Clinical value of non-coding RNAs in cardiovascular, pulmonary, and muscle diseases

Abstract

Although a majority of the mammalian genome is transcribed to RNA, mounting evidence indicates that only a minor proportion of these transcriptional products are actually translated into proteins. Since the discovery of the first non-coding RNA (ncRNA) in the 1980s, the field has gone on to recognize ncRNAs as important molecular regulators of RNA activity and protein function, knowledge of which has stimulated the expansion of a scientific field that quests to understand the role of ncRNAs in cellular physiology, tissue homeostasis, and human disease. Although our knowledge of these molecules has significantly improved over the years, we have limited understanding of their precise functions, protein interacting partners, and tissue-specific activities. Adding to this complexity, it remains unknown exactly how many ncRNAs there are in existence. The increased use of high-throughput transcriptomics techniques has rapidly expanded the list of ncRNAs, which now includes classical ncRNAs (e.g., ribosomal RNAs and transfer RNAs), microRNAs, and long ncRNAs. In addition, splicing by-products of protein-coding genes and ncRNAs, so-called circular RNAs, are now being investigated. Because there is substantial heterogeneity in the functions of ncRNAs, we have summarized the present state of knowledge regarding the functions of ncRNAs in heart, lungs, and skeletal muscle. This review highlights the pathophysiologic relevance of these ncRNAs in the context of human cardiovascular, pulmonary, and muscle diseases.

INTRODUCTION

Although exonic regions of protein-coding genes comprise only a minor proportion of the mammalian genome, the vast majority of the mammalian genome is nevertheless transcribed to RNA (154, 307, 308). In light of this fact, it comes as no surprise that there exists a growing number of identified RNA species that do not encode a cognate polypeptide and that play essential roles in the regulation of a number of basic cellular processes and tissue-specific physiologic functions. RNAs that do not encode for proteins are collectively referred to as “non-coding RNAs” (ncRNAs), and they are categorized according to their length, as opposed to their function. Of these ncRNAs, the most widely studied are the microRNAs (miRNAs). miRNAs are small ncRNAs, ~22 nucleotides (nt) in length. They function as RNA silencers and mediate posttranscriptional regulation of gene expression. miRNAs are abundant in many mammalian cell types and have documented roles in the maintenance of normal tissue homeostasis and disease pathophysiology (122, 188, 311, 312). The primary transcripts of miRNAs (pri-miRNAs) are transcribed by RNA polymerase II and subsequently enzymatically processed by the Drosha microprocessor complex within the nucleus (143). The cropping of the pri-miRNA by the Drosha complex results in the formation of hairpin-shaped miRNA precursors (pre-miRNAs), which are exported from the nucleus to the cytoplasm via the exportin-5/Ran-GTP complex. In the cytoplasm, pre-miRNA is further processed into miRNA duplexes (21–24 nt long) by another microprocessor complex, consisting of the RNase III, Dicer, and the human immunodeficiency virus transactivation response RNA-binding protein (TRBP; 53). The miRNA duplex then associates with Argonaute (Ago) family proteins to form the miRNA-RNA-induced silencing complex (RISC). The miRNA-RISC facilitates the dissociation of miRNA duplex and target strand selection/complementation. Because of miRNA base complementation with its cognate mRNA sequence, gene silencing of target mRNAs occurs. Mechanistically, microRNAs posttranscriptionally regulate gene expression by binding to the 3′-untranslated region (UTR) of their target mRNAs, resulting in either repression of protein translation or mRNA degradation/deadenylation (36). miRNA signaling constitutes an auxiliary mode by which cells can regulate gene expression programs, and their dysregulation has been associated with the pathophysiology and progression of various diseases. The role of miRNAs as posttranscriptional regulators of human disease has been the driving force behind numerous studies that seek to exploit circulating miRNAs as prospective biomarkers (4, 148, 250, 299, 320, 333).

Whereas the functions of miRNAs have been the focus of intense investigation over the past two decades, the potential functions of longer ncRNAs are largely unknown and are just now being investigated. Longer ncRNAs, those >200 nt in length, are referred to as “long non-coding RNAs” (lncRNAs). Although they were once considered nonfunctional nucleic acids (91), it is suggested that a number of these longer RNA species may play a role in various biological processes. Furthermore, as technologies employed in their detection have advanced, more and more lncRNAs are constantly being discovered and reported, although their functions have yet to be uncovered.

Adding to this extensive and seemingly growing list of various ncRNA species, some investigations have identified circular RNAs (circRNAs) as an additional source of regulatory RNAs. circRNAs are by-products of splicing events, a process termed “back-splicing” (25, 127, 128). As early as the 1990s, it was known that circRNAs, with their lariat structure, are especially stable and persist for extended periods within the cell’s cytoplasmic compartment (56, 236). Although their precise functions remain largely unidentified, some studies suggest that circRNAs may function as miRNA sponges to sequester miRNAs that are available to participate in the translational inhibition of target protein-coding genes (97, 110, 180, 211, 359). However, recent comprehensive bioinformatics analyses (104) and biological validation experiments (218) indicate that it is rare for circRNAs to act as miRNA sponges. Instead, direct interactions with RNA-binding proteins (RBPs) are more common (2, 75, 279). Indeed, some RBPs [e.g., Muscleblind (12) and Quaking (59)] are involved in the biogenesis of circRNAs. Whether circRNAs are mere by-products of splicing events or are functional regulators of cellular processes warrants further investigation.

Since there are a large number of reports about ncRNAs (e.g., miRNAs, lncRNAs, and circRNAs), especially in the fields of immunity and cancer, we focused this review on heart, lungs, and skeletal muscle. It is noteworthy that disorders of the heart, lung, and skeletal muscle are often encountered in combination in diseases such as heart failure, pulmonary hypertension, and myopathies.

CARDIOVASCULAR DISEASE

Cardiovascular disease (CVD) remains the number one cause of death and disability in the United States and much of the developed world (136, 202). CVD is both a significant health burden and a significant economic burden, which is anticipated to only worsen with a growing incidence of diabetes and obesity. Despite significant advances in our general understanding of the fundamental cellular and molecular mechanisms underlying cardiac disease pathophysiology, the recent emergence of ncRNAs suggests that CVD etiological mechanisms may have greater complexity than originally thought. To this end, the following section seeks to provide a nonexhaustive general overview of ncRNAs in the heart and CVD (Fig. 1).

Fig. 1.

Summary of non-coding RNA in the heart. Each transcript (i.e., mRNA) could be a target for microRNAs (miRNAs) to suppress translation to control a variety of cellular processes in the heart. These miRNAs could also be targeted by long non-coding RNAs (lncRNAs), which function as miRNA sponges to sequester the available miRNAs. When transcribed from DNA, exons could join by splicing machinery, which excludes introns as well as exons in some cases. These excluded exons and/or introns could undergo back-splicing to join a downstream splice donor site reversely with an upstream splice acceptor site, resulting in a covalently closed transcript, circular RNA (circRNA). APF, autophagy-promoting factor; CARL, cardiac apoptosis-related lncRNA; CHRF, cardiac hypertrophy-related factor; UTR, untranslated region.

miRNAs and heart development.

Cardiac development is a complex process of tightly coordinated events, which involve distinct and sequential phases of proliferation, migration, differentiation, programmed cell death, and structural remodeling (205). miRNAs have been implicated in the modulation of gene expression patterns during cardiac development via regulating transcription factor expression and function (205). In zebrafish, which is an excellent model organism to study heart development (14), miRNA expression has been shown to increase during development in a tissue-specific manner with most expression occurring after organogenesis (205, 332). In the heart, expression of miRNA-1 (miR-1) and miR-133 increases during development from the embryonic to the neonatal stage (47, 260). Importantly, in embryos injected with miR-1, the cardiac tissue was completely absent; meanwhile, the cardiac tissue was highly disorganized in miR-133-injected embryos (47). miR-1 and miR-133 belong to a common bicistronic transcript. Deletion of both miRNAs in mice causes fatal ventricular-septal defects in approximately half of double-mutant embryos or neonates. The absence of miR-133a expression results in ectopic expression of smooth muscle genes in the heart and aberrant cardiomyocyte proliferation through increased expression of serum response factor (SRF) and cyclin D2 (178). Many targets of miR-1 have also been identified that are essential to heart development, including histone deacetylase 4 (HDAC4), SRF, heart and neural crest derivatives expressed 2 (Hand2), gap junction protein-α1 (GJA1), and potassium voltage-gated channel subfamily J member 2 (KCNJ2; 47, 152, 346). In addition, overexpression of miR-133 can reinduce fetal genes, including those encoding atrial natriuretic factor, skeletal muscle and cardiac α-actin (Acta1 and Actc1, respectively), and α- and β-myosin heavy chain (Myh6 and Myh7, respectively), leading to actomyosin chain rearrangement and subsequent cytoskeletal reorganization and to perinuclear localization of atrial natriuretic factor protein (37), which further supports its role in regulating cardiac maturation and development. Intriguingly, miR-208a and miR-208b are involved in the regulation of myosin heavy chain isoform switching at different cardiac developmental stages (239). miR-208b and miR-499 play a dominant role in the specification of muscle fiber identity by activating slow-twitch myofiber gene programs and repressing fast-twitch myofiber gene programs (313). Together, these findings indicate that miRNAs play essential roles in heart development.

miRNA signature and function in cardiac progenitor cells.

Heart development is a dynamic and complex process that requires the interactions of multiple lineages of cardiac cells. The differentiation of cardiac progenitor cells into these lineages requires a tight regulation to enable effective, coordinated interactions. miRNAs participate in various aspects of cardiac progenitor cell biology, such as cell proliferation, lineage commitment, and migration, by modulating cardiac gene expression. miR-499, a cardiac-specific miRNA, can induce differentiation of human cardiomyocyte progenitor cells (289). Various miRNAs regulate proliferation-related signaling pathways in cardiac progenitor cells (3). Several miRNAs (e.g., miR-21, miR-218, miR-548c, miR-509, and miR-23b) can target negative cell proliferative regulators including phosphatase and tensin homolog (PTEN) and secreted frizzled-related protein-2 (sFRP2), leading to cardiac progenitor cell proliferation, whereas miR-1, miR-200b, and miR-204 inhibit cardiac progenitor cell proliferation by modulating proliferation-related transcription factors, including HDAC, Hand2, GATA binding protein-4 (GATA-4), and activating transcription factor 2 (ATF2; 162). Recently, the miR-322/503 cluster has been found to have the highest enrichment in the mesoderm posterior 1 (Mesp1) lineage of cardiac progenitor cells. This cluster is specifically expressed in the developing heart tube and drives precocious cardiomyocyte formation by targeting an RNA-binding factor, CUG-binding protein Elav-like family member 1 (Celf1), which otherwise leads embryonic stem cells toward neural fates (282). In contrast, miR-142-3p is highly expressed in undifferentiated embryonic stem cells (ESCs). Ectopic expression of miR-142-3p suppresses cardiomyocyte formation through downregulation of the expression of cardiac mesodermal marker gene Mesp1 and the downstream cardiac progenitor marker genes NK2 homeobox 5 (Nkx2.5), T-box 5 (Tbx5), and myocyte enhancer factor 2C (Mef2c; 52). In addition, miR-29a induces vascular smooth muscle cell differentiation from mouse ESCs by negatively regulating YY1 transcription factor (YY1), which inhibits muscle cell differentiation and muscle-specific gene expression (130). Therefore, miRNAs in lineage-committed cells may play controlling roles in the cell fate determination by cross talking with other lineages. Taken together, miRNAs are potent regulators of early cardiac fate and heart development.

miRNAs and cardiovascular diseases.

The initial stages and progression of CVD involve miRNAs, and changes in their expression profiles could be useful for the diagnosis, prognosis, and treatment of a diverse spectrum of CVDs. Identifying miRNAs that contribute to or suppress hypertension, cardiomyopathies, arrhythmias, atherosclerosis, and ischemic heart disease could support the development of therapies for these diseases. In vitro and in vivo experiments reveal many positive regulators in cardiac hypertension, including miR-145, miR-130/301, miR-21, miR-17, miR-9, miR-17/92, miR-22, miR-487, miR-210, and miR-155, whereas the negative regulators are miR-204, miR-206, miR-193, miR-328, miR-424/503, miR-124, miR-126, miR-666, miR-708, and miR-376c (284). For cardiomyopathy, miR-196a, miR-499, miR-146a, and miR-208 were found to promote cardiac remodeling and cardiomyopathy (360). For cardiac arrhythmias, miR-208a, miR-328, miR-1, miR-133, and miR-212 play important roles in contributing to this disease (205). For atherosclerosis, miR-126, miR-181b, and miR-10a oppose the progression of atherosclerosis (205). In contrast, miR-33 has a positive effect on this disease (261). For ischemic heart disease, miR-1, miR-133, miR-26, miR-29, and miR-21 mediate arrhythmia, cell death, hypertrophy, and fibrosis (292). Meanwhile, miR-21, miR-24, and miR-210 promote angiogenesis and survival (292). Present studies have revealed miRNAs as a major factor contributing to heart disease and related CVD. Understanding the expression and downward effects of miRNAs will provide new insights into the mechanisms involved in CVD and potentially lead to new therapies.

Expression of miRNAs detectable in the peripheral blood of patients suffering from left ventricular afterload or conduction disorders can benefit the early detection and diagnosis of disease. In a prospective longitudinal cohort study of pregnant women in the first trimester (10–13 wk), the upregulation of miR-516-5p, miR-517, miR-520h, and miR-518b was found to be associated with a risk of later development of gestational hypertension (118). The human cytomegalovirus (HCMV)-encoded miRNA hcmv-miR-UL112 and lethal-7b (let-7b) were also found to have potential in prognostics for hypertension (105, 171). It is also essential to identify the expression of specific miRNAs dependent on the time of onset of myocardial infarction. Increased miR-1, miR-133a, miR-133b, and miR-4995p levels and decreased miR-122 and miR-375 levels were found in plasma samples from patients with ST segment elevation myocardial infarction collected several hours after the onset of myocardial infarction symptoms and coronary reperfusion (64). The changes are time dependent (64). Therefore, the evaluation of miRNAs provides not only new insights into the pathophysiology of cardiovascular diseases but also the potential of cardiac biomarkers to diagnose diseases and even provide a real-time glimpse of the progression of the disease.

lncRNAs in heart.

A number of studies have been published to screen for lncRNAs in the cardiovascular field (93, 269, 273, 306, 307, 315). Furthermore, recent investigations have begun to shed light on prospective biological activities of lncRNAs. For instance, some lncRNAs [e.g., autophagy-promoting factor (APF; 323), cardiac apoptosis-related lncRNA (CARL; 325), cardiac hypertrophy-related factor (CHRF; 324), myocardial infarction-associated transcript (MIAT; 367)] have been suggested to act as molecular sponges that can effectively sequester miRNAs. Such activities give lncRNAs the ability to rapidly fine-tune a cellular transcriptional program via mediating the availability of miRNAs. In addition, a recent study shows that the lncRNA antisense Igf2r RNA (Airn) binds an RNA-binding protein, insulin-like growth factor 2 mRNA-binding protein-2 (Igf2bp2), to control the translation of genes in cardiomyocytes (117). Besides posttranscriptional control, other lncRNAs [e.g., Braveheart (140, 345), cardiac mesoderm enhancer-associated non-coding RNA (CARMEN; 241), and cardiac hypertrophy-associated epigenetic regulator (Chaer; 328)] have been implicated to bind and regulate epigenetic factors [namely, polycomb repressive complex 2 (PRC2)], which give them the ability to activate or repress genetic programs via directing enzyme-mediated modifications in chromatin structure. More recently, a promoter-associated lncRNA called Upperhand (Uph) was reported to be important for its transcription in cis to promote cardiac transcription factor Hand2 expression via GATA-4 binding and superenhancer activity (7). As of late, it is estimated that the number of expressed lncRNAs surpasses that of protein-coding genes (212). What is more, hundreds of lncRNAs have been detected in the heart (151, 241–243), and only a small fraction of these have been functionally studied. In fact, potential connections to CVD are still in the early stage (85, 100, 109, 117, 149, 163, 174, 176, 185, 216, 249, 314, 323, 324, 344), which emphasizes the importance and novelty of more intensive lncRNA research in the heart.

circRNAs in heart.

One interesting feature of circRNAs is that they are considerably more stable than noncircular RNAs (e.g., mRNAs), principally because circRNAs lack RNase-accessible free ends. Because of their stability, circRNAs are able to accumulate and exist in plasma for longer periods. As such, several studies have reported the detection of circRNAs in the circulating plasma, which highlights the possible use of circRNAs as biomarkers for various human diseases/conditions, including CVD (147, 192, 194). For example, using peripheral blood of patients with type 2 diabetes (T2D) compared with healthy controls, the circRNA Hsa_circ_0054633 was identified as a potential diagnostic biomarker of prediabetes and T2D (358). Furthermore, a recent study identified circRNA ankyrin repeat domain 36 (circANKRD36) by comparing peripheral blood from patients with T2D and healthy controls. circANKRD36 was significantly upregulated in patients with T2D, which was positively correlated with elevated glucose and glycosylated hemoglobin levels as well as signs of inflammation based on the expression of interleukin-6 (IL-6; 86). Taken together, more systematic and functional studies of circRNAs in the heart and their relationship to CVD will greatly enhance our understanding of the pathophysiological mechanisms of heart failure.

PULMONARY DISEASES

Research on ncRNAs has expanded into a huge number of diseases, and respiratory diseases are no exception. In addition to playing a pivotal role in the pathological mechanisms of pulmonary diseases, expression levels of ncRNAs in whole blood, plasma, serum, urine, sputum, or exhaled breath condensate are of great interest as potential biomarkers for diagnosis, prognosis, and stratification of patients for personalized therapeutic strategies. The following section seeks to provide a nonexhaustive overview of the clinical usefulness of ncRNAs as biomarkers and therapeutic targets in selected pulmonary diseases.

Lung cancer.

In the last decade, ncRNAs have emerged as key therapeutic targets and biomarkers in lung cancer. Thus, a great deal of data has been gathered demonstrating that abnormal expression of miRNAs, lncRNAs, and more recently circRNAs contributes to cancer initiation and progression through a range of different mechanisms (Fig. 2). The demonstration that miRNAs constitute appealing anticancer drug targets is perhaps best exemplified by let-7 miRNA (Table 1). Indeed, reduced expression of let-7 miRNA has repeatedly been observed in non-small cell lung cancer (NSCLC), resulting in enhanced cell proliferation and survival through derepression of Ras and c-Myc oncogenes (131, 270). Intranasal delivery of let-7 was shown to exert both preventive (81) and curative (305) effects in a genetic lung cancer model, providing a strong rationale for testing let-7 miRNA replacement therapy in human lung cancer. In parallel, because of the high incidence and mortality rates of lung cancer, circulating ncRNAs have gained considerable interest as a novel biomarker. Using a screening approach, various circulating ncRNA signatures have been shown to predict lung cancer diagnosis, histological subtype, treatment response, and survival (123). Although the quest for circulating ncRNA signatures is marked by a low concordance among studies, a serum miRNA signature called “miR-Test” was developed to detect early lung cancer in a high-risk group composed of heavy smokers over 50 yr of age (224). In this study, the overall sensitivity, specificity, and accuracy of the miRNA signature were 77.8, 74.8, and 74.9%, respectively, indicating that miR-Test may represent a valuable screening tool to complement imaging-based diagnostic approaches. The lncRNAs urothelial carcinoma-associated 1 (UCA1) and metastasis-associated lung adenocarcinoma transcript 1 (MALAT1) have also been shown to exert oncogenic functions in lung cancer cells (278, 337, 353). Because UCA1 expression in tumor tissues correlates with that obtained in plasma in patients with cancer, UCA1 levels may represent a promising tool for lung cancer diagnosis (319). Similarly, serum exosomal MALAT1 was reported to be highly expressed in patients with NSCLC and positively associated with tumor stage and lymphatic metastasis (353). Nevertheless, the clinical validations of their diagnostic and prognostic utilities in human lung cancer remain unproven.

Fig. 2.

Nonexhaustive list of non-coding RNAs (ncRNAs) implicated in selected respiratory diseases and serving as potential disease biomarkers. BAL, bronchoalveolar lavage; COPD, chronic obstructive pulmonary disease; Ki67, proliferation marker protein Ki-67; let-7c, lethal-7c; lncH19, long ncRNA (lncRNA) H19 imprinted maternally expressed transcript; lncHOTAIR, lncRNA HOX transcript antisense RNA; lncMALAT1, lncRNA metastasis-associated lung adenocarcinoma transcript 1; lncMEG3, lncRNA maternally expressed 3; lncPAXIP1-AP1, lncRNA PAXIP1 antisense RNA 1; lncTUG1, lncRNA taurine-upregulated gene 1; lncUCA1, lncRNA urothelial carcinoma-associated 1; LnRPT, lncRNA regulated by PDGF and transforming growth factor-β; miR, microRNA; TA, tracheal aspirate.

Table 1.

Nonexhaustive list of miRNAs involved in pulmonary diseases

miRNA	References	Roles	Partners/Targets	Diseases
let-7	(81, 131, 270, 305)	Reduced let-7 derepresses Ras and c-Myc, which enhances cell proliferation and survival	Ras, c-Myc	Lung cancer
miR-17	(40, 190, 257)	Increased miR-17 in PAH decreases Mfn2 expression, which promotes proliferation and apoptosis resistance of PAH PASMCs; inhibition of miR-17 also upregulates p21 (CIP1/WAF1) in rodent PH models	Mfn2, p21 (CIP1/WAF1)	PAH
miR-19a	(106, 286, 287)	miR-19a is increased in epithelia of subjects with severe asthma; miR-19a promotes bronchial epithelial cell proliferation by targeting TGFBR2	TGFBR2	Asthma
miR-21	(177, 182, 189, 274, 339)	miR-21 is increased in the lungs of mice with bleomycin-induced fibrosis as well as in the lungs of patients with IPF; increased miR-21 is found in bronchial epidermal cells and serum from patients with asthma and animal models of asthma; miR-21 directly inhibits PTEN expression and promotes airway smooth muscle cell proliferation and migration in asthma; upregulation of miR-21 is found in a rodent model of allergic airway inflammation; miR-21 targets IL-12p35	Smad7, PTEN, IL-12p35	IPF, asthma, allergic airway inflammation
miR-25	(115)	Increased miR-25 expression decreases MCU in PAH, which increases cytosolic calcium level and lowers intramitochondrial calcium level; this promotes a glycolytic shift in metabolism, increases mitochondrial fission, and enhances rates of cell proliferation	MCU	PAH
miR-34a-3p	(49)	Downregulation of miR-34a-3p increases expression levels of MiD49 and MiD51, which increases mitochondrial fission and cell proliferation in PAH	MiD49, MiD51	PAH
miR-34a-5p	(300)	Upregulated in animal models of BPD and in tracheal aspirates and lungs from human neonates who subsequently develop BPD	Angiopoietin-1	BPD
miR-124	(39, 350, 357)	Decreased miR-124 levels increase PTBP1 and cause a proglycolytic increase in the PKM2-to-PKM1 isoform ratio in pulmonary vascular endothelial cells and fibroblasts in PAH; miR-124 is downregulated in patients with NSCLC; miR-124 suppresses proliferation and glycolysis in NSCLC by targeting AKT-GLUT1/HKII	PTBP1, Akt	Lung cancer, PAH
miR-126	(207, 245)	miR-126 expression is upregulated in allergic asthma, which promotes TH2 cytokine production; increased level of miR-126 in plasma is predictive of asthmatic status	Unknown	Asthma
miR-138	(115)	Increased miR-138 decreases MCU in PAH, which increases cytosolic calcium level and lowers intramitochondrial calcium level; this promotes a glycolytic shift in metabolism, increases mitochondrial fission, and enhances rates of cell proliferation	MCU	PAH
miR-204	(39, 60, 285)	Downregulation of miR-204 upregulates SHP2 and activates the Src kinase and NFAT; miR-204 has therapeutic efficacy in a preclinical model of PAH; miR-204 is downregulated in NSCLC cell lines; miR-204 inhibits human NSCLC metastasis through suppression of NUAK1	SHP2, NUAK1	Lung cancer, PAH
miR-210 (hypoxamiR)	(42, 256, 331)	Hypoxia induces expression of miR-210; miR-210 is elevated in late-stage NSCLC and vascular and endothelial tissue affected by PH; elevated miR-210 impacts energy metabolism by repressing genes implicated in mitochondrial respiration, including SDHD and ISCU1/2	SDHD, ISCU1/2	Lung cancer, PAH
miR-218	(57, 58, 341)	miR-218 is reduced in serum and lung tissue from patients with COPD and in both cell-free bronchoalveolar lavage fluid and sputum from mice exposed to cigarette smoke; miR-218 represses TNFR1-mediated activation of NF-κB in smoking-induced bronchiolitis of COPD	TNFR1	COPD
miR-876–3p	(153)	Reduced in tracheal aspirates from BPD-susceptible neonates and animal model of BPD; expression levels may be a biomarker for BPD; augmentation of miR-876-3p decreases neutrophilic inflammation and rescues alveolar simplification in a murine BPD model	Unknown	BPD

BPD, bronchopulmonary dysplasia; CIP1/WAF1, cyclin-dependent kinase (CDK)-interacting protein-1/wild-type p53-activated fragment 1; COPD, chronic obstructive pulmonary disease; GLUT1, glucose transporter 1; HKII, hexokinase II; hypoxamiR, hypoxia-induced microRNA; IPF, idiopathic pulmonary fibrosis; ISCU1/2, iron-sulfur cluster assembly protein-1/2; let-7, lethal-7; MCU, mitochondrial calcium uniporter; Mfn2, mitofusin 2; MiD49 and MiD51, mitochondrial dynamics protein of 49 and 51 kDa, respectively; miRNA (miR), microRNA; NFAT, nuclear factor of activated T cells; NF-κB, nuclear factor-κ-light chain enhancer of activated B cells; NSCLC, non-small cell lung cancer; NUAK1, NUAK family SNF1-like kinase 1; PAH, pulmonary arterial hypertension; PASMCs, pulmonary artery smooth muscle cells; PH, pulmonary hypertension; PKM1 and PKM2, pyruvate kinase isozyme M1 and M2, respectively; PTBP1, polypyrimidine tract-binding protein-1; PTEN, phosphatase and tensin homolog; SDHD, subunit D of succinate dehydrogenase complex; SHP2, tyrosine-protein phosphatase nonreceptor type 11; Smad7, mothers against decapentaplegic homolog (SMAD) family member 7; TGFBR2, transforming growth factor-β receptor 2; TH2, T helper type 2; TNFR1, tumor necrosis factor receptor 1.

Pulmonary arterial hypertension.

Pulmonary arterial hypertension (PAH) is a complex and multifaceted disease characterized by persistent elevation of the mean pulmonary artery (PA) pressure as a result of sustained pulmonary vasoconstriction and vascular remodeling, a process driven by exaggerated cell proliferation and survival of PA resident cells. As a consequence of a narrowed PA and increased pulmonary vascular resistance, the right ventricular afterload increases and right heart failure eventually develops, which leads to high mortality (302). Given that PAH and cancer cells share a hyperproliferative and apoptosis-resistant phenotype (29, 258), fueled in part by a metabolic switch toward glycolysis (298), it is not surprising that multiple ncRNAs described as aberrantly expressed in cancer cells are also documented in PAH (Fig. 2). For instance, elevated expression of hypoxia-induced miR-210 is found in lung cancer (256) and PAH cells (331), impairing energy metabolism by repressing genes implicated in mitochondrial respiration. Similarly, downregulation of miR-204 and downregulation of miR-124 are features of both PAH (39, 60) and lung cancer cells (285), accounting for enhanced signal transducer and activator of transcription 3 (STAT3) activation and increased glycolysis (39, 357). In direct connection, the lncRNA MALAT1, reported to sponge miR-204 and miR-124 (167, 181), is upregulated by hypoxia and drives PA smooth muscle cell (PASMC) proliferation (31). On the basis of preclinical studies, several miRNAs have also been identified as key regulators of pathological vascular remodeling. Indeed, the downregulation of miR-424 and miR-503 and the subsequent elevation in expression of their downstream target [i.e., fibroblast growth factor 2 (FGF2)] in PA endothelial cells (PAECs) were reported to exert a proproliferative effect in an autocrine/paracrine manner. More importantly, restoration of their expression improved pulmonary hypertension in multiple animal models (138). Increased levels of miR-130/301 were identified as a common feature of PAH-PAECs and PAH-PASMCs, leading to reduced expression of the transcription factor peroxisome proliferator-activated receptor-γ (PPARγ) and culminating in enhanced vasoconstriction and proliferation of PA cells (21, 22). Furthermore, inhibition of miR-130/301 reversed Sugen/hypoxia-induced PAH in mice, whereas forced expression of miR-130 resulted in increased vessel wall thickness (22). Naturally, miRNAs implicated in the abnormal behavior of PA resident cells in PAH are not limited to those mentioned above. We invite the reader to refer to recent publications for more comprehensive coverage of the miRNAs specifically implicated in PAH (214, 234).

The repertoire of lncRNAs implicated in pulmonary hypertension (PH) has expanded greatly over the last 2 years. Among them, downregulation of lncRNA regulated by PDGF and transforming growth factor-β (LnRPT; 45) and upregulation of PAXIP1 antisense RNA 1 (PAXIP1-AP1; 126) were recently documented affecting multiple PH transcriptional programs. Despite the recognition of ncRNAs as key regulators of PAH pathogenesis with numerous studies clearly showing beneficial effects of miRNA-based therapy in preclinical models, no clinical trials using ncRNAs as therapeutic targets are currently underway in PAH. Furthermore, very few studies have documented a relationship between deregulation of an ncRNA in the blood of patients with PAH and disease severity or survival (263), and thus their utility as biomarkers in PAH management remains to be confirmed.

Role of miRNAs in mitochondrial and metabolism abnormalities seen in PAH and cancer.

Similar to cancers, it is established that the hyperproliferative and apoptosis-resistant phenotype of PAH is related in part to impaired mitochondrial dynamics and a metabolic shift from oxidative metabolism to aerobic glycolysis (9, 298). Mitochondrial dynamics is a term that encompasses mitochondrial fission (division), fusion (union), and mobility. Although mitochondria have a discrete genome, mitochondrial dynamics are primarily regulated by nuclear-encoded proteins. Dysregulation of these proteins, most of which are GTPases, leads to a shift in the balance of fission and fusion that favors fission and leads to a hyperfragmented mitochondrial network. This phenotype is ubiquitous in hyperproliferative, apoptosis-resistant cells (9, 220).

Excessive mitochondrial fission is a shared pathological feature of cancer cells and PAH cells (i.e., smooth muscle cells and fibroblasts; 9, 44, 304), and its inhibition leads to cell cycle arrest and apoptosis (49, 203, 262). Mechanistically, the increase in mitotic fission results from the upregulation of the GTPase dynamin-related protein-1 (Drp1; 203, 262) and its binding partners, including mitochondrial dynamics protein of 49 kDa (MiD49) and 51 kDa (MiD51; 13, 96, 125, 244). miRNAs play important roles in mitochondrial dynamics including fission. For example, Li et al. (166) showed that miR-30 indirectly inhibited Drp1 expression by downregulating its transcription regulator p53, whereas overexpression of miR-30a and miR-30b reduced Drp1-mediated mitochondrial fission in cardiomyocytes (166). Interestingly, miR-30b levels are elevated in patients with PAH compared with healthy subjects, and the degree of elevation (though modest) was proportional to the severity of the PH defined by mean PA pressure measured by right heart catheterization (329). However, the authors did not assess mechanistic impact of this miRNA (if any) on mitochondrial morphology, dynamics, or Drp1 activation. Thus, the relevance of miR-30 to PAH remains unclear.

For Drp1 to mediate mitochondrial fission, it must interact with several binding partners (e.g., MiD49 and MiD51). Interestingly, MiD49 and MiD51 expressions are upregulated in PAH-PASMCs compared with those from individuals without PAH (49, 203). This contributes to increased mitochondrial fragmentation. Reducing MiD expression is sufficient to decrease rates of cell proliferation and increase apoptosis (49). We have recently demonstrated that the downregulation of miR-34a-3p, in both human and experimental PAH, accounts for the increase in MiD expression (Fig. 3A). Therapeutically, miR-34a-3p mimic nebulization in a rodent PAH model reversed pulmonary vascular disease and improved right ventricular functions (49), supporting the role of the miR-34/MiD/fission axis in the etiology of PAH. Finally, we demonstrated the biomarker potential of miR-34a-3p in two independent PAH cohorts (49).

Fig. 3.

Schematic diagram illustrating two different pathways involved in microRNA (miRNA)-regulated mitochondrial protein dysregulation in pulmonary arterial hypertension (PAH). A: decreased expression of miRNA-34a-3p (miR-34a-3p) in PAH leads to increased expression of mitochondrial dynamics protein of 49 kDa (MiD49) and MiD51, binding partners of dynamin-related protein-1 (Drp1; 49). Increased MiD expression increases Drp1-dependent mitotic fission and promotes cell proliferation. SR, sarcoplasmic reticulum. [Reprinted from Chen et al. (49), with permission of the American Heart Association.] B: increased levels of miR-25 and miR-138 in PAH-pulmonary artery smooth muscle cells (PASMCs) directly inhibit the expression of mitochondrial calcium uniporter (MCU; 115). The loss of MCU expression, exacerbated by increased expression of mitochondrial calcium uptake protein-1 (MICU1), reduces the function of the MCU complex. This simultaneously overloads cytosolic calcium while depriving the mitochondria of calcium. The former triggers PASMC migration and proliferation (and vasoconstriction), whereas the latter affects mitochondrial metabolism, inhibiting pyruvate dehydrogenase and promoting a shift to uncoupled glycolysis (the Warburg phenomenon). In aggregate these epigenetic changes promote cell proliferation and apoptosis resistance. CREB1, cAMP responsive element-binding protein-1; UTR, untranslated region. [Reprinted from Hong et al. (115), with permission of the American Thoracic Society. Copyright © 2019 American Thoracic Society.]

Mitochondrial fusion is mediated by mitofusin 1 (MFN1), mitofusin 2 (MFN2), and optic atrophy protein-1 (OPA1; 124, 271), all of which are critically regulated by miRNAs. The expression of fusion mediators, notably MFN2, is decreased in PAH (262, 268). Mfn2 is regulated by miR-17, a member of the miR-17/92 cluster. miR-17 has been demonstrated to directly target the 3′-UTR of the MFN2 gene and is associated with several cancers, and thus it is considered to be an oncomiR (107). miR-17 level is significantly increased in rodent and human PAH (40, 190), and by decreasing MFN2, it promotes proliferation and apoptosis resistance of PAH-PASMCs (190). Interestingly, although miR-17 inhibition in rodent PH models improved hemodynamics, its effects were not exclusively mediated by MFN2 but also associated with upregulation of p21 [cyclin-dependent kinase (CDK)-interacting protein-1 (CIP1)/wild-type p53-activated fragment 1 (WAF1)], a cell cycle inhibitor (257), and through the modulation of the bone morphogenetic protein receptor type II (BMPR2) axis, a gene that is mutated in up to 80% of familial PAH cases and 10% of idiopathic PAH cases. BMPR2 is generally downregulated in PAH, even in the absence of a BMPR2 mutation (83, 99). Another miRNA that has been demonstrated to directly target the 3′-UTR of MFN2 is miR-761. miR-761 is upregulated in hepatocellular carcinoma compared with controls. Inhibition of miR-761 repressed tumor growth and metastasis by upregulating MFN2 (363), but its role in PAH has never been explored.

Epigenetic control of mitochondrial metabolism in both PAH and cancer.

Not only do cancer and PAH share a mitochondrial morphologic phenotype, but also they share a common Warburg-like metabolic dysfunction characterized by a shift from oxidative glucose metabolism to uncoupled aerobic glycolysis (11, 175). The shift away from oxidative metabolism reduces the likelihood of mitochondrial-mediated apoptosis and promotes cell proliferation and lactate production. In PAH, pyruvate dehydrogenase kinase (PDK)-mediated inhibition of pyruvate dehydrogenase (PDH) contributes to induction of aerobic glycolysis in both the lung vasculature and right ventricle (RV; 11). Different PDK isoforms are involved differentially in target organs in PAH. Of the four PDK isoforms (PDK1–PDK4), PDK4 is the key regulator of PDH in the RV, and PDK1 is responsible in pulmonary vasculature (200, 248). Inhibition of PDK by a small molecule, dichloroacetate, is effective in rectifying the Warburg metabolic imbalance and is beneficial in multiple animal models of PAH and in some patients with PAH (102, 210, 215). Although no miRNAs have been identified to target PDK in PAH, multiple miRNAs have been validated to target PDKs in cancer. In pancreatic and esophageal cancers, downregulation of miR-375 leads to the upregulation of its target, PDK1, whereas in osteosarcoma, decreased miR-379 increases PDK1 (172, 173, 361). There are also several miRNAs targeting PDK4 in cancer. Significant reduction or absence of miR-211 is observed in human melanoma cells, and its direct target, PDK4, is significantly upregulated. Administration of miR-211 suppressed the invasiveness and aggressiveness of melanoma by suppressing PDK4 expression (208).

In PAH-PAECs and PA fibroblasts, the activity of pyruvate kinase (PK), the terminal step in glycolysis, is altered by increased expression of polypyrimidine tract-binding protein-1 (PTBP1). PTBP1 regulates splicing and inhibits the expression of the PK isozyme M1 (PKM1) isoform, relative to the PKM2 isoform. miR-124 targets PTBP1 in cancer cells, PAECs, and blood outgrowth endothelial cells from patients with PAH [reviewed by Archer (10)]. Low levels of miR-124 cause the observed increase in PTBP1 and cause a proglycolytic increase in the PKM2-to-PKM1 ratio. This glycolytic shift (another mechanism of achieving Warburg metabolism) promotes rapid cell proliferation and apoptosis resistance (39, 350). miR-124 mimics might be a therapeutic strategy for modulating pyruvate kinase and treating PAH and cancer.

Epigenetic control of the mitochondrial calcium uniporter complex and effects on metabolism.

Another noncanonical yet important role of mitochondria in cell biology is calcium regulation. The endoplasmic reticulum (ER) physically connects with mitochondria at the mitochondria-associated membranes, allowing for transfer of Ca²⁺ from ER to mitochondria (233). The mitochondrial calcium uniporter complex (MCUC) is a multiprotein calcium channel in the inner mitochondrial membrane (92). Within the MCUC, the mitochondrial calcium uniporter (MCU) serves, with essential MCU regulator (EMRE), as the pore-forming subunit. MCUC function is crucial for cytosolic and mitochondrial calcium balance (18). The MCUC allows uptake of calcium from the cytosol, supplying calcium to the mitochondria’s calcium-dependent dehydrogenases that mediate oxidative metabolism (66). MCUC-mediated calcium influx also buffers cytosolic calcium levels. Increased cytosolic calcium level in cancer cells and PAH-PASMCs is partially due to the dysfunction of the MCUC, due in large part to miRNA-mediated downregulation of MCU expression. Decreased MCU expression and function have been reported in both PAH and cancer (115, 316). In PAH-PASMCs, downregulation of MCU expression and upregulation of the inhibitory mitochondrial calcium uptake protein-1 (MICU1) subunit are noted (115). Both of the changes impair MCUC function and contribute to increased cytosolic calcium concentration and decreased mitochondrial calcium concentration. Reduced concentrations of mitochondrial calcium inhibit calcium-dependent dehydrogenases, such as PDH. Simultaneously, MCU downregulation increases cytosolic calcium, which promotes mitochondrial fission and vasoconstriction. In PAH-PASMCs, increased expression of miR-25 and miR-138, both of which are proven to bind to the 3′-UTR of MCU, leads to decreased MCU subunit expression. Conversely, nebulized anti-miRNAs against miR-25 and miR-138 restored MCU expression, reduced cell proliferation, and regressed established PAH in the monocrotaline-induced preclinical model of PAH (115; Fig. 3B). Likewise, in human colon cancer, where upregulation of miR-25 downregulates MCU, inhibition of miR-25 increased intramitochondrial calcium levels and increased apoptosis (201).

Prediction of novel mitochondria-relevant miRNAs in PAH.

The Archer laboratory has adopted an unbiased in silico approach to identify novel miRNAs that are predicted to bind to each of the eight mitochondrial-relevant genes identified in Table 2. To do this, we utilized miRWalk2.0 (79), a database of predicted miRNA-gene interactions. Default parameters result in miRNAs that are predicted with significance (P < 0.05) to bind to sites within 2 kb of the 3′-UTR of specified target genes. We identified a total of 2,512 putative miRNAs that are predicted to bind to at least 1 of these mitochondria-related genes (Supplemental Table S1; Supplemental Material for this article is available online at https://doi.org/10.6084/m9.figshare.7808732). Importantly, we show that the majority of these miRNAs (2,387 miRNAs in total) are predicted to bind more than one target gene. When we screened for the miRNAs experimentally validated by others, we confirmed that all the miRNAs that we discussed, except miR-182, are predicted to bind to multiple mitochondria-related gene targets. We were interested to see whether any of these epigenetic regulators of nuclear gene expression were also predicted to bind to mitochondrially encoded genes. Of the miRNAs predicted to bind to 1 or more nuclear-encoded, mitochondrial-relevant genes, 2,512 are also predicted to bind to regions on the mitochondrial genome (Supplemental Table S2).

Table 2.

Genes and validated miRNAs involved in mitochondrial fission/fusion, calcium regulation, and mitochondrial metabolism of cancer and PAH

Gene Symbol	Entrez Gene ID	Chromosome	Map	Validated in PAH	Validated in Cancers
DNM1L (Drp1)	10059	12	12p11.21		miR-30 (166)
MIEF2 (MiD49)	125170	17	17p11.2	miR-34a-3p (49)
MIEF1 (MiD51)	54471	22	22q13	miR-34a-3p (49)
MFN2	9927	1	1p36.22	miR-17 (190)	miR-761 (363)
MCU	90550	10	10q22.1	miR-25 (115), miR-138 (115)	miR-25 (201)
PDK1	5163	2	2q31.1		miR-375 (172, 361), miR-379 (173)
PDK4	5166	7	7q21.3		miR-182 (218, 349), miR-211 (208)
PKM	5315	15	15q22	miR-124-3p (39, 350)

DNM1L, dynamin 1-like; Drp1, dynamin-related protein-1; MCU, mitochondrial calcium uniporter; MFN2, mitofusin 2; MiD49 and MiD51, mitochondrial dynamics protein of 49 and 51 kDa, respectively; MIEF1 and MIEF2, mitochondrial elongation factor 1 and 2, respectively; miRNA (miR), microRNA; PAH, pulmonary arterial hypertension; PDK1 and PDK4, pyruvate dehydrogenase kinase 1 and 4, respectively; PKM, pyruvate kinase M1/2.

Finally, we wanted to place our predicted miRNAs into some biological context, and so we filtered the list of putative miRNAs against our previously published miRNA microarray data from the PASMCs of human patients with PAH compared with individuals without PAH (49, 115). We present here a list of 25 putative miRNA targets that are known to be downregulated in human PAH by >2-fold change and 2 that are known to be upregulated by >2-fold change, which are also predicted to bind to multiple genes related to mitochondrial dynamics and metabolism, as well as several mitochondrially encoded genes (Table 3). More work is required to understand these miRNAs and their influence on the molecular network since those miRNAs that are upregulated are targeting the same genes as those miRNAs that are downregulated. It is plausible that these miRNAs are attempting to rescue the dysfunction in mitochondrial dynamics in cancer and PAH and are therefore also relevant targets.

Table 3.

List of 25 putative miRNA targets in human PAH

	Nuclear Genes Involved in Mitochondrial Dynamics and Metabolism
Predicted miRNA	PKM	PDK1	PDK4	MiD51	MiD49	MFN2	MCU	Drp1	Count	Mitochondrially Encoded Targets	miRNA Microarray: PASMCs from PAH Compared with Controls
hsa-miR-146a-5p			P	P			P	P	4	ND5	−248.37
hsa-miR-29c-3p			P	P	P				3	CYTB	−9.33
hsa-miR-660-5p				P		P		P	3	COX2	−5.49
hsa-miR-224-3p	P	P	P			P	P		5	COX3	−5
hsa-miR-452-5p	P	P	P	P	P	P			6	RNR2	−4.6
hsa-miR-22-5p	P	P	P	V (112)	P	P	P	P	8	RNR2	−4.59
hsa-miR-126-3p					P			P	2		−3.82
hsa-miR-148b-3p		P	P			P		P	4	ND2	−3.38
hsa-miR-181c-5p		P	P	P	P	P	P		6		−3.05
hsa-miR-30e-5p		P	P	P			P	P	5	RNR2	−2.95
hsa-miR-188-5p	P	P	P		P		P		5	RNR2	−2.76
hsa-miR-154–5p	P		P	P	P	P		P	6	ND6	−2.73
hsa-let-7i-5p				V (139)	P	P	P	P	5	ND5	−2.62
hsa-miR-34a-5p			P	P	P	V (112)	P	P	6	ND4	−2.52
hsa-miR-628-3p			P	P	P	P	P	P	6	TRNA	−2.52
hsa-miR-21-5p			P	P	P	P	P	P	6	COX1	−2.51
hsa-miR-425-3p			P	P	P	P			4	ND6	−2.35
hsa-miR-145-5p	P	P	P	P	P				5	RNR1	−2.33
hsa-miR-21-3p		P	P	P		P			4		−2.3
hsa-miR-148a-3p		P	P			P		P	4	ND2	−2.29
hsa-miR-193a-3p	P			P				P	3	ND1	−2.27
hsa-miR-143-5p	P	P		P	P	P	P	P	7		−2.11
hsa-miR-6782-5p	P					P	P		3	ND3	−2.07
hsa-miR-34a-3p	P	P	P	P	P		P		6	ND1	−2.04
hsa-miR-6075					P	P			2	ND5	−2.01
hsa-miR-6831-5p			P	P					3	ND4	2.19
hsa-miR-3911		P	P	P		P	P	P	7	ND1	2.11

The threshold of 2-fold change was applied. P and V, predicted and validated microRNA (miRNA, or miR) targets, respectively. COX1–3, cyclooxygenase-1–3, respectively; CYTB, cytochrome b; Drp1, dynamin-related protein-1; MCU, mitochondrial calcium uniporter; MFN2, mitofusin 2; MiD49 and MiD51, mitochondrial dynamics protein of 49 and 51 kDa, respectively; ND1–6, NADH-ubiquinone oxidoreductase chains 1–6, respectively; PAH, pulmonary arterial hypertension; PASMCs, pulmonary artery smooth muscle cells; PDK1 and PDK4, pyruvate dehydrogenase kinase 1 and 4, respectively; PKM, pyruvate kinase M1/2; RNR1, mitochondrial-derived peptide MOTS-c; RNR2, humanin; TRNA, mitochondrially encoded tRNA-Ala.

Idiopathic pulmonary fibrosis.

Idiopathic pulmonary fibrosis (IPF) is commonly considered to result from repeated alveolar epithelial injuries promoting epithelial-mesenchymal transition and driving uncontrolled proliferation and persistent activation of myofibroblasts. Consequently, exaggerated deposition of connective tissue matrix proteins occurs leading to irreversible deterioration of the normal lung architecture, reduced lung compliance, and compromised gas exchange (334). Liu et al. (177) demonstrated that increased expression of miR-21 is a common feature of experimental and human IPF. Notably, sequestration of miR-21 by intratracheal delivery of miR-21 antisense prevented bleomycin-induced lung fibrosis in mice (177). Similarly, in vivo silencing of miR-21 inhibited cardiac fibrosis and attenuated cardiac dysfunction in a model of heart failure induced by pressure overload (303) as well as ameliorated renal fibrosis induced by unilateral ureteral obstruction (43). Collectively, these findings identify miR-21 as a master regulator of fibrosis. Subsequent studies have addressed circulating expression levels of miR-21 as a potential biomarker in IPF. As observed in lung tissues, patients with IPF exhibit elevated serum miR-21 expression that correlates with disease severity, as assessed by forced vital capacity and radiological features (170, 196). Along with miR-21, aberrant expression of numerous ncRNAs [e.g., miR-101, miR-29, miR-155, and lncRNA H19 imprinted maternally expressed transcript (H19)] was found to be associated with the fibrotic process (62, 120, 169). For some, their altered expression profile in the serum of patients with IPF correlated with forced vital capacity and fibrosis severity on imaging (169).

Chronic obstructive pulmonary disease and asthma.

Chronic obstructive pulmonary disease (COPD) and asthma are two highly prevalent and heterogeneous disorders characterized by airflow limitation and airway remodeling. Both diseases are characterized by an exaggerated, but distinct, inflammatory response in which deregulated expression of ncRNAs plays a significant role. Among others (301, 310), miR-218 was found to be downregulated in lung tissue from smokers without airflow limitation and patients with COPD compared with never-smokers. Similar findings were noted in cigarette smoke-exposed mice (58). In a separate study, reduced miR-218 expression in cell-free bronchoalveolar lavage fluid and sputum was found to be the common denominator in mice exposed to cigarette smoke and in the induced sputum supernatant from smokers with COPD, respectively (57). Similarly, low levels of miR-218 were detected in serum from smokers with COPD compared with smokers without COPD (341). In vivo administration of an miR-218 inhibitor markedly exacerbated cigarette smoke-induced inflammation (58). In bronchial epithelial cells, miR-218 was reported to repress tumor necrosis factor receptor 1 (TNFR1)-mediated activation of nuclear factor-κ-light chain enhancer of activated B cells (NF-κB), leading to blockade of mucin 5AC, oligomeric mucus/gel-forming (MUC5AC), IL-6, and IL-8 production (341). Taken together, these findings indicate that miR-218 may serve as a novel biomarker and therapeutic target in COPD.

In asthma, many studies have been devoted to profile differentially expressed ncRNAs in body fluids and bronchial epithelial brushing from healthy controls and subjects with moderate and severe asthma or between the various inflammatory subtypes (111, 245, 336). High levels of miR-21 were detected in bronchial epidermal cells and serum from patients with allergic asthma compared with healthy controls (274, 339), as well as in multiple animal models mimicking asthma (189). Its inhibition using antagomiR significantly improved the development of allergic airway inflammation in mice (157). Similar to myofibroblasts in IPF, functional analyses showed that miR-21 promotes airway smooth muscle cell proliferation and migration by directly repressing PTEN expression (182). It also regulates the polarization of adaptive immune responses and activation of T cells (189). miR-126 and the miR-17-92 cluster member miR-19a are also implicated in asthma through the promotion of T helper type 2 (TH2) cytokine production as well as proliferation of bronchial epithelial cells (106, 207, 245, 286, 287). Further studies are needed to confirm the therapeutic potential of their modulation and their utility as noninvasive biomarkers in asthma.

Neonatal chronic lung disease.

Bronchopulmonary dysplasia (BPD), a common complication of low-birth weight preterm infants who require prolonged mechanical ventilation and oxygen supply for acute respiratory distress, remains a major problem in pediatric pulmonology units. Characterized by impaired lung development with simplified alveoli and dysmorphic pulmonary vascular growth, infants with BPD are at high risk to develop pulmonary hypertension. They also experience more frequent respiratory infections and display increased susceptibility to develop asthma and COPD in adulthood (28). Besides the urgent need to develop effective therapies to repair and restore lung architecture, a major challenge is the lack of reliable methods for predicting the development of BPD. Although considerable attention has been paid to ncRNAs in many respiratory diseases in adulthood, evidence of their implication in normal lung development and neonatal lung diseases remains limited (78, 238, 264). Most of the studies are preclinical, having analyzed the expression profile of ncRNAs in rodents at various stages of lung development or in response to lung injuries that mimic the morphological changes of human BPD (229). Nevertheless, miR-34a has recently emerged as a pivotal player in the development of human BPD. Indeed, upregulation of miR-34a was reported in various animal models of BPD as well as in tracheal aspirates or lungs from human neonates who subsequently develop BPD (300). Importantly, overexpression of miR-34a in control animals was sufficient to elicit a BPD phenotype, whereas its inhibition exerted therapeutic effects (300). An exosomal miRNA signature predictive of severe BPD has been identified (153). Among the components of this signature, reduced concentration of miR-876-3p in tracheal aspirates collected from extremely low birth weight infants at the first day of life was found to be the most sensitive biomarker. Further functional analysis demonstrated that supplementation in miR-876-3p decreases neutrophilic inflammation and rescues alveolar simplification in the hyperoxia-based mouse model of BPD, underscoring its potential therapeutic value (153).

Challenges in pulmonary diseases.

The implication of ncRNAs in the pathophysiology of pulmonary diseases is a highly productive area of research. Although several miRNA-based therapies (i.e., miRNA replacement or inhibition) have shown promising results as single agents in various preclinical animal models of pulmonary diseases, several challenges remain to be addressed before their implementation into clinical practice. Taking into account the extensive list of miRNAs (and, more recently, lncRNAs) documented as being implicated in each lung disease, the prioritization of ncRNAs to be targeted in clinical testing represents a major challenge.

Moreover, different miRNAs can target the same gene, and one miRNA can silence a subset of different genes involved in the same biological process. In this context, a combination therapy of two or more miRNAs is likely to achieve a better efficacy.

Despite the fact that the lungs offer the advantage of being easily accessible through nebulization, major barriers remain, such as 1) stability and integrity of miRNAs, 2) selective delivery to the target cells, and 3) off-target effects. To overcome these issues, several strategies are currently being developed. Indeed, chemical modifications of oligonucleotides encapsulated in nanostructured lipid or polymer carriers are used to protect the oligonucleotide from enzymatic degradation, thus enhancing bioavailability while minimizing toxicity and immunogenicity. Moreover, surface modifications of nanostructured lipid carriers can increase delivery in diseased cells and limit adverse side effects on healthy cells. Because the multiple molecular functions exerted by lncRNAs are only beginning to be appreciated (142), more work needs to be done before any therapeutic interventions in humans are performed.

Apart from their potential therapeutic value, the ability of ncRNAs to be easily detected in body fluids, sputum, and exhaled breath condensate has generated interest in the potential use of such factors to improve diagnosis, predict outcome, and guide treatment in pulmonary diseases (209, 213). Although ncRNA signatures abound, there is a poor overlap between them. Explanations for these discordances include the lack of standardized methods in the collection of samples and in detection, the types of input material used for ncRNA detection, and the small sample size. Although reproducibility in independent validation cohorts by different research groups is mandatory before moving into the clinic, there is no doubt that ncRNAs hold considerable promise as biomarkers for pulmonary diseases.

ncRNAS AND SKELETAL MUSCLES

Skeletal muscles represent 30–40% of body weight (103) and provide power for human activities. Muscle development and regeneration are thus of primary importance throughout life. Skeletal muscle is formed from myogenic progenitor cells expressing the paired box transcription factors paired box 3 (Pax3) and paired box 7 (Pax7) and destined to become myoblasts (20, 32). These myoblasts then undergo proliferation and differentiation and fuse into multinucleated myotubes under the regulation of myogenic regulatory factors, including myogenic differentiation 1 (Myod1/MyoD), myogenin, myogenic factor 5 (Myf5), and myogenic factor 6 [Myf6/muscle-specific regulatory factor 4 (MRF4); 19, 23]. Importantly, CVD (72, 135) and pulmonary diseases (17, 195, 199) discussed above are frequently associated with dysregulated myogenesis and regeneration. Even in adults, however, skeletal muscles remain an active organ with significant plasticity. Elucidating the factors and molecular regulatory pathways governing myogenesis and satellite cell behavior is thus the first step toward successful use of these cells in therapeutic strategies for muscle diseases. miRNAs, lncRNAs, and, more recently, circRNAs are emerging as novel families of regulators of muscle development and regeneration. The following section seeks to provide a nonexhaustive overview of ncRNAs during the muscle development and regeneration, especially focusing on muscle diseases.

miRNAs in skeletal muscle development.

During myogenesis, cells commit to myogenic lineage and progress along the myogenic pathway by proliferation and terminal differentiation, ultimately resulting in multinucleated myofibers (33). The full process of myogenesis is regulated and controlled by complex networks of signaling pathways. Not surprisingly, miRNAs have emerged as key regulators of skeletal muscle development (322), and the conditional deletion of Dicer in muscle causes perinatal lethality with reduced skeletal muscle mass and abnormalities in muscle fiber morphology (240). Many muscle-specific miRNAs, called myomiRs, have been shown to control myoblast or satellite cell proliferation and differentiation. The most studied myomiRs are the miR-1, miR-206, and miR-133 families, which account for nearly 25% of the total miRNA expression in skeletal muscle (Table 4). These families are encoded by three loci in humans, and members of each family differ in only a few nucleotides outside the seed region. They are thus assumed to be partly functionally redundant. miR-1 promotes myogenesis by targeting HDAC4, a transcriptional repressor of muscle gene expression. Conversely, miR-133 enhances myoblast proliferation by repressing SRF (47), whereas miR-1 and miR-206 downregulate Pax7 and induce myoblast differentiation (137). Surprisingly, despite their predominant role in myogenesis, the double knockout of the miR-1 and miR-133 families has no overt effect on skeletal muscle development (340), and muscle is grossly normal in a knockout of the miR-206/133b cluster (26). Conversely, mice lacking miR-133a develop an adult-onset centronuclear myopathy (179). Other muscle-specific miRNAs, such as miR-208 and miR-499, as well as many non-muscle-specific miRNAs, are also required for the differentiation of muscle through interaction with myogenic factors (Table 4).

Table 4.

Nonexhaustive list of miRNAs involved in myogenesis and skeletal muscle regeneration

miRNAs	References	Roles	Partners/Targets
miR-1 (myomiR)	(5, 47, 48, 187, 295)	Targets gap junction protein, Cx43, and HDAC4; also downregulates Pax7; miR-1 also targets YY1, which inhibits miR-1/133 expression, forming a feedback circuit; promotes satellite cell differentiation; also affects angiogenesis by modulating the levels of VegfA, which signals to adjacent endothelial cells	Cx43, HDAC4, Pax7, VegfA
miR-133 (myomiR)	(89, 348)	Promotes myoblast proliferation by targeting SRF, whereas it inhibits proliferation by repressing CCND1, inducing G1 phase arrest and targeting FGFR1 and PP2AC, which regulate phosphorylated ERK MAP kinase levels	SRF, CCND1, FGFR1, PP2AC, ERK MAP
miR-206 (myomiR)	(48, 66, 295)	Downregulates Pax7; affects angiogenesis by modulating the levels of VegfA, which signals to adjacent endothelial cells; promotes satellite cell differentiation	Pax7, VegfA
miR-15b-5p	(27, 276)	Downregulation of INSR and IRS1 and upregulation of PIK3R1	INSR, IRS1, PIK3R1
miR-17-92	(259)	Promotes myoblast proliferation
miR-20a/b	(191)	Represses myoblast proliferation
miR-24	(296)	TGF-β-mediated downregulation of miR-24; inhibits myogenesis through reduced expression of Mef2d, Myf5, MyoD, myogenin, and MyHC	Mef2d, Myf5, MyoD, myogenin, MyHC
miR-26a	(67, 335)	Promotes myoblast differentiation by negatively regulating the BMP signaling pathway (Smad1/Smad4 expression); the upregulation of miR-26a also inhibits Ezh2, a known suppressor of skeletal muscle cell differentiation	Smad1/Smad4, EZH2
miR-27b	(61, 281)	Promotes satellite cell differentiation and mitochondrial biogenesis through downstream genes of Foxj3 (Mef2c, PGC-1α, NRF1, and mtTFA)	Pax3, Foxj3
miR-29a	(95)	Promotes satellite cell proliferation	FGF2
miR-29b	(165)	Promotes muscle atrophy by targeting IGF-1 and PI3K, ultimately increasing atrogin-1 and Murf-1	IGF-1, PI3K
miR-31	(61)	Maintains satellite cell quiescence
miR-34c	(63, 327)	Represses myoblast proliferation; tightly related to in vivo skeletal muscle mitochondrial function in humans	YY1
miR-106b-5p	(355, 356)	Negatively regulates mitofusin-2 and PGC-1α expression
miR-128a	(226)	Regulates genes involved in insulin signaling, including INSR, IRS1, and PIK3R1
miR-135a-5p	(114)	Downregulates INSR and IRS2	INSR, IRS2
miR-149	(223)	Inhibits PARP-2 and promotes mitochondrial biogenesis via SIRT1/PGC-1α	PARP-2, SIRT1/PGC-1α
miR-150-5p	(63)	Tightly related to in vivo skeletal muscle mitochondrial function in humans
miR-155	(280)	Represses the expression of MEF2A to inhibit myoblast differentiation	MEF2A
miR-181	(232)	Required for skeletal myoblast terminal differentiation	HoxA11
miR-195	(272)	Targets cell cycle genes, including Cdc25 and Ccnd	Cdc25, Ccnd
miR-208 (myomiR)	(313)	Specifies muscle fibers
miR-214	(88, 132)	Targets Ezh2, the catalytic subunit of PRC2, promoting differentiation through negative feedback; promotes differentiation through myogenin and MyHC	EZH2, myogenin, MyHC
miR-320a	(63)	Tightly related to in vivo skeletal muscle mitochondrial function in humans
miR-351	(51)	Promotes myoblast proliferation
miR-374b	(193)	Negatively regulates myoblast differentiation	Myf6
miR-378	(94)	Promotes myogenic differentiation
miR-410	(290)	Promotes myogenic differentiation
miR-431	(338)	Promotes satellite cell differentiation	Pax7
miR-433	(290)	Promotes myogenic differentiation
miR-486 (myomiR)	(48, 66)	Coregulates Pax7 with miR-206 during myoblast differentiation
miR-489 (myomiR)	(55)	Maintains the quiescent state of satellite cells by targeting the oncogene Dek, thus preventing cell cycle entry	Dek
miR-497	(272)	Targets cell cycle genes, including Cdc25 and Ccnd	Cdc25, Ccnd
miR-499 (myomiR)	(129, 313)	Specifies muscle fibers; has suppressive effect on satellite cell adipogenic differentiation through PRDM16 downregulation	PRDM16
miR-501	(221)	Promotes myogenic differentiation
miR-675	(68)	Promotes myogenic differentiation
miR-715	(186)	Inhibits myogenic differentiation

BMP, bone morphogenetic protein; CCND1, cyclin D1; Cdc25, cell division cycle 25; Cx43, connexin-43; Dek, DEK proto-oncogene; EZH2, histone-lysine N-methyltransferase EZH2; FGFR1, FGF receptor 1; Foxj3, forkhead box j3; HDAC4, histone deacetylase 4; HoxA11, homeobox protein Hox-A11; INSR, insulin receptor; IRS1, insulin receptor substrate 1; Mef2, myocyte enhancer factor 2; miRNA (miR), microRNA; myomiR, muscle-specific miRNA; mtTFA, transcription factor A, mitochondrial; Murf-1, muscle-specific RING finger 1; Myf5 and Myf6, myogenic factor 5 and 6, respectively; MyHC, myosin heavy chain; MyoD, myogenic differentiation 1; NRF1, nuclear respiratory factor 1; PARP-2, poly(ADP-ribose) polymerase-2; Pax7, paired box 7; PGC-1α, peroxisome proliferator-activated receptor-γ coactivator-1α; PI3K, phosphatidylinositol 3-kinase; PIK3R1, PI3K regulatory 1; PP2AC, protein phosphatase 2A catalytic subunit; PRC2, polycomb repressive complex 2; PRDM16, histone-lysine N-methyltransferase PRDM16; SIRT1, sirtuin 1; Smad1 and Smad4, mothers against decapentaplegic homolog (SMAD) family members 1 and 4, respectively; SRF, serum response factor; TGF-β, transforming growth factor-β; VegfA, vascular endothelial growth factor A; YY1, YY1 transcription factor.

miRNAs in skeletal muscle regeneration.

Following muscle injury, three distinct overlapping phases are commonly distinguished: 1) an inflammatory reaction, 2) the activation and differentiation of satellite cells, and 3) the maturation of newly formed myofibers and remodeling of regenerated muscle. Satellite cells, a self-renewing population of adult stem cells, have remarkable regenerative abilities in adult muscles (23). They are located outside the plasma membrane of multinucleated muscle cells and beneath the basal lamina surrounding each myofiber (275). In response to muscle injury, satellite cells leave their quiescent state, become activated, proliferate, and migrate, ultimately differentiating and fusing to form new fibers to restore the damaged muscles, whereas a small portion of the activated satellite cells will rapidly exit the cell cycle and return to a quiescent state to replenish the satellite cell pool (144, 160). miRNAs modulate satellite cells’ quiescence. Satellite cells lacking Dicer escape quiescence and become activated precociously (55), whereas miR-489, miR-195, and miR-497 maintain the quiescent state of these cells by targeting DEK proto-oncogene (Dek), cell division cycle 25C (Cdc25), and cyclin D (Ccnd), thus preventing cell cycle entry (55, 272). Conversely, other miRNAs are involved in satellite cells’ proliferation and differentiation (76). miR-1, miR-206, and miR-133 are upregulated in activated satellite cells and promote their differentiation potentially mediated through direct targeting of Pax7 and Pax3 (48, 113).

miRNAs in skeletal muscle metabolism function.

Skeletal muscle is a highly metabolic tissue, accounting for up to ~80% of glucose disposal under insulin-stimulated conditions and playing a fundamental role in glycemic homeostasis (368). This phenomenon is mostly mediated through solute carrier family 2, facilitated by glucose transporter member 4 (GLUT4), which is rapidly translocated to the plasma membrane in response to the hormone (1, 141). Recently, the role of miRNAs in peripheral glycemic homeostasis has gained much attention. So far, miR-21a-5p, miR-29a-3p, miR-29c-3p, miR-93-5p, miR-106b-5p, miR-133a-3p, miR-133b-3p, miR-222-3p, and miR-223-3p have been reported to directly or indirectly regulate GLUT4 expression (82). Skeletal muscle is also characterized by high-energy needs and high mitochondrial content (90). More recently, miRNAs located inside mitochondria (mitomiRs) have been documented (294), regulating multiple aspects of mitochondrial functions, including mitochondrial biogenesis, energy metabolism, and electron transport chain subunits (see pulmonary diseases, Prediction of novel mitochondria-relevant miRNAs in PAH; 168, 223, 281, 343). Although only a few studies have investigated miRNA expression within human mitochondria, they have identified at least 200 miRNAs located within the mitochondria of cultured myotubes (15), including miR-320a, miR-196b-3p, miR-150–5p, and miR-34c-3p, which are tightly related to in vivo skeletal muscle mitochondrial function in humans (63).

lncRNAs in muscle development and regeneration.

lncRNAs are emerging as a novel family of regulators of myogenesis and satellite cell biology. A nonexhaustive list of lncRNAs implicated in myogenesis and muscle regeneration is shown in Table 5. After birth, the lncRNA H19 is maintained in skeletal muscle, where it encodes the conserved miR-675-3p and miR-675-5p, which target mothers against decapentaplegic homolog (SMAD) family member 1 (Smad1), SMAD family member 5 (Smad5), and cell division cycle 6 (Cdc6) and thus promotes skeletal muscle differentiation and regeneration (68, 251). The lncRNA H19 is also highly expressed in satellite cells (204), where it regulates the quiescent satellite cell pool and may promote satellite cell differentiation by suppressing sirtuin 1 (Sirt1) and forkhead box O1 (FoxO1; 342). The lncRNA H19 was also shown to enhance muscle insulin sensitivity, at least in part by activating protein kinase AMP-activated catalytic subunit-α1 (PRKAA1/AMPK; 98). Long intergenic non-protein-coding RNA, muscle differentiation 1 (linc-MD1), also a muscle-specific lncRNA, acts as a competing endogenous RNA by sequestering miR-133 (41). lncMyoD, an intergenic lncRNA encoded next to the MyoD gene, is directly activated by MyoD and plays a role in myoblast differentiation by binding to IGF2 mRNA-binding protein-2 (Igf2bp2/IMP2) and negatively regulating IMP2-mediated translation of proliferation genes (101). Finally, MyoD upstream noncoding (MUNC), a muscle-specific lncRNA located upstream of MyoD, acts on multiple promoters to increase myogenic gene expression during differentiation (228). As miRNAs, lncRNAs also regulate mitochondrial function and are deregulated in pathological states (184). Some are also encoded by the mitochondrial genome (155), although our understanding of their putative role in the mitochondria remains limited, especially for skeletal muscles.

Table 5.

Nonexhaustive list of lncRNAs involved in myogenesis and skeletal muscle regeneration

lncRNAs	References	Roles	Partners/Targets
CERNA	(227)	Regulates chromatin remodeling and Pol II recruitment at the promoter of MyoD
Dum	(326)	Promotes myoblast differentiation and muscle regeneration through recruiting DNA methyltransferase to repress Dppa2 gene	DNMTs
DBE-T	(30)	Binds and recruits Ash1L to the FSHD locus to derepress its nearby genes in cis	TrxG protein Ash1L
DWORF	(235)	Enhances SR Ca2+ uptake and myocyte contractility through displacing SERCA inhibitors and enhancing SERCA activity	SERCA
Gtl2 (Meg3)	(364)	Regulates skeletal muscle development by abolishing expression of imprinted Dlk1 genes with PRC2	PRC2
H19	(68, 98, 134, 342)	Regulates muscle differentiation by sponging let-7 and encoding miR-675; promotes the differentiation of bovine skeletal muscle satellite cells by suppressing Sirt1/FoxO1; enhances muscle insulin sensitivity, at least in part by activating AMPK	Ago2, let-7, Sirt1, FoxO1, AMPK
lncRNA-Six1	(35)	Cis-regulates its neighbor gene Six1 and promotes the expression of muscle growth-related genes
linc-MD1	(41)	Induces differentiation by acting as a decoy of miR-133 and miR-135 to increase the expression of MAML1 and MEF2C	miR-135, miR-133
lnc-mg	(366)	Promotes myogenesis by blocking miR-125b to control protein level of IGF2	miR-125b
lncMD	(297)	Promotes myogenesis by sequestering miR-125b to enhance protein level of IGF2 in cattle	miR-125b
lncMyoD	(101)	Promotes myoblast differentiation by competing with c-Myc for IMP2 binding	IMPs
lnc-SEMT	(330)	Promotes muscle development and growth, acting as a molecular sponge for miR-125b controlling IGF2 abundance	miR-125b, IGF2
linc-RAM	(347)	Promotes myogenic differentiation by enhancing the formation of MyoD-Baf60c-Brg1 complex	MyoD
Myolinc	(217)	Regulates myogenesis by recruiting the DNA/RNA-binding protein TDP-43 to the promoters of muscle marker genes (e.g., MyoD)	TDP-43
linc-YY1	(362)	Promotes myoblast differentiation and muscle regeneration by interacting with YY1 to evict YY1/PRC2 from the target loci	YY1
m1/2-sbsRNAs	(321)	Regulates myogenesis by base pairing with 3′-UTR of SINE-containing mRNAs and triggering SMD	SINE-containing mRNA 3′-UTRs
Malat1	(50)	Acts as a ceRNA for miR-133; represses myoblast differentiation and muscle regeneration by recruiting Suv39h1 to MyoD-binding loci	miR-133, miR-181a, Suv39h1
MDNCR	(164)	Promotes myoblast differentiation and inhibits cell proliferation by sponging miR-133a	miR-133
MLN	(6)	Interacts with SERCA to control muscle relaxation by regulating Ca2+ uptake into the SR	SERCA
MUNC (DRRRNA)	(227, 228)	Promotes muscle differentiation via regulating MyoD binding to DRR and promoter of myogenin to facilitate recruitment of the transcription machinery to the regions	MyoD
Myomixer/minion/myomerger	(8)	Is required for muscle formation during embryogenesis and interacts with Myomaker to confer fusogenic activity	Myomaker
Sirt1 AS	(318)	Promotes myoblast proliferation by stabilizing Sirt1 mRNA by competing with miR-34a	Sirt1 mRNA
SRA	(38)	Promotes muscle differentiation by assembly of coregulators p68/p72/MyoD	p68, p72, MyoD
Yam-1	(186)	Inhibits myoblast differentiation via activating the expression of miR-715, which targets Wnt7b

Ago2, Argonaute 2; Ash1L, absent, small or homeotic discs 1L; Baf60c, mammalian chromatin remodeling complex BRG1-associated factor 60C; Brg1, BRM/SWI2-related gene 1; ^CERNA, RNA transcript corresponding to core enhancer; ceRNAs, competing endogenous RNAs; DBE-T, D4Z4 binding element transcript; Dlk1, delta-like 1 homolog; DNMTs, DNA (cytosine-5)-methyltransferases; Dppa2, developmental pluripotency associated 2; ^DRRRNA, RNA transcript corresponding to distal regulatory regions; Dum, long non-coding RNA (lncRNA) Dppa2 upstream binding muscle; DWORF, dwarf open reading frame; FoxO1, forkhead box O1; FSHD, facioscapulohumeral muscular dystrophy; Gtl2, gene trap locus 2; H19, H19 imprinted maternally expressed transcript; IMP2, IGF2 mRNA-binding protein-2; let-7, lethal-7; linc-MD1, long intergenic non-protein-coding RNA (lincRNA) muscle differentiation 1; linc-RAM, lincRNA activator of myogenesis; linc-YY1, lincRNA YY1 transcription factor; lncMD, lncRNA muscle differentiation; lnc-mg, lncRNA myogenesis-associated; lncMyoD, lncRNA myogenic differentiation 1; lncRNA-Six1, lncRNA six homeobox 1; lnc-SEMT, lncRNA sheep enhanced muscularity transcript; Malat1, metastasis-associated lung adenocarcinoma transcript 1; MAML1, mastermind-like protein-1; MDNCR, lncRNA muscle differentiation-associated; MEF2C, myocyte enhancer factor 2C; Meg3, maternally expressed 3; miR, microRNA; MLN, myoregulin; MUNC, MyoD upstream noncoding; Pol II, RNA polymerase II; PRC2, polycomb repressive complex 2; sbsRNAs, Staufen1-binding site RNAs; SINE, short interspersed element; Sirt1 AS, sirtuin 1 antisense RNA; SMD, Staufen1-mediated mRNA decay; SR, sarcoplasmic reticulum; SRA, steroid receptor RNA activator; Suv39h1, histone-lysine N-methyltransferase SUV39H1; TDP-43, transactive response (TAR) DNA-binding protein-43; TrxG, trithorax group; UTR, untranslated region; Yam-1, YY1‐associated muscle.

circRNAs in skeletal muscle.

More recently, thousands of circRNAs were shown to be actively regulated in myoblasts (352), some of them controlling myoblast proliferation and differentiation (159). The predicted interactions between numerous circRNAs and miRNAs support the argument that circRNAs may act as microRNA sponges to regulate gene expression in skeletal muscle (352). Despite these findings, the expression profile of circRNAs in skeletal muscle and their biological function remain largely unknown in humans.

miRNAs in exercise.

Exercise has a powerful stimulus to activate satellite cells and propel protein synthesis in muscle, influencing gene expression and activating signal pathways. The change in gene expression is partly due to the change in miRNA expression. Furthermore, the type of exercise can have a direct impact on the nature of the miRNA response (74). For example, miR-1 is upregulated, whereas there is no change in the expression of miR-133a and miR-206, after resistance training (73); in contrast, expression levels of miR-9, miR-23a, miR-23b, and miR-31 are decreased, and the muscle-enriched miR-1, miR-133a, miR-133b, and miR-181a are increased (267) after submaximal endurance exercise.

miRNAs in normal aging muscles.

Decrease in mass, strength, and contraction rate of skeletal muscle is a general hallmark of human aging and results from the blunted rate of skeletal muscle protein synthesis (246). This is reflected by the decreased myogenic capacity of myoblasts with age, paralleling the disrupted expression of numerous miRNAs, lncRNAs, and circRNAs in aging muscles (46, 158, 293). Comparative analysis of miRNA expression profiles identified reduced abundance of miR-431, as well as miR-181a and its target gene Sirt1 (293), and miR-143, a regulator of IGF-binding protein-5 (Igfbp5), in aged myoblasts. Overall, it also seems that the deregulation of miRNAs contributes to the age-related changes in satellite cell function.

Muscle atrophy and shift in muscle fiber type distribution.

Muscle atrophy is a debilitating systemic response to denervation, long-term inactivity, excessive fasting, aging, and a variety of chronic diseases, including COPD (199), congestive heart failure (CHF; 72, 135), and PAH (17, 195). Muscle atrophy is most commonly associated with increased expression of atrophy-linked ubiquitin ligases, including muscle-specific RING finger 1 (MURF-1) and atrogin-1 (24). Not surprisingly, miRNAs have been shown to play a role in different models of muscle atrophy in vitro and in vivo, including miR-1, miR-133, miR-23a, miR-21, miR-27, miR-628, miR-431, and miR-206 (146, 291, 317) and, more recently, miR-29b (165; Table 4). In addition to myocyte cross-sectional area, which influences muscle function, physiologic, metabolic, and molecular parameters of skeletal muscle are also influenced by the heterogeneous myofiber composition, categorized into fast-twitch (mostly type IIB) and slow-twitch (type I) myofibers reflecting the expression of myosin heavy chain isotypes (277). Interestingly, miR-29b upregulation was also associated with a decrease in the proportion of type I fibers and an increase in type IIB fiber (165), a critical feature of skeletal muscle disuse that is also documented in CHF, COPD, and PAH (65, 195, 197). Consistently, miR-29b upregulation was also a feature of skeletal myopathy in experimental PAH, together with the deregulation of 17 other miRNAs (225). Recently, several studies have also focused on the analysis of epigenetic modifications in the muscles of patients with COPD. Expressions of miR-1, miR-206, and miR-27a were shown to be increased in the quadriceps of patients with COPD with muscle weakness and wasting (255), whereas miR-1 levels were lower in patients with preserved body composition (161). Intriguingly, recent data suggest that skeletal muscle maladaptation during heart failure could be mediated by myocardium-derived miRNAs released to the circulation (230). Nonetheless, the specific role of miRNAs in skeletal muscle atrophy and fiber type proportion observed in specific chronic diseases remains largely unknown.

Muscle ischemia, impaired muscular O₂ utilization, and defective angiogenesis.

Angiogenesis is a complex process that is regulated by the delicate balance of gene expression. The formation process of new angiogenic vessels involves the proliferation, migration, and differentiation of endothelial cells largely dictated by vascular endothelial growth factor (VEGF), which, in turn, activates VEGF receptor 2 to stimulate tip cell migration. Recently, the lncRNA Malat1 was shown to be significantly upregulated by oxygen-glucose deprivation and hypoxia in endothelial cells (351) and to regulate cell-autonomous angiogenesis in vivo (354). Interestingly, many chronic cardiorespiratory diseases, including COPD, CHF, and PAH, are also characterized by skeletal muscle capillary rarefaction. In PAH, this reduction in skeletal muscle capillary density resulted in impaired muscular O₂ utilization (69, 195, 198, 231, 252). Mechanistically, this capillary rarefaction in patients with PAH was shown to be caused by the downregulation of miR-126, whereas ectopic restoration of miR-126 in PAH animal models restored capillary density and endurance capacity (252). Similar miR-126-mediated capillary rarefaction was also observed within the right ventricle of patients with PAH and PAH animal models, contributing to right ventricular failure (253).

Insulin resistance, diabetes, and mitochondrial impairment.

Insulin receptor gene expression is downregulated in miR-135a-5p-transfected C₂C₁₂ cells, together with insulin receptor substrate 1 (IRS1; 114). Several other miRNAs have been described as repressors of GLUT4 expression or modulators of the machinery involved in GLUT4 translocation. Moreover, the expression of these miRNAs, including miR-29a-3p, miR-29c-3p, miR-93–5p, miR-222-3p, miR-223-3p, and miR-106b-5p, has been described as altered in humans or experimental models with insulin resistance, thus revealing their potential participation in the pathophysiology of diabetes (82). ncRNAs may thus be involved in the impaired mitochondrial content and function in skeletal muscle (197, 199, 266), as well as insulin resistance (254), observed in numerous chronic diseases, including myopathies, diabetes, COPD, and PAH.

miRNAs as potential biomarkers for skeletal muscle disorders.

Recently, miRNAs with high expression in muscle have been detected in plasma and serum in the presence of muscle disorders (150), suggesting that the expression levels of circulating miRNAs may be used as new biomarkers for muscular diseases. For example, miR-1, miR-133a, and miR-206 in serum have been shown to be useful and reliable biomarkers for muscular dystrophy (222). Similarly, miR-1 was inversely associated with fat-free mass, whereas levels of miR-499 correlated with strength and quadriceps type I fiber proportion in a cohort of patients with COPD (71). Hence, serum miRNA expression level has great potential as a biomarker for diagnosing, monitoring disease progression, or evaluating therapeutic efficacy of a series of muscle conditions (34).

CONCLUSIONS

As evident from a number of articles, ncRNAs are widely and abundantly expressed in various cell/tissue types and play an important role in the maintenance of cellular homeostasis and function. The capacity of ncRNAs to regulate the expression (transcription and/or translation) of protein-coding genes makes them reliable targets for pathognomonic mutations that directly and/or indirectly contribute to disease initiation and/or progression, a topic that was actively discussed in this review. There are other classes of ncRNAs not covered in this review that should be considered for further studies. Generally speaking, ribosomal RNA (rRNA) constitutes ~80% of the total RNA in a mammalian cell, whereas 15% is transfer RNA (tRNA; 183). The remaining ~5% consists of protein-coding genes and ncRNAs (excluding rRNAs and tRNAs) mentioned in this review. As rRNAs and tRNAs are essential constituents of the cellular translational machinery, it is not unexpected to find perturbations in their expression, structure, and/or function to be directly linked to various diseases (16, 80, 121, 156, 237, 247, 283, 288, 365). Thus, it is imperative to also consider these ncRNAs, which are often disregarded in the present strategies for RNA-sequencing (RNA-seq) experiments, in future investigations (Fig. 4). Compared with conventional RNA-seq strategies [both poly(A) and rRNA-depleted RNA-seq], the library preparation of small RNA-seq (sometimes called miRNA-seq) does not exclude rRNAs and tRNAs, as only size selection (below 50–100 nt) is used. Hence, in principle, sequencing reads from small RNA-seq data should only contain small RNAs [e.g., small Cajal body-specific RNAs (scaRNAs), small nucleolar RNAs (snoRNAs), and tRNA halves; 87, 108, 116], assuming the small RNA-seq data are analyzed properly. There are bioinformatics tools available to analyze such small RNAs, including DARIO (87) and Fragment Location Annotation Mapper (FlaiMapper; 116) for snoRNAs with a length of 60–300 nt, which consist of scaRNA, snoRNAs, H/ACA box (“SNORA”), and small nucleolar RNA, C/D box (“SNORD”), as well as fragments from mature tRNA, which include “tRNA halves” (~35 nt) and tRNA-derived RNA fragments (“tRFs”; ~20 nt). Another bioinformatics tool, piPipes (108), detects PIWI-interacting RNAs (“piRNAs”; ~26–31 nt long) from small RNA-seq data. These small RNAs are important factors in regulating translation and guiding chemical modifications of other RNAs (e.g., rRNAs and tRNAs) in the germ line (77, 84, 145, 206). In addition, the emerging field of epitranscriptomics (covalent modifications of RNAs; 70, 119, 133, 265, 309) will greatly benefit from the future findings in ncRNAs as RNA modifications are common in ncRNAs as in the case of rRNAs. Taken together, further research in ncRNAs will yield significant findings that could explain disease mechanisms and possibly manipulate such ncRNAs to alleviate and eventually cure such diseases.

Fig. 4.

RNA-sequencing (RNA-seq) methods. Small RNA-seq aims to identify RNAs whose lengths are less than ~50–100 nt. In contrast, there are two major methods of generating libraries for RNA-seq, which are based on poly(A) selection and ribosomal RNA (rRNA) depletion. Both methods are aimed at removing rRNAs. Poly(A) selection will result in the identification of protein-coding genes and long non-coding RNAs (lncRNAs) with poly(A) tails [~60% of total lncRNAs (54)], whereas rRNA depletion can identify the rest of the lncRNAs and circular RNAs (circRNAs), in addition to those identified in the former method. The presence of circRNAs is detected only with the latter method as circRNAs arise from exons and/or introns that are spliced out, which are devoid of poly(A) tails. miRNAs, microRNAs; piRNAs, PIWI-interacting RNAs; scaRNAs, small Cajal body-specific RNAs; snoRNAs, small nucleolar RNAs.

GRANTS

This study was supported in part by a Canadian Institutes of Health Research foundation grant, National Institutes of Health Grants R01-HL-071115 and 1RC1-HL-099462, a Tier 1 Canada Research Chair in Mitochondrial Dynamics, and the William J. Henderson Foundation (to S. L. Archer); Canadian Vascular Network Scholar Award 2016–2018 (to D. Wu); NIH Grant R01-HL-141081 (to J. B. Moore); Canadian Institutes of Health Research (to S. Provencher, S. Bonnet, O. Boucherat, and R. Paulin); Pulmonary Vascular Research Institute and Heart and Stroke Foundation of Canada grants and a Canada Research Chair in pulmonary vascular disease (to S. Bonnet); a Fonds de recherche du Québec-Santé (FRQS) senior grant (to S. Provencher) and an FRQS junior grant (to O. Boucherat); NIH Grants P01-HD-083132, R01-HL-118861, R01-HL-128209, and R21-NS-103017 (to L. Zhang); NIH Grant P30-GM-127607, the V. V. Cooke Foundation (Kentucky), and start-up funding from the Mansbach Foundation, the Gheens Foundation, and other generous supporters at the University of Louisville (to S. Uchida).

DISCLOSURES

No conflicts of interest, financial or otherwise, are declared by the authors.

AUTHOR CONTRIBUTIONS

S.B., O.B., R.P., D.W., C.C.T.H., S.L.A., R.S., J.B.M., S.P., L.Z., and S.U. prepared figures; S.B., O.B., R.P., D.W., C.C.T.H., S.L.A., R.S., J.B.M., S.P., L.Z., and S.U. drafted manuscript; S.B., O.B., R.P., D.W., C.C.T.H., S.L.A., R.S., J.B.M., S.P., L.Z., and S.U. edited and revised manuscript; S.B., O.B., R.P., D.W., C.C.T.H., S.L.A., R.S., J.B.M., S.P., L.Z., and S.U. approved final version of manuscript.