Review

Introduction

At the cellular level, heart tissue constitutes cardiomyocytes, cardiac fibroblasts, vascular and neuronal cells, as well as inflammatory cells under certain pathological conditions. In the healthy heart, cardiac fibroblasts are the most prevalent cell type, accounting for up to 70% of cells, depending on the species in question [1, 2]. Although cardiac fibroblasts have been much less well studied than cardiomyocytes, it is becoming increasingly apparent that the fibroblasts (and their differentiated phenotype, myofibroblasts) are integral to the development, normal function and repair of the heart, as well as contributing to adverse myocardial remodelling, fibrosis and heart failure progression [3, 4]. Through physical and biochemical communication with myocytes and other cell types in the heart and the cardiac extracellular matrix (ECM), fibroblasts are well placed to sense and respond to stress or injury to the myocardium.

Fibroblasts are a heterogeneous population of cells, reflecting both their multiple developmental origins and their exposure to differential physical and chemical microenvironments. Fibroblasts derived from different anatomical sites have been proposed to effectively represent distinct differentiated cell types as they exhibit unique transcriptional signatures that probably reflect phenotypic differences [5]. Such diversity has made precise characterisation of fibroblasts challenging, and there remains no truly unique single marker that unequivocally identifies a cell as a fibroblast [6].

Although fibroblasts have the capacity to proliferate, migrate and regulate ECM turnover to maintain cardiac homeostasis, they are also able to undergo differentiation into a more contractile and synthetic myofibroblast phenotype to aid with cardiac repair following myocardial infarction (MI) [7–9]. Myofibroblasts are not normally found in the healthy myocardium, but are the most prevalent cell type in the infarct scar and are the main effectors of fibrogenesis [10]. Myofibroblasts are characterised by increased expression of particular contractile proteins (for example, α-smooth muscle actin, SMemb, vimentin), focal adhesion proteins (for example, paxillin, tensin, αVβ3 integrin), cell surface receptors (for example, transforming growth factor beta (TGF-β) type II receptor, angiotensin AT1 receptor, Frizzled-2), structural ECM proteins (collagen I, collagen III, fibronectin extra domain A splice variant (FN-ED-A)) and matricellular proteins (for example, periostin, osteopontin, tenascin C) [7–9]. Cardiac myofibroblasts are also highly proliferative, and those isolated from infarcted myocardium exhibit a higher rate of proliferation than cardiac fibroblasts from remote areas [11, 12]. Although myofibroblasts are able to actively migrate to the infarcted region of the heart [13], a process regulated by Wnt/Frizzled signalling [14, 15], they also appear to become less migratory as expression levels of contractile proteins increase [11, 16]. Together these phenotypic changes confer increased tensile and ECM-secretory characteristics on the cells, enabling them to effectively facilitate the wound healing process.

Beneficial and detrimental roles of myofibroblasts

Appreciating the dual roles of cardiac myofibroblasts in the myocardial remodelling process is important, as they can be perceived to be both beneficial and detrimental depending on their prevalence and their temporal and spatial location. The infarct scar is not a simple acellular structure comprising structural ECM molecules; on the contrary, it contains myofibroblasts that maintain a viable, dynamic scar important for maintaining myocardial integrity against a background of continuous mechanical forces associated with the pum** of the heart [17]. Myofibroblasts are essential for rapid and robust (that is, strong and flexible) scar formation following MI. Interference with myofibroblast recruitment can result in infarct expansion, ventricular wall thinning, dilatation, systolic dysfunction and propensity to rupture [7] (Figure 1). Conversely, myofibroblast persistence can contribute to fibrosis and adverse myocardial remodelling, particularly if the myofibroblasts remain active in otherwise healthy areas of the heart away from the original site of injury (reactive fibrosis) [7]. Areas of increased ECM protein deposition can disturb the electrical conductance of the myocardium, thus increasing the likelihood of arrhythmias [18]. Moreover, direct coupling of cardiomyocytes to myofibroblasts, as opposed to fibroblasts, may also promote arrhythmias [19, 20]. Fibrosis in the remote myocardium inevitably leads to increased myocardial stiffness, resulting in systolic and diastolic dysfunction, neurohormonal activation and, ultimately, heart failure [21, 22] (Figure 1).

Figure 1
figure 1

Summary of the influence of myofibroblast density on post-myocardial infarction remodelling. Low myofibroblast density in the infarct area results in a poorly structured, expansive and vulnerable scar that is prone to rupture or leads to systolic dysfunction and subsequent adverse myocardial remodelling. Although high myofibroblast density is important for a robust, contractile scar, excessive myofibroblast numbers (particularly in the remote myocardium away from the original infarct) drives fibrosis and myocardial stiffness, resulting in contractile dysfunction, arrhythmia and heart failure progression.

Full size image

Origin of myofibroblasts

The differential origin of myofibroblasts in the remodelling heart has become a hot topic in recent years [6, 23]. Although once assumed to be solely derived from differentiation of resident fibroblasts, it is now apparent that cardiac myofibroblasts can also be derived from a multitude of alternative cellular precursors. These precursors include epithelial cells (through a process termed epithelial–mesenchymal transition), endothelial cells (through endothelial–mesenchymal transition; EndMT), mesenchymal stem cells, bone marrow-derived circulating progenitor cells (fibrocytes), smooth muscle cells and pericytes [6, 23]. The recruitment of myofibroblasts from such diverse origins underlines their importance in the cardiac repair process, and probably represents optimised responses to different types of stress or injury. However, reports on the precise proportions of cells derived from different sources in different experimental models have varied considerably, so consensus has yet to be reached on the relative importance of myofibroblasts derived from resident cardiac fibroblasts versus extra-cardiac sources [6]. Another important aspect is whether these data are recapitulated in the human scenario. Nevertheless, a picture is now emerging that the source of myofibroblasts in the remodelling heart may depend heavily upon the nature of the initiating stimulus or injury. For example, whereas resident mesenchymal stem cells have been identified as important contributors to the myofibroblast population that drives post-MI scar formation, fibrocyte-derived myofibroblasts may be more important for interstitial fibrosis in the absence of MI [24]. Such knowledge opens up the exciting prospect that selective targeting of distinct myofibroblast populations could be used to protect essential repair mechanisms following MI, whilst reducing remote fibrosis and subsequent adverse myocardial remodelling.

Factors stimulating myofibroblast differentiation

Phenotypic conversion of resident cardiac fibroblasts to myofibroblasts requires integration of both mechanical and biochemical stimuli. Fibroblasts are mechanosensitive and are therefore able to detect the loss of integrity of the ECM that occurs following MI. In response to increased mechanical stress and platelet-derived growth factor, fibroblasts adopt a partially differentiated phenotype known as the proto-myofibroblast [8]. Conversion of the proto-myofibroblast to the fully differentiated myofibroblast occurs in response to additional biochemical signals, particularly increased levels of active TGF-β and FN-ED-A [8], the levels of which are elevated in the damaged region of the heart post MI [25, 26]. Such a phenotypic conversion is also promoted when cardiac fibroblasts are grown in vitro on rigid plastic surfaces; hence studies on cultured cardiac fibroblasts are generally indicative of myofibroblast behaviour [16, 27]. TGF-β is normally present in the interstitium in a latent form, which can be rapidly activated by protease-mediated cleavage of the latency-associated peptide [28]. However, it has also been demonstrated that TGF-β activation can be stimulated directly by mechanical strain without the need for protease activity [29], and this mechanosensitive mechanism probably plays an important role in early myofibroblast conversion.

A number of additional stimuli that promote differentiation to the myofibroblast phenotype have been reported, including specific cytokines, growth factors and ECM molecules; several of which elicit their effects through up regulation of TGF-β activity and/or signalling [30]. There is also emerging evidence for an important role for the transient receptor potential family of ion channels in regulating cardiac myofibroblast differentiation. For example, the TRPM7 channel [89].

Molecular tools for manipulating miR levels (through inhibition or mimicry) have been an area of rapid development and ongoing refinement [88]. As discussed above, several promising miR targets have been identified that appear to regulate myofibroblast differentiation and/or function (Figure 2). Preclinical studies manipulating miR-21 and miR-29 have shown beneficial effects on post-MI cardiac remodelling in rodents. Specifically, a miR-29 mimetic has proven successful in a murine model of cardiac fibrosis [56] and miR-21 inhibition increased survival after MI [55].

Progressive expansion of our knowledge concerning dysregulation of miRs in cardiac (myo)fibroblast phenotype and function will undoubtedly lead to strategies that optimise targeted delivery of miR therapeutics. The ability to deliver therapies directly to selected cell types is indeed a realistic option for future medicine.

Conclusions

Cardiac myofibroblasts represent a unique, yet developmentally diverse, population of cells that play key roles in post-MI infarct healing, but also in adverse cardiac remodelling, fibrosis and progression to heart failure. Improved understanding of not only the origins of myofibroblasts in the post-MI heart, but also the capacity to assign specific roles and regulatory mechanisms to them, creates optimism for the future that this multifunctional cell type can be manipulated therapeutically to optimise infarct scar formation, whilst ameliorating reactive fibrosis.