By definition, stem cells are undifferentiated cells that are clonogenic, self-renewing, and pluripotent or multipotent, capable of giving rise to all or multiple cell types in the body (
27). Typically, stem cells can be divided into three broad categories: embryonic stem cells, derived from the inner cell mass of blastocysts; induced pluripotent stem cells, derived from adult differentiated cells through reprogramming by the introduction of pluripotent transcription factors; and adult stem cells, such as skeletal myoblasts, hematopoietic stem cells, mesenchymal stem cells, endothelial stem cells, and cardiac stem cells present in adult tissues.
Since 2000, a number of different cell types have been investigated to treat patients with AMI or chronic ischemic cardiomyopathy in clinical trials, including skeletal myoblasts, bone marrow mononuclear cells, circulating progenitor cells, and mesenchymal stem cells (
27,
36,
74,
94). However, these cell sources have presented mixed results, and no cell type has been proved to be the best candidate for cardiac repair and regeneration in clinical trials conducted to date (
36,
94). On the other hand, it is important to note that autologous CSCs, including c-kit
+ cells and cardiosphere-derived cells, are already being investigated in phase I clinical trials. Encouragingly, the results of two clinical trials were recently released in Lancet, reporting the safety and efficacy of the use of resident CSCs in patients with ischemic heart disease (
9,
58). With better understanding of endogenous CSCs and advances in technology for generating iPSCs, it is hoped that CSCs and iPSCs could play crucial roles in the future of cardiovascular regenerative medicine. Each of these stem cells has its own advantages and disadvantages, which are briefly summarized in
Table 1 below. Certainly, time will decide which stem cell type is the optimal choice in the end.
Embryonic Stem Cells
Embryonic stem cells (ESCs), the prototypical stem cell, can develop into all cell types in the body, including pancreatic β-cells, neural cells, and cardiomyocytes (
11,
103). The isolation of mouse embryonic stem cells was first reported in 1981 (
31). Mouse ESCs (mESCs) were originally used to investigate embryonic development and establish genetically modified mice (
129). Over time, researchers expanded their interest to the field of the regenerative therapy involving a stem cell-based approach (
129). In 1998, the first human ESCs (hESCs) were isolated by Thomson et al. (
110) and have subsequently been attracting significant interest as a potential cell source for regenerative medicine due to their pluripotent capability and proliferation potential.
A number of studies showed that mESC- and hESC-derived cardiomyocytes can survive and improve heart function when injected into infarcted myocardium in murine models (
46,
48,
71,
123). However, there are several undesirable limitations with the practical application of hESCs, such as ethical problems, teratoma formation, and immunological rejection, which have hampered the initiation of clinical trials in patients with cardiovascular disease (
10,
92). It is clear that a better understanding of molecular and genetic pathways for ESC differentiation and cardiac development could prevent contamination with undifferentiated ESCs, preventing teratoma formation when transplanted into the body (
54,
94). Alternatively, to overcome the ethical issues and immune rejection, induced pluripotent stem cells might present a more attractive alternate, as they are of autologous origin (
103).
Induced Pluripotent Stem Cells
Recently, induced pluripotent stem cells (iPSCs) have been generated using a novel technology, which involves the introduction of transcription factors related to pluripotency into adult terminally differentiated cells, such as dermal fibroblasts, causing them to revert to an embryonic stem cell-like stage (
116). Takahashi et al. established that overexpression of four transcription factors [sex determining region Y box 2 (Sox2), octamer binding transcription factor 4 (Oct4), myelocytomatosis viral oncogene homolog (c-Myc), and Krüppel-like factor 4 (Klf4)] could convert mouse skin fibroblasts into pluripotent stem cells (
104). Furthermore, differentiation of iPSCs into functional murine cardiomyocytes was demonstrated (
66,
84). In 2007, Yu et al. successfully reprogrammed human somatic cells to iPSCs using four genes including Sox2, Oct4, Nanog, and Lin28 (
128), and these human iPSCs have been shown to have the potential to differentiate into functional cardiomyocytes (
77,
130).
Importantly, despite subtle epigenetic differences associated with reprogramming, iPSCs strongly resemble ESCs in terms of morphology, differentiation capacity, gene expression profile, and teratoma formation (
18). The use of iPSCs avoids the ethical dilemmas and immunological problems of ESCs, as they are derived from an autologous source; however, there remain concerns for clinical application because their generation employs viruses and oncogenes (
54). Thus, for safety reasons, the development of approaches using nonvector pluripotent induction (
45,
81,
127) without the need for oncogenes (
76) might pave the way for clinical applications in the future (
33,
103,
127).
Skeletal Myoblasts
Skeletal myoblasts or satellite cells, giving rise to skeletal muscle, were extensively studied in animal models of myocardial infarction before entering the clinical arena, due to the advantages of their myogenic commitment, high expansion capacity in culture, good resistance to ischemia, and autologous origin (
28,
54,
74).
Reports from experimental studies have shown that implanted myoblasts resulted in ventricular wall thickening and increased contractility, thereby improving function in the infarcted myocardium (
34,
41,
105,
107). Based on the positive results, skeletal myoblasts were the first cell type to be examined in a human trial for cardiac repair (
68). After the beneficial effects were shown in several pilot studies, further randomized controlled trials (RCT) using autologous skeletal myoblasts were conducted including MAGIC (Myoblast Autologous Grafting in Ischemic Cardiomyopathy), MARVEL-1 (To Assess Safety and Efficacy of Myoblast Implantation Into Myocardium Post Myocardial Infarction) and SEISMIC [Safety and Effects of Implanted (Autologous) Skeletal Myoblasts (MyoCell) Using an Injection Catheter]. The MAGIC study by Menasche et al. was terminated early because of a failure to reveal clinical efficacy and more events of arrhythmias in cell-treated patients undergoing coronary artery bypass grafting (CABG) for ischemic cardiomyopathy (
67). In contrast, the MARVEL-1 trial by Lainscak et al. showed a mean increase of 90.9 m during a 6-min walk test 6 months after myoblast transplantation via catheter in patients with congestive heart failure, compared with a mean decrease of 3.7 m in the placebo group (
49). Recently, the SEISMIC trial by Duckers et al. reported that injection of autologous skeletal myoblasts in patients with HF is safe and relieves symptoms based on a trend toward improved exercise tolerance in the cell-treated group despite no significant effect in left ventricular ejection fraction (LVEF) (
29). However, despite improving cardiac function when transplanted into ischemic myocardium, these cells were unable to transdifferentiate into cardiomyocytes and integrate electromechanically with the host myocardium, thereby increasing the risk of sustained ventricular tachycardia, a life-threatening arrhythmia (
54,
69). To suppress ventricular arrhythmia in patients receiving skeletal myoblast therapy, prophylactic cardioverter-defibrillator implantation and/or amiodarone may need to be used (
10). Collectively, in light of no cardiomyocyte regeneration, failure to integrate with host myocardium, potential lethal arrhythmia, and mixed results, further research is required prior to future clinical applicability.
Bone Marrow-Derived Stem Cells
The bone marrow is a heterogeneous tissue, consisting of different subpopulations, including hematopoietic stem cells (HSCs) and endothelial progenitor cells (around 2–4%), very rare mesenchymal stem cells (MSCs) (0.001–0.01% of the nucleated cells), and large proportions of committed progenitor cells and their specifically differentiated progeny (
2,
27,
74,
86). Technically, bone marrow-derived cells are easily accessible either by bone marrow aspiration or by isolation from peripheral blood after mobilization with cytokines such as stem cell factor (SCF) and/or granulocyte colony-stimulating factor (G-CSF) (
33). These cells have attracted great attention as candidates for cell-based therapy because of their autologous origin, safety, ease of isolation, and reduced immunogenicity (MSCs) (
4,
122). The use of bone marrow mononuclear cells (BMMNCs), endothelial progenitor cells (EPCs), purified progenitor cells (CD34
+ or CD133
+), and MSCs in experimental and clinical studies has provided informative data related to human CVDs. (
14,
122). However, as they are multipotent, it should be noted that bone marrow-derived cells could differentiate into a variety of cell types when transplanted, thereby carrying a potential risk of bone, cartilage, and adipose tissue formation in the heart (
101). It has been reported that unselected bone marrow cells resulted in substantial intramyocardial calcification and MSCs caused bone formation after transplantation into infarcted hearts in animal models (
12,
125).
HSCs, identified by the expression of cell surface antigens such as CD34, CD133, c-kit (CD117), and stem cell antigen-1 (Sca-1), are lineage negative (Lin
–) (
120,
121). These cells can be obtained from the bone marrow, umbilical cord, and peripheral blood, giving rise to all blood cell types (
120). HSCs have been extensively studied and used to treat a variety of hematological disorders in the clinic, such as anemia, leukemia and lymphoma (
33). A study by Orlic and colleagues showed that Lin
–/c-kit
+ bone marrow cells injected into infarcted myocardium of mice were able to generate new cardiomyocytes (
82). However, other studies have been unable to demonstrate cardiomyocyte transdifferentiation of HSCs and cardiac function improvement in animal models of myocardial infarction (
5,
75,
79). Furthermore, over the past years, several clinical trials of human HSCs have shown insignificant or no benefits in terms of ejection fraction (
94,
120).
MSCs, identified by the surface marker expression of CD90, CD105, and CD73, are precursors of nonhematopoietic tissues, such as bone marrow, muscle, cartilage, adipose tissue, and heart, and have the capacity to give rise to fibroblasts, osteoblasts, chondroblasts, and adipocytes in vitro (
15,
26,
119). Experimental results suggested that MSCs were able to transdifferentiate into cardiomyocyte-like cells under special culture conditions and in normal or injured myocardium in animals (
7,
15,
57,
111). In addition, it was shown that MSCs injected into infarcted myocardium could increase regional blood vessel density, prevent scar expansion, promote regional wall motion, and prevent ventricular remodeling (
95,
112). Interestingly, a recent report by Hatzistergos et al. demonstrated that allogeneic bone marrow MSCs stimulated the proliferation and differentiation of c-kit
+ CSCs when injected into the swine model of MI (
37). Clinical trials conducted in the past years have demonstrated safety and feasibility and displayed improvement in left ventricular function (
17,
118); however, these benefits are inconsistent (
1). Currently, other clinical trials of MSCs in patients with IHD are still under way. For example, PROMETHEUS (Prospective Randomized Assessment of Mesenchymal Stem Cell Therapy in Patients Undergoing Surgery) is evaluating the safety and effectiveness of injecting MSCs into the heart in postinfarct patients who are undergoing CABG (
http://www.clinicaltrials.gov/ct2/show/NCT00587990). The POSEIDON-Pilot Study (The Percutaneous Stem Cell Injection Delivery Effects on Neomyogenesis Pilot Study) aims to study and compare the safety and therapeutic efficacy of allogeneic MSCs with autologous MSCs in patients with chronic ischemic cardiomyopathy (
http://www.clinicaltrials.gov/ct2/show/NCT01087996). So far, it remains controversial whether MSCs have the potential to truly transdifferentiate into cardiomyocytes (
36,
85,
89).
BMMNCs (CD34
+ and CD133
+) are the most common bone marrow cell type used in clinical trials for patients with AMI and ischemic cardiomyopathy over the last decade (
101). In the early cohort and randomized pilot studies, BMMNCs were reported to result in a modest increase, of between 1% and 5%, in left ventricular ejection fraction (EF) at short-term follow up (3 to 6 months) (
14,
94). However, some clinical trials have not demonstrated the consistent results at long-term follow up (longer than 12 months) (
20). For example, the REPAIR-AMI (Reinfusion of Enriched Progenitor Cells and Infarct Remodeling in Acute Myocardial Infarction), a RCT including 204 patients, showed that intracoronary delivery of BMMNCs following AMI increased EF by 2.5% (
p = 0.01) compared with the control group at 4 months follow-up. The significant effect on EF was lost at 12 months follow-up; however, decreased mortality was observed in the treatment group (
93). Importantly, there are two recent clinical trials: BALANCE for AMI (Clinical Benefit and Long-Term Outcome After Intracoronary Autologous Bone Marrow Cell Transplantation in Patients With Acute Myocardial Infarction) and STAR-heart for ischemic cardiomyopathy (the acute and long-term effects of intracoronary stem cell transplantation in 191 patients with chronic heart failure) using BMMNCs, both showing a significant improvement in ejection fraction and survival rate in the treated group receiving BMMNC transplantation at 5 years follow-up, suggesting that BMMNC therapy results in significant and long-standing improvements in LV function and mortality in patients with AMI and chronic IHD (
102,
126). There are clinical trials using BMMNCs still in progress, such as TIME, REGENERATE-AMI, and REGENERATE-IHD. The TIME study aims to assess the safety and effect of timing of administration of BMMNCs in patients with AMI (
http://www.clinicaltrials.gov/ct2/show/NCT00684021) (
113). The REGENERATE-AMI study was designed to investigate the efficacy of early BMMNC transplantation within 6 h after successful primary angioplasty in AMI patients (
http://www.clinicaltrials.gov/ct2/show/NCT00765453). The REGENERATE-IHD is comparing the efficacy of BMMNC treatments using three different delivering methods, including indirect approach by mobilization using G-CSF, intracoronary infusion, and intramyocardial injection, in patients with ischemic HF (
http://www.clinicaltrials.gov/ct2/show/NCT00747708). It is hoped that more clinical studies will provide further insights into the therapeutic efficacy and help solve issues regarding bone marrow-derived cell transplantation in patients with IHD, including optimal cell type, cell dosing, and timing and route of delivery.
Resident Cardiac Stem Cells
The heart has traditionally been regarded as a terminally differentiated organ without the ability to regenerate itself. Recently, this dogma has been challenged by the discovery of a subpopulation of Lin
– and c-kit
+ cardiac stem cells (CSCs) resident in the rat heart, reported by Anversa and colleagues in 2003 (
8). Furthermore, Anversa et al. developed methods for isolation and expansion of c-kit
+ human CSCs (hCSCs) from small myocardial specimens. When injected into immunocompromised rats and mice, these cells differentiated into cardiomyocytes and improved the LV performance of infarcted hearts (
6). Other groups have also identified likely CSC populations using different cell surface markers such as Sca-1 (
80) and ATP-binding cassette transporter Abcg2 (
62) or by the transcription factor insulin gene enhancer protein (Isl-1) (
50). Furthermore, Messina et al. described a method to culture CSCs (grown as multicellular clusters, termed cardiospheres) to generate a mixed population expressing c-kit, Sca-1, and fetal liver kinase-1 (Flk-1) (
70). In addition, another source of endogenous resident cardiac progenitor cells with regenerative potential for the adult heart is the epicardium, with several groups reporting the discovery of epicardium-derived myocardial and vascular progenitors in embryonic mouse and adult human heart (
56,
97,
98,
114,
132). Importantly, these CSC populations have been found to have the potential to differentiate into cardiomyocytes and, in some cases, also into smooth muscle and endothelial cells (
63).
Marban et al. modified the protocol described by Messina's group to substantially expand cardiosphere-derived cells (CDCs) in vitro and showed myocardial regeneration and functional improvement when these cells were injected into the infarcted mouse heart (
99). In contrast to other populations of CSCs, cardiospheres and CDCs have been reported to contain a mixed population consisting of c-kit
+ cardiac progenitor cells and cells expressing CD90 (mesenchymal-related) and CD31/CD34 (endothelial progenitor-related) markers (
24,
61,
70,
99). It is possible that the cardiac progenitor cells could readily engraft, differentiate, and function when transplanted into the injured myocardium in the presence of cardiac mesenchymal stem cells and endothelial progenitor cells via synergistic paracrine effects (
16,
61).
Taken together, it is plausible that CSCs provide a desirable candidate cell for cardiac therapy in the clinical setting due to their endogenous origin and potential to develop into the three main cardiac lineages (
23). To date, human trials of endogenous CSCs in patients with ischemic heart disease include SCIPIO (Cardiac Stem Cell Infusion in Patients with Ischemic Cardiomyopathy) and CADUCEUS (Cardiosphere-Derived Autologous Stem Cell to Reverse Ventricular Dysfunction). The SCIPIO trial by Bolli et al. was designed to examine the safety and efficacy of intracoronary delivery of autologous CSCs, which are expanded c-kit-expressing cells from right atrial appendages, in patients with ischemic cardiomyopathy. The initial results, published in the November 2011 issue of Lancet, are encouraging, confirming the safety and feasibility, and providing the evidence which shows that intracoronary infusion of autologous c-kit
+ CSCs leads to a significant improvement in LV systolic function and a substantial reduction in scar size at 1 year of follow-up (
9). Similarly, the CADUCEUS trial, led by Marban et al., aimed to investigate the effects of autologous CDC transplantation via the intracoronary route in patients with a recent MI and ischemic left ventricular dysfunction. The results, published in Lancet early in 2012, were that intracoronary infusion of autologous CDC contributed to significant increases in viable myocardium, regional contractility, and regional systolic wall thickening despite no significant change in LVEF, which might be explained by the fact that EF at baseline was only moderately impaired (39%), leaving little room for improvement by 6 months (
58). Because of the positive findings, further research with longer follow-up and larger, phase II studies are required to confirm the true and persistent clinical benefits of c-kit
+ CSCs and CDCs.