Cardiovascular disease remains the number one, non-communicable killer disease, which recorded a mortality rate of 17.5 million in 2012, and was accounted for 46.2% of all reported deaths worldwide in 2014 (1). Myocardial infarction (MI) is a common cause of heart failure (HF) due to a consequence of partial or complete occlusion of the coronary artery, which diminishes the delivery of oxygen and nutrient supply to the myocardium where the vessel serves (2). Approximately 25% of myocardial infarcted patients suffer from severe left ventricular dysfunction and are at risk of progressive heart remodeling (3). Conventional pharmacological approaches with drugs, such as thrombolytic agent, β-blocker, and angiotensin-converting enzyme inhibitor, is often the first non-invasive treatment option offered to patients. However, in more severe cases, ST-elevated myocardial infarction (STEMI), a more invasive balloon angioplasty, and stent insertion may be recommended to achieve myocardial reperfusion. Highly invasive coronary artery bypass grafting procedure is only recommended if severe, irreversible coronary occlusion is evident. These approaches had shown to alleviate the symptoms of the disease and improve the patients’ quality of life. Nevertheless, none of these therapies were able to remove the fibrotic scar or replace the lost myocardium with new functional cardiomyocytes. The presence of the akinetic tissue restricts the overall cardiac performance, forcing the remaining myocytes to increase contractility to maintain adequate cardiac output. These events trigger abrupt alterations in cardiac architecture and cause cardiomyocyte hypertrophy, further myocyte loss, thinning of the ventricular wall, weakening of contractility, and an eventual cease in function of the cardiomyocytes (4). To date, heart transplantation is the only curative option. Although there are survivors from successful heart transplantations, the long waiting time, high patient-to-donor ratio, high incidence of post-procedural complications, and limited number of transplantable hearts prompt an urgent need for an alternative solution. Stem cell-based therapies are fast becoming an attractive and highly promising treatment for heart disease and failure. The most common types of stem cell candidates, which had been tested in clinical trials thus far, are derived mainly from the bone marrow. In this review, we will discuss the basic discovery and current progress of the candidate cells in human cardiac regenerative therapy, and the potential to combine multiple cell types for regenerating complex components that make up the myocardium (Figure 1). Finally, we touch on an emerging prospective application in heart tissue engineering.

The discovery of recipient-derived cardiomyocytes in sex-mismatched donor hearts after bone marrow transplants spiked the interest of using bone marrow cells for cardiac cell therapy (5–7). Bone marrow mononuclear cells (BMNCs) were the first hematopoietic cells selected for this purpose, because of their availability and feasibility to be isolated from patients through bone marrow aspiration (8). In fact, the in vitro procedure involved minimal manipulation for clinical transplantation, making it the most favorite cell candidate in initial cardiac repair clinical trials. Nevertheless, most clinical studies observed a marginal, yet clinically significant, improvement in cardiac function after injection with BMNCs (Table 1). Despite evidences that showed the BMNCs contribute to angiogenesis (9) and neovascularization (10) by secreting paracrine factors, their capability of cardiomyogenic differentiation in vivo remains skeptical. The earliest study, where lineage-negative (Lin⁻), c-kit-positive (c-kit⁺), EGFP + HSCs were injected into the contracting wall bordering the infarct in mice, showed newly formed myocardium, comprised cardiomyocytes and vasculature, occupying 68% of the infarcted portion of the ventricle 9 days after transplanting the bone marrow cells (11). These findings failed to be replicated by others. Murry et al. (12) tracked the fate of HSCs (c-kit⁺, Lin⁻) after 145 transplants into normal and injured adult mouse hearts and found no trans-differentiation of HSCs into cardiomyocytes (12). Moreover, Balsam and colleagues showed that when GFP⁺Lin⁻c-kit⁺ HSCs were injected into infarcted mouse hearts, abundant GFP⁺ cells were detected in the myocardium at 10 days, with few cells detectable at 30 days (13). It was found that the GFP⁺ cells did not express cardiac tissue-specific markers, but expressed the hematopoietic marker CD45 and myeloid marker Gr-1, representing mature hematopoietic fates.

More recently, van Berlo et al. (40) generated c-kit^cre-IRES-eGFP knocked-in mice to revisit the fate of c-kit⁺ cells in development and following injury (40). They found that most eGFP-c-kit⁺ cells were mainly non-myocytes in the developing and injured adult heart. Indeed, c-kit⁺ cells largely adopted an endothelial lineage phenotype in the developing or infarcted heart, and rarely became cardiomyocytes (41, 42). While these models set out to tag all c-kit⁺ cells in the organism, questions were raised over the fidelity of the model and reporter gene to successfully recombine the endogenous, resident cardiac stem, and progenitor cells, which also express c-kit (43).

Mesenchymal stem cells, or also known as mesenchymal stromal cells (MSCs), are a subset of bone marrow-derived stem cells that have plastic adherence characteristics, express CD105, CD73, and CD90 but not CD34, CD45, CD14 or CD11b, CD79α or CD19, and HLA-DR, and possess the ability to form adipocytes, chondrocytes, and osteoblasts in vitro (44). As MSCs express low MHC Class I and are lacking MHC Class II (45), the phenotype confers the capability of evading host immune responses and hence enables the cells for allogeneic transplantation (45). Several in vivo studies showed improvements in myocardial function despite low rates of MSC engraftment and differentiation (46, 47). Although trans-differentiation of MSCs into cardiomyocytes was achievable by using demethylating chemicals (48, 49) or by coculturing with rodent myocytes in vitro (50, 51), the event in vivo had been reportedly low (52). Furthermore, electrophysiological analysis revealed that differentiated myocytes did not possess similar electrical properties to a functional cardiomyocyte (53). Hence, the main regenerative function of MSCs was largely confined to its secretome, which contained a plethora of factors with cardioprotective effects, or stimulants that activate endogenous repair mechanisms including the resident cardiac stem and progenitor cells (54, 55).

Many trials had been conducted to examine the therapeutic efficacy of MSCs in regenerating damaged human hearts at different severities, either with autologous or allogeneic cell sources (Table 2). In POSEIDON, transendocardial-administered allogeneic BM-MSCs attenuated the progressive heart remodeling, reduced the scar mass, and improved the early enhancement defect and sphericity index in ischemic cardiomyopathic patients, and the effects were greater with a lower cell dose (20 million), as compared to a higher dose (200 million) (56). The injected allogeneic MSCs did not trigger immune responses in recipients, and the observed benefits were mostly similar to autologous MSCs (56). However, both allogenic and autologous MSC-treated groups did not show significant improvements in ejection fraction. In contrast, the phase 2, placebo-controlled randomized MSC-HF trial reported encouraging results, which demonstrated that HF patients who received a high number of intramyocardially delivered autologous MSCs showed greater functional improvements in the ischemic heart after 12 months (57). They also suggested a possible correlation between cell dose and disease severity. Through a longer, 2-year follow-up, the phase 1 pilot study MESAMI revealed similar benefits from intramyocardial MSC injection in patients with chronic ischemic cardiomyopathy, albeit with a smaller sample size of 10 (58).

Cardiospheres and CDCs represent a mixed cell population, which employs an assortment of heterogeneous cells and this heterogeneity sparks the idea of employing synergistic effects between various cells to aid CSCs to perform better for cardiac regeneration.

Mononuclear bone marrow cells had been shown to benefit the injured myocardium after their administration, but the effect was then concluded as not sustainable. Paracrine signaling is a generally accepted explanation for the mechanism of repair, regeneration, and modest improvement in cardiac function. Loffredo et al. (86) conducted a sophisticated experiment using bitransgenic MerCreMer ZEG mice to study the degree of new myocyte formation after induced injury in the heart, following BMDC transplantation (86). In this model, all cardiomyocytes were permanently shifted to express GFP from β-galatosidase (β-gal) by a pulse treatment of 4-OH tamoxifen, and all new myocytes were identified as non-GFP expressing β-gal-positive cells. The study revealed that the number of new myocytes was greater in subjects treated with c-kit⁺ bone marrow MNCs. This coincided with increased resident GATA4⁺Nkx2.5⁺ cardiac progenitors, which was not observed when the subjects were given bone marrow MSCs. Moreover, Hatzistergos et al. (54) used GFP-transduced MSCs and when transendocardially injected them into the infarcted heart of the Yorkshire swine and showed increased GFP⁻ c-kit cells by 2- and 15-fold in the infarcted and border regions, respectively (54). These cells coexpressed MDR-1 and GATA4, suggesting that they were of endogenous CSC origin. These findings were consistent with the in vitro data which showed that greater c-kit⁺ CSCs were mobilized from heart explant cultures in the presence of MSC feeder layers (87). In addition to the activation of the endogenous pool of CSCs, MSCs prompted cardiomyocyte proliferation, which correlated with an increased number of cardiomyocytes expressing serine 10 phosphorylated histone H3, a mitotic marker indicative of cell cycling (54). The regenerative capability of MSCs was further confirmed with a study by Suzuki et al. (88), which discovered the ability of MSCs in mobilizing CD133 and c-kit⁺ bone marrow cells as well as stimulating myocyte proliferation in chronic hibernating myocardium (88). Indeed, the administration of MSCs was found to drive the increment of c-kit⁺/CD133⁻, c-kit⁺/CD133⁺ progenitor cells, and Ki67⁺ and phospho-histone H3⁺ cardiomyocytes.

The synergistic effects between MSCs and heart-derived, resident c-kit⁺ CSCs were further confirmed in two studies where cotransplantation of both cell types showed greater amelioration in improving cardiac performance and scar size post-infarction (89, 90). Both transepicardial and transendocardial administrations of either xenogenic or autologous MSCs showed greater scar reductions and global heart function restorations as compared to single-cell administration in swine model, which illustrates the interaction between MSCs and CSCs in enhancing the regeneration of the heart post-infarction.

Bone marrow-derived stem cells remain the most common, first-generation cell candidate used in clinical transplantation. A striking report by Nowbar et al. (107), who conducted a weighted regression and meta-analysis to study 49 trial reports using autologous bone marrow stem cells and outlined the discrepancies between these trials, concluded that only 10% of the human studies were performed without errors with none showing benefits from BMNCs (107). In contrast, Fisher et al. (108) performed a systemic review that excluded all the non-randomized trials, of which contributed to the majority of discrepancies outlined in Nowbar’s report, and suggested that autologous bone marrow stem cell treatment can improve HF patients’ quality of life and exercise capacity (108). Their findings are in line with a recent meta-analysis which included 48 randomized-controlled trials by Afzal et al. (109) and also agreed that bone marrow-derived cells (both BMNCs and MSCs) improved heart function in ischemic heart disease patients (109). Nonetheless, it is widely accepted that the therapeutic benefits from bone marrow-derived cells are mainly attributed to a paracrine mechanism that activates endogenous healing. Reconstituting injured myocardium with cardiomyocytes may require second-generation cardiogenic cells, the more defined, homogeneous cardiac-derived stem/progenitor cells or pluripotent stem cells, some of which have been used for clinical trials (73, 74, 76, 77, 110) (Figure 2). Careful selection of cell candidates, mode of delivery, employment of cell engraftment and enhancement strategies, in-depth investigation of mechanisms of efficacy, and clinically meaningful endpoints in future experimental studies can help to advance cardiac cell therapy (111).

Although several stem cells have been proposed to regenerate the heart, there is no consensus on the best cell type to be used in cellular therapy and the search for establishing a gold standard is still ongoing. Given the complexity of the heart, and the emptiness of the infarcted area, the regeneration process will require multiple coordinations from different therapeutic cells with synergistic functions, together with an established extracellular matrix scaffold. Some in vivo studies have investigated these approaches, but it has not been widely explored. It is important to realize that most experiments are conducted in two-dimensional culture systems and little is known about the survival and performance of these interactions in the three-dimensional structure. These questions lead back to the fundamental investigation of determining the optimal cell types for the engineering of tissue constructs, and their functional behaviors in three-dimensional cultures. Ott et al. (112) demonstrated a new concept of producing bio-engineered hearts by using the natural hearts from rats (112) by decellularizing the heart scaffold using detergents, then re-cellularizing through introducing neonatal cardiac cells and endothelial cells (112).

With the invention of induced-pluripotent stem (IPS) cells, the mass generation of human cardiomyocytes is no longer difficult. The challenging aspect, however, is reintroducing the cells into the construct, finding the means of ensuring their long-term survival and identifying the factors that drive their maturation. Lei Yang’s laboratory generated cardiovascular progenitors from IPS cells and attempted to reintroduce these cells into the decellularized mouse heart scaffold. The group demonstrated ex vivo proliferation, migration, and differentiation in the three-dimensional construct, but failed to regrow the myocardium to acquire sufficient strength for pumping fluid like the native heart (113). Another similar study by Guyette et al. (114) repopulated decellularized human hearts with cardiomyocytes derived from IPS cells. This study showed that the cardiomyocytes successfully engrafted onto the cardiac scaffolds and showed electrical conductivity and thus set the ground for the translational value of using acellular human heart matrix for complete myocardial regeneration in the future (114). A complex three-dimensional construct is an extremely promising approach for heart regeneration. However, the research is still in its infancy, and more studies are required before this technique can be translated into clinical applications.

JT and GE-H contributed to the conception and design of the review. YL, WN, and JT prepared, drafted, and wrote the manuscript. JT and GE-H wrote, critically revised, proofread, and approved the manuscript.

JT received a research grant from CryoCord Sdn Bhd. All funders have no role in manuscript writing and funding this project. The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.