The cardiovascular system has been an especially rich source of miRNAs with roles in disease, perhaps reflecting the susceptibility of the heart and blood vessels to injury and the dependence of mammals on persistent cardiovascular function (Small and Olson, 2011
). miRNA profiling in mouse models of cardiovascular disease and in human biopsies has revealed signature patterns of miRNAs diagnostic for numerous cardiovascular disorders including heart failure, cardiomyopathy, myocardial infarction, atherosclerosis, ischemia, and angiogenesis. Gain- and loss-of-function studies in mice have also validated the importance of specific miRNAs in the pathophysiology of many of these disorders.
As in cancer, numerous miRNAs function as mediators of pathogenic stress-related signaling pathways in settings of cardiovascular disease (). For example, cardiac stress commonly results in fibrosis due to excessive extracellular matrix (ECM) production and collagen deposition, leading to stiffening of the ventricular chambers and cardiac arrhythmias (Hill and Olson, 2008
). miR-29 is down-regulated under conditions of cardiovascular disease associated with fibrosis and excessive ECM production (Boon et al., 2011
; Cushing et al., 2011
; van Rooij et al., 2008
), implicating this miRNA as a repressive mediator of fibrotic disease (). This miRNA targets a broad collection of mRNAs encoding multiple collagen isoforms and other ECM proteins, such that its downregulation during cardiovascular disease enhances fibrosis. TGF-β signaling in fibroblasts, a key driver of fibrosis, triggers miR-29 down-regulation (van Rooij et al., 2008
). Due to the ability of miR-29 to impede excessive ECM production, delivery of this miRNA may represent an effective therapeutic approach for fibrosis in the cardiovascular system and elsewhere. However, there are settings where miR-29 overexpression appears to be detrimental. Aneurysms, characterized by ballooning of the vasculature, arise from a loss of ECM components and therefore may be exacerbated by miR-29 activity. Accordingly, inhibition of miR-29 with antisense oligonucleotides stimulates ECM production and diminishes aortic dilatation in the setting of aortic aneurysm in mice (Boon et al., 2011
Figure 3 Examples of miRNAs that function in cardiac stress signaling pathways. (A) miR-29 and miR-15 family members act as mediators of stress signaling pathways that regulate fibrosis and cardiomyocyte proliferation and survival, respectively. (B) miR-208a and (more ...)
Members of the miR-15/16 family also serve as important stress signal mediators through their roles in pathways that regulate cardiomyocyte proliferation and survival in response to injury (). During neonatal development, the miR-15 family is upregulated in the heart, coinciding with irreversible withdrawal of cardiomyocytes from the cell cycle and loss of regenerative potential (Porrello et al., 2011
). This family of miRNAs is further upregulated following myocardial infarction (MI), causing irreversible loss of cardiomyocytes and cardiac dysfunction (Cimmino et al., 2005
; Hullinger et al., 2011
; Linsley et al., 2007
; van Rooij et al., 2008
). In keeping with a critical role for these miRNAs in ischemic cardiac pathology, inhibition of miR-15/16 family members with LNA-modified oligonucleotides confers protection against cardiomyocyte apoptosis following MI in rodents (Hullinger et al., 2011
miRNAs have also been documented to perform stress signal modulation functions in the setting of cardiovascular disease, as illustrated by miR-208a (). Numerous forms of cardiac stress, including myocardial infarction, hypertension, and chemotherapy, result in pathological remodeling of the heart and consequent loss of pump function, culminating in heart failure, arrhythmias, and death (Hill and Olson, 2008
). A hallmark of heart disease is a switch from the adult alpha-myosin heavy chain (α–MHC) isoform to expression of the fetal β-MHC gene, with concomitant diminution in cardiac function. miR-208a is encoded by an intron of the α-MHC gene and is expressed specifically in the heart along with its host gene (Callis et al., 2009
; van Rooij et al., 2007
). miR-208a knockout mice are protected from pathologic cardiac remodeling under conditions of chronic stress, indicating that miR-208a is required for the stress response which contributes to many cardiac diseases (). The actions of miR-208a appear to be mediated by a set of transcriptional repressor proteins that govern cardiac gene expression in response to stress. Thrap1/MED13, a component of the Mediator complex, is among the strongest targets of miR-208a (Callis et al., 2009
; van Rooij et al., 2007
). Upregulation of Thrap1 (and other targets), as occurs in the absence of miR-208a, is believed to impose a blockade to activation of downstream stress-responsive genes, including β-MHC. Systemic delivery of miR-208a inhibitors delays the onset of cardiac dysfunction in hypertensive rats, confirming the key role of this miRNA in pathological cardiac remodeling and suggesting the therapeutic potential of miR-208a inhibitors in heart disease (Montgomery et al., 2011
Another example of stress signal modulation by a miRNA is provided by miR-126, an endothelial cell-specific miRNA encoded by an intron of the EGFL7
gene (Fish et al., 2008
; Kuhnert et al., 2008
; Wang et al., 2008
), which encodes an endothelial growth factor (). miR-126 knockout mice have fragile, leaky vessels and display impaired angiogenesis in response to injury and angiogenic signaling (Kuhnert et al., 2008
; Wang et al., 2008
). Deletion of miR-126 impairs angiogenic signaling at least in part due to the consequent upregulation of the miR-126 targets Spred1 and PIK2R2 which function as negative regulators of the pro-angiogenic MAPK/ERK and PI3K/Akt signaling pathways, respectively.
As in cancer, negative feedback loops involving miRNAs play important roles in maintenance of cardiovascular function. A classic example involves the regulation of serum response factor (SRF), a key transcriptional regulator of cardiomyocyte growth and differentiation. SRF activates transcription of miR-133a in cardiomyocytes and miR-133a, in turn, negatively regulates SRF expression (Chen et al., 2006
; Liu et al., 2008
) (). The importance of this negative feedback loop is demonstrated by the pathologic consequences that result from deletion of the two miR-133a-encoding loci in mice (miR-133a-1 and miR-133a-2). Loss-of-function of these miRNAs results in lethal structural heart defects in approximately half of mutant animals (Liu et al., 2008
). Those that survive exhibit aberrant expression of smooth muscle genes in adult cardiomyocytes and ultimately succumb to dilated cardiomyopathy which can be partially attributed to abnormally high SRF activity.
In addition to negative feedback loops, a related mechanism of stress-signal dampening occurs when a stimulus that leads to a specific phenotypic response simultaneously induces an inhibitor of that phenotype. This type of regulation, referred to as incoherent feed-forward, is represented by the regulation of angiogenesis by miR-92a. Ischemia is a potent inducer of angiogenesis which serves to restore blood-flow to poorly perfused tissue sites. Yet ischemia has also been shown to cause the upregulation of miR-92a, which functions as a negative regulator of blood vessel growth. Systemic delivery of antisense oligonucleotides directed against miR-92a in a mouse model of hindlimb ischemia increases blood vessel growth and improves recovery from ischemic damage (Bonauer et al., 2009
). The anti-angiogenic actions of miR-92a have been ascribed to its repression of α5 integrin, which exerts pro-angiogenic activity, but undoubtedly other targets also contribute to the effects of miR-92a on angiogenesis. It is noteworthy that miR-92a is encoded by the miR-17-92 cluster, which encodes miR-17, -18a, -19a, -20a, -19b and -92a. miR-18 and -19 have been reported to promote blood vessel growth and tumor angiogenesis through their ability to downregulate the anti-angiogenic factors thrombospondin-1 (Tsp-1) and connective tissue growth factor (CTGF) (Dews et al., 2010
). Therefore, the ultimate consequence of expression of this miRNA cluster on angiogenesis is likely dependent on which of its targets plays a dominant role in a given tissue or disease context.
Numerous cardiovascular miRNAs establish positive feedback loops that influence phenotypic switching during disease (). Among this class of miRNAs is miR-21, which is upregulated by MAPK/ERK signaling in response to cardiac stress (Patrick et al., 2010
; Thum et al., 2008
). miR-21 targets Sprouty2 (Spry2), a negative regulator of the MAPK/ERK cascade, thereby further enhancing signaling through this pathway. Systemic delivery of cholesterol-modified miR-21 antagomirs has been reported to diminish MAPK signaling in cardiac fibroblasts and thereby completely prevent and reverse cardiac dysfunction, fibrosis, and hypertrophy in response to pressure overload (Thum et al., 2008
). Surprisingly, genetic deletion of miR-21 in mice does not diminish heart disease in response to pressure overload or multiple other stresses (Patrick et al., 2010
). Likewise, inhibition of miR-21 with LNA-modified oligonucleotides does not prevent pathological cardiac remodeling in response to stress. These disparate findings may indicate that the therapeutic activity of cholesterol-modified miR-21 inhibitors is due to selective uptake into a key cell type or off-target effects of these molecules. Alternatively, these observations raise the possibility that compensatory mechanisms may overcome the absence of miR-21 in the setting of chronic genetic deletion.
Cardiac stress also induces the expression of the miR-23a/27a/24-2 cluster via activation of the NFAT transcription factor (Lin et al., 2009
), which serves as an effector of the calcium-sensitive phosphatase calcineurin (Molkentin et al., 1998
). miR-23a, in turn, has been shown to repress expression of the muscle-specific ubiquitin ligase MuRF1, establishing a positive feedback loop that perpetuates stress-dependent cardiac hypertrophy (). Knockdown of miR-23a using miRNA inhibitors confers resistance to cardiac stress and diminishes hypertrophy through dampening of this positive feedback loop. Calcineurin/NFAT signaling also induces expression of miR-199a (da Costa Martins et al., 2010
), which negatively regulates the inhibitory NFAT kinase Dyrk1a, thereby further amplifying pathogenic signaling in the heart (). Remarkably, inhibition of miR-199a has been reported not only to prevent cardiac hypertrophy and fibrosis in response to stress, but also to reverse these pathological processes through elevation of Dyrk1a activity (da Costa Martins et al., 2010
Expression of the miR-23a/27a/24-2 cluster is also induced by VEGF-MAPK signaling, establishing another positive feedback loop involving these miRNAs (Cheng et al., 2009
; Zhou et al., 2011
). These miRNAs enhance vessel growth by promoting angiogenic signaling through targeting of the MAPK inhibitors Spry2 and Sema6A, which restrain angiogenesis (). These miRNAs have been implicated in pathological angiogenesis of the retina, such that their inhibition represents a potential therapeutic application for this disorder.
Finally, miRNAs have been described that target both positive and negative regulators of a cardiovascular stress response and thereby buffer pathway activity. This is exemplified by the miR-143/145 cluster which regulates phenotypic switching of smooth muscle cells (Boettger et al., 2009
; Cheng et al., 2009
; Cordes et al., 2009
; Xin et al., 2009
). Vascular injury and atherosclerosis are accompanied by de-differentiation and excessive proliferation of vascular smooth muscle cells, causing occlusion of the vascular lumen. Expression of miR-143/145 is elevated upon differentiation of smooth muscle cells and is diminished in injured and atherosclerotic vessels. Restoration of miR-145 expression in injured arteries through adenoviral delivery impedes smooth muscle proliferation and neointimal growth. Paradoxically, mice with genetic deletion of miR-143/145 are similarly resistant to pathological vascular remodeling in response to injury (Xin et al., 2009
). Thus, both loss and gain of these miRNAs in vivo
evoke similar phenotypes. The pro-differentiation and anti-proliferative actions of miR-143/145 have been attributed to negative regulation of genes involved in actin dynamics, cytoskeletal remodeling and actin gene regulation. Among these targets is the transcriptional repressor KLF4, which inhibits smooth muscle cell differentiation and promotes proliferation (). Opposing these actions is the myocardin coactivator MRTF-B, which stimulates smooth muscle differentiation, and is also targeted by miR-145. Although miR-143 and 145 have been reported to be essential for smooth muscle cell differentiation in vitro
(Cordes et al., 2009
), knockout mice lacking either or both of these miRNAs are born without obvious vascular abnormalities (Xin et al., 2009
). However, contractility of the blood vessels of these mice is compromised, suggesting a subtle shift toward a less differentiated phenotype (Boettger et al., 2009
). Taken together, the available data support a model whereby miR-143/145 do not play a major role in the initial differentiation of smooth muscle cells. Rather, through the balanced regulation of positive and negative regulators of smooth muscle differentiation and proliferation, this miRNA cluster maintains the phenotypic plasticity of this cell type, facilitating an appropriate (and sometimes pathologic) response to injury.