Cardiac Regeneration: Evidence for Transdifferentiation
The ultimate objective for myocardial cellular transplantation is for donor cells to engraft in the recipient tissue and ultimately differentiate into new, functional cardiomyocytes and vascular cells (smooth muscle and endothelial cells). This is especially the case given the heart's limited capacity for self-regeneration and the central role that cardiomyocyte death and replacement fibrosis play in contributing to the development of heart failure .
In vitro differentiation of MSCs into cells resembling cardiomyocytes prompted early expectation of their capacity to regenerate these cells in vivo. Although not universally successful , exposure of MSCs to the DNA demethylating chemical 5-azacytidine has been the commonest strategy for inducing their cardiac differentiation in vitro. Under this condition, immortalized stromal cell lines and primary MSCs, from different species and different tissue sources, have been reported to modify their phenotype, with adoption of myotube morphology, expression of immature action potentials, and a variety of cardiac-specific genes (e.g., myocyte enhancer factor (MEF)-2A and MEF-2D) and peptides (e.g., myosin, desmin, actinin, and atrial and brain natriuretic peptides) [60, , , –64]. Functional differentiation has been indicated by the formation of intercellular connections via intercalated discs and by evidence of spontaneous cell contractility. Typically, these changes occur within 2–4 weeks of exposure to 5-azacytidine. Because of the potential genotoxicity of this chemical, other in vitro approaches to cardiomyocyte transdifferentiation have included culture in medium enriched with dexamethasone and ascorbic acid , bone morphogenetic protein-2, and fibroblast growth factor-4  and coculture with cardiomyocytes . It is unknown whether in vitro differentiation of MSCs into cardiomyocytes will enhance the reparative effects of these cells once they are transplanted in vivo .
Mesenchymal stromal cells can also produce cardiac connexin-43 and in theory, unlike skeletal myoblasts, can electromechanically couple to host cardiomyocytes in vivo . Evidence for in vivo differentiation has come from immunohistological analysis in small animal studies of xenogeneic MSC transplantation [43, 53] and large animal studies of autologous  or allogeneic [6, 44] cell therapy. Generally, authentication of differentiation has been incomplete, as engrafted MSCs have de novo expression of some myocyte (e.g., desmin, troponin T, phospholamban) or vascular (e.g., factor VIII, α-smooth muscle actin) markers, without acquiring the full phenotypic complement [6, 44, 70, –72]. Adding to skepticism about the differentiative capacity of transplanted cells, have been confounders of traditional immunofluorescent and immunohistochemical analysis, including tissue artifact and cellular fusion. Several groups have reported that engrafted BM-derived cells actually undergo fusion with endogenous cardiac cells, rather than true differentiation [73, 74]. However, this phenomenon has also been contradicted by other reports confirming the diploid chromosomal content of new cardiomyocytes generated after cell transplantation [75, 76]. Sophisticated new cell labeling techniques, such as direct immunofluorescence with quantum dots, should significantly improve the assessment of cellular differentiation in vivo and have recently been applied to demonstrate cardiomyocyte differentiation of c-kit-positive BMCs in a murine study .
The modest evidence for in vivo differentiation observed with MSC therapy to date may partly be due to the impure, heterogeneous nature of cells obtained through plastic-adherence isolation and the underwhelming retention of multipotent stem cells after ex vivo expansion. Compounding this are the extremely modest rates of intramyocardial retention, engraftment, and survival of cells following their administration by current delivery techniques . Cell survival may be especially compromised in the presence of unresolved myocardial ischemia or inflammation, and there is also evidence from research with skeletal myoblasts that host rejection may be further accentuated if cells have been cultured in media containing xenogeneic ingredients such as fetal bovine serum . A significant challenge for this field of research is the refinement of cellular biology and delivery technology to maximize the engraftment, survival, and function of cells in vivo.
Cardiac Repair: Paracrine Mechanisms
The variable observations relating to cell transdifferentiation have prompted a rethinking of the mechanisms that account for the functional benefits observed in studies of cardiac cell therapy. Increasingly, it is believed that current cell therapies assist the heart predominantly by facilitating endogenous repair processes, rather than through actual regeneration of lost cardiac and vascular cells. Paracrine actions may underlie much of this reparative potential, including the capacity for cell transplantation to induce neovascularization, reduce infarct size and scar formation, and improve myocardial contractility [80, 81] (Fig. 1). Mesenchymal stromal cells are able to influence other cells in their vicinity by direct cell-to-cell interaction and by release of a wide array of soluble growth factors and cytokines . These soluble factors are influenced by the developmental status and properties of the MSCs themselves and by the local milieu in which they find themselves. Relevant signaling and growth factors identified from both in vitro and in vivo experiments include stromal cell-derived factor-1 (SDF-1/CXCL12), hepatocyte growth factor (HGF), insulin-like growth factor-1 (IGF-1), basic fibroblast growth factor, hypoxia inducible factor-1α (HIF-1α), vascular endothelial growth factor (VEGF), angiopoeitin-1, monocyte chemoattractant protein-1, interleukins-1 and -6, placental growth factor, plasminogen activator, and tumor necrosis factor-α [80, –82]. Of these factors, HIF-1α is a transcription factor that is closely linked with cellular expression of VEGF and is activated and stabilized under hypoxic conditions . Recent results indicate that in addition to its role in angiogenesis, it may also help mediate the protective effects that MSCs exert on cardiomyocytes during in vitro hypoxic culture .
Figure Figure 1.. Proposed mechanisms of action mediating cardiovascular repair. Paracrine mechanisms contributing to the reparative effects of MSC transplantation may be mediated by the release of soluble cytokines/growth factors, as well as direct contact between transplanted cells and resident cardiac cells. In vivo transdifferentiation of MSCs into functional cardiomyocytes or vascular cells is probably limited with current transplantation strategies. Abbreviation: MSC, mesenchymal stromal cell.
Download figure to PowerPoint
It is not clear what influence the cellular composition of MSCs (i.e., the relative proportions of immature and mature cell types) has on the expression of these paracrine mediators. The targets of these paracrine effects are also uncertain but probably include both mature cells and resident progenitor cells in the recipient heart. Following MSC transplantation, myocardial injection sites have been shown to contain Ki67-positive cardiomyocytes. Although these proliferating cells could represent MSCs undergoing transdifferentiation, it is equally plausible that they represent endogenous cardiomyocytes or cardiac precursor cells undergoing regeneration . This is supported by discoveries showing the existence of resident cardiac stem/progenitor cells within the adult hearts of different mammalian species [86, –88]. These cardiac stem cells may be under the influence of growth factor signals (e.g., IGF-1 and HGF) that direct their migration to sites of injury, where they proliferate and differentiate . In theory, these processes could be facilitated by the paracrine actions of exogenous MSC therapy. Mature cardiomyocytes could also benefit from the protective effects of MSCs against hypoxic and ischemic insult , whereas resident endothelial cells and endothelial progenitor cells may be promoted to participate in angiogenesis and vasculogenesis [90, 91]. An additional mechanism that might contribute to cell-mediated cardiac repair is the digestion of scar tissue and the modification of extracellular matrix proteins .