Cardiac fibrosis is a fundamental constituent of most cardiac pathologies and represents the upshot of nearly all types of cardiac injury. Generally, fibrosis is a scarring process, characterized by accumulation of fibroblasts and deposition of increasing amounts of extracellular matrix (ECM) proteins in the myocardium. Therapeutic approaches that control fibroblast activity and evade maladaptive processes could represent a potential strategy to attenuate progression towards heart failure. Currently, cell therapy is actively perceived as an alternative to traditional pharmacological management of myocardial infarction (MI). The majority of the studies applying stem cell therapy following MI have demonstrated a decline in fibrosis. However, it was not clearly recognized whether the decline in cardiac fibrosis was due to replacement of dead cardiomyocytes or because of the direct effects of paracrine factors released from the transplanted stem cells on the ECM. Therefore, the main focus of this review is to discuss the impact of different types of stem cells on cardiac fibrosis and associated cardiac remodelling in a variety of experimental models of heart failure, particularly MI.
Cardiac fibrosis is a fundamental constituent of most cardiac pathologies and represents the upshot of nearly all types of cardiac injury . Generally, fibrosis is a scarring process, characterized by accumulation of fibroblasts and deposition of increasing amounts of extracellular matrix (ECM) proteins in the myocardium . Fibrosis appears in two distinct forms: (A) replacement fibrosis, which occurs in response to injurious stimuli causing cardiomyocyte death, such as myocardial infarction (MI), which activates a reparative response with subsequent replacement of dead cardiomyocytes and the formation of a collagen-based scar; and (B) reactive interstitial or perivascular fibrosis, which develops as a result of insults that do not lead to marked cardiomyocyte loss, such as pressure or volume overload, hypertrophic cardiomyopathy, dilated cardiomyopathy, brief repetitive ischaemic events, and diabetes or obesity-induced cardiomyopathy. This initial reactive fibrosis arises as an adaptive response to keep the pressure-generating ability of the heart; however, it may finally evolve into a state of replacement fibrosis [1, 3, 4].
Originally, fibrosis is caused by the activation of inflammatory and reparative pathways as a result of deleterious stimuli . Reported data propose that inflammatory and fibrogenic pathways can be evoked by short or sub-lethal stimuli with insignificant cardiomyocyte loss. However, the activation of these pathways following persistent insults such as MI is well recognized . Fibrogenesis is an important process in the infarcted myocardium repair mechanism, yet it contributes to the pathogenesis of the detrimental cardiac remodelling in response to increased myocardial wall stretching and hypoxia, which are physiological factors characterizing MI. The reparative response following MI encompasses three overlapping phases: the inflammatory phase; the proliferative (granulation) phase; and the maturation phase [6-8]. During the initial inflammatory phase, activated macrophages and inflammatory cells produce a variety of cytokines and growth factors (fibrogenic mediators) such as transforming growth factor (TGF)-β, platelet-derived growth factor (PDGF), fibroblast growth factor (FGF), tumour necrosis factor (TNF)-α, connective tissue growth factor (CTGF), and monocyte chemoattractant protein (MCP)-1 to stimulate the proliferation and phenotypic differentiation of fibroblasts into myofibroblasts [1, 6, 9]. In fibrotic hearts, the myofibroblasts are generated through proliferation and activation of resident cardiac fibroblasts, from endothelial cells via a cellular transition process known as the endothelial–mesenchymal transition, or through recruitment of bone marrow (BM)-derived progenitor cells . In the granulation phase, myofibroblasts infiltrating the infarct area contribute to the injured ventricle repair via sequential matrix degradation by ECM-degrading enzyme metalloproteinases (MMPs) and synthesis of ECM proteins. Extinction of the fibrotic reaction to the injury seems to be coupled with myofibroblast apoptosis leaving a mature scar consisting of cross-linked collagen and other matrix components. Nevertheless, in the case of MI, myofibroblasts keep their profibrotic assets and positively contribute to the adverse cardiac remodelling and progressive function loss [10-13]. Post-MI remodelling includes both the infarcted and the non-infarcted areas of the heart and causes enhanced interstitial fibrosis in the spared non-infarcted area .
Cardiac fibrosis has significant impacts on cardiac structure and function. Fibrosis may cause slippage of cardiomyocytes, resulting in ventricular wall thinning and consequent ventricular dilation . Additionally, enhanced ECM synthesis and reduced degradation lead to enlarged mechanical stiffness and cardiac dysfunction . Furthermore, increased deposition of ECM among cardiomyocyte layers may interrupt their electrical coupling, resulting in weakened cardiac contraction and an increased risk of arrhythmia . Moreover, diminished oxygen and nutrient flow as a result of inflammation and fibrosis in the perivascular areas may aggravate the adverse remodelling reaction .
Despite the development of a variety of treatment options, heart failure management has failed to inhibit myocardial scar formation and replace the lost cardiomyocyte mass with new functional contractile cells. This shortage is complicated by the limited ability of the heart for self-regeneration. Accordingly, novel management approaches have been introduced into the field of cardiovascular research, leading to the evolution of gene- and cell-based therapies . Interestingly, the majority of the studies applying stem cell therapy following MI have demonstrated a decline in fibrosis. However, it was not clearly recognized whether the decline in cardiac fibrosis was due to the replacement of dead cardiomyocytes or because of the direct effects of paracrine factors released from the transplanted stem cells on the ECM. Therefore, the goal of this review is to discuss the impact of different types of stem cells on cardiac fibrosis and associated cardiac remodelling in a variety of experimental animal models of heart failure, particularly MI.
Stem cell therapy and cardiac fibrosis
Even with the progression in pharmacological and surgical strategies, heart failure following MI remains the main cause of death in developed countries. As mentioned above, MI leads to a marked cardiomyocyte loss, which is replaced by fibrotic tissue that develops a permanent scar with a clear deterioration in heart function. In an attempt to replace the lost cardiomyocytes following ischaemia, to decrease fibrous tissue formation, and to improve cardiac performance, cellular therapy has been pursued as an alternative to classic pharmacological management . Examples of the effects of different stem-cell transplantations on cardiac fibrosis are shown in Table 1.
Table 1. Examples of the effects of stem cell therapy on cardiac fibrosis
Transplantation of ESCs has been shown to inhibit cardiomyocyte apoptosis and myocardial fibrosis, with subsequent improvements in adverse cardiac remodelling and cardiac function following MI  and in doxorubicin-induced cardiomyopathy . These beneficial effects have been attributed to decreased collagen deposition, enhanced myocardial regeneration, and possible paracrine factors released from ESCs such as the anti-apoptotic proteins cystatin C, osteopontin, and clusterin, and the anti-fibrotic tissue inhibitor metalloproteinase (TIMP)-1 . ESCs were also able to stimulate the endogenous cardiac regeneration via activating the differentiation of resident cardiac stem cells (CSCs) into cardiac myocytes . Paracrine factors such as hepatocyte growth factor (HGF) and insulin growth factor (IGF)-1 have been suggested to play a significant role in this activation process . Furthermore, HGF and IGF-1 possess cardioprotective effects through anti-apoptotic, mitogenic or anti-fibrotic activities. In particular, the anti-fibrotic properties of HGF via inhibiting TGF-β expression are beneficial in heart failure [22, 23].
Bone marrow-derived stem cells (BMSCs)
Most recent studies have shown that transplanted BMSCs following MI were differentiated into myocytes and endothelial cells in the ischaemic heart, and ultimately reduced both infarct size and fibrosis . Also, conditioned medium from BM-derived culture modified cells inhibited fibroblast collagen production and TGF-β signalling in vitro and markedly diminished cardiac fibrosis in a model of chronic cardio-renal disease. This in vivo effect has been suggested to be mediated by the secretion of diffusible anti-fibrotic factor(s) that inhibited the activity of the TGF-β receptor complex . Interestingly, a recent study showed that post-MI administration of BM-derived mononuclear cells decreased fibrosis and apoptosis and improved cardiac function . BM-derived mononuclear cells have been shown to regulate cardiac miRNAs via reducing the expression of the profibrotic miR-21 and the pro-apoptotic miR-34a. This study further demonstrated that BM-derived mononuclear cells secrete IGF-1, which inhibits miR-34a processing and blocks apoptosis of cardiomyocytes both in vitro and in vivo .
Mesenchymal stem cells (MSCs) in particular have been extensively investigated as a potential therapeutic approach for cardioprotection, due to their distinctive characteristics. A growing body of evidence supports the concept that MSCs decrease myocardial fibrosis and thereby attenuate cardiac remodelling . Xu et al showed that MSCs selectively down-regulate the ECM expression in the infarcted myocardium. In MI rats, MSCs significantly attenuated mRNA expression of collagen types I and III, TIMP-1, and TGF-β, a powerful initiator of ECM component production in both infarcted and non-infarcted regions . Likewise, in a rat model of dilated cardiomyopathy and in a global heart failure model induced by isoproterenol, MSC transplantation significantly attenuated myocardial fibrosis, as evidenced by decreased collagen volume fraction, decreased expression of collagen types I and III ,and lower levels of MMP-2 and MMP-9. Additional beneficial effects have been proposed to be mediated by paracrine signalling from the MSCs through the anti-fibrotic HGF [22, 27]. Paracrine anti-fibrotic effects of MSCs via direct action on cardiac fibroblast were also evident. MSC-conditioned medium attenuated cardiac fibroblast proliferation and viability by up-regulating anti-proliferation-related genes such as elastin, myocardin, DNA-damage inducible transcript 3, and adrenomedullin [9, 28, 29]. This was coupled with decreased type I and III collagen expression and suppressed type III collagen promoter activity.
Another important role of MSCs in attenuating cardiac fibrosis is the anti-inflammatory property. The possibility that MSCs have the ability to attenuate inflammatory responses has been raised after Di Nicola et al reported that MSCs strongly suppress T lymphocytes . Interestingly, co-culturing of T lymphocytes from post-MI mice with cardiac fibroblasts has been shown to increase pro-collagen expression, indicating that the in vivo repression of T lymphocyte accumulation and/or function may also prevent fibrosis [31, 32]. Thereafter, the anti-inflammatory effects of MSCs have been confirmed in several reports. For example, post-MI transplantation of MSCs was found to decrease protein production and gene expression of inflammatory cytokines such as TNF-α, IL-1β, and IL6 – an effect that has been correlated with decreased type I and III collagen deposition, as well as gene and protein expression of MMP-1 and TIMP-1 . Another study reported that MSC administration attenuated the increase in CD68+ inflammatory cells and myocardial MCP-1 expression in acute myocarditis . Similarly, in a mouse model of global heart failure induced by isoproterenol, an NF-κB-mediated signalling pathway has been proposed to be responsible for the inhibitory effects of MSCs on the expression of MMP-9 and TIMP-1 and collagen deposition .
Additionally, in cardiomyopathic hamsters, extracardiac administration of MSCs mediated cardiac repair with decreased fibrosis along with suppressed expression of collagens, MMPs, and TIMPs, most probably through engagement of the skeletal muscle JAK–STAT3 axis. The trophic cascade started by skeletal muscle JAK–STAT3 signalling enhanced growth factor levels in multiple tissues, resulting in increased circulating growth factors such as HGF and VEGF. The synergistic actions of these trophic factors further trigger the myocardial repair mechanisms coordinated by the Akt, ERK, and JAK–STAT3 signalling pathways [36, 37].
Vascular smooth muscle cells (VSMCs), human umbilical cord blood-derived cells (HUCBCs), and skeletal myoblast cells (SMCs)
It has also been reported that VSMCs exert their protective effects on the progression of cardiomyopathy in a paracrine manner. Cell transplantation in cardiomyopathic hamsters was demonstrated to enhance the ratio of TIMP-3 to MMP-9 and the ratio of TIMP-2 to MMP-2, indicating a reduced capacity for matrix degradation and preserved myocardial matrix homeostasis . A similar paracrine effect has been described for the human umbilical cord blood mononuclear cells. Administration of these cells to cardiomyopathic hamsters decreased fibrosis and increased cardiac function, most probably due to released growth factors and anti-inflammatory cytokines . Equally, a paracrine hypothesis has been illustrated for the beneficial effects of SMCs in infarcted myocardium through releasing angiogenic growth factors and matrix modulators such as MMPs and TIMPs .
In line with these studies, our group has previously demonstrated that transplantation of MSCs , HUCBCs , and SMCs  in ischaemic myocardium attenuated cardiac fibrosis and decreased infarct size. Our data demonstrated an increase in myocardial oxygenation at the site of implantation in the infarct heart, which correlated positively with cell engraftment, increased VEGF expression, promotion of neovascularization, and improved cardiac function . High levels of VEGF were found to provide cardioprotective effects and to promote functional collateral vessels, which help in the salvaging of ischaemic myocardium and reduce the infarct area . In addition, the reduction in myocardial fibrosis may also serve as an important indicator for improved heart function, since late reperfusion of infarcted vascular beds attenuates left ventricular remodelling including infarct expansion .
Cardiac progenitor cells (CPCs)
CPCs are striking as they regularly reside in the niche of the heart and apparently are responsible for refilling the pool of cardiac myocytes and coronary vessels under normal conditions. Administration of CPCs into ischaemic rat hearts has been shown to decrease infarct size and fibrosis, not only via differentiation into cardiomyocytes and vascular cells, but also via inducing the proliferation of resident CPCs in the infarct zone, possibly through a paracrine mechanism .
Induced pluripotent stem cells (iPSCs)
Recently, iPSCs have been generated from cardiac ventricular specific cell types instead of the commonly used mouse and human fibroblasts. These cells were effective in reducing fibrosis and the infarct zone following MI , and in diabetes-induced cardiomyopathy , probably through autocrine or paracrine mechanisms that remain to be determined. More recently, it has been reported that transcription factors such as GATA4, HAND2, MEF2C, and TBX5 (GHMT) can reprogram cardiac fibroblasts into functional cardiac-like myocytes in vitro and in vivo. Furthermore, exogenous GHMT expression in non-cardiomyocytes of the heart post-MI also decreases fibrosis and improves cardiac function. Regardless of the underlying mechanism, this strategy presents a prospective means of recovering cardiac function in vivo, evading the obstacles related to cellular transplantation .
Stem cell modifications and cardiac fibrosis
Myocardial fibrosis can be modulated not only via the paracrine factors released from stem cells, but also by modifying these cells through overexpression of cytokines or growth factors, pretreatment with pharmacological or modulating factors, or anoxic preconditioning [19, 50]. For instance, ESCs overexpressing TIMP-1 significantly reduced cardiac fibrosis compared with unmodified ESCs after MI, which may be attributable to inhibition of MMP-9 through TIMP-1 overexpression . Likewise, anoxic preconditioning enhanced the survival of BMSCs via up-regulating Akt and eNOS through paracrine mediators and it increased their ability to attenuate cardiac fibrosis after MI . Modified BMSC therapy might act in a supportive paracrine manner that promotes the mobilization and growth of CSCs, resulting in infarct sizereduction .
Additionally, MSCs overexpressing Akt led to a marked decrease in cardiac fibrosis, compared with unprimed cells [53, 54]. The therapeutic benefits of Akt-MSCs, at least in the acute phase of MI, seem to be due to diffusible factors released by the cells such as VEGF, fibroblast growth factor (FGF)-2, HGF, IGF-I, and thymosin B4, which act in a paracrine fashion to limit infarct size and improve ventricular function [53, 54]. Also, compared with unmodified cells, MSCs transfected with VEGF exhibited significantly higher levels of myocardial angiogenesis and cardiomyocyte regeneration, and prevented progressive scar formation and heart dysfunction in a rodent model of myocardial ischaemia . Consistently, our group has reported similar results in human cord blood-derived haematopoietic cells overexpressing VEGF and PDGF . Additionally, in comparison with unmodified cells, post-MI transplantation of MSCs transfected with stromal cell-derived factor (SDF)-1α significantly reduced collagen I and III expression and MMP-2 and −9. SDF-1-modified MSCs improved the tolerance of engrafted MSCs to hypoxic injury in vitro and increased their viability in infarcted hearts, thus helping to attenuate cardiac fibrosis and preserve the contractile function, possibly through enhancing paracrine signalling from MSCs via anti-fibrotic factors such as HGF . Furthermore, in vivo transplantation of MSCs overexpressing either the anti-apoptotic Bcl-2  or tissue transglutaminase  was shown to increase the survival rate of the cells following MI, probably via enhanced anti-apoptotic or adhesion properties of the modified MSCs, respectively. This was associated with restoration of cardiac function and decreased infarct size and degree of fibrosis compared with unmodified cells. Increased VEGF expression and neovascularization were also suggested to play a role in the salvaging of ischaemic myocardium and decreasing the infarct area .
Pretreatment of MSCs with a combination of growth factors (IGF-1, FGF-2, and bone morphogenetic protein-2) enabled these cells to survive in the infarcted myocardium in vivo more than unmodified MSCs. Accordingly, surviving MSCs showed higher efficiency in promoting cardiac repair and reducing infarct size. These effects have been attributed to direct differentiation and paracrine and cytoprotective action mediated by direct cell-to-cell contact through the formation of gap junctions, which are proposed to propagate cell survival and death signalling . In yet another study, pretreatment of MSCs with melatonin caused an increase in their anti-fibrotic activity. It was also shown that melatonin stimulates cytokine secretion by MSCs, and this could lead to an increase in the production and release of paracrine factors by MSCs, thereby regulating the anti-fibrotic properties in cardiac fibroblasts. However, this was not the case; in fact, the increase in the anti-fibrotic activity observed after in vivo injection of melatonin-treated MSCs was related to the improvement of MSC survival . We have recently demonstrated that pharmacological treatment of MSCs with either trimetazidine (Vastrel®)  or hyperbaric oxygen [60, 61] results in significantly higher efficiencies in attenuating cardiac fibrosis and improving cardiac function than the untreated cells. These improvements in the infarcted hearts were linked to up-regulation of survival signaling such as pAkt and Bcl-2 , increased graft survival, VEGF, NOS3 activation, and angiogenesis [60, 61], which may contribute to the enhanced anti-fibrotic effects of the modified cells. Moreover, preconditioning of adipose-derived stem cells with a phosphodiesterase inhibitor, sildenafil (Viagra®), increased their viability and enhanced the release of pro-angiogenic/pro-survival growth factors such as VEGF, FGF, IGF and angiopoietin-1 both in vitro and in vivo . The in vivo effect was also associated with increased vascular density and decreased cardiac fibrosis. Overall, the protective effect of modified cells was suggested to be due to increased duration and release of paracrine factors in the ischaemic myocardium . Figure 1 summarizes the effects of stem cell modifications on cardiac fibrosis following MI.
Conclusions and future perspectives
Stem cell therapy appears to be a promising strategy to attenuate cardiac fibrosis and to stop progression towards heart failure (Figure 2). In addition to cell differentiation and increased myocardial angiogenesis as reported in several studies, attenuation of fibrosis seems to be due to the direct effects of paracrine factors released from different subpopulations of stem cells on the ECM. However, the exact mechanism(s) has not yet been determined and needs further investigation. Therefore, studies focused particularly on mechanisms by which the stem cell treatment attenuates cardiac fibrosis may represent a hot area of research in the years to come. Recognizing these mechanisms and factors will help in developing effective therapies to understand the role of non-pharmacological treatment strategies in attenuating cardiac fibrosis and decrease the mortality rates in millions of people who are suffering from heart failure.
We would like to acknowledge funding from AHASDG (0930181N) and CCTS (UL1RR025755).
Author contribution statement
All authors participated fully in the writing of this review.