Stem cell therapy: pieces of the puzzle.

Acute ischemic injury and chronic cardiomyopathies can cause irreversible loss of cardiac tissue leading to heart failure. Cellular therapy offers a new paradigm for treatment of heart disease. Stem cell therapies in animal models show that transplantation of various cell preparations improves ventricular function after injury. The first clinical trials in patients produced some encouraging results, despite limited evidence for the long-term survival of transplanted cells. Ongoing research at the bench and the bedside aims to compare sources of donor cells, test methods of cell delivery, improve myocardial homing, bolster cell survival, and promote cardiomyocyte differentiation. This article reviews progress toward these goals.

Cardiovascular disease remains the leading cause of death worldwide. Current therapies seek to prevent atherosclerosis through risk factor modification and to manage the consequences of thrombosis after injury has occurred. Regenerative medicine offers a new paradigm for treatment of heart disease. Stem cell therapies may improve ventricular function after injury through either direct or indirect means, by engraftment and differentiation into cardiac and vascular cells or by secretion of paracrine factors that promote tissue survival and recovery. Thus far, the results of animal studies and clinical trials have been encouraging, despite limited evidence for the long-term survival of transplanted cells.

In contrast to conventional therapies where defined pipelines lead from target identification to drug discovery and development, cardiovascular cell therapies are evolving by a much more diffuse process, more akin to a group of investigators working to solve a puzzle. Pieces of this cell therapy puzzle may now be defined, so that they can be refined and assembled through ongoing study and collaboration. These pieces include (1) the cell types available for transplant, (2) the methods of cell delivery, (3) the means of myocardial homing, (4) the pathways of cell survival, and (5) the goals of cell differentiation.

Stem Cell Populations with Potential for Cardiac Regeneration

Over the past decade, many cell types have been evaluated in an effort to find the best source for cardiac regeneration. Of these, the most extensively studied have been embryonic stem cells, mesenchymal stem cells, skeletal myoblasts, and bone marrow-derived progenitor cells, while more recent possibilities have included induced pluripotent stem cells and resident cardiac stem cells.

Embryonic stem cells (ESCs) can be obtained from the inner cell mass of a pre-implantation blastocyst and expanded in vitro almost indefinitely without loss of pluripotency [1-3]. When allowed to differentiate as embryoid bodies, ESCs give rise to most somatic lineages, including cardiovascular lineages [4-8]. Culture of embryoid bodies with specific growth factors or small molecules can drive differentiation toward cardiovascular phenotypes, thereby enriching the pool of ESC-derived cardiac cells available for transplant [8-13]. Both pluripotent ESCs and committed ESC-derived cardiac cells have been tested in rodent models with encouraging results in terms of engraftment, survival, and improvement in ventricular function [14-16]. However, no clinical trials have been attempted, for concerns of teratoma formation [16, 17], graft-versus-host disease [18], and bioethics. While improved in vitro differentiation of ESCs may eliminate the first of these concerns, recent discoveries indicating that adult somatic cells can be reprogrammed to yield ESC-like cells may eliminate the second and third of these issues.

Induced pluripotent stem cells (iPSCs) can be generated from murine and human adult somatic cells by overexpression of transcription factors critical for maintenance of ESC pluripotency. While initial iPSC protocols required retro- or lentiviral transfer of four factors, namely Octamer-binding transcription factor 3/4 (Oct 3/4), Sry-related HMG-box transcription factor (Sox2), Krüppel-like factor 4, and cellular myelocytomatosis oncogene [19, 20] or Oct 3/4, Sox2, Nanog, and Lin28 [21], subsequent reports have reduced the risk of tumorigenesis introduced during the de-differentiation process by utilizing non-integrating technologies [22-25] or by transferring fewer transcription factors [26-28] while still achieving pluripotency. Once established, iPSCs differentiate much like ESCs, yielding functional cardiovascular lineages from embryoid bodies in vitro [28-32] and contributing to normal cardiovascular development from chimeric blastocysts in vivo [28, 31, 33]. In an initial proof-of-principle study, four-factor iPSCs successfully restored myocardial structure and function after coronary artery ligation in mice [33]. Thus, iPSCs may provide an unlimited supply of autologous donor cells, once technological advancements eliminate the risk of teratoma or other tumor formation.

Mesenchymal stem cells (MSCs) can be separated from bone marrow and adipose tissues based on their adherence to a culture dish [34]. MSCs most readily differentiate into osteoblasts, chondrocytes, and adipocytes [35, 36] but can be induced to differentiate into cardiomyocytes under special conditions [37-39]. They can be rapidly expanded in culture, thereby allowing for autologous transplantation, and they may be less immunogenic than other cell populations, thereby allowing for allogenic transplantation [40, 41]. But, their propensity to differentiate into non-cardiac tissues (e.g., heterotopic ossification) currently limits their use [42, 43]. Nonetheless, at least one clinical trial has shown improved ventricular function at 3 and 6 month follow-up after intracoronary infusion of autologous MSCs postinfarction [44].

Skeletal myoblasts, also known as satellite cells, can be harvested from muscle biopsies, expanded in culture, and autologously reimplanted, albeit on a timescale more appropriate for chronic heart failure therapy than for early postinfarction management. As these cells are further differentiated than ESCs, they are less prone to teratoma formation; they are also more resistant to ongoing ischemia and more apt to function in a contractile capacity postengraftment. However, true cardiomyocyte differentiation has not been observed in vivo. As a result, although skeletal myoblasts may be incorporated into an infarct site, their contractions are dyssynchronous due to failure of electromechanical coupling with the surrounding myocardium [45, 46]. Thus, no significant benefit has been seen in large-scale clinical trials, while a trend toward more frequent arrhythmias has been observed [47].

Bone marrow-derived mononuclear cells, including hematopoietic and endothelial progenitor cells, can be mobilized with cytokine therapy or harvested by standard collection techniques. In culture, they can differentiate into cardiomyocytes, endothelial cells, and smooth muscle cells [48-51], and after transplant, they can supply a broad range of paracrine factors with proangiogenic, positive remodeling, and antiapoptotic properties [52]. A meta-analysis of 18 randomized and non-randomized trials involving 999 patients with acute myocardial infarction or chronic ischemic cardiomyopathy found that intracoronary infusion of adult bone marrow-derived stem cells improved left ventricular ejection fraction by 3.7%, decreased infarct scar size by 5.5%, and lowered left ventricular end-systolic volume by 4.8 ml (p < 0.01 for each index) [53]. Among these pooled studies, the largest randomized trial, REPAIR-AMI, was particularly notable for not only demonstrating improvement in ventricular function but also for reducing the combined endpoint of death, myocardial infarction, or revascularization at 12 months when compared with placebo [54]. Nonetheless, controversy remains regarding the fates of bone marrow-derived cells used for cardiac regeneration [55].

Spatiotemporal Aspects of Stem Cell Delivery

Identification of the most safe and efficacious means of stem cell delivery remains an important challenge, both in terms of route and timing of administration. Current methods of stem cell delivery include non-invasive mobilization from resident niches (especially bone marrow) with cytokine therapy, non-selective intravenous or intracavitary injection, selective intracoronary infusion, and intramyocardial injection from either surgical or percutaneous access.

Although some intravenously injected bone marrow-derived stem cells do home to the infarct border zone, only a small fraction (< 5%) of intravenously injected cells is retained in this intended destination. Rather, the majority of intravenously injected cells are trapped by the lungs, liver, kidneys, and spleen [56, 57]. More proximal but still non-selective delivery via intracavitary or intra-aortic (with the distal aorta occluded) injection marginally improves cardiac uptake [56, 57], suggesting that closer may be better. Such reasoning underlies the majority of clinical trials which have used intracoronary delivery techniques. Indeed, intracoronary delivery appears superior to intravenous delivery in man [58, 59], although overall cell retention remains low as rapid washout follows balloon deflation [60]. Clinical trials assessing the biodistribution of infused cells have recently been reviewed [61].

In animal studies, intramyocardial delivery yields better retention than intracoronary delivery [57, 62, 63]. But so far, no clinical trial has directly compared these two modalities, although such results should be forthcoming [64]. In theory, intramyocardial delivery may provide more precisely targeted therapy without need for patent coronary arteries, while intracoronary delivery may benefit from operator familiarity and lower cost. Since intramyocardial injections carry risk of myocardial perforation and electrical instability shortly after infarction, these two strategies in fact may be complementary, with intracoronary cell delivery provided early and intramyocardial cell delivery provided late.

Despite the potential importance of correctly timing stem cell administration, few studies have directly addressed this issue. On one hand, stem cell survival may be compromised shortly after infarction by transient microvascular obstruction, recurrent ischemia, or reperfusion injury. On the other hand, stem cell homing to the infarct border zone may be enhanced shortly after injury by transient upregulation of cytokines such as stromal cell-derived factor 1 (SDF-1), vascular endothelial growth factor (VEGF), hepatocyte growth factor (HGF), and monocyte chemotactic protein-3 (MCP-3) [65-68]. In rodents, this latter phenomenon appears to have the upper hand, as the benefits of MSC therapy are seen only when administered within 4 days postmyocardial infarction, paralleling myocardial SDF-1 expression [67].

In man, the optimum timing of stem cell therapy may be different. So far, stem cell administration within 1 day of reperfusion therapy has not shown clinical benefit [69]. Therefore, the NIH Cardiovascular Cell Therapy Research Network's TIME trial will compare an early time point of 3 days postmyocardial infarction to an intermediate time point of 7 days postmyocardial infarction for intracoronary stem cell infusion [ [70]; http://ccct.sph.uth.tmc.edu/cctrn/Public/PublicHome.aspx]. This trial will run in parallel with a related study called LATE-TIME, which will focus on a late time point of 2–3 weeks postmyocardial infarction for intracoronary cell infusion. Beyond that time point, a recent substudy of the MYSTAR trial found no significant differences in the long-term improvements in infarct size and ventricular function observed after bone marrow mononuclear cell delivery at 3–6 weeks versus 3–4 months [71].

Means of Improving Myocardial Homing

Treatments designed to improve stem cell homing to diseased myocardium have the potential to markedly enhance both the safety and efficacy of cardiovascular cell therapy. By increasing cardiac retention, stem cell capture by non-cardiac tissues may be reduced, thereby limiting potential side effects. This may allow stem cell dosages to be increased, even as outcomes on a per cell basis are improved. Alternatively, decreased stem cell numeric requirements may reduce cell culture and delivery costs.

Many current strategies to improve cardiovascular homing aim to enhance adhesion protein and cytokine interactions between donor cells and host tissues. Ex vivo conditioning of donor cells, for example, by incubation of endothelial progenitor cells with a β2-integrin activating antibody, can improve retention and transcapillary migration at sites of injury, where intracellular adhesion molecule-1 (ICAM-1) and fibrinogen are upregulated [72]. Alternatively, in vivo conditioning of the host vasculature, for example, by infusion of adenosine, can promote cardiac retention of endothelial progenitor cells by upregulating coronary artery expression of P-selectin [73]. Increased display of this cell adhesion protein captures endothelial progenitor cells that constitutively express P-selectin glycoprotein ligand-1 [73, 74]. Adenosine can be safely infused into the human coronary circulation [75]; thus, it could be an adjuvant to many existing cell therapy protocols.

Other potential homing agents remain in pre-clinical development. These include master regulators of ischemic responsiveness such as integrin-linked kinase (ILK) and hypoxia inducible factor-1α (HIF-1α), as well as secreted homing factors such as SDF-1, MCP-3, and VEGF. In response to hypoxia, ILK is stabilized by heat shock protein 90. Increased ILK activity then leads to increased expression of SDF-1 and ICAM-1 through HIF-1α- and nuclear factor κB-mediated signaling [76]. This results in endothelial progenitor cell recruitment to ischemic tissues [76]. While this beneficial property of ILK has been demonstrated only in a hindlimb ischemia model thus far, its cardiac application is likely not far off. Already, it is known that SDF-1 recruits endothelial progenitor cells expressing CXCR4 to injured myocardium [65], and ICAM-1 binding to its receptor, β2 integrin, on endothelial progenitor cells is required for engraftment after myocardial infarction [77]. Therefore, given the functional improvements seen with ILK gene therapy after coronary artery ligation in rats [78], utilization of ILK in stem cell therapy would seem a logical step.

Indeed, encouraging results have been obtained when local myocardial therapy has been combined with stem cell transplantation. Downstream of ILK, HIF-1α regulates both SDF-1 and its receptor CXCR4, as well as ICAM-1 [76, 79, 80]. Combining intramyocardial injection of adenovirus-encoded HIF-1α with transplantation of skeletal myoblasts results in improved cell engraftment and cell survival after coronary artery ligation in rats [81]. This results in marked improvement in ejection fraction (by 27%) at 1 month posttherapy over baseline at 1 week postinfarct. In comparison, monotherapy with HIF-1α or myoblasts alone serves only to prevent further decline of the ejection fraction over this interval [81].

Further down this ischemic response pathway, both SDF-1 and its receptor CXCR4 have been used successfully as adjuncts to cardiac cell therapy. In mice, endogenous SDF-1 is upregulated for less than 1 week after myocardial infarction [65-67]. Re-establishment of SDF-1 expression at later time points can enhance stem cell recruitment and retention when the ischemic border zone might be less hostile to cell engraftment and survival. Indeed, intramyocardial delivery of scaffold-bound SDF-1 after coronary artery ligation in rats and mice results in enhanced c-Kit+ cell homing and improved ventricular function [82, 83].

Cell-based gene therapy can provide a similar biological scaffold. While it may seem counterintuitive for a stem cell to deliver its own homing factor, considerable benefit has been obtained when genetically modified stem cells recruit additional native and transplanted cells to injured myocardium. For example, improved cell engraftment and survival, as well as improved ventricular function, has been observed after transplantation of SDF-1−expressing skeletal myoblasts (as opposed to unmodified skeletal myoblasts) at both early and late time points after myocardial infarction in rats [84, 85]. Similarly, transplantation of SDF-1-expressing cardiac fibroblasts promotes homing of endogenous CD117+ stem cells to injured myocardium [65]. This results in improved ventricular function even without cardiomyocyte regeneration [65]. Rather, improved engraftment and differentiation of transplanted SDF-1-expressing MSCs (as opposed to unmodified MSCs) into myofibroblasts, but not true cardiomyocytes, results in improved ventricular function by providing trophic support for cardiomyocyte preservation—but not regeneration—after infarction [86]. Similar results are obtained after transplantation of CXCR4-overexpressing MSCs [87], indicating that improved stem cell homing, engraftment, and paracrine activity can be obtained via cell-based gene therapy with both stem cell homing factors and their receptors.

Other growth factor–chemokine-receptor pairs, which may be utilized to improve outcomes of cardiac cell therapy, include VEGF−Flk-1, HGF−c-Met, MCP-3−CCR1/2, and growth-related oncogene-1−CXCR1/2 [66, 68, 88–91]. Of these, the most extensively studied factor is VEGF. VEGF is transiently upregulated after myocardial infarction [66] and serves to mobilize endothelial progenitor cells while promoting neovascularization [92, 93]. VEGF is both induced by SDF-1 and an inducer of SDF-1's receptor CXCR4 [94-96]. Early clinical trials demonstrated that intramyocardial delivery of naked plasmid-encoded VEGF gene therapy is safe but results in only marginal clinical benefit [97, 98]. Subsequent human and large animal studies of combination intramyocardial VEGF gene delivery plus stem cell therapy have yielded mixed results [99-101]. Meanwhile, VEGF-transfected stem cells have consistently outperformed untransfected control cells after coronary artery ligation in rodents [102-107]. And in those studies that have compared cell-based gene therapy to direct gene delivery, VEGF-transfected skeletal myoblasts and VEGF-transfected MSCs have improved ventricular function to a greater degree than either direct adenoviral injection or liposomal plasmid delivery, respectively [106, 107]. Taken together, these results suggest that cell-based gene therapy may be more effective than either unmodified stem cell therapy or acellular gene therapy alone. The degree to which this principle is generalizable remains to be seen.

Enhancement of Transplanted Cell Survival

Despite the structural and functional benefits of stem cell therapy in animal and human studies of acute myocardial infarction and chronic heart failure, limited transplanted cell survival has been observed, especially in the long-term. Less than 1% of transplanted cells survive in patient hearts [108], and even in the idealized scenario of syngeneic neonatal cardiomyocyte intramyocardial injection in healthy rats, only 15% of transplanted cells survive for 12 weeks [109]. Nonetheless, recent discoveries demonstrative of the heart's capacity for regeneration and repair suggest that these challenges are not insurmountable. Measurement of 14C in the DNA of individuals born prior to Cold War era nuclear testing estimates that nearly half of all cardiomyocytes are exchanged during a normal life span [110]. Since 14C is incorporated equally by the division of resident cardiomyocytes and circulating progenitor cells, carbon dating cannot distinguish the cellular origins of such impressive turnover. Rather, the phenomenon of cardiac chimerism seen after sex-mismatched heart or bone marrow transplant confirms that extra-cardiac progenitor cells are capable of long-term engraftment and survival in the healthy human heart [111]. Thus, the challenge for cardiac cell therapy is to strengthen donor cells for the harsh microenvironment of the infarcted heart.

Ischemic preconditioning has long been recognized as a means of improving myocardial ischemic tolerance [112]. In addition to activating survival pathways in cardiomyocytes, ischemic preconditioning recruits endothelial progenitor cells to serve as mobile reservoirs of cardioprotective cytokines [113]. Proponents of cardiac cell therapy now seek to harness this property to enhance donor cell survival. Hypoxic preconditioning of mesenchymal stem cells and cardiac progenitor cells prior to transplant improves not only their recruitment, survival, and differentiation at the infarct border zone but also overall ventricular structure and function [114, 115]. Similar benefits might be obtained by hypoxic preconditioning of endothelial progenitor cells, based on results from an ischemic hindlimb model [116]. This is because, in each of these cell types, hypoxic preconditioning upregulates a diverse array of survival, angiogenic, and migratory proteins, including HIF-1α, Akt-1, Bcl-2, Ang-1, VEGF and its receptor Flk-1, SDF-1's receptor CXCR4, and HGF's receptor c-Met [114, 115, 117]. Together, these factors allow donor cells not only to endure but also to heal the injured myocardium.

Extending this concept, hypoxic preconditioning can be mimicked by genetic modification of donor cells to overexpress hypoxia-responsive survival factors. Akt-1 is activated by hypoxia and other stimuli, including cytokines. As a general mediator of survival signaling, Akt-1 is both necessary and sufficient for cell survival in a variety of settings [118]. Blockade of Akt-1 activation with the phosphoinositol 3-kinase (PI3K) inhibitor wortmannin attenuates the cardioprotective effects of endothelial progenitor cell therapy after ischemic injury in pigs [119]. On the other hand, genetic modification of MSCs to overexpress Akt-1 results in marked improvement in intracardiac MSC retention and survival in rat and pig infarct models [120, 121]. In both species, transplantation of Akt-1-overexpressing MSCs also normalizes ventricular function.

Follow-up studies have demonstrated that Akt-1-overexpressing MSCs release paracrine factors that promote myocardial survival and recovery after infarction [122, 123]. By means of functional genomics, secreted frizzled related protein 2 (Sfrp2) was identified as the key Akt-1–MSC-expressed paracrine factor mediating myocardial survival and repair after ischemic injury [123]. When Sfrp2 is suppressed in Akt-1−MSCs, the ability of these cells to repair injured myocardium is lost [123]. Alternatively, genetic modification of MSCs to specifically overexpress Sfrp2 promotes their retention and survival within the infarct border zone, as well as facilitates restoration of cardiac structure and function after infarction [124]. These benefits may be attributable to Sfrp2-mediated inhibition of Wnt binding to the frizzled receptor, which results in accumulation of β-catenin and upregulation of antiapoptotic and proangiogenic signaling.

Other paracrine factors upregulated in Akt-1−MSCs include thymosin β4 (Tβ4) and insulin-like growth factor-1 (IGF-1) [122]. Tβ4 is a secreted, G-actin sequestering, 43 amino-acid peptide that stimulates the migration and survival of cardiomyocytes and endothelial cells [125], as well as the migration and differentiation of epicardial coronary vascular progenitors [126]. After coronary artery ligation in mice, intramyocardial injection of Tβ4 protein activates ILK and Akt-1, limits scar formation, and improves ventricular function [125]. Exemplary of Tβ4's importance in cell therapy, shRNA-mediated knockdown of Tβ4 in endothelial progenitor cells blocks the cardioprotective actions of these cells after ischemia-reperfusion in mice [127].

In addition to stimulating myocardial growth and contractility, IGF-1 activates the PI3K/Akt-1 pathway to promote cardiomyocyte survival and proliferation in response to injury [128, 129]. IGF-1 protein has been combined with cell therapy to improve donor cell survival and differentiation in host myocardium [130, 131]. Extending these benefits, IGF-1 cell-based gene therapy has improved donor cell engraftment and proliferation and enhanced host cell recruitment to ischemic myocardium [132, 133]. This has been accompanied by decreased infarct size, increased angiogenesis, and improved ventricular function. Molecular studies have attributed IGF-1's therapeutic benefit, in part, to activation of PI3K/Akt-1 and Bcl-2 and to release of SDF-1 and VEGF [132, 133].

Downstream of IGF-1 and PI3K/Akt-1, but not Sfrp2 [123], Bcl-2 has been identified as a key antiapoptotic protein with potential utility in cardiac cell therapy. Bcl-2 preserves mitochondrial integrity during ischemia and is upregulated by ischemic preconditioning [134]. Cell lines derived from ESCs which overexpress Bcl-2 self-renew continuously, even in the absence of serum and feeder cells [135]. This gives a marked survival advantage to Bcl-2 overexpressing donor cells upon cardiac transplant. Indeed, MSCs, smooth muscle cells, and cardiomyoblasts genetically modified to overexpress Bcl-2 have all demonstrated reduced cell loss and improved ventricular function after transplant into ischemic myocardium [136-138]. But while such results may be encouraging, cell-based gene therapy utilizing cell survival factors is not without its challenges, as these proteins may predispose to tumorigenesis. Much work still lies ahead to regulate the expression of these factors before their clinical application.

Attaining Cardiomyocyte Regeneration

Myocardial regeneration after injury has long been the goal of cardiac cell therapy. However, recent experience has shown that cell therapy is capable of providing significant structural and functional benefit with only minimal cardiomyocyte regeneration. As we have discussed, donor cells can recruit host cells to the site of injury, especially if they are modified to overexpress stem and progenitor cell homing factors. Donor cells can secrete paracrine factors to promote host tissue preservation and repair, especially if they are modified to overexpress survival or growth factors [52, 122, 123, 133], and donor muscle cells can provide contractile reserve, especially if they are modified to overexpress connexins essential for electromechanical coupling [139]. Given these capabilities, one might legitimately ask if the last piece of the cardiac cell therapy puzzle—cardiomyocyte regeneration—even needs to be solved.

We would argue that cardiomyocyte regeneration remains a worthy and attainable goal. As we have discussed, recent measurements of 14C incorporation into the heart indicate that nature provides for cardiac regeneration throughout a normal human lifetime [110]. Our challenge is to nurture that process after injury. Genetic fate-mapping studies have shown that nearly 20% of cardiomyocytes bordering a murine myocardial infarction are derived from endogenous stem or precursor cell differentiation [140]. Cell therapy aims to increase that ratio. This may be accomplished by utilization of more committed cardiac progenitor cells isolated from the adult heart or derived in culture from pluripotent stem cells.

The adult heart is home to a side population (SP) of stem cells identified by exclusion of the vital dyes Hoechst 33342 and rhodamine 123 [141, 142]. Isolated cardiac SP cells that express stem cell antigen-1 (Sca-1+), but are negative for the endothelial marker CD31, are capable of differentiation into functional cardiomyocytes after co-culture with adult cardiomyocytes [143] or treatment with a cocktail of bone morphogenetic protein 2, fibroblast growth factors 4 and 8, the Wnt antagonist Dickkoff-1, and 5-azacytidine [144]. Transplantation of both pre-differentiated and undifferentiated Sca-1+/CD31− cells after myocardial infarction in mice results in efficient homing, engraftment, and differentiation into cardiomyocytes and endothelial cells, as well as improved ventricular function [144-146].

A second resident progenitor population comprises c-Kit (also known as CD117) positive cells, which are located in small clusters within the adult heart [147]. Isolated c-Kit+ cells do not differentiate into cardiovascular cell types in culture but showed impressive regenerative potential in some studies after transplantation into the injured rat heart, where they gave rise to cardiomyocytes, endothelial cells, and smooth muscle cells, while also improving ventricular function [147-150]. However, these results have not been reproduced in all studies [55]. Nonetheless, a rare Sca-1+/c-Kit+ cardiac cell type that differentiates into cardiomyocytes upon oxytocin treatment has been identified in mice [151].

From a technical standpoint, cardiac progenitor cells can be isolated from mouse hearts and human endomyocardial biopsies by enzymatic digestion to release round cells that form so-called cardiospheres in suspension [152, 153]. Cardiosphere-derived cells express endothelial and stem cell markers, show contractile activity in culture, and can differentiate into cardiomyocytes, endothelial cells, and smooth muscle cells. Transplantation of these cells improves ventricular function after myocardial infarction in mice and swine, due to both direct cardiomyocyte regeneration and indirect paracrine action [153-155]. To test whether these presumably committed cardiac progenitor cells have benefit in man, enrollment in the first clinical trial of “Cardiosphere-derived autologous stem cells to reverse ventricular dysfunction” (i.e., the CADUCEUS trial) is now underway [156].

An alternative to laborious isolation of committed cardiac progenitor cells from the adult heart is differentiation of committed cardiac progenitor cells from embryonic stem cells, induced pluripotent stem cells, or other multipotent cellular resources such as adipose tissue-derived MSCs. After study of many different growth factor cocktails, a growing consensus now agrees that modulation of bone morphogenetic protein (BMP) signaling may be the key to reliable ex vivo cardiomyocyte differentiation. However, consensus is lacking with regard to identification of the critical BMP agonists and antagonists, as well as their appropriate dosing. Treatment of murine ESCs or MSCs with BMP2 in the presence of a second growth factor such as transforming growth factor-β1, fibroblast growth factor 2 or 4, or leukemia inhibitory factor promotes cardiomyocyte differentiation and improves cardiac structure and function following transplantation into infarcted myocardium [8, 9, 11, 157]. Similar treatment of human ESCs with BMP4 also promotes cardiomyogenesis [13]. However, it is also known that the endogenous BMP antagonist Noggin and the synthetic BMP inhibitor dorsomorphin also increase cardiomyocyte differentiation [10-12], in apparent contrast to studies that blocked the inductive properties of BMP2 and BMP4 with Noggin [8, 13]. A unifying hypothesis for these observations suggests that BMP signaling is necessary for at least two steps in cardiomyocyte development: mesodermal induction and cardiomyocyte differentiation [158, 159], between which transient block of BMP signaling is essential [10]. Thus, the optimal timing and dosage of BMP agonist and antagonist treatments may be critical. Finally, recent evidence that adult cardiomyocytes may be induced to proliferate could provide a complementary strategy for cardiac regeneration [160].

Ongoing research at the bench and the bedside will surely continue to perfect these pieces of the cardiac cell therapy puzzle: comparing sources of donor cells, testing methods of cell delivery, augmenting myocardial homing, bolstering cell survival, and promoting cardiomyocyte differentiation, and while much more work needs to be done, we have confidence that stem cell therapies in conjunction with current treatment modalities will ultimately reduce the mortality and improve the quality of life of patients with cardiovascular disease.