the regulatory effector of miR-148-3p [19]. Notably, miRNA distribution has been reported to differ between the right and left ventricles in mammals, and the functional mechanisms of right ventricular fibrosis are less clear than those of the left ventricle. Boon et al. [20] showed that miR-34a was induced in the aging heart and that in vivo silencing or gene deletion reduced age-related apoptosis of cardiomyocytes. MiR-34a-targeting PNUTS induced the DNA damage response and telomere shortening to improve functional recovery after acute myocardial infarction (AMI) [20]. Whether a miRNA-based therapeutic strategy to prevent left ventricular fibrosis is effective on the right ventricle remains unknown, and the miRNAs to target have yet to be determined.

Atrial fibrillation (AF) is the most common cardiac arrhythmia and a result of atrial remodeling. Atrial remodeling, characterized by persistent biventricular enlargement and fibrosis of myocardial tissue, is caused by atrial volume or pressure overload and eventually progresses to diastolic insufficiency or even sudden death by thromboembolism. MiRNAs are significantly differentially expressed in AF. For example, miR-21, miR-23b-3p and miR-27b-3p expression has been found to be significantly increased in atrial tissue of patients with AF [21, 22, 43], but the levels of miR-132 and miR-443 were decreased [23, 24]. The main target of miR-21 is TGF-$\beta$. The results of an experimental study performed with human fibroblasts by Tao et al. [21] confirmed that downregulated miR-21 increased the expression of WW structural domain-blinding protein-1 to inactivate the TGF-$\beta$1/Smad2 signaling pathway, inhibit the proliferation of CFs, and reduce collagen I and III levels and the collagen volume fraction. Importantly, miR-29b, miR-30c and miR-10a also regulate the TGF-$\beta$/Smad pathway, suggesting that miRNAs may be potential therapeutic targets for the prevention of atrial fibrosis [25, 26, 27]. Pan et al. [28] established a coculture model with atrial fibroblasts and adipocytes and found that miR-21-3p regulated FGFR1, FGF21, and PPAR$\gamma$ to control epicardial adipocyte brownout (EAT) and ameliorate glucose-induced atrial fibrosis, suggesting that modulating EAT may be a new strategy to prevent or treat atrial fibrosis or AF in DCM. Reducing miR-205 expression in AF rats attenuated atrial fibrosis by targeting prolyl-hydroxylase $\alpha$ polypeptide III (P4H$\alpha$3) to inhibit CF proliferation and migration and inactivating the JNK pathway [29]. Other miRNAs are also involved in atrial remodeling, including miR-133, miR-590 and miR-146b-5p, and all were downregulated in a canine model of AF [30, 44]. The expression of miR-146b-5p, whose target gene is TIMPs, was upregulated in atrial cardiomyocytes during AF. The inhibition of miR-146b-5p expression inhibited the acquisition of a cardiac fibrosis phenotype in MI mouse models. The expression of fibrotic markers MMP9, TGF-$\beta$1 and COL1A1 was significantly downregulated, while that of IMP4 was significantly upregulated by miR-146b-5p inhibition in the cardiomyocytes of MI hearts. In contrast, in vitro, miR-146b-5p regulated TIMP/MMP9-mediated ECM synthesis, increasing collagen synthesis in a human-induced pluripotent stem cell-derived atrial cardiomyocyte fibroblast coculture cell model [30].

Age exacerbates mortality from cardiovascular disease in elderly individuals. During cardiac aging, the accumulation of senescent cells and the deposition of ECM with collagen lead to a progressive decline in cardiac functions. A 3% increase in ECM volume and a 50% increase in the risk of all-cause mortality have been reported [73]. Studies have shown that ncRNA expression correlates with aging. For example, miR-1468-3p expression has been shown to be increased in healthy elderly hearts and to promote cardiac fibrosis by enhancing TGF-$\beta$1/p38 signaling [37]. In MI mouse models and heart biopsy samples from patients with aortic stenosis, the lncRNA Wisper was found to be associated with cardiac fibrosis, with Wisper regulating the CF gene expression program, and to be essential for the proliferation, migration and survival of CFs [53]. CircRNA-Yap inhibited cardiac fibrosis by interacting with tropomyosin-4 and $\gamma$-globular actin interactions, inhibiting actin polymerization and subsequent fibrosis [68]. CircRNA-000203 enhanced the expression of three fibrosis-related genes, COL1A2, COL3A1 and $\alpha$-SMA, by inhibiting miR-26b-5p expression in mouse CFs [72]. Multiple ncRNAs are regulated in the aging heart and may serve as potential targets for age-related cardiac fibrosis therapy.

Viral myocarditis (VMC) is local or diffuse damage to the myocardial parenchyma or interstitium caused directly by viral infection or indirectly by immune system dysfunction and is accompanied by cardiomyocyte destruction, reparative fibrosis and ultimately HF. NcRNAs regulate the viral life cycle and immune and inflammatory responses by targeting viral or host genes. MiR-155, lncRNA AK085865 and lncRNA MEG3 regulated macrophage polarization and reduced myocardial injury, which was correlated with a reduction in the development of myocardial fibrosis [35, 74, 75]. Studies in mice with autoimmune myocarditis revealed that silencing miR-21a-5p resulted in a significant reduction in TNF$\alpha$, IL-6 and collagen I expression, reducing excessive infiltration of damaging myocardial cells and inhibiting myocardial fibrosis formation [31]. In an isoproterenol-induced myocardial fibrosis model of VMC, the lncRNA ROR upregulated c-myc expression and increased serum IL-6 levels, thereby facilitating the proliferation and differentiation of CFs. This suggested that lncRNA ROR plays an important role in chronic VMC [62]. Differential expression of circRNAs has been found in patients with fulminant myocarditis and in mice, and a bioinformatics analysis revealed the involvement of severely dysfunctional immune signaling pathways, including TNF and T-cell receptors, in VMC [76, 77]. Compared to miRNAs, other lncRNAs and circRNAs, ncRNAs have not been studied in detail in the field of mechanistic development of VMC and may be potential biomarkers and therapeutic targets for VMC in the future.

In recent decades, the mechanisms of drug-induced cardiotoxicity, mainly DNA damage, excessive reactive oxygen species production, mitochondrial dysfunction, endoplasmic reticulum-mediated apoptosis, and disruption of calcium homeostasis have been studied. With the deepening of research, it has been shown that ncRNAs play key roles in the mechanisms of cardiotoxicity induced by drugs such as doxorubicin (DOX) [78, 79]. For example, silencing miR-34a upregulated B lymphocytoma-2, and sirtuin 1 attenuated DOX-induced cardiotoxicity, reducing the apoptosis rate, attenuating the inflammatory response, and inhibiting senescence and fibrosis in rats [32]. MiR-133b attenuated polypyrimidine tract bundle-binding protein 1 and transgelin 2 to regulate apoptosis and mediate cardiac fibrosis induced by adriamycin, implying that miR-133b may be a potential biomarker of adriamycin-induced cardiac injury [33]. Endogenous miR-495-3p protects cells against DOX-induced cardiotoxicity by activating the AKT pathway in vivo and in vitro [36]. In addition, the expression of the lncRNA KCNQ1OT1 downregulated fusion-type sarcoma in oocytes and reduced the myocardial fibrosis area in DOX-treated mouse models [61]. Increasing the expression of circ-Nlgn decreased cardiac function and induced cardiac fibrosis by upregulating Gadd45b, Sema4C and RAD50 and activating p38 and pJNK in circNlgn transgenic mouse hearts. Silencing circ-NIgn prevented DOX-induced expression of fibrosis-associated molecules. The Nlgn173 protein translated by circ-NIgn can bind and activate H2AX and the production of $\gamma$H2AX, which results in the upregulation of IL-1b, IL-2Rb, IL-6, EGR1 and EGR3. Later studies showed that silencing these molecules in the signaling pathway could prevent cardiomyocyte apoptosis, reduce CF proliferation and inhibit collagen production, mitigating the side effects of DOX in cancer patient treatment [69].

Heart transplantation (HT) is the best treatment option for end-stage HF. However, HT can cause cardiac ischemia reperfusion injury. In a mouse model of ectopic HT, 59 miRNAs were dysregulated in transplanted hearts with ischemia reperfusion injury compared to undamaged transplanted hearts [80]. In contrast, circ-Foxo3 is a newly discovered molecular regulator that protects cardiac grafts from prolonged ischemia reperfusion injury during HT. MircFoxo3 also indirectly affects miR-433 and miR-136 expression [70]. In addition, many ncRNAs can be used as noninvasive detection indicators, such as for the assessment of myocardial injury after HT. MiR-133b has been found to be an important marker of myocardial injury and to be associated with hemodynamic changes evident early after transplantation [34]. Circ-FSCN1 silencing produced tolerogenic dendritic cells, which prevented alloimmune rejection in HT, prolonged patient survival and reduced myocardial fibrosis [71]. However, whether functional interventions based on specific ncRNAs can block the development of long-term fibrosis after HT and thus improve survival after transplantation still needs to be explored in the future.