1.1 Introduction

Cardiovascular diseases remain to be one of the leading causes of mortality worldwide and represent an enormous health and economic burden (Whelan et al. 2010). Identifying strategies to regenerate the adult human heart after injury has spurred a furiously paced experimental race toward this goal.

Historically, the mammalian heart has been considered to be a postmitotic organ, without any capacity for cell turnover and regeneration post-injury (Laflamme and Murry 2011). Instead of regenerating muscle, a scar is formed to maintain the integrity of the mammalian heart following injury; hypertrophy of the remaining myocardium takes place, but the loss of myocardium can eventually lead to the development of heart failure (Jessup and Brozena 2003). Mechanical approaches for treatment of heart failure aimed to counteract the weakening of the heart muscle following injury include left ventricular assist devices (Terracciano et al. 2010). Other approaches include neurohormonal inhibition, which is widely used in clinical practice (Sharpe et al. 1991). These approaches are beneficial for patients with heart failure, but the potential to completely regenerate lost myocardium remains an important goal.

Early discoveries showed that endogenous cardiac regeneration can occur in some vertebrate organisms such as the newt and zebrafish (Oberpriller and Oberpriller 1974; Poss et al. 2002). Recently, the neonatal mouse heart was reported to regenerate in response to injury in a manner similar to lower vertebrates (Porrello et al. 2011, 2013). The regenerative response has been attributed to the ability of cardiomyocytes to proliferate with restoration of functional myocardium (Jopling et al. 2010; Kikuchi et al. 2010; Porrello et al. 2011). Cardiomyocyte cell cycle activity is maintained throughout the adult life of vertebrates, but rapidly declines with age in mammals (Li et al. 1996; Poss et al. 2002; Walsh et al. 2010). The lessons learned from lower vertebrates as well as the neonatal mouse suggest that endogenous heart regeneration can occur, and understanding this process could allow new therapeutic approaches to regenerate the human heart.

Early mouse studies showed that cardiomyocyte turnover in the adult murine heart occurs at low levels, around 1% annually (Soonpaa and Field 1997). Cardiomyocyte turnover in the adult mouse heart during aging and following injury was demonstrated at high resolution using mouse genetic lineage tracing and multi-imaging mass spectrometry showing similar levels of myocyte turnover (Senyo et al. 2013; Hsieh et al. 2007). To measure the levels of cell turnover in the adult human heart, a landmark study exploited the rise of 14C levels during the cold war testing of nuclear weapons, which created an opportunity to trace the levels of 14C from human heart samples and thus enabled the researchers to determine the rate of cardiomyocyte turnover in the adult human heart (Bergmann et al. 2009). Similar to the murine heart, the adult human heart showed cardiomyocyte turnover at extremely low levels, around 1% annually (Bergmann et al. 2009). Although the cardiomyocyte refreshment is insufficient for a substantial regenerative response following injury, this indicates that the heart is much more resilient than previously considered. Surprisingly, a recent study reported that the human neonatal heart can regenerate after a myocardial infarction (Haubner et al. 2015). The similarities between the neonatal mouse and neonatal human heart, as well as the adult mouse and adult human heart, suggest that regenerating the adult heart will be feasible.

1.2 Cell Therapy for Cardiac Repair

The development of many tools in regenerative medicine has inspired cardiovascular investigators to utilize these methods to regenerate the human heart to restore contractile function following injury. Stem cells have generated particular excitement for their potential for cell-based cardiac repair (Garbern and Lee 2013). The plasticity of stem cells and their ability to differentiate into multiple cell types has generated hope for the future of regenerative medicine. The past decade has witnessed numerous studies that used different cell types with varying abilities for cardiac repair, which led to many clinical trials. The results from these trials continue to generate controversy regarding the impact of cellular therapy, but it is clear that cell therapy may have an important future for human heart regeneration.

The collective knowledge of cellular plasticity in the mammalian heart as well as the explosion of the stem cell field fueled the hope to either harness the endogenous potential of the mammalian heart or utilize the potential of exogenous stem cells that can differentiate into functional myocardium. In the following paragraphs, we will discuss the utility of different types of exogenous stem cells, as well as the potential of different endogenous cardiac progenitors for cellular transplantation, in addition to cellular reprogramming, to regenerate the adult human heart.

1.2.1 Skeletal Myoblasts

Skeletal myoblasts were among the initial cell types to be introduced for clinical cardiac cell therapy. Skeletal myoblasts were reasonable candidates due to their resistance to ischemia, as well as their differentiation potential (Durrani et al. 2010). In addition, early results showed the promise of skeletal myoblasts in heart repair following injury in multiple experimental animal models (Durrani et al. 2010). However, it was shown that myoblasts fail to integrate with the host myocardium and thus fail to beat in sync with the heart (Leobon et al. 2003). Furthermore, the first multicenter, randomized, placebo-controlled human clinical trial for myoblast autologous grafting in ischemic cardiomyopathy (MAGIC) did not enhance cardiac contractile function (Menasche et al. 2008). These results led to a reduced enthusiasm toward the use of skeletal myoblasts, and regenerative approaches moved onward toward more promising cell types.

1.2.2 Bone Marrow-Derived Stem Cells

Bone marrow-derived stem cells have the capacity to differentiate into multiple cell types including vascular and cardiac cell fates both in vitro and in vivo (Hirschi and Goodell 2002). The detection of Y-chromosome-positive cardiomyocytes in female hearts that were transplanted into male patients suggested that bone marrow-derived stem cells can differentiate into cardiomyocytes (Quaini et al. 2002). Over a decade ago, bone marrow-derived stem cells that express the surface marker c-kit emerged as candidates for regenerating the heart following injury through transdifferentiation into cardiomyocytes (Orlic et al. 2001). The differentiation potential of c-kit + cells in vivo led to controversy. Although the initial report suggested transdifferentiation of these cells into cardiomyocytes, subsequent studies by other groups found no evidence of transdifferentiation into cardiomyocytes, but rather showed the formation of more mature hematopoietic cell lineages following transplantation (Murry et al. 2004; Balsam et al. 2004). Improved ventricular function was detected following bone marrow-derived stem cell injections in multiple studies; several reports suggested that this effect is due to a paracrine effect through enhancing proliferation and differentiation of endogenous cardiac progenitors, thus promoting cardiac repair indirectly (Loffredo et al. 2011; Hatzistergos et al. 2010; Urbich et al. 2005; Mathieu et al. 2009; Kinnaird et al. 2004; Gnecchi et al. 2006; Hong et al. 2014).

REPAIR-AMI was the first randomized, blinded clinical trial to use autologous bone marrow cells through intracoronary infusion for acute myocardial infarction (MI) patients (Schachinger et al. 2006). There was a significant improvement in the left ventricular function following bone marrow transplantation, an effect that persisted up to 2–5 years after transplantation but with no impact on survival, though the trial had insufficient power to study survival (Assmus et al. 2010, 2014). However, a subsequent trial (TIME) using autologous bone marrow cells in ST-segment elevation MI (STEMI) showed no effect on improving cardiac function (Traverse et al. 2012). These mixed results generated debate on the impact of bone marrow-derived cells on improving function and survival of MI patients (Marban and Malliaras 2012). Retrospective evaluation of these clinical trials revealed some of the discrepancies and potential pitfalls to be avoided for proper assessment of the value of these cells as a clinical treatment (Simari et al. 2014; Nowbar et al. 2014). A large phase 3 clinical trial to assess the value of bone marrow cells in myocardial infarction patients is currently underway (BAMI trial).

1.2.3 Endothelial Progenitors

Endothelial progenitor cells (EPCs) are a small population of adult hematopoietic CD34+ progenitors that were identified in 1997 and which have the capacity to differentiate into endothelial cells (Asahara et al. 1997). Preclinical studies of EPCs showed promising results in enhancing recovery following ischemia in different tissues, mainly by enhancing neovascularization (Kawamoto and Losordo 2008). Following myocardial infarction, EPC transplantation may enhance functional recovery and myocardial integrity in vivo (Iwasaki et al. 2006; Kawamoto et al. 2001). This requires the homing of the EPCs to the site of ischemia followed by proliferation and differentiation into functional endothelial cells (Hristov et al. 2007). In addition to neovascularization, endothelial cells enhance cardiomyocyte survival and organization and contraction of surrounding cardiomyocytes through paracrine signaling (Narmoneva et al. 2004). These data suggest that endothelial cells can promote cardiac repair through different mechanisms.

Early phase clinical trials using cell transplantation of EPCs showed promising results for functional recovery following a cardiac insult (Vrtovec et al. 2013; Stamm et al. 2007). Similarly, pharmacological mobilization of EPCs using granulocyte colony-stimulating factor (G-CSF) showed enhanced cardiac function post-injury (Achilli et al. 2010). A major impediment to the understanding the full potential of EPCs is the different isolation methods for EPCs between different groups; thus the identity and purity of the EPC populations have not been consistent. Owing to the limited size of previous trials, large studies would be required to establish the efficacy of EPCs for heart repair.

1.2.4 Mesenchymal Stem Cells

Another subset of progenitors within the bone marrow is mesenchymal stem cells (MSCs). MSCs are multipotent and can differentiate into adipocytes, chondrocytes, and osteoblasts (Pittenger et al. 1999). MSCs are found in multiple tissues, and they can be expanded to the large numbers necessary for transplantation. Furthermore, MSCs appear less immunogenic due to the absence of MHC-II complex and may have lower probability of rejection (Kuraitis et al. 2011). Allogeneic MSCs showed therapeutic benefits following transplantation in the injured rodent and swine heart (Williams and Hare 2011). Initially there was evidence that allogeneic MSCs can differentiate into cardiomyocytes in vivo following engraftment in the adult murine and swine heart (Toma et al. 2002; Quevedo et al. 2009). Subsequent studies showed that MSCs probably accomplish beneficial effects via paracrine mechanisms (Mirotsou et al. 2007; Gnecchi et al. 2005, 2008). MSC transplantation in large animals showed activation and differentiation of cardiac stem cells (Hatzistergos et al. 2010). The POSEIDON trial showed improved patient outcome and ventricular remodeling, but no significant improvement in ventricular function (Hare et al. 2012). A randomized phase 3 clinical trial using MSCs for ischemic heart failure is currently ongoing (CHART-1), and the results from this trial will shed light on the future of MSCs in the clinic.

1.2.5 Endogenous Cardiac Stem Cells

1.2.5.1 C-kit + Cardiac Progenitors

Following the developments in the hematopoietic stem cell (HSC) field, c-kit, the receptor for stem cell factor, was described as a surface marker of HSC stemness. It was reported that the heart has an endogenous cardiac progenitor cell (CPC) population that is c-kit + without any hematopoietic lineage marker expression (Lin-) (Beltrami et al. 2003). These cells were described as clonogenic and multipotent due to their ability to differentiate into cardiomyocytes, endothelial cells, and smooth muscle cells in vitro and in vivo. Expansion of the c-kit + CPCs ex vivo and injection of the cells in vivo following MI showed a dramatic regenerative effect on the heart (Beltrami et al. 2003). Furthermore, a similar population of c-kit + CPCs was reported in the adult human heart and could repopulate the infarcted murine myocardium (Bearzi et al. 2007). The ability of c-kit + CPCs to enhance myocardial regeneration was demonstrated by different groups in small and large animal studies (Linke et al. 2005; Fischer et al. 2009; Angert et al. 2011). In contrast, other groups showed that c-kit + CPCs from adult hearts do not differentiate into cardiomyocytes ex vivo (Zaruba et al. 2010; Jesty et al. 2012). One study reported that c-kit + CPCs are not only necessary but also sufficient for myocardial regeneration following cardiac injury (Ellison et al. 2013). These conflicting results regarding the differentiation potential and functional impact of c-kit + CPCs on myocardial regeneration were addressed by a very well-designed lineage tracing study aimed to label the putative c-kit + CPCs in the heart to trace their lineage during aging and following injury (Van Berlo et al. 2014). Although this study showed that c-kit + CPCs could differentiate into cardiomyocytes, this occurred at negligibly low rates, which suggested that c-kit + CPCs would have no impact on myocyte replenishment following injury (Van Berlo et al. 2014). In contrast, c-kit + CPCs produced a high percentage of cardiac endothelial cells, suggesting that the endogenous c-kit + CPCs are more likely to be endothelial progenitor cells rather than true cardiomyocyte progenitors.

Although it is still unclear how c-kit + CPCs can enhance myocardial repair, their impact over the past decade has led to a number of clinical trials to assess their therapeutic potential. An early clinical trial to test the safety of c-kit + CPCs was a phase 1, randomized clinical trial for CPC intracoronary infusion in patients with ischemic cardiomyopathy (SCIPIO) (Bolli et al. 2011). This trial showed that CPC injection is safe with no adverse effects up to 1 year, with an improvement in LV function. The phase 1 trial was very small, however.

1.2.5.2 Cardiosphere and Cardiosphere-Derived Cells

Cardiosphere-derived cells represent another subset of cardiac progenitor cells in both murine and human hearts that are multipotent and can differentiate into different cardiac cell types (Messina et al. 2004). Cardiosphere-derived cells (CDCs) can be isolated from cells cultured from endomyocardial biopsies, and injection of CDCs in a large animal model of infarct has led to enhancement of cardiac function (Smith et al. 2007). Cardiosphere-derived cells were isolated from human hearts as well, and intracoronary injection of human CDCs in a pig infarct model improved cardiac function and reduced scar formation (Johnston et al. 2009). A phase 1 clinical trial using cardiosphere-derived autologous stem cells to reverse ventricular dysfunction (CADUCEUS) for intracoronary injection in myocardial infarction patients suggested the safety of these cells for clinical use, with potential benefits on cardiac function (Makkar et al. 2012). Further mechanistic studies of cardiosphere-derived cells showed that intracoronary injection of CDCs post MI stimulate endogenous cardiomyocyte proliferation, as well as recruitment of endogenous progenitors (Malliaras et al. 2013). These dual mechanisms may explain the beneficial outcomes following CDC cell therapy. Larger studies will be necessary to reveal the impact of cardiospheres and CDCs on this cell type as a candidate for myocardial regeneration in humans.

1.2.5.3 Side Population Cells

Side population (SP) cells are a population of cells characterized by their ability to exclude the Hoechst dye, since they express the ATP-binding cassette transporter proteins. These cells were first characterized as a hematopoietic stem cell population in the bone marrow (Goodell et al. 1996). To determine whether SP cells from the bone marrow can enhance cardiac repair, SP cells were transplanted in mice following an ischemia reperfusion injury and shown to have some therapeutic benefit and enhance myocardial repair (Jackson et al. 2001). Cardiac SP cells were further isolated from the developing and adult mouse heart, with the capacity to differentiate into cardiomyocytes, endothelial cells, and smooth muscle cells (Martin et al. 2004). Intravenous infusion of cardiac SP cells into rats that underwent myocardial infarction demonstrated that cardiac SP cells were able to migrate and home to the injured myocardium, differentiate into different cardiac cell types, and enhance heart regeneration (Oyama et al. 2007). These preclinical studies suggest a potential benefit for SP cells for cell therapy, although human clinical trials have not been performed with SP cells injected into the heart.

1.2.5.4 Sca1+ Cardiac Progenitors

Stem cell antigen1 (Sca1) is a cell surface marker expressed on the surface of multiple tissue-specific resident stem cells (Holmes and Stanford 2007). A Sca1+ cardiac stem cell population has been identified in the adult mouse heart and can differentiate into cardiomyocytes in vitro following treatment with 5-azacytidine as well as oxytocin (Oh et al. 2003; Matsuura et al. 2004). Intravenously injected Sca1+ cells were able to home to injured myocardium, differentiate into cardiomyocytes or fuse with host cells, and enhance repair following ischemia reperfusion injury (Oh et al. 2003). Similarly, intramyocardial transplantation of Sca1+ cardiac stem cells improved LV function post MI, although this effect was probably mediated through a paracrine effect by increased neovascularization and enhanced cardiomyocyte function (Wang et al. 2006). Lineage tracing of Sca1+ cells in the heart suggested that Sca1+ cells contribute to cardiomyocyte renewal in the adult heart, as well as in response to injury (Uchida et al. 2013). A major impediment to the clinical potential of Sca1+ cardiac stem cells is the lack of the Sca1 antigen in humans, which limits the use of these cells for human therapy.

1.2.5.5 Islet1+ Cardiac Progenitors

Islet1 (Isl1) is a transcription factor that marks a cardiac progenitor population from the second heart field that can differentiate into multiple cardiac lineages during heart embryonic development (Moretti et al. 2006; Cai et al. 2003). Interestingly, an Isl1+ cardiac progenitor population was identified in the early postnatal heart that can differentiate into mature cardiac lineages (Laugwitz et al. 2005). However, the Isl1+ progenitor population only persists in the sinoatrial node in the adult murine heart and is nearly absent from the left ventricle either at baseline or following MI (Weinberger et al. 2012). Isl1+ cardiac progenitors have not been studied for human cell therapy.

1.2.5.6 Epicardial Progenitors

Epicardial progenitor cells that express the transcription factor Wt1 have an important role during murine heart development, as they contribute to the formation of functional cardiomyocytes (Zhou et al. 2008). This role of epicardial progenitors led to the search for a similar progenitor population in the adult mammalian heart. Lineage tracing of Wt1 in the adult heart showed that Wt1 epicardial progenitors are present in the adult mouse heart and can give rise to bona fide cardiomyocytes following MI (Smart et al. 2011). Priming the Wt1 epicardial progenitors with thymosin β4 before MI may be an important step for these progenitors to give rise to cardiomyocytes, as priming them with thymosin β4 after MI does not seem to contribute to the differentiation of epicardial progenitors into cardiomyocytes (Zhou et al. 2012). However, embryonic and adult epicardial progenitors seem to be different subpopulations, as adult epicardial progenitors are more heterogeneous and have a different expression profile at the molecular level than embryonic progenitors (Bollini et al. 2014). Interestingly, a recent protocol described the derivation of primary human epicardial-derived cells from right atrial appendage biopsies, which can serve as a platform to further identify the therapeutic potential of epicardial progenitors for adult cardiac cell repair (Clunie-O’Connor et al. 2015).

1.2.6 ES-Derived Cardiomyocytes

Generation of differentiated, mature, and functional cardiomyocytes from pluripotent ES cells is a promising approach to replenish lost myocardium following injury (Xu et al. 2002). The development of directed differentiation protocols of pluripotent embryonic stem cells (ES) into cardiomyocytes has witnessed significant advances (Mummery et al. 2012). Purified human ES cell-derived cardiomyocytes (hESC-CM) can be derived when cultured with activin A and bone morphogenetic protein 4, which can improve cardiac function of the infarcted rat heart (Laflamme et al. 2007). Transplantation of cardiovascular progenitors derived from hESC led to engraftment in the infarcted hearts of nonhuman primates (Blin et al. 2010).

To establish the electrophysiological properties of hESC-CM, purified hESC-CM were transplanted in guinea pigs following injury (Shiba et al. 2012). These grafts led to reduced arrhythmias and were able to electrically couple with the host myocardium and thus efficiently enhance myocardial function following cryoinjury (Shiba et al. 2012). Recently, a large animal study in macaques showed that hESC-CM was able to remuscularize and regenerate the infarcted monkey heart (Chong et al. 2014). These recent promising results show the significant potential of hESC-CM. However, the ability of the transplanted hESC-CM to integrate efficiently in syncytium and prevent arrhythmias is still a concern (Chong et al. 2014). These issues will need to be addressed before hESC-CMs can be used for clinical trials. While ethical concerns might hamper ES use in the clinic, induced pluripotent stem cells (iPSCs) could replace hESC as the source of cardiomyocytes for cell therapy.

1.2.7 Induced Pluripotent Stem Cells and Reprogramming

1.2.7.1 iPSCs

Reprogramming adult mouse and human fibroblasts into a pluripotent state by transduction of four transcription factors, OCT4, SOX2, KLF4, and c-MYC (OSKM), was a revolutionary moment in biomedicine (Takahashi and Yamanaka 2006; Takahashi et al. 2007). iPSCs resemble ES cells morphologically and molecularly, and thus they provide an alternative to ES cell use, in addition to the advantage of generating patient-specific cell lines for autologous regenerative therapies. Although there was an initial concern toward using iPSCs clinically due to the use of oncogenes and viral vectors which can lead to teratoma formation, new methods and protocols are emerging that utilize small molecules, episomes, or proteins for reprogramming, which will increase the safety of the generated iPSCs (Zhou et al. 2009; Lin et al. 2009; Okita et al. 2008). Human iPSCs (hiPSCs) have been successfully used to generate numerous cell types including cardiomyocytes (Karakikes et al. 2015; Zhang et al. 2009). More importantly, hiPSCs provide a novel platform to dissect the underlying mechanisms of disease in patients (Bellin et al. 2012; Wang et al. 2014; Davis et al. 2012). Furthermore, intramyocardial transplantation of hiPSC-derived cardiomyocytes (hiPSC-CM) in a large animal model following MI led to a significant improvement of ventricular function and reduction of scar size while abrogating ventricular arrhythmias (Ye et al. 2014). Although further studies are required to truly understand the optimal way to use hiPSC-CMs, the recent developments indicate that this approach holds significant promise for future cell therapy (Okano et al. 2013).

1.2.7.2 Direct Reprogramming into Cardiomyocytes

Reprogramming fibroblasts into pluripotent cells led to a race toward reprogramming one cell type to another differentiated cell type. Transdifferentiation of fibroblasts to cardiomyocytes is an appealing approach as it could use the fibroblasts in the scar region to generate new myocardium. Using multiple combinations of cardiac transcription factors, a combination of three transcription factors, GATA4, MEF2C, and TBX5 (known as GMT), was able to reprogram mouse fibroblasts into induced cardiac-like myocytes in vitro (Ieda et al. 2010). To further examine whether direct reprogramming can occur in vivo, GMT and GHMT (H for HAND2) retroviral injections successfully reprogrammed cardiac fibroblasts into cardiomyocytes in vivo that resulted in an improved cardiac function and reduced scar following MI (Qian et al. 2012; Song et al. 2012). Furthermore, reprogramming fibroblasts into cardiomyocytes was achieved using microRNAs both in vitro and in vivo with improved cardiac regeneration (Jayawardena et al. 2012, 2015). Similar to iPSCs, reprogramming fibroblasts into cardiomyocytes occurs at low efficiency and may lead to the formation of immature cardiomyocyte-like cells rather than bona fide mature cardiomyocytes. Interestingly, recent studies showed that reprogramming efficiency could be enhanced significantly via upregulation of Akt1, as well as through inhibition of pro-fibrotic signaling (Zhou et al. 2015; Zhao et al. 2015). Further studies are necessary in order to generate mature and functional cardiomyocytes, as well as to fully understand the molecular mechanisms of reprogramming before moving forward to the clinic.

1.3 Future of Cell Therapy

Cardiac cell therapy has witnessed enormous achievements over the past decade. Cardiac cell therapy appears to be safe, with minimal adverse effects, while showing potential therapeutic benefits. However, cardiac cell therapy is not yet a clear success, as some analyses revealed no therapeutic benefit in acute myocardial infarction patients (Gyongyosi et al. 2015; Fisher et al. 2015). There is no consensus on which cell type will prove to be most effective. A recent study aimed at comparing hESC-CMs, cardiovascular progenitors (CVPs), and bone marrow mononuclear cells in a nude rat model of myocardial infarction (Fernandes et al. 2015). Interestingly, hESC-CMs and CVPs showed comparable improvement of cardiac repair, while bone marrow cells were less efficient (Fernandes et al. 2015). Comparing different cells is an important step in order to understand the optimal cell therapy for humans.

The results from large outcome trials are highly anticipated in order to determine the impact and the usefulness of cells in cardiac therapy. Although the mechanism of action of different cell types may vary, whether through direct differentiation into new myocardium, neovascularization, or paracrine effects, we need to expand our understanding at the molecular level. Clinical trials are the gold standard for assessing any treatment, but owing to the controversies within the cell therapy field, it is important to take a step back and progress at both the bench and the bedside. The lessons learned from heart regeneration in lower vertebrates and neonatal mice should improve our understanding of the promising approaches to regenerating the adult heart. Cardiac cellular therapy does lead to the complete regenerative response seen in animal models of endogenous heart regeneration, and thus there are many lessons to be learned from nature.