Article Text

other Versions

Epicardium-derived cells: a new source of regenerative capacity
  1. Joaquim M Vieira,
  2. Paul R Riley
  1. Molecular Medicine Unit, University College London Institute of Child Health, London, UK
  1. Correspondence to Paul R. Riley, Molecular Medicine Unit, University College London Institute of Child Health, 30 Guilford Street, London, WC1N 1EH, United Kingdom; p.riley{at}


Cardiovascular regenerative medicine aims to counter muscle loss post ischaemic disease with the identification of new cellular sources for cardiomyocyte replacement. A number of embryonic and adult cell models have been explored preclinically and in patient trials, but modest outcome, coupled with issues with impaired graft survival and limited/immature (trans-) differentiation alongside host rejection, has left the door open for more therapeutically efficacious sources of myocardial regeneration. Due to its fundamental role in heart development, the epicardium emerges as an obvious candidate. Here, recent findings are reviewed that show adult epicardium-derived cells as a new source of regenerative capacity for heart repair.

  • Heart failure
  • cardiomyocyte replacement
  • EPDCs
  • cardiac remodelling

Statistics from

Cardiovascular disease (CVD) remains the major cause of mortality and morbidity in the Western world.1 Cardiovascular disease eventually leads to heart failure and its main forms are coronary heart disease (CHD) and stroke. In the UK, CVD accounts for almost 198 000 deaths each year and CHD is the biggest killer, with one in every five men and one in every seven women dying from the disease.1 Coronary heart disease is not only the single most common cause of death in the UK; it is also very costly, imposing a huge annual burden on UK economy. The costs of healthcare alone are over £1.7 billion a year.1 Thus, there is an urgent need to lessen the clinical and economical burden of heart failure.

The heart, unlike a number of other organs or tissues in the human body (such as blood, bone marrow, liver and skin), has a reduced capacity for self-repair after severe damage.2 3 Consequently, acute or gradual loss of functional cardiomyocytes following ischaemic heart disease and acute myocardial infarction (MI) leads to diminished cardiac output and performance, and underlies progressive heart failure. Therapies available at present comprise lifestyle modification, drug treatments, surgery and heart transplantation. The latter remains the only existing cure for end-stage heart failure, but is limited by host immune rejection and reduced supply of donor hearts.4 As a result, in the last decade there has been an intensive effort in cardiovascular regenerative medicine to develop cell-based strategies for cardiac repair. Such strategies include the isolation and transplantation of adult progenitor cells from a number of sources (eg, bone marrow, adipose, skeletal and cardiac tissue) or the generation of pluripotent cells by the reprogramming of somatic cell types (induced pluripotent stem cells; iPS).5 However, whereas some of these cellular replacement strategies revealed disappointing outcomes in clinical trials (eg, bone marrow stem cells; see below), others are unlikely to be used in human testing for ethical and safety reasons (eg, embryonic stem cells and iPS; see below). Hence, there remains an urgent need to identify the most promising cardiovascular stem cells; either an exogenous source for transplantation and engraftment or a resident population for appropriate activation and induction towards a cardiomyocyte fate.

The heart contains both muscle and non-muscle cell types. Hence, heart regeneration requires production of functional cardiomyocytes and the development of a network of blood vessels to support and nourish these newly formed cardiac muscle cells. Basic research into the cellular and molecular mechanisms controlling mammalian heart development could provide key clues for cardiovascular regenerative medicine. For instance, during heart development the epicardium, an epithelial sheath that completely envelops the developing heart, has the potential to contribute adventitial fibroblasts, components of the coronary vasculature and cardiomyocytes by epithelial-to-mesenchyme transition of epicardium-derived cells (EPDCs).6–9 However, the ‘stemness’ potential of EPDCs is significantly reduced throughout embryonic development, and, in the adult, the epicardium becomes relatively quiescent.10 As a result, EPDCs have therapeutic potential in terms of coronary vasculature and cardiomyocyte regeneration pending reactivation of the dormant adult epicardium. Chemical approaches and small molecules have provided the key to many biological discoveries,11 and are therefore viewed as a promising advance to target and unlock the regenerative capacity of the epicardium.

In the subsequent sections, the current status of cardiovascular regenerative medicine will be reviewed, with a particular focus on the potential of the epicardium as a novel source of adult progenitors for replenishment of lost cardiovascular cells following CVD.

Non-cardiac adult progenitor cell-based therapies

In the last decade, a significant approach in cardiovascular regenerative medicine has been the use of adult stem/progenitor cells derived from bone marrow, skeletal muscle and adipose tissue, in transplantation towards cardiac repair.12–17 These sources are relatively accessible and provide a high number of progenitor cells. Moreover, the risk of immune rejection of these transplanted progenitors is minimal, as they can be produced in an autologous manner from patient biopsies.

Initial studies in rodents demonstrated that the infusion of bone marrow-derived cells into the injured heart led to vascular and cardiac muscle regeneration promoting cardiac repair.18–20 Even though some of these findings were difficult to reproduce in subsequent studies leading to controversy in the field,19 21–23 and the underlying molecular mechanism was unclear, the apparent ‘preclinical promise’ led to a number of clinical trials, where autologous-derived bone marrow cells were used to treat patients with acute or chronic MI.12 In all trials, primary angioplasty, the principal form of MI treatment, was used alone or in combination with bone marrow stem cell administration, via intracoronary infusion.24 Overall, the clinical trials revealed that limited engraftment of the infused cells was a major issue, as only approximately 2–7% of the infused cells were successfully found in the heart 24 h post transplantation.25 26 In order to improve the trial efficacy, other studies were designed where, instead of infusing autologous unfractionated bone marrow mononuclear cells, enriched fractions containing bone marrow-derived stem cells (BMSC), haematopoietic stem cells (HSC), endothelial progenitor cells (EPC) or mesenchymal stem cells (MSC) were used. Furthermore, higher doses of infused cells were administered, given an apparent positive correlation between the cell dosage and improvement in left ventricular ejection fraction (LVEF).27 28 On the whole, a modest benefit (increase in cardiac wall motion, reduction in infarct size and 3–4% increase in LVEF) was seen in meta-analyses of all available trials using bone marrow stem cells.27 29 There was marked heterogeneity between trials and LVEF, cardiac wall motion and infarct size were the clinical outcomes with most results, being widely used to define the success of a clinical trial. Nevertheless, it should also be mentioned that some studies have found improvement in patient outcome despite small or no improvement on LVEF.30 Therefore, other clinical outcomes/parameters need to be defined in order to assess the success of future trials. Of note, the preliminary results arising from clinical trials using isolated HSC/EPCs were equally disappointing, with marginal clinical improvement on a par with that arising from infusion of total BMSCs.31 Finally, the elevated LVEF observed during short-term follow-up did not seem to persist long term,32 suggesting that, instead of leading to cardiomyocyte replacement, bone marrow stem cells may only modulate cardiac repair in a paracrine fashion, for instance by limiting fibrosis in the affected area.

As an alternative to bone marrow, researchers in the field have considered the use of autologous muscle-derived skeletal myoblasts for deriving de novo cardiomyocytes. Skeletal myoblasts are progenitor cells that give rise to skeletal muscle cells, but are also thought to have the potential to differentiate into cardiac muscle.33 These progenitors can be isolated from patient muscle biopsies and intracardiac transplantation of skeletal myoblasts is already undergoing clinical trials.13 14 Phase I trials demonstrated that autologous skeletal myoblast transplantation into post-MI scars was a feasible approach, with improved cardiac function, notably LVEF. However, autologous skeletal myoblast transplantation also led to an increased incidence of ventricular arrhythmias, which significantly detracted from its therapeutic potential.14 34 35 The exact mechanism for the incidence of ventricular arrhythmias is not fully understood, but may relate to the fact that nascent skeletal muscle cells are electrically isolated after differentiation. Hence, transplanted skeletal myoblasts may not be able to form electromechanical junctions with cardiomyocytes when engrafted into the heart.36 It should also be pointed out that, so far, no study has unequivocally proven the ability of skeletal myoblasts to differentiate into cardiomyocytes.

Finally, adipose tissue may also be considered as a source for adult progenitor/stem cells in cardiovascular regenerative medicine.16 17 This tissue is rich in haematopoietic, endothelial and mesenchymal progenitors and can be easily obtained from patients by liposuction. Preliminary studies of intracoronary cell infusion conducted in rodent and porcine models of heart failure revealed that adipose tissue-derived progenitor cells when engrafted in the peri-infarcted regions, preserve heart function and augment local angiogenesis and cardiac nerve sprouting following myocardial infarction.16 17 However, they were not observed to exhibit significant cardiomyocyte differentiation. Therefore, transplanted adipose tissue-derived progenitor cells are more likely to exert their beneficial effects in a paracrine manner by providing trophic factors that stimulate neovascularisation in the injured heart. Despite the fact that adipose tissue is not a source for cardiomyocyte replacement, the beneficial effect observed in the animal studies has led to ongoing human clinical trials.37

Embryonic and induced pluripotent stem cell-based therapies

The modest clinical outcome of following trials using non-cardiac adult progenitor cell-based therapies, in particular the inability of these cells to regenerate lost or damaged myocardium has led to an intense effort to direct human embryonic stem (ES) cells and, more recently, human- iPS cells to form cardiomyocytes. Embryonic stem cells are pluripotent stem cells derived from the inner cell mass of the blastocyst, an early stage embryo.38 They are characterised by two distinctive properties: their pluripotency and their ability to self-renew indefinitely. Embryonic stem cells are, therefore, able to differentiate into all derivatives of the three primary germ layers (ectoderm, endoderm and mesoderm) and can be maintained in culture for long periods.38 39 Significantly, cardiogenic differentiation of ES cells can be invoked by stimulation with specific growth factors. For instance, stimulation of human ES cells with activin A and bone morphogenic protein 4 (BMP4) improves cardiac differentiation,40 and BMP2 treatment combined with an inhibitor of fibroblast growth factor receptor promotes the cardiogenic fate of human ES cells.41 So far, transplantation studies of human ES cell-derived cardiomyocytes into rodent models of heart failure have demonstrated some short-term (4 weeks post transplantation) functional improvement.40 41 However, cardiomyocytes derived from human ES cells resemble immature fetal cardiomyocytes due to their electrophysiology, calcium handling, force generation, contractile protein expression and myofibrillar structure.42–45 In addition, the use of human ES cells in clinical trials is severely compromised by ethical issues, immune tolerance and teratoma formation.

The generation of patient-specific iPS cells by retroviral-mediated nuclear reprogramming of somatic cells by pluripotency genes (OCT3/4, SOX2, KLF4 and c-MYC) provides an exciting alternative to ES cells.46 47 Induced pluripotent stem cells also have the potential to differentiate into all derivatives of the three primary germ layers and can differentiate into functional cardiomyocytes.48 Intramyocardial delivery of human iPS cells in a murine model ensured functional and structural repair on infarcted myocardium.49 As iPS cells can be generated in an autologous fashion, ethical issues and immune rejection are largely negated. Unfortunately, the risk of teratoma formation remains with iPS, and tumours due to retroviral vector integration should also be considered.50 Despite significant recent advances in non-viral delivery of reprogramming factors, for example use of piggyBAC vectors,51 the future clinical applications of iPS cell-based therapies will largely depend on: A. the elimination of the risks associated with both exogenous genetic manipulations and possible endogenous genetic alterations, during the slow and inefficient reprogramming process; B. the generation of homogeneous populations of lineage-specific cell types; and C. establishment of mature derivatives and functional differentiation of cardiovascular cell types. As the field currently stands, the utility of both human ES and iPS cell-derived cardiomyocytes is arguably limited to toxicology testing and pharmacological profiling of cardiovascular drug regimens and gaining insight into the cellular basis for CVD, as opposed to viable cell transplantation approaches towards cardiac repair and regeneration.

Resident cardiac progenitor cell-based therapies

As an alternative paradigm, researchers in cardiovascular regenerative medicine have turned to the heart as the ultimate source of regenerative capacity. Although, the mammalian heart itself has classically been seen as a postmitotic organ, recent findings suggested that it retains some capacity to renew cardiomyocytes throughout postnatal life.3 However, unlike non-mammalian vertebrate hearts, there is little or no significant cardiac muscle regeneration after an injury such as acute MI.2 A significant body of research has focused on non-mammalian species, such as zebrafish (Danio rerio), in order to understand how ‘natural’ heart regeneration can be blocked or enhanced.52–56 The zebrafish can fully regenerate its heart after amputation of up to 20% of the ventricle.53 Recent experimental evidence suggests that heart regeneration may result from limited dedifferentiation, followed by proliferation of pre-existing cardiomyocytes, rather than from progenitor cells,55 56 and from organ-wide, developmental activation of the adult zebrafish epicardium, the expansion of which supports the regenerating myocardium.54 In humans, cardiomyocyte dedifferentiation also occurs to some extent on injury, via activation of what is commonly referred to as hibernating myocardium,57 but it is not sufficient to enable cardiac muscle regeneration. Overall, the studies in zebrafish revealed a previously unappreciated role for the epicardium not only in adult heart repair, but also during continuous growth of the adult heart as epicardium-derived cells were incorporated into the compact layer of the myocardium.58

Epicardium derived-cells

The outermost epithelial layer of the heart known as epicardium originates from the proepicardium, a transient extracardiac structure formed by pericardial coelomic mesothelium at the venous pole (inflow region) of the embryonic heart.9 Differentiation of epicardial cells begins after looping of the developing heart is completed. During epicardial formation, a subset of epicardial cells detaches, undergoes epithelial-to-mesenchyme transition and migrates into the subepicardium, and subsequently into the myocardium.9 Initial studies using quail-chick chimeras and retroviral labelling revealed that these epicardium-derived cells (EPDCs) have the potential to differentiate both in vitro and in vivo into interstitial fibroblasts, coronary smooth muscle cells and coronary endothelium. In addition, under certain in vitro conditions EPDCs may also form cardiomyocytes.8 More recently, studies in mice confirmed some of the original observations reported in chickens, but also identified differences in EPDC fate between the two species. First, the endothelial compartment of the coronary vasculature is mostly derived from endothelial sprouts of the sinus venosus, the venous inflow tract of the embryonic heart, rather than from EPDCs.59 And second, unlike the chick, murine epicardium has a significant contribution to the cardiomyocyte lineage during heart development.6 7 In humans, little is known about the differentiation potential of EPDCs.60 Intramyocardial injection of human EPDCs in a mouse model of MI can improve cardiac function, suggesting that there is a genuine therapeutic potential for EPDCs to restore heart function after injury.61 Also, cells expressing the cell surface cytokine receptor c-kit and the glycoprotein CD34, known markers of stem cells, were identified in the subepicardial space of fetal and adult human hearts.62 Significantly, in vitro, these cells can acquire a myocardial phenotype, based on their gene expression. A similar population of cells exists in murine subepicardium, which, after MI, proliferates and differentiates into myocardial and smooth muscle cell lineages, as assessed by their gene expression profile.62 Taken together, these studies establish a basis for translating EPDC myocardial potential into adult heart regenerative therapies.

One obstacle limiting the use of adult EPDCs in cardiomyocyte replacement therapies is their relative quiescence. Indeed, throughout embryonic development, the capacity of EPDCs to differentiate reduces dramatically, both in vitro and in vivo.7 10 Yet, we and others have recently shown that thymosin β 4 (Tβ4), an actin monomer-binding protein releases adult epicardium from a quiescent state and restores pluripotency, by an unknown mechanism.10 63 During heart development, Tβ4 is expressed in the myocardium and provides a paracrine stimulus to EPDCs to promote their inward migration and differentiation into coronary smooth muscle cells. Translating this essential role for Tβ4 in coronary vessel development to the adult, we found that treatment of cultured adult heart explants with Tβ4 stimulated extensive outgrowth of epicardin-positive epicardial cells, which, as they migrated away from the explant, differentiated into procollagen type I, smooth muscle actin α, and vascular endothelial growth factor receptor 2-expressing cells indicative of fibroblasts, smooth muscle, and endothelial cells respectively.10 The ability of Tβ4 to promote coronary vessel development and potentially induce new vasculature in the adult is essential for cardiomyocyte survival and could contribute significantly towards the reported Tβ4-induced cardioprotection and repair in the adult heart.64 Remarkably, we found that Tβ4-treated adult epicardial explants were also able to differentiate into cardiac troponin T-expressing cardiomyocytes (Riley and coworkers, unpublished observations, 2010). This exciting finding suggests that cardiomyocyte replacement may be achieved via the adult epicardium on treatment with Tβ4. Studies to confirm the ex vivo findings in the heart proper are currently ongoing.

A chemical approach to unlock adult EPDC progenitor potential

A therapeutic ideal for heart failure would be the production of functional myocardium and the development of a network of blood vessels to support and nourish these newly formed cardiomyocytes. Adult EPDCs have the potential to generate all these cell lineages, pending reactivation of the embryonic developmental programme. Hence, an area of cardiovascular regenerative medicine for future development is the identification and characterisation of molecules that, like Tβ4, are able to restore resident adult EPDC progenitor potential (figure 1). This could be achieved via chemical genetics using small molecules and drug-like compounds. This approach has several advantages, as small molecules typically provide a high degree of temporal control over protein function, inducing rapid inhibition or activation, and their effects are often reversible.11 Besides, small molecule effects can be finely tuned by varying their concentrations, and a single molecule can potentially modulate multiple specific targets within, or across, protein families and signalling pathways.11 Furthermore, several natural and synthetic compounds with cardiogenic potential, including flavonoids and dimethyl sulphoxide (DMSO) have been identified.65 Finally, as adult EPDCs can be maintained in culture,66 this facilitates the application of phenotypic screens as a powerful way of examining multiple markers and functional changes using automated high-content imaging technologies in a high-throughput manner.11

Figure 1

A chemical approach to unlock adult epicardium-derived cells (EPDC) progenitor potential and promote heart repair. A therapeutic ideal for acute myocardial infarction (MI) and subsequent scar formation would be the treatment of the injured heart (B) with small-molecules (A) with the potential to reactivate the adult epicardium, leading to EPDC delamination and migration into the subepicardium (C). These adult EPDC would then give rise to de novo cardiomyocytes, which account for the loss of cardiac muscle following MI, and components of the coronary vasculature (smooth muscle cells), leading to neovascularisation (C).


In the past years, cardiovascular regenerative medicine has advanced significantly in the identification of putative sources for cardiomyocyte replacement, either exogenous or cardiac resident. Yet, when translated into clinical trials, exogenous sources tested revealed largely disappointing results with only modest improvement in cardiac function, most likely due to paracrine actions rather than cardiac differentiation. While we await the outcome of ongoing longer term follow-up in terms of mortality rates and improved survival,67 cardiac resident progenitor cell-based therapies offer a viable alternative, especially when based on predetermination to acquire a cardiomyocyte fate. During embryogenesis, EPDCs give rise to several cardiovascular cell lineages, including cardiac muscle cells and, therefore, can be seen as a new source of regenerative capacity. Future research is required to fully characterise this multilineage potential. For instance, by providing unequivocal evidence for adult EPDC cardiomyogenic differentiation in vivo and characterising the efficiency, frequency and electrophysiology properties of the EPDC-derived cardiomyocytes. In addition, this potential needs to be translated into therapies for acute MI in human patients (figure 1), for example via application of Tβ4 or small molecule-stimulated EPDCs. Significantly, the success of this therapeutic translation will largely depend on the presence of a functional pool of EPDCs in the human heart throughout ageing. As patients 60 years and above have the highest incidence of CHD, it will be critical to assess EPDC number and multilineage potential throughout the human lifespan.


We would like to thank all colleagues who contributed to the recent technical and scientific advances in cardiovascular regenerative medicine, including the authors of many papers not cited here due to space restriction.


View Abstract


  • Funding British Heart Foundation.

  • Competing interests None.

  • Provenance and peer review Commissioned; externally peer reviewed.

Request permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.