Cardiac stem cells for myocardial regeneration: promising but not ready for prime time

Abstract

Remarkable strides have been made in the treatment of ischemic heart disease in decades. As the initial loss of cardiomyocytes associated with myocardial infarction serves as an impetus for myocardial remodeling, the ability to replace these cells with healthy counterparts would represent an effective treatment for many forms of cardiovascular disease. The discovery of cardiac stem cells (that can differentiate into multiple lineages) highlighted the possibility for development of cell-based therapeutics to achieve this ultimate goal. Recent research features cardiac stem cell maintenance, proliferation, and differentiation, as well as direct reprogramming of various somatic cells into cardiomyocytes, all within the context of the holy grail of regeneration of the injured heart. Much work remains to be done, but the future looks bright!

Graphical Abstract

Remarkable strides have been made in the prevention and treatment of ischemic heart disease in the last 70 years. Predisposing factors for this disease were initially defined from large observational studies, such as the Framingham Heart Study; upon these risk factors various behavioral and pharmacological interventions were subsequently brought to bear [1]. In addition to establishing the paradigm of primary (and primordial) prevention of ischemic heart disease, tremendous strides have been made in the treatment of acute myocardial infarction (AMI). Arguably the first major advance occurred with the advent of the coronary care unit in 1961 [2]: with immediate identification and treatment of the electrical complications of the disease, mortality associated with AMI was cut in half (from ~30%). The era of myocardial reperfusion started in 1975, when Chazov and colleagues utilized intracoronary streptokinase for fibrinolysis [3]. Other mortality-reducing pharmacological interventions soon followed, such as the use of beta-blockade, ACE-inhibitors, antiplatelet agents, and statins [4,5]. Subsequently, with the widespread adoption of primary percutaneous angioplasty and later stenting for coronary reperfusion [6,7], in-hospital mortality from AMI is currently thought to be on the order of 3.5%.

Our ability to treat the acute phase of ischemic heart disease came at the cost of a burgeoning population with chronic sequelae of AMI: namely, congestive heart failure (CHF). Indeed, with nearly 800,000 new cases of AMI in the US annually, there are approximately 5.7 million individuals with CHF, accounting for $30 billion in healthcare costs in 2012 [8]; this figure is forecast to grow exponentially with improvements in acute therapies, as well as the ageing of a population persistently exposing itself to traditional risk factors. As the initial loss of cardiomyocytes associated with AMI serves as an impetus for myocardial remodeling that culminates with CHF, the ability to replace these cells with electrically- and mechanically-healthy counterparts would represent an effective treatment for many forms of cardiovascular disease [9]. Indeed, the discovery of cardiac stem cells (that can self-expand and differentiate into multiple lineages, including cardiomyocytes) highlighted the possibility for development of cell-based therapeutics to achieve this end. At this time, however, understanding of this field precludes obtaining and integrating the ~10⁹ cardiomyocytes that would be required to replace those lost in an infarcted organ. In this review, we aim to highlight what is understood regarding cardiac stem cell maintenance, proliferation, and differentiation, as well as direct reprogramming of various somatic cells into cardiomyocytes, all within the context of the holy grail of regeneration of the injured heart.

Cardiac stem cells.

Until recently, the heart was viewed as a “post-mitotic” organ, incapable of functionally-significant regeneration [10]. Over the past 14 years, however, this notion has been challenged by a number of reports suggesting a low level of basal cardiomyocyte renewal in the adult human heart [11–16]. The source of these cardiac stem and / or progenitor cells (CSCs) has been debated. In 2003, the Anversa group published the first report of an endogenous CSC with regenerative potential from the mammalian heart based on the stem-cell factor (tyrosine kinase) receptor c-kit [12]. Subsequently, a number of other markers have been utilized to identify potential CSCs, including expression of stem cell antigen (sca-) 1 [11,15–17], the transcription factor Isl-1 [18–20], the ability to efflux DNA-binding dyes through an ATP-binding cassette transporter (side population, or SP cells) [21–25], or the propensity to grow in self-adherent clusters (cardiosphere-derived cells, or CDCs) [26–31]. Whether these cell types represent independent flavors of CSC, or rather one type of CSC in various stages of differentiation, has been questioned [32].

Cardiac stem cells expressing the c-kit receptor have been the most-studied flavor of CSC. In the seminal work from Beltrami and colleagues, these CSCs were demonstrated to be self-renewing, clonogenic, and multipotent, giving rise to cardiomyocytes, smooth muscle cells, and vascular endothelial cells [12]. Further in vivo evidence for the regenerative potential of these cells was supplied by this same group. Dawn and colleagues demonstrated that a population of c-kit+ CSCs gave rise to cardiac myocytes when injected intravascularly in a rat model of AMI, a process associated with a reduction in infarction size and attenuation in LV systolic dysfunction [33]. Bearzi and colleagues then showed in 2007 that the adult human heart possesses c-kit+ CSCs, which when injected locally in the infarcted myocardium of immunodeficient mice or immunosuppressed rats, generates a chimeric heart [34]; in these instances, the c-kit+ CSCs differentiated predominantly into cardiomyocytes [33,34]. Recently, however, two groups utilized the Kit locus for lineage tracing analysis in murine models of myocardial injury and demonstrated that endogenous c-kit+ cells generate cardiomyocytes at functionally-insignificant levels; endothelial cells, on the other hand, were readily generated [35,36].

In 2003, Oh and colleagues identified a population of CSCs in adult mice by their expression of Sca-1 [16]. This population could be directed toward a cardiomyocyte lineage with exposure to 5’-azacytadine; when infused systemically in a model of MI, these cells would selectively localize to the infarcted myocardium, where the authors documented differentiation into cardiomyocytes [16]. Unlike c-kit, Sca-1 is not expressed in humans. However, using an antibody directed against the murine Sca-1 epitope, the Doevendans group successfully isolated a population of human self-renewing multipotent cardiac progenitor cells from both fetal and adult samples [17]. In vivo evidence for the regenerative potential of these cells was supplied in 2009, when Smits and colleagues demonstrated that human fetal CSCs expressing the Sca-1-like antigen attenuated LV systolic dysfunction in an immunodeficient murine model of MI [37].

2003 was also the year that Cai and colleagues identified isl-1, a LIM homeodomain transcription factor, as a marker for cardiogenic precursor cells that delineate formation of the second heart field [18]. This transcription factor is highly conserved and is expressed in rodents, pigs and humans. Isl-1 is downregulated once cells enter a differentiation program and is invariably absent in fully-differentiated cells. We revealed that both murine and human embryonic stem-cell derived isl-1+ cardiac progenitor cells are self-renewing and can be directed into cardiomyocytes, smooth muscle cells, and vascular endothelial cells [20,38]. Recently, Bartulos and colleagues demonstrated that both murine and human isl-1+ cardiac progenitor cells engrafted in a murine model of MI, associated with reduced infarct area, increased blood vessel density, and improved LV systolic function [39].

Most reports on CSCs include data of application in a model system of MI. Overall in model systems, application of CSCs is associated with a modest improvement in LV ejection fraction (~12%) and appears independent of the CSC marker studied [40]. Owing largely to extreme need, evaluation of the application of endogenous CSCs has proceeded rapidly into Phase 1 trials. In 2011, Bolli and colleagues evaluated c-kit+ cells in patients with ischemic cardiomyopathy [41]. The following year, Makkar and colleagues evaluated autologous cardiosphere-derivedcell infusion into the infarct-related artery in post-MI patients with LV dysfunction. The latter authors found no change in LV function with CSC infusion but an improved scar-burden measured by MRI [42]. There is conjecture that this finding may not be the direct result of CSC engraftment, but instead a beneficial effect of paracrine signaling mediated by the infusate. These data suggest that CSC therapies may be nearing the realm of clinical utility and additional clinical studies evaluating cardiosphere-derived cells (ALLSTAR and DYNAMIC) and c-kit+ cells (with mesenchymal stem cells, CONCERT-HF) are ongoing.

Cardiomyocyte proliferation.

Cardiomyocyte proliferation has been known to occur in certain adult animals for nearly half a century. In 1974, Oberpriller and colleagues subjected the adult newt to a surgical injury of the LV apex resulting in a loss of ~20% cardiomyocytes, which could be completely repaired through endogenous mechanisms [43]. A similar phenomenon was later demonstrated in the zebrafish [44]. More recently in the latter organism, this process leading to cardiac regeneration was elucidated, demonstrating a process of dedifferentiation of cardiomyocytes adjacent to the injury site, which preceded activation of a cardiac differentiation program (including GATA4) and organ repair [45–47].

Mammalian cardiomyocytes, on the other hand, lose the majority of their regenerative potential after the early postnatal period, and until the late years of the twentieth century, the organ was regarded by most as “post-mitotic.” Several studies have led to a recent gradual shift in this paradigm. In the mid 1990s using labeled thymidine and thymidine analogs, which are integrated into genomic DNA during synthesis and can be used to annotate cell cycle events, multiple groups demonstrated progressive loss of cardiomyocyte DNA synthesis in rats within 10 days of birth [10,48,49]. In the adult heart, this cardiomyocyte regeneration continues to diminish: using a similar approach in the adult murine heart, Soonpaa and colleagues measured a 0.0005% labeling frequency of cardiomyocytes expressing a transgenic marker for the α-MHC-promoter (2h after tritiated thymidine injection) [50]. These data were substantiated in humans with C-14 dating, a technique devised by the Frisen group. An elegant pulse-chase experiment, C-14 dating relies on the transient increase in the biospheric quantity of this isotope, which is a product of above-ground nuclear testing. In 2009, Bergman and colleagues estimated the annual rate of cardiomyocyte turnover to be ~1% at 25 years of age and ~0.45% at 75 years of age [14]. However, numerically-trivial, the identification of a process of cardiomyocyte turnover in the adult heart is likely of critical importance, as augmenting an existing biological process is infinitely easier than initiating such a process de novo. Recently, one study revealed that partial reprogramming can erase cellular markers of aging in mouse and human cells, which might have the potential to unlock cardiac regeneration in adult hearts [51].

It is unclear whether the relative loss of regenerative potential in the adult mammalian heart is due to an intrinsic cell cycle block or to loss of proliferation stimuli. Indeed, manipulating developmental signals and cell cycle checkpoints have been documented to simulate cardiomyocyte proliferation. Administration of neuregulins, which are known to regulate cardiomyocyte proliferation and differentiation in development, has been shown to promote adult cardiomyocyte proliferation [52]. Modification of the Hippo pathway has been demonstrated to enhance neonatal and adult mouse cardiomyocyte cycling [53]. Additionally, over-expression of cyclin genes and knockdown of p21/p27 increased cell cycle activity of cardiomyocytes. Knockout of the regulators of the E2F transcription factors, the pocket proteins Rb and p107, increased cardiomyocyte numbers [54,55]. Chemical inhibition and genetic knockout of GSK3β have been reported to increase cardiomyocyte cycling. The implications of activation of these cell cycle factors are likely far-reaching (beyond activation of cell division):over-expression of the adenoviral oncogene E1A or E2F-1 induced cardiomyocyte cycling led to apoptosis [54,55]. Understandably, the clinical implications of these data remain unclear.

Transdifferentiation.

Transdifferentiation, or lineage reprogramming, is a process whereby one mature somatic cell is transformed into another without an intermediate pluripotent or progenitor state. This technique may possess some advantages over others for cardiac regeneration, such as shorter times for myocyte generation, less chance of tumor formation, and the avoidance of certain ethical issues. Fibroblasts constitute the majority of cells in the heart under physiological conditions and replace the lion’s share of cardiomyocytes following AMI: therefore, they represent the perfect substrate for myocardial transdifferentiation. In 2010, Ieda and colleagues used 3 transcription factors (Gata4, Mef2c, and Tbx5) to transdifferentiate murine cardiac fibroblasts into cardiomyocytes in vitro and in vivo [56]. The addition of Hand2 to the previous 3 transcription factors further improved efficiency of cardiac reprogramming [57]. MiRNAs have also emerged as a tool for transdifferentiation, both alone and in combination with various transcription factors. Jayawardena and colleagues demonstrated that miRNAs 1, 133, 208, and 499 could transdifferentiate murine fibroblasts to cardiomyocyte-like cells both in vitro and in vivo [58]. Nam and colleagues demonstrated that miRNAs 1 and 133, in combination with GATA4, HAND2, TBX5, and MYOCD, improved transdifferentiation efficiency of human fibroblasts into cardiomyocytes [59]. Additional evidence suggests that certain small molecules, including DNA methyltransferase inhibitors, histone methyltransferase inhibitors, histone deacetylase inhibitors, glycogen synthase kinase-3 beta (GSK-3β) inhibitors may replace some reprogramming factors [60–63]. On a similar note, the Kamp group has successfully reprogrammed murine fibroblasts into proliferative induced cardiac stem/progenitor cells (iCPCs), where addition of the transcription factor isl-1 facilitated stable reprogramming and produced more proliferative colonies [64]. Although transdifferentiation for cardiomyocyte generation (and generation of iCPCs) is nascent technology, it may hold promise for future applications in regenerative medicine.

In conclusion, there remains a tremendous need for cell based therapies for cardiac regeneration, which is sure to increase over time. Recent research has identified multiple potential approaches to achieving this goal (Figure 1). On the one extreme, it is conceivable that with understanding of the master genes that control the CSC state, in vivo genesis of whole organs using human donor CSCs in cardiogenesis-disabled host animals may be possible [65,66]. It is the authors’ opinion, however, that autologous stem cell infusion will be the first of such therapies to enter routine clinical use, although better techniques for optimizing cell engraftment and limiting electrical inhomogeneity will necessarily precede widespread use. Much work remains to be done, but the future looks bright!

Figure 1. Cell-based therapeutic approaches for heart repair fall into two main categories:

1) Transplantation: human cardiac stem cells (CSCs) or cardiomyocytes (CMs) are derived from various sources, expanded in vitro, and subsequently transplanted into patients in any form (e.g. cell patches, mini tissues, etc); in the extreme, human-animal chimerism provides whole hearts from human donor CSCs for transplantation.

2) In vivo heart repair: exogenous factors re-activate quiescent CSCs to differentiate into various cardiac cell types (including cardiomyocytes) or promote division of endogenous quiescent CMs to regenerate lost cardiomyocytes. Alternatively, myocardial regeneration is achieved using various factors to directly reprogram non-myocytes into CMs.

Highlights.

• The discovery of cardiac stem cells shines the light on heart repair.

• Cardiomyocyte proliferation is an alternative approach for heart regeneration.

• Conversion of somatic cells into cardiomyocytes may have important therapeutic implications.

• Human stem cells can contribute to forming organs in animals.

• Human-animal chimeras hold great potential for heart transplantation.