Neovascularization: The Main Mechanism of MSCs in Ischemic Heart Disease Therapy

Mesenchymal stem cell (MSC) transplantation after myocardial infarction (MI) has been shown to effectively limit the infarct area in numerous clinical and preclinical studies. However, the primary mechanism associated with this activity in MSC transplantation therapy remains unclear. Blood supply is fundamental for the survival of myocardial tissue, and the formation of an efficient vascular network is a prerequisite for blood flow. The paracrine function of MSCs, which is throughout the neovascularization process, including MSC mobilization, migration, homing, adhesion and retention, regulates angiogenesis and vasculogenesis through existing endothelial cells (ECs) and endothelial progenitor cells (EPCs). Additionally, MSCs have the ability to differentiate into multiple cell lineages and can be mobilized and migrate to ischemic tissue to differentiate into ECs, pericytes and smooth muscle cells in some degree, which are necessary components of blood vessels. These characteristics of MSCs support the view that these cells improve ischemic myocardium through angiogenesis and vasculogenesis. In this review, the results of recent clinical and preclinical studies are discussed to illustrate the processes and mechanisms of neovascularization in ischemic heart disease.


INTRODUCTION
Ischemic heart disease (IHD) is characterized by reduced blood supply to the heart and is the leading cause of death and disability worldwide. Long-term myocardial ischemia and acute massive myocardial infarction often result in decreased left ventricular function. Although the development of new drugs and the use of stent implantations have benefited numerous patients with coronary heart disease, some patients still have no effective treatment due to issues associated with diffuse coronary artery lesion, postoperative restenosis and heart failure after myocardial infarction (MI).
The foundation of IHD treatment is the reconstruction of vessels and the recovery of blood flow. Over the past decades, with the introduction of the concept of therapeutic angiogenesis, more and more studies have demonstrated that neovascularization can effectively improve the blood supply of ischemic myocardium. There are two primary mechanisms by which neovascularization occurs: vasculogenesis and angiogenesis. Vasculogenesis is the in situ assembly of endothelial progenitors into capillaries, while angiogenesis is a process through which new blood vessels form from pre-existing vessels through sprouting and intussusception (1). Cytokine-based therapeutic angiogenesis from the bench to clinical trials has been a major focus of medical research, and the efficacy of vascular endothelial growth factor (VEGF) blockers has led to the approval of anti-angiogenesis drugs for cancer and eye disease. Conversely, the use of angiogenesis factors, such as VEGF and basic fibroblast growth factor (bFGF), has been shown to promote notable increases in collateral vessel and myocardial perfusion in ischemic myocardium, reduced infarct size and improved cardiac function (2), demonstrating the theoretical and experimental promise of this approach in treating ischemic diseases. Unfortunately, despite the exciting results obtained using angiogenesis factors to treat IHD, gene therapy is also limited by its restricted efficacy and resistance (3). For example, VEGF also accelerates angiogenesis in atherosclerotic plaques and promotes plaque growth, which may eventually lead to plaque instability, while it promotes angiogenesis in ischemic tissue, an observation referred to as the famous Janus phenomenon (4). Angiogenesis greatly improves blood flow in myocardial ischemia, but the safety of growth factor-based angiogenesis therapy is an issue that remains to be overcome. Thus, how to avoid the risks associated with angiogenesis therapy is a problem that must be considered.
Stem cell-based therapies provide a promising new method for the formation of new blood vessels. MSCs have become the most promising seed cells for the treatment of IHD, with advantages of rapid self-renewal, multidifferentiation potential, and weak immunogenicity in autologous transplantation. Clinical and preclinical studies have shown that MSCs therapy effectively limits the infarcted area and improves heart function. However, the mechanisms associated with the activities of MSCs in IHD therapy remain controversial. We primarily attribute the cardiac protective effect of MSCs to their ability to promote neovascularization for the following two reasons. First, MSCs secrete soluble paracrine factors that contribute to angiogenesis and vasculogenesis. Second, MSCs are able to differentiate into ECs, pericytes and smooth muscle cells (SMCs), which form the foundation of vessels, processes that both participate in the protective ability of MSCs toward IHD. In this review, we focus on the mechanisms and clinical applications of MSCs in IHD therapy through neovascularization to provide reference for the application of stem cells in IHD.

COMPARISON OF MSCS FROM DIFFERENT SOURCES
MSCs can be isolated from bone marrow, adipose tissue, umbilical cord blood, peripheral blood and almost every tissues in adults. Although MSCs can be harvested from different sources, regardless of their origin, they all have the capability of differentiating into adipocytes, osteoblasts and chondroblasts in vitro under specific conditions and can adhere to plastic under culture conditions. Furthermore, the surface of MSCs displays CD73, CD90, and CD105 but lack CD34, CD45, HLA-DR, CD14 or CD11b, CD79a or CD19. The International Society for Cell Therapy proposed the three criteria described above as identification standards for MSCs (5). Although MSCs from different sources share many of the same biological features, there are also some differences between distinct MSC populations. Bone marrow-derived MSCs (BMSCs), adipose-derived MSCs (AMSCs) and umbilical cord-derived MSCs (UCMSCs) are the

Differentiation Capacity
MSCs have the ability to differentiate into adipocytes, osteoblasts and chondroblasts. The amount of calcium deposits and sulfated proteoglycans stained by Alizarin red and Alcian blue, respectively were both higher in BMSCs than that observed in AMSCs, indicating that BMSCs have a higher capacity toward osteogenic and chondrogenic differentiation than AMSCs. While similar adipogenic differentiation potential was observed between these two types of cells (6), some studies have reported that AMSCs are more prone to adipogenic differentiation than BMSCs (7). Baksh et al. (8) observed that compared to BMSCs, UCMSCs underwent osteogenic differentiation more rapidly, exhibited higher alkaline phosphatase activity, and generated significantly more fat-containing cells when grown under adipogenic conditions by day 21. The differentiation ability of stem cells is affected by donor sex, age, isolation and culture conditions, etc. (9). Thus, which types of MSCs have a greater ability to differentiate into adipocytes, osteoblasts and chondroblasts remains disputed. In addition, MSCs also have the ability to differentiate into ECs, pericytes and SMCs, which are necessary components of blood vessels (10)(11)(12)(13). Lu et al. (14) showed that MSCs from adipose tissue may have significantly greater ability to promote angiogenesis both in vitro and in vivo than UCMSCs and endometrial MSCs.  (6). Therefore, the proliferative ability of UCMSCs is the strongest, followed by AMSCs and BMSCs.

Migration Capacity
MSCs play an important role in posttraumatic tissue repair and cell therapy, and their migration ability is a key factor affecting their therapeutic efficacy. The migration capacity of BMSCs and placenta-derived MSCs (PMSCs) was observed to be 5.9and 3.2-fold higher than that of UCMSCs, respectively. These results were consistent with the observed levels of migrationenhancing proteins in UCMSCs, including cathepsin B, cathepsin D and prohibitin, which were significantly lower than those observed in BMSCs and PMSCs, while the levels of migrationinhibiting proteins such as plasminogen activator inhibitor-1 and manganese superoxide dismutase were higher (16). Vimentin also contributed to the higher migration capability of BMSCs than UCMSCs (17). In contrast, UCMSCs exhibited an enhanced migration capacity toward factors released by hepatocellular carcinoma compared with BMSCs (18).

PARACRINE FUNCTION OF MSCS THROUGHOUT THE NEOVASCULARIZATION PROCESS
The mechanism of MSC therapy is still controversial, because few MSCs can be found in myocardium after injection in vivo study. Wang et al. (25) showed that most intravenously injected MSCs remain in the lungs and liver, with only a small portion reaching the myocardial tissue. Similarly, Uemura et al. (26) observed only a few GFP-labeled MSCs in the periinfarct myocardium. Even so, clinical and preclinical studies still indicated the cardiac function of ischemic heart was improved, and infarct size and the number of apoptotic cardiomyocytes were significantly reduced after MSCs intervention. Which suggested the efficacy of MSCs did not benefited from themselves in some degree. Paracrine hypothesis was firstly advanced by Gnecchi et al. (27). They found genetically modified BMSCs overexpressing the Akt1 released paracrine factors that exert cytoprotective effects on cardiomyocytes exposed to hypoxia and limited infarct size and improved ventricular function (27, 28). Furthermore, high VEGF, bFGF, IGF-1 and SDF-1 expression in hypoxiapreconditioned MSCs medium was examined, the results of which indicated that the paracrine function of MSCs may play more important role than their differentiation ability (26). Recently, it has been reported that MSCs secreted a wide array of cytokines that exerted beneficial angiogenesis in ischemic tissue, including PDGF, thrombopoietin, and angiogenin (29, 30). These factors are all involved in the neovascularization process, including MSC mobilization, migration, homing, adhesion and retention, and the differentiation of ECs. Especially VEGF and bFGF, which both have high affinity toward heparin and participate in angiogenic processes such as migration and amplification of ECs, are also necessary substances to induce the transformation of stem cells into ECs (10, 31). MSCs overexpressing Akt and angiopoietin-1 showed higher Flk1 and Flt1 positivity and promoted intrinsic Flk1 + and Flt1 + cell mobilization into the infarcted heart (32). Huang et al.
(33) observed that overexpression of miR-126 promoted the differentiation of MSCs toward ECs through activation of the PI3K/Akt and MAPK/ERK pathways and the release of VEGF and bFGF factors. Therefore, paracrine factors secreted by MSCs may have pivotal functions throughout the neovascularization process. The role of various secretory factors in the neovascularization process will be discussed below.

INVOLVEMENT OF MSCS IN THE NEOVASCULARIZATION PROCESS
The process by which MSCs promote neovascularization involves in a number of steps. First, once ischemia occurs which also follows stress change, MSCs can perceive the associated changes and are mobilized from their niches to migrate and adhere to ischemic tissue to proliferate and differentiate. Notably, MSCs secrete various factors, including chemokines and growth factors, and this paracrine function is carried out throughout the neovascularization process (Figure 1). The completion of all biological processes depends on the cooperation of different types of cells, and the neovascularization process requires the collaboration of ECs, endothelial progenitor cells (EPCs) and pericytes. In addition, exosomes derived from MSCs act as a messenger that participate in cell-to-cell communication.

Perception of Hypoxia by MSCs
Despite the benefits of MSC transplantation in cardiac tissue, detailed in vivo observations have shown that MSCs only survive for a brief period after engraftment due to harsh microenvironmental conditions (including ischemia, inflammation and anoikis) in the infarcted myocardium (34). Once tissue is damaged, signals, including those involve in ischemia-associated pathways, cell-cell interaction and stress mobilize stem cells to migrate from the stem-cell niche to damaged tissues, where they adhere, self-renew and differentiate. Once ischemia occurs, MSCs have the ability to secrete a number of growth factors through their paracrine function to promote new tube formation of ECs to provide new blood for ischemic tissue.
However, this environment contributes to the mobilization of MSCs from their niches.
MSCs originated from the bone marrow microenvironmental niche exhibit low oxygen tension. O 2 is a necessary factor in the maintenance of cell life as the final receptor in the intracellular aerobic respiration electron transport chain and is a substrate of some enzymes. Once the supply of O 2 is insufficient, the hypoxia signal will be rapidly transmitted to nucleus and initiate related gene expression to maintain oxygen homeostasis and the balance of energy metabolism between the cells and organism. Hypoxia inducible factor 1 (HIF-1), which has a dimeric complex composed of HIF-1a and HIF-b subunits, is oxygen-sensitive and the most important transcription factor affecting gene regulation under hypoxia (35). Once ischemia occurs, HIF-1 increases the expression of angiogenesis-associated genes, including VEGF, its receptors Flk-1 and Flt-1, bFGF and the fibrinogen system (36, 37). At the same time, HIF-1 improves the expression of proteases, such as membrane type matrix metalloproteinases, which hydrolyzes extracellular protein to promote cell migration, matrix reconstruction and the formation of tubule-like structures (38).
Hypoxia is also a basic aspect of the microenvironment that determines the differentiation of MSCs. Compared with a normoxia group, VEGF expression in embryonic and MSCs under hypoxia was observed to be significantly increased (39-41). Likewise, the in vivo administration of hypoxia-inducible VEGF-engineered MSCs was shown to induce ischemiaresponsive VEGF production and lead to a significant increase in myocardial neovascularization after myocardial infarction in rats (39).

Cell-Cell Interactions
In 1997, Asahara et al. (42) identified and named a small population of CD34 + cells as "EC progenitors." Indeed, EPCs are involved in a number of processes during angiogenesis, including mobilization, differentiation into ECs, homing, paracrine function and others (43, 44). Coculture of EPCs and MSCs significantly increased the transcription levels of endothelial specific markers, including vWF, CD31, VEcadherin, Flk-1 and Flt-1 (45) and enhanced tube-like formation (46) through platelet derived growth factor (PDGF), Notch and TACE/TNF alpha signaling (45, 47). Joensuu et al. (48) noted that in cocultures of human MSCs and peripheral blood mononuclear cells, the previously nonadherent cells attached and started to elongate and form tube-like structures within 1 week concomitant with VEGFR1 upregulation, and platelet endothelial cell adhesion molecule 1 (PECAM-1) and endoglinpositive vessel-like structures were observed after 20 days. In addition, MSC-EC interactions were observed to decrease endothelial permeability induced by lipopolysaccharide through hepatocyte growth factor (HGF) by restoring the integrity of endothelial monolayers and remodeling endothelial intercellular junctions (49). VEGF secreted by stem cells from apical papilla is also used by human umbilical vein endothelial cells to increase the number of endothelial tubules, tubule lengths, and branching points (50).

Laminar Shear Stress and Pulsatile Stress
There are many force-sensitive molecules on the cell surface, such as cilia, integrins, ion channels and plaque proteins. Integrins connect the cytoskeleton and extracellular matrix through adhesive plaque and transform the force signals into intracellular biological signals through this plaque (51, 52). Considering the key role of shear force in the differentiation of ECs, researchers reported that such mechanical stimulation in cell culture in vitro was equally effective for the differentiation of stem cells into ECs (53). MSCs are highly reactive to mechanical stimuli in the environment, and different types of stress on the same MSC population will lead to different differentiation results (54). After generating canine BMSCs under shear stress provided by a pulsatile bioreactor for 4 days, the expression of endothelial cell markers, such as PECAM-1, VE-cadherin and CD34 was observed to be significantly increased (55). Fisher et al. (56) noted that AMSCs could form cords but failed to take up acetylated low density lipoprotein (acLDL) or express molecular markers after being cultured in endothelial cell growth supplement. Only the subsequent exposure of stem cells to shear stress did the cells exhibit realignment, acLDL uptake and CD31expression, indicating that stem cells differentiation to ECs requires the synergism of biochemical and shear force.

Dynamic Process of MSCs to Repair Ischemic Tissue
Mobilization MSC mobilization is key for its involvement in tissue repair following their sensing of hypoxia, stress or other signals. An anoxic environment is one of the factors that induces stem cells to migrate out of their niches. Prolyl hydroxylase (PHD) and factor inhibiting HIF-1 (FIH) are key oxygen sensors in MSCs. HIF-1α upregulation by double knockdown of PHD and FIH synergistically increases stem cell mobilization and myocardial angiogenesis and improves cardiac function (57). The high concentration of growth factors outside of stem-cell niches may be another factor causing MSCs to mobilize from their original niches. Stromal cell-derived factor-1 (SDF-1 α)/Cxc chemokine receptor 4 (CXCR-4) are part of the most important chemotactic axis regulating MSC mobilization and migration. VEGF and insulin-like growth factor-1 (IGF-1)-overexpressing MSCs accelerate BMSC mobilization via the activation of SDF-1 α/CXCR4 signaling to promote myocardial repair (58, 59). Wan et al. (60) showed that active transforming growth factor β (TGF-β) also control the mobilization and recruitment of MSCs to participate in vascular repair. In addition, high-intensity exercise may be a potent stimulus that promotes circulating mesenchymal cells mobilization in patients with stable coronary artery disease (61).

Migration and Homing of MSCs
Homing and migration comprise a key step after MSC mobilization. Microenvironmental interactions between hypoxia and MSCs may control the ability of MSCs to migrate and their migration direction. In hypoxic tissue, SDF-1 and CXCR-4 are also important factors for cell migration. Ischemic myocardial and vascular tissues secrete SDF-1 to attract CXCR-4-expressing cells, particularly their therapeutic progenitors. Yu et al. (62) showed that SDF-1/CXCR-4 may mediate the migration of BMSCs toward heart MI through activation of PI3K/Akt signaling. Growth factors play an important role in the process of MSC migration. The stimulation of SDF-1α expression in infarcted hearts by VEGF-overexpressing MSCs was observed to result in the massive mobilization and homing of BMSCs (59). TGF-β1, HGF, IGF-1 and endothelial nitric oxide synthase (eNOS) also promoted the migration and homing BMSCs to the ischemic myocardium (63-65). Schmidt et al.

Proliferation and Survival
Although MSCs transplantation is a promising therapeutic approach for IHD, the low viability of MSCs after transplantation needs to be improved. Hypoxic preconditioning may improve the functional survival and therapeutic efficiencies of engrafted BMSCs, at least in part through autophagy regulation (68). Some growth factors, including increased VEGF, TGF-β, IGF-1, SDF-1a and angiogenin were shown to enhance MSC survival and vasculogenesis in an MI model (69). Preconditioning with other factors, such as protein kinase C epsilon (εPKC), CTRP9, dimethyloxalylglycine and connexin-43 improves the retention and survival of transplanted MSCs in rat MI through the SDF-1/CXC and PI3K/AKT pathways (66, 70-72). Qu et al. (73) showed that atorvastatin, a hypolipidemic agent, has a protective effect on cardiomyocytes against apoptotic cell death in infarct and peri-infarct areas and could also increase the survival rate of implanted BMSCs in acute myocardial ischemia.

Adhesion and Retention
MSCs need to stay and adhere to ischemic tissue to play their important role in ischemic tissue repair. Molecular imaging studies have shown that <5% of MSCs engraft in ischemic tissues after being intravenously injected, with most of them dying within few hours after administration (74). This poor engraftment may be attributed to the constant blood flow and the harsh environmental conditions after acute ischemic injury. Since the long-term efficacy of cell therapy is proportional to the number of retained cells, this low retention and viability needs to be improved. Increasing the ability of MSCs to adhere to the ischemic tissues is key to improving their retention and viability. IGF-1 can increase the adhesion of MSCs and prolong their survival under hypoxia in vitro through PI3K activation (75). MSC adhesion can also be promoted by increasing the expression of integrin-linked kinase, periostin, and 2, 4-dinitrophenol (76)(77)(78). Reactive oxygen species (ROS) inhibit the cellular adhesion of engrafted MSCs, indicating that the elimination of ROS may be a novel strategy for improving the survival of engrafted MSCs

INTERACTIONS AMONG MSCS AND ECS, EPCS AND PERICYTES
Multiple cell types are known to be involved in the processes of angiogenesis and vasculogenesis, including MSCs, ECs, EPCs and pericytes. In particular, ECs are indispensable for angiogenesis and the relationship between MSCs and ECs mainly attribute to the following aspects. First, MSCs secret growth factors which can repair the injured but not dead ECs through their paracrine function. Second, the interaction and crosstalk between MSCs and existing ECs promotes the formation of new ECs and the repair of injured ECs. Third, MSCs have the potential to differentiate into new ECs although it is controversial. Last, MSCs also interact with EPCs and pericytes to influence EC formation and function (Figure 2). The ability of MSCs to limit infarct size may attributed to their pro-angiogenesis activity through existing ECs (Figure 2). The stimulated angiogenic activity of ECs is associated with the secretion of various growth factors and cytokines, including VEGF, HGF, IL-6, TGF-β1 and monocyte chemoattractant protein-1 (81). Lu et al. (82) showed that nestin(+) BMSC transplantation improved cardiac function in a mouse AMI model by recruiting resident cardiac ECs to the infarcted border region. BMSCs also rescued injured ECs through modulation of mitophagy or activation of signaling pathways such as PI-3K/AKT/m-TOR/eNOS and p38/MAPK (83,84). Hypoxia also influences the interactions between the endothelium and MSCs (85).
In addition to the interactions between MSCs and ECs, studies showed MSCs had the ability to differentiate into ECs to promote angiogenesis in some degree although it was still controversial. For example, Otto et al. (86) did not observe MSC transdifferentiation into cardiomyocytes, ECs or SMCs and that the transdifferentiation of MSCs into cardiomyocytes or vascular cells did not significantly contribute to the improvement of cardiac function. Conversely, Silva et al. (87) showed that BMSCs promoted the angiogenesis of dog ischemic myocardium by differentiating into ECs, which accelerated the establishment of collateral circulation. Studies support MSCs own the potential to differentiated into ECs according to the below reasons. First, MSCs are multipotent stem cells derived from the mesoderm. Theoretically, MSCs can be differentiated into all mesoderm derived cells, and since ECs are mesoderm-derived cells, MSCs have the potential to differentiate into ECs. Second, MSCs express molecular markers of early ECs, such as VEGF receptor 2 (VEGFR-2/Flk-1/KDR) and bFGF, indicating that ECs can be derived from mesenchymal colonies and that MSCs arise from precursors with angiogenic potential (31, 88). Last, a series of in vivo and in vitro experiments proved that MSCs can differentiate into ECs. Oswald et al. (10) successfully used 2% fetal bovine serum supplemented with 50 ng/mL VEGF to induce BMSCs to differentiate into ECs in vitro. They observed that differentiated cells increased the expression of endothelialspecific markers, such as KDR and VEGF receptor 1 (VEGFR-1/Flt-1), and formed capillary-like structures. Furthermore, the process of MSC differentiation into ECs may require the synergy of bFGF, IGF, epidermal growth factor (EGF) (89,90). ERK signaling may also involve in the differentiation of porcine AMSCs into ECs (90).
Furthermore, MSCs also function with EPCs to promote tissue repair. As a precursor of ECs, EPCs also differentiate into ECs and promote ischemia angiogenesis through their paracrine function (91,92). MSCs could attract and promote the migration and vascularization of EPCs, which may depend on a positive feedback loop between CXCR-2 and CXCR-4 (93,94). The viability and ability of MSCs to promote nerve regeneration is also improved by EPCs through PDGF-BB/PDGFR-β signaling (95). Rossi et al. (96) found MSCs and EPCs into the hind limbs of ischemia model together accelerated ischemic muscle recovery through an endoglindependent mechanism. Consequently, MSCs, ECs and EPCs may have a synergistic effect in ischemic tissue repair (Figure 2).
Pericytes, also known as mural cells, wrap around ECs in arterioles, capillaries and venules to regulate the maturation of ECs, stabilize the microvascular wall and promote angiogenesis. Although pericytes are surrounded by a basement membrane, they contact the ECs with through a "peg and socket" mechanism through holes in the basement membrane. Studies have shown that pericytes also communicate with ECs via paracrine signaling to improve tissue repair (97,98). It is notable that pericytes also have stem cell-like properties and exhibit the morphology, mitotic activity and surface antigens of MSCs (99) and are seemingly able to differentiate into adipocytes, chondrocytes, osteoblasts, neurons, astrocytes and oligodendrocytes, leading them to be identified as MSCs (100)(101)(102). However, it is still debated whether pericytes are MSCs. Guimaraes-Camboa et al.
(103) challenged this concept and suggested that mural cells do not intrinsically behave as MSCs during aging and repair in multiple adult organs using a transgenic cell line. Over 2 years, the study showed that Tbx18 lineage-derived cells maintained their perivascular identity in the brain, heart, muscle and fat, indicating that mural cells do not exhibit an overt potential to give rise to other cell types. In contrast, MSCs can serve as a potential source of pericytes and induce vasculogenesis as mentioned previously (13,104,105), but similar to the multidifferentiation potential of MSCs, there needs to be standard guidelines for assessing pericyte differentiation in future studies. Furthermore, MSCs secrete various growth factors, including PDGF, which serves as a biomarker and crucial factor controlling the differentiation and recruitment of pericytes (106)(107)(108). These findings indicate that MSCs may regulate the recruitment of pericytes to injured tissue to participate in angiogenesis, but the associated mechanisms between MSCs and pericytes need to be further elucidated.

MSCS AND MSC-DERIVED EXOSOMES
Exosomes are a type of extracellular microvesicle secreted by multiple eukaryotes. Compared with cell therapy, MSC-derived exosomes (MSC-exos) have lower immunogenicity and are safer and more efficient, providing a new strategy for tissue regeneration via cell-free therapy (109,110). MSC-exos are a type of message carrier that harbor a modifiable content of microRNAs, mRNAs and proteins, mediating communication between cells and functioning as key mediators of the paracrine effect of MSCs (111,112). The pro-angiogenesis function of MSC-exos has been demonstrated in a number of studies. For instance, exosomes from MSCs overexpressing Akt, HIF-1α or CXCR-4 were shown to accelerate EC proliferation, migration and tube-like structure formation in vitro, as well as blood vessel formation to improve cardiac function in an MI model (113)(114)(115). MSC-exosomes may also have anti-inflammatory activities in MI model (116). Currently, the application of MSCexos primarily focuses on preclinical experiments. One of the key problems for exosome clinical therapy is how to collect and purify enough exosomes so that they can be used safely. Andriolo et al. (117) developed a GMP-class method for the mass preparation of stem cell-derived exosomes to enable them to be used in future clinical applications. Indeed, as they are secreted by MSCs, MSC-exos have similar biological properties to MSCs to some extent. MSC-exos also have paracrine functions and mediate communication between MSCs and ECs, and they are also influenced by microenvironmental stress conditions, such as hypoxia and irradiation (118,119). Furthermore, as they harbor a part of and not the entirety of MSC contents, MSC-exos are not an MSC "mini-me" and cannot replace MSCs in some respects, including their multiple differentiation and proliferation abilities.

APPLICATION OF MSCS IN IHD CLINICAL AND PRECLINICAL PRACTICE
The results of numerous clinical and preclinical studies have indicated that MSC transplantation is safe, significantly improves cardiac function and decreases infarct size and fibrosis in ischemic patients, which may be associated with the survival, retention, angiogenesis, paracrine action and the anti-apoptosis activities of MSCs (Tables 2, 3). Although cellular therapies hold great promise for the treatment of human IHD and have good safety, the efficacy of MSCs remains disputable, especially when used in clinical trials. Meta-analyses of randomized clinical trials showed that the transplantation of BMSCs resulted in limited improvement on cardiac function for MI patients (171,172).
As it was showed in Table 2, some clinical trials proved MSC transplantation did not improve LVEF although it may limit infarct size. Different from clinical trials, MSC transplantation in animal experiments showed significantly elevated LVEF in most studies ( Table 3).
Clinical patients are different from animal ischemia models, and the efficacy of MSCs in clinical practice is influenced by many different factors, such as (1) disease etiology and severity of patients, and (2) the type, number, delivery route and time, retention, survival, proliferation and differentiation of MSCs. Another meta-analysis showed that MSCs are more effective in patients with lower baseline left ventricular ejection fraction (LVEF) (≤50%), and the effects of cells that were transferred at 3-7 days post-AMI was superior to those transferred within 24 h or more than 7 days in improving LVEF and decreasing LV end-systolic and diastolic dimensions (173), which suggested transplantation time was a key factor to influence cardiac function. Compared with clinical trials, animal experiments are easier to obtain positive results because of their simplicity, such as the MI model can be established uniformly by ligation of the left anterior descending coronary artery. Compared with MSCs intervention alone, pretreated MSCs with some growth factors together may get more efficacy ( Table 3).
It is notable except for growth factors, more attention has been paid to natural botanical medicines. EGb761, an extract of Ginkgo biloba, was shown to exhibit a biphasic effect on hypoxia/serum deprivation-induced BMSC apoptosis, and its effect was closely associated with the PI3K/Akt and caspase-9 signaling pathways (174). Salvia miltiorrhiza is a widely used traditional Chinese medicine in cardiovascular diseases, and its constituent Tanshinone IIA was observed to decrease infarct size by increasing the recruitment of BMSCs to the infarct region by upregulating the SDF-1/CXCR-4 axis in a rat MI model (175). In addition to botanical medicines, chemicals such as statins, as the most commonly used lipid-lowering agents, exert activity toward a wide spectrum of cellular functions in addition to their lipidlowering effects, including anti-inflammatory, anti-apoptotic, anti-fibrotic and pro-angiogenesis effects (176,177). The results of multiple studies have suggested that atorvastatin has the ability to increase the survival rate of implanted BMSCs in an MI model, and combined with MSCs, it also ameliorated the cardiac milieu by reducing inflammatory cell infiltration, myeloperoxidase

STRATEGIES AND FUTURE DIRECTIONS
MSCs display robust reparative properties through their paracrine and differentiation abilities that can limit apoptosis, enhance neovascularization and direct positive tissue remodeling. However, some problems with MSCs remain and must be solved before they can have widespread use.
MSCs are important infiltrating cells that are also drived by blood and vasoconstriction. So the first problem is the low survival and retention of transplanted cells in vivo which limits their overall effectiveness in clinical usage. Consequently, identifying strategies to improve cell survival and retention in vivo is a priority. However, cell transplantation is affected by many factors, each of which may have an impact on the survival of transplanted cells, and there is still no consistent recommendation for each factor. The microenvironment of transplanted cells directly affects the survival of stem cells. The blood supply in the marginal area of myocardial infarction is well known to directly affect the survival rate and recovery of cardiac function after cell transplantation. One important goal of cell transplantation is to promote angiogenesis in the ischemic area and reduce the generation of myocardial scars. Studies have shown that hypoxia-induced stem cells release a variety of factors to improve the microenvironment through anti-inflammatory and anti-fibrosis effects and by promoting angiogenesis (170).
Hypoxia or other growth factors used to precondition stem cells may allow MSC survival and retention to be improved, but additional comparisons and a set of standards are needed to identify the most powerful factors. Another important factor limiting the clinical application of stem cells is the shortage of effective monitoring methods for stem cells. The successful implementation of cell therapies requires a better understanding of cell fate after transplantation. Currently, there are three primary labeling methods for stem cells, including reporter genes, fluorescent dyes and nanoparticles, which require optical imaging, MRI and radionuclide imaging to trace the transplanted stem cells, respectively or in combination, with each technique having its advantages and disadvantages (74,184). Thus, there is an urgent need to develop a nontoxic and noninvasive tracer technology that exhibits long term stability and that can also be used to dynamically monitor the survival status of transplanted cells with respect to processes such as migration and differentiation in vivo.

AUTHOR CONTRIBUTIONS
WS contributed to the design and manuscript writing. QX, RY, and YY collected and assembled data. WC contributed the review design and financial support. KC was responsible for proofreading and final approval of the review. All authors contributed to the article and approved the submitted version.