| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Cardiovascular Research Institute, Department of Medicine, New York Medical College, Valhalla, New York
ABSTRACT I. INTRODUCTION II. EXOGENOUS PROGENITOR CELLS AND CARDIAC REPAIR A. Human Embryonic Stem Cells B. Hematopoietic Stem Cells C. Mesenchymal Stem Cells D. Controversy on Bone Marrow Cell Transdifferentiation and Cardiac Repair III. CARDIAC PROGENITOR CELLS A. Identification B. Origin of Cardiac Stem Cells C. Cardiac Stem Cell Niches IV. CARDIAC PROGENITOR CELLS AND MYOCARDIAL REPAIR A. Myocardial Regeneration and Ischemic and Nonischemic Cardiomyopathy B. Cell Fusion and Myocardial Regeneration V. FUTURE DIRECTIONS GRANTS ACKNOWLEDGMENTS REFERENCES
 |
ABSTRACT |
|---|
 Top NextReferences |
|---|
|
I. INTRODUCTION |
|---|
|
TopPreviousNextReferences |
|---|
|
In spite of the presence of adult stem cells in self-renewing organs, tissue repair involves scar formation (280, 410, 512). Understanding the factors that dictate the evolution of a lesion into scarring instead of regeneration is at its beginning. Some insights come from embryonic development. Tissue repair in embryos is rapid, efficient, and scar free (143). Skin wounds in the early mammalian embryo heal with restitutio ad integrum, whereas wounds in late gestational fetuses and adult mammals result in scarring. This difference may be related to the intrinsic characteristics of embryonic fibroblasts or to external stimuli. Fibroblasts in embryos are less prone to synthesize and release collagen, attenuating the amount of fibrosis following injury (287). The presence of hyaluronic acid and fibronectin in the amniotic fluid may also interfere with scar formation during wound healing (269). Moreover, the inflammatory response is attenuated; a low number of poorly differentiated inflammatory cells accumulate in the region of damage, and the growth factors present at the site of healing in embryos are different from those in the adult organism (143).
The expression of transforming growth factor-
(TGF-) isoforms, which have powerful profibrotic effects, and their receptors (TBRI, TBRII) vary in the skin of embryos at different gestational ages (92, 143). TGF-3 is more abundant in early embryo, while the amount of TGF-1 and TGF-2 increases progressively with development. Similarly, TBRI and TBRII increase with maturation. The different ratios of TGF- isoforms in the skin appear to condition the evolution of healing triggering profibrotic or antifibrotic signals in prenatal and postnatal life (5). Adult wounds in mice, rats, and pigs, exposed to the same profile of cytokines present in the embryo, experience a more effective repair process with reduction in scar formation (143). Although scar-free regeneration of damaged organs is well known in amphibians and zebrafish (103, 385, 451), there is only one example in mammals in which repair occurs in the absence of fibrosis. Cartilage, skin, hair follicles, and cryoinjured myocardium regrow in the adult MRL mouse (197, 262). However, ischemia-reperfusion injury leads to comparable infarct size in wild-type and MRL mice (3), suggesting that the native capacity to reconstitute dead myocardium after infarction may be limited in this model.
In summary, modulation of the inflammatory response may improve tissue repair in the adult organism. However, the removal of these extrinsic signals may not be sufficient for long-term regeneration of the injured organ. Resident stem cells in proximity to the lesion can change the properties of the microenvironment by secreting cytokines that contribute to cell homing, proliferation, and differentiation (50, 239, 240, 291, 482). Enhancement of the intrinsic growth reserve of the organ constitutes the foundation of regenerative medicine and regenerative cardiology. This review discusses current understanding of the role that endogenous and exogenous progenitor cells may have in the treatment of ischemic and nonischemic heart failure.
|
II. EXOGENOUS PROGENITOR CELLS AND CARDIAC REPAIR |
|---|
|
TopPreviousNextReferences |
|---|
Novel methodological approaches applied to the analysis of the myocardium have challenged the paradigm introduced more than 60 years ago that the heart is a postmitotic organ and myocytes are terminally differentiated cells that participate in cardiac function throughout life (7, 97, 288, 329, 339, 340, 354). A subset of parenchymal cells expresses the molecular components mediating the entry and progression of these cells through the cell cycle, and karyokinesis and cytokinesis. The recognition that myocyte replication, hypertrophy, and death occur in the pathological heart has significantly enhanced our understanding of the dynamic nature of the myocardium and the critical role that these variables have in the preservation of cardiac performance or in the onset of ventricular dysfunction (22, 25, 27, 34, 326, 327). Contrary to the general belief, the restricted regenerative capacity of the myocardium does not represent the initial causal event of impaired cardiac function. The nonischemic failing heart, in its early phases of decompensation, has a number of myocytes that often exceeds the number of cells in a normal heart. Alternatively, the modest reduction in myocyte number chronically is not consistent with the severe deterioration in ventricular performance. A typical example is found in hypertensive cardiomyopathy or chronic aortic stenosis in animals and humans (29, 30, 497). The initiation and evolution of cardiac failure appears to depend mostly on two other crucial factors that are strictly interrelated in the determination of pump function: the accumulation of old, poorly contracting cells and the formation of multiple foci of myocardial scarring (52, 53, 67, 98, 484, 497).
A severe reduction in cell number, however, occurs acutely after ischemic injury that inevitably results in scar formation not only in the heart but also in other organs, whether their parenchymal cells are highly proliferating, slowly cycling, or terminally differentiated (263, 280, 410, 512). The need to overcome this biological obstacle has favored the development of strategies aiming at the replacement of dead cells with viable cells. Several interventions have been used in an attempt to promote cardiac regeneration experimentally. These protocols have employed different cell types, including fetal tissue and fetal and adult cardiomyocytes (264, 395, 444, 544), skeletal myoblasts (306, 323, 466), embryonic-derived myocytes and endothelial cells (109, 142), bone marrow-derived immature myocytes (194), fibroblasts (165), smooth muscle cells (272), and bone marrow c-kit positive and negative progenitor cells (144, 217, 226, 228, 356, 357, 539, 545). Unexpectedly, these multiple approaches had a rather uniform outcome that consisted of variable degrees of improvement in cardiac contractile function. Most likely, the implanted cells formed a passive graft, which reduced negative remodeling by decreasing the stiffness of the scarred portion of the ventricular wall. The elastic properties of the implanted cells could be the major cause for enhanced heart function. An active graft, which dynamically contributed to myocardial performance, was observed only in a few cases (144, 228, 356, 357, 539). The possibility has also been advanced that the implanted cells exert a paracrine effect on resident cardiac stem cells (CSCs) activating myocardial regeneration (50, 539). Despite these efforts, however, the most appropriate form of cell therapy for actual restitutio ad integrum of the damaged myocardium has not yet been identified.
It is intuitively apparent that if the adult heart possesses a pool of primitive multipotent cells (51, 201, 293, 296, 338), these cells must be tested first before more complex and unknown cells are explored. The attraction of this approach is its simplicity. Cardiac regeneration would be accomplished by enhancing the normal turnover of myocardial cells. However, major difficulties exist in the acquisition of myocardial samples and the isolation and expansion of CSCs in quantities that can be employed therapeutically. Until protocols capable of resolving these limitations are implemented, CSC therapy remains a difficult task. Alternatively, CSCs may be activated locally within the myocardium, but the growth factor-receptor systems that regulate CSC proliferation and differentiation are largely unknown (277, 335). Although the importance of resident CSCs in organ homeostasis is obvious and, in the future, CSCs may become the most appropriate form of cell therapy for the diseased heart, currently, emphasis has been placed on the identification of exogenous sources of highly proliferating cells that can acquire the cardiac phenotypes.
Embryonic stem cells (ESCs), which are derived from the inner mass of the blastocyst (337, 369, 397, 477), possess unique properties. These cells can be grown in vitro and propagated indefinitely in their undifferentiated state while retaining a normal karyotype. Additionally, they maintain over time the property of multilineage commitment and, therefore, can differentiate in vitro into all cell types present in an organism (49, 88, 337, 525). Because of their enormous potential, human ESC lines have been obtained from the human blastocyst (369, 477) with the expectation of a future successful application of their broad therapeutic potential to patients (124, 233, 370). However, a controversy exists between the therapeutic relevance of ESCs and autologous adult progenitor cells (PCs). The debate concerning the efficacy of ESCs and adult PCs is unfortunate because the utilization of ESCs in clinical practice is at least a decade away while adult PCs present minimal risks and are the only applicable form of cell therapy for myocardial regeneration today. This does not exclude that ESCs may become in the future a powerful alternative to adult PCs.
Human ESCs cultured in suspension form spontaneously cellular aggregates called embryoid bodies (EBs). EBs are composed of cells derived from the three germ layers (118). These teratoma-like structures consist of semiorganized tissues composed of cardiomyocytes, striated skeletal muscle cells, neuronal rosettes, and hemoglobin-containing blood islands. However, they lack critical features of embryonic patterning (216, 231). The morphology of solid EBs can be affected by the process of cavitation that results in the formation of cystic EBs. During their formation, markers of the mesoderm (
-globin), ectoderm (neurofilament 68), and endoderm (
-fetoprotein) are expressed (187, 216). Synchronously pulsing EBs contain myogenic cells positive for -cardiac actin, desmin, myogenin, and smooth muscle actin (44, 234). Consistent with these in vitro observations, in vivo delivery of human ESC lines to skeletal muscle or the testis gives rise to teratomas (397, 459, 477). Teratomas include endodermal, mesodermal, and ectodermal structures formed by mature cells (Fig. 2). Conversely, teratocarcinomas contain undifferentiated cells and possess a stem cell compartment (16, 88). Although the malignant tumorigenic potential of ESCs is not well defined yet, it is worrisome that ESCs and the embryonal carcinoma (EC) cells, i.e., the stem cell of teratocarcinomas, share several markers including Oct4 and Nanog (88, 89, 193, 342, 363). Moreover, the self-renewal property of ESCs and EC cells is modulated by similar pathways (88, 191, 458). Conversely, the ability to induce formation of human EBs containing cells of neuronal, hematopoietic, and cardiac origins is important for the recognition of early human embryonic development. In the last 15 years, this accessible in vitro system has been exploited to elucidate differentiation events in several tissues (118, 126, 397, 477). The study of ESCs may provide a unique understanding of the mechanisms of embryonic development that may lead to therapeutic interventions in utero and the correction of congenital malformations. This may become the most exciting outcome of ESC research.
|
superfamily of proteins (50, 244). Although some degree of functional improvement was observed, the therapeutic utilization of ESCs can be associated with dramatic side effects. Teratomas are benign tumors, but the cardiac localization confers on them a malignant biological behavior. Most importantly, ESCs injected intravenously can engraft and colonize all organs with the possible development of neoplastic lesions (509). In an attempt to avoid these complications, ESCs have been partially differentiated in vitro before their implantation in the injured heart (235, 284, 313, 535, 542). Some degree of cardiomyogenic commitment was claimed to enhance engraftment of the cells into the myocardium, attenuate the probability of ESCs to acquire undesired cell lineages, and thereby reduce the risk of teratoma formation (397, 477, 492). Although immature myocytes have been obtained from ESCs and their morphological, phenotypical, and functional properties characterized (196, 234, 235), the residual growth potential of the committed cells was not established. Additionally, the purity of the preparation remains a major problem since pluripotent ESCs may persist among the enriched population of developing myocytes and injection of this cell pool may engender tumor formation. There are no rigorous methodologies for ESC differentiation into cardiomyocytes in vitro (62, 196, 198, 234, 235, 519, 548). It is emblematic that eight distinct growth factors have forced ESCs to acquire only in part a specific cell phenotype while other undesired cells were consistently present in the preparations (417). None of the growth factors was capable of driving the commitment to a single cell type, making it impossible, at present, to obtain a uniform population of mature progeny from ESCs in culture.
However, the plasticity of ESCs is greater than that of adult stem cells. The implantation of beating EBs reverses complete atrioventricular block in pigs (235). The morphology of the grafted cells resembled those of embryonic cardiomyocytes being of small size with disorganized myofibrils located at the periphery of the cells. The mechanical and electrical behavior of beating EBs indicates that human ESCs give rise to cells that display functional properties typical of the embryonic human myocardium. When the electrophysiological analysis of the newly formed cells was associated with the analysis of their gene expression profile, the conclusion was reached that cardiomyocytes derived from ESCs maintain a phenotype reminiscent of the embryonic heart tube and hardly myocytes develop a chamberlike myocardial phenotype. Whether these immature myocytes can reach adult characteristics in vivo remains an unanswered question (147).
The immunorejection of the allogeneic ESCs or their differentiated progeny constitutes an additional limitation inherent in the use of ESCs (128). Because of their early stage of development, ESCs were considered "immune-privileged" and therefore not recognizable by the immune defenses of the recipient (477). Conversely, recent evidence suggests that even in their undifferentiated state, human ESCs express discrete levels of HLA class I antigens that increase as the cells mature (127, 294). The immunosuppressive regimen is poorly tolerated, and this treatment severely impairs a patient's quality of life. The immunodeficient state of mice that develop teratomas after ESC implantation (397, 477, 509) closely reflects the immunosuppressed condition that may have to be reached clinically to avoid rejection of ESCs. However, these limitations should not discourage the promoters of ESCs. Intense research in this area is expected to resolve these current biological problems and change the skepticism concerning the future use of ESCs for human disease.
Finally, the potential application of human ESCs poses profound ethical concerns. A fundamental issue involves the notion of the moral status of human embryos. However, when the ethical principles are taken to the extreme, they threaten to paralyze public funding for ESCs (523), leaving experimentation to the private sector and, thereby, precluding the possibility to effectively monitor practices or the search for alternatives (149, 167, 285). The privatization of ESCs has reached a point that protocols of cryopreservation have been developed in spite of the fact that we are far from the actual documentation of the efficacy and safety of ESCs. Active research on human ESCs is highly desirable, but the duty to heal sick individuals cannot overcome the moral imperative to treat human beings as subjects and not as objects. An interesting viewpoint has recently been proposed (254). In parallel with the definition of organismic death of the adult, which occurs when the criteria for brain death are met, an embryo can be considered organismically dead when it has lost a fundamental function. This consists of continued, integrated cellular division, growth, and differentiation. This premise implies that irreversible arrest of cell division rather than the death of each and every cell corresponds to the organismic death of the embryo (189, 254). Nearly 60% of the in vitro fertilized embryos fail to cleave at 24 h (261). Often, this growth inhibition reflects severe genetic abnormalities but not necessarily all the cells of the embryo are abnormal. In these mosaic embryos, at least one normal diploid blastomere is found (254). Thus the ethical framework for organ donation can be applied to the harvesting of human ESCs from dead embryos. This would provide a common basis in which the need to protect human dignity and progress in biomedical research are no longer in conflict.
Before discussing bone marrow cell (BMCs) transdifferentiation, a few comments concerning the developmental origin of the hematopoietic system and the heart may be relevant. Blood cells derive from the mesoderm (328, 516), and the hematopoietic progenitors accumulate first in blood islands in the early somites, migrate then to the fetal liver and, ultimately, reach the bone marrow, which represents their definitive localization in late embryogenesis and adulthood (328, 431, 516). Concurrently, the primitive heart is generated within the anterior lateral mesoderm (100, 133, 138). Eisenberg et al. (134) have shown that the early mesoderm has the capability of creating both blood and cardiac cells. The specification of organs that arise from adjacent regions in the developing embryo may differ by a single developmental decision, affecting the expression of a relatively small number of master switch genes (438, 485–487). This modification in cell fate has been defined as direct transdifferentiation (56, 132, 426) to contrast the transdifferentiation that requires a dedifferentiation step (256, 257, 316) or the transdifferentiation that occurs through the involvement of a stem cell (10, 27, 178, 327, 392, 438, 487). If this were also the case for the bone marrow and the heart, differentiated BMCs could undergo a direct form of transdifferentiation and generate new myocardium. Although this remains a possibility, committed CD45 positive myeloid cells appear to be able to give rise to skeletal muscle cells or hepatocytes (79, 80). However, so far, only c-kit-positive immature myelomonocytic progenitors can contribute to muscle fibers while more mature CD11b positive myelomonocytic cells do not transdifferentiate into skeletal muscle (125).
Embryonic hematopoietic progenitors acquire the cardiomyocyte phenotype in the presence of retinoic acid, dexamethasone, PGE2, interleukin (IL)-2, fibroblast growth factor (FGF)4, and bone morphometric protein (BMP)4 (136). Although the efficiency of transdifferentiation is low, the committed embryonic progenitors express transcription factors and contractile proteins of cardiac muscle cells. The creation of cardiomyocytes is markedly increased when lineage negative Sca-1-positive adult BMCs are cocultured with embryonic mesoderm. The new myocytes integrate functionally with the embryonic tissue (100). Together, these lines of evidence support the notion that BMCs can form cardiac and skeletal muscle cells in vitro and in vivo. This possibility has been strengthened by recent observations in which c-kit-positive fetal liver hematopoietic progenitor cells have been employed successfully to regenerate myocytes and coronary vessels in the infarcted heart (258). These hematopoietic progenitors were obtained from cloned embryos derived from somatic cell fusion between nuclei of LacZ-positive fibroblasts and enucleated oocytes of a different mouse strain (256, 257, 259).
The newly formed myocardium replaced 38% of the infarct at 1 mo (258). The rebuilt tissue expressed LacZ and was composed of myocytes and vessels connected with the primary coronary circulation. Myocytes were functionally competent and expressed contractile proteins, desmin, connexin43, and N-cadherin. These structural characteristics indicated that the developing new myocytes were coupled electrically and mechanically (258). Similarly, the formed coronary arterioles and capillary structures contained blood cells and contributed, therefore, to tissue oxygenation. Cardiac replacement resulted in an improvement of ventricular hemodynamics, attenuation of cardiac remodeling, and reduction of diastolic wall stress. These beneficial effects were obtained by stem cell transdifferentiation and commitment to the cardiac cell lineages. In fact, myocardial growth was independent from fusion of the injected progenitor cells with preexisting partner cells. Although problems currently plague nuclear transplantation, including the potential for epigenetic and imprinting abnormalities, fetal hematopoietic progenitor cells derived from cloned embryos are sufficiently normal to repair dead myocardium in vivo.
The adult bone marrow contains several cell populations. In addition to differentiated cells, such as stroma cells, vascular cells, adipocytes, osteoblasts, and osteoclasts, primitive cells reside in the bone marrow. This class of undifferentiated cells with stem cell properties is heterogeneous (383, 431, 457, 516), being composed of HSCs and mesenchymal stem cells (MSCs). These two subsets of cells have been employed in the repair of damaged organs giving rise to tissues distinct from the organ of origin (178, 206, 245, 253, 295, 315, 390, 392, 473, 490). However, a mixed population of bone marrow cells rather than a highly purified stem cell pool has commonly been used for the restoration of injured organs (64, 68, 131, 211, 275, 357, 360, 388, 475). In the context of this review, we have defined this heterogeneous cell population as BMCs, i.e., a cell preparation that contains an enriched pool of HSCs together with variable degrees of MSCs and endothelial progenitor cells.
In 1961, Till and McCullock discovered clonogenic bone marrow cells that gave rise to multilineage hematopoietic colonies in the spleen (48, 436, 479). Till and McCullock (480) proposed that these cells were multipotent HSCs that possess the properties of self-renewal and multilineage differentiation. The definition of stem cells has not changed, and bone marrow HSCs are functionally defined by their capacity to self-renew, form clones, and differentiate into mature blood cell types (431, 516, 517). Although HSCs were discovered more than four decades ago, their isolation has become possible only more recently when surface markers were identified (445, 493, 517). Different epitopes characterize HSCs in different species. Mouse HSCs correspond to a subpopulation of BMCs that are c-kit+Sca-1+Thy-1-low (358, 493). Operationally, this population is multipotent and repopulates the bone marrow of irradiated mice (493). However, the c-kit+Sca-1+Thy-1-low cells are functionally heterogeneous. Based on the identification of additional surface markers and clonal analysis, the Mac-1-negative-CD4-negative cells within this category are enriched for long-term reconstituting cells (320). Conversely, the Mac-1-low-CD4-negative pool contains short-term repopulating cells. Finally, the Mac-1-low-CD4-low cells correspond to transient multipotent progenitors together with B lymphocyte progenitors (319). The CD34 antigen has been used to sort human HSCs (331) which, however, are also present in the CD34 negative fraction (123, 181). AC133 is another marker of human HSCs (57) but, so far, there is no specific epitope for the true stem cell.
The discovery that adult HSCs or BMCs retain a remarkable degree of developmental plasticity and may have the potential to differentiate across boundaries of lineage and tissue (9, 10, 150, 438, 487, 502) has divided the scientific and clinical community (324, 507, 508, 518). Therefore, the therapeutic efficacy of BMCs for the damaged heart has been questioned (112, 137, 295, 309, 508). Consistently, the injection of BMCs improves the performance of the pathological heart (144, 217, 228, 356, 539), but the mechanism by which the administration of BMCs results in enhanced cardiac function remains controversial (135, 137, 200, 247, 295). In spite of these uncertainties, clinical trials have been completed and some are ongoing (39, 69, 158, 163, 186, 273, 372, 373, 415, 421, 448, 449, 452, 528). The safety of this approach has been documented, and double-blind clinical studies are in progress. Limitations in the analysis of myocardial regeneration in humans due to the difficulty in obtaining cardiac biopsies, together with contrasting findings in animals, have prompted different interpretations of the positive outcome of BMC administration. Three possibilities have been advanced. They include the development of coronary vessels that rescue hibernating myocardium, de novo formation of myocytes and vascular structures, or the activation and growth of resident progenitor cells via a paracrine effect mediated by the implanted BMCs (17, 50, 135, 144, 217, 226, 228, 239, 356, 357, 460, 539, 545). These mechanisms of action of BMCs are not mutually exclusive and may be operative in the rescue of the injured heart. However, it is relevant to discuss available information to establish what has been shown so far and what has to be done to document unequivocally the actual role of BMCs in the management of cardiac diseases. This analysis can only be performed by comparing protocols and data accumulated in animal studies.
The efficacy of adult BMCs for myocardial regeneration after infarction was documented four years ago (356). BMCs of male mice heterozygous for EGFP were collected, immunodepleted for lineage markers, and then sorted for the stem cell antigen c-kit. Enriched lineage negative c-kit-positive cells were injected in the border zone of infarcted mice where they colonized to the dead tissue and gave rise to contracting myocardium occupying 68% of the original infarct. The newly formed EGFP-Y-chromosome positive cells corresponded to functionally competent myocytes and vascular cells organized in coronary arterioles and capillary structures. Overall, the aspect of the newly formed myocardium was that of a rather immature tissue with myocytes small in size, 
500 µm3, and vessels with a small lumen and a thick multilayered wall. Myocardial regeneration with amelioration of cardiac performance was obtained only in 40% of the treated mice. Coronary ligation in mice is a complex procedure with an inherent variability in infarct size and a 50% probability of correct injection (228, 356). The mouse heart beats 600 times/min and has a left ventricular (LV) wall that is <1 mm thick. These factors make the injection of cells within the LV wall highly problematic. When these technical difficulties were overcome by the mobilization of BMCs with the systemic administration of stem cell factor (SCF) and granulocyte-colony stimulating factor (G-CSF), myocardial regeneration was obtained in all infarcted treated animals. The new myocardium was composed of contracting myocytes and patent coronary vessels (357).
These studies prompted other investigators to document the ability of BMCs to repair the damaged heart (144, 217, 226, 228, 334, 411, 539, 545). Initially, the hypothesis tested was whether spontaneous mobilization of BMCs occurs after infarction and whether these circulating cells home to the region of injury rescuing the dead tissue (217). Although the degree of engraftment and transdifferentiation in coronary vessels was low and markedly exceeded that of cardiomyocytes, this report confirmed the plasticity of BMCs. The difference in the degree of myocardial reconstitution between this and the previous studies can be attributed to several factors. They include the pathologic model, ischemia-reperfusion injury in irradiated mice (217) versus permanent coronary occlusion in nonablated mice (228, 356, 357), together with the modality of intervention, spontaneous mobilization of BMCs (217) versus intramyocardial injection of BMCs (228, 258, 356) or cytokine-mediated BMC mobilization (357), and the type of BMCs, c-kit enriched BMCs versus peripheral blood cells. The bone marrow of irradiated mice was repopulated with a subset of cells capable of excluding the Hoechst 33342 dye (217). These cells are positive for both c-kit and Sca-1 and are probably more enriched for true HSCs than the lineage negative c-kit positive cells (112) implemented in other studies (228, 356, 357). In spite of several unresolved issues, these observations are consistent with the notion that BMCs can adopt the cardiogenic fate by forming myocytes and coronary vessels.
More recently, the possibility that cell therapy of the infarcted heart exerts its beneficial effects not only by reconstitution of dead myocardium but also by the activation of resident progenitor cells in the spared portion of the ventricular wall has been carefully examined (228, 291, 539). To properly evaluate the formation of cardiomyocytes and coronary vasculature in the region bordering the infarct and distant from the infarct, the accumulation of newly formed myocytes and vascular structures was measured utilizing markers of the cell cycle and morphometric methods. Although similar protocols were employed in the analysis of myocyte and vessel growth, positive and negative results were obtained concerning the paracrine effects of the administered BMCs (228, 539). However, an enriched population of mouse c-kit-positive BMCs was used in one case (228) and a novel human bone marrow stem cell in the other (539). Both cell populations differentiated into cardiac cell lineages and repaired the infarcted heart, but the latter also promoted a robust regenerative response in the surviving myocardium. Thus a specific human BMC has the ability to transdifferentiate in cardiac muscle cells, smooth muscle cells, and endothelial cells in vitro and in vivo (144, 539). Additionally, this unique cell population can induce endogenous neovascularization and cardiomyogenesis (539). Thus these findings support the notion that BMCs adopt the cardiac phenotype and potentiate the growth reserve of the adult heart. Collectively, these observations point to the therapeutic import of BMCs for cardiac diseases in humans (17, 39, 69, 158, 163, 239, 273, 372, 373, 415, 421, 448, 449, 452, 528).
There is no accepted definition of MSCs. Originally, MSCs were thought to correspond to the putative marrow cells that self-renew and give rise to one or more mesenchymal tissues (314). However, the marrow stromal cell population that contains the MSC pool can also differentiate into tissues other than those that originate in the embryonic mesoderm, raising questions about the appropriateness of the term mesenchymal for this type of stem cell (413, 505, 529). In the majority of studies, the adherent fibroblastic cells obtained from the unfractionated mononuclear cell class of the bone marrow are termed mesenchymal stem or progenitor cells (8). But this heterogeneous cell population is recognized to be too "crude" to represent the purified pool of MSCs (383). A common and widely used characterization of MSCs is that they are the nonhematopoietic multipotent stem cells of the bone marrow (314, 383).
The notion of MSCs was introduced in 1961 by Friedenstein (154) who documented that marrow stroma cells contain osteogenic progenitors. Subsequently, the culture conditions necessary for the expansion and differentiation of this cell category were developed. Marrow stromal cells were found to be highly proliferative, clonogenic, and capable of forming colonies of different size and density, composed of several mesenchymal lineages (86, 87, 155, 453, 465). In the late 1980s and in the 1990s, MSCs acquired a significant biological role with the work performed by Caplan (84), who suggested that the actual MSC is an adherent fibroblastic cell isolated by Percoll density centrifugation. This postulated MSC expressed antigens reactive with the monoclonal antibodies SH2 and SH3 that recognize CD105 and CD73, respectively (195). The disadvantage of this traditional method of isolation is the inevitable contamination from HSCs and the cellular heterogeneity of the preparation. Over the past decade, Caplan's definition has been questioned as being not restrictive enough for the ideal MSC. Although CD105 is a relevant epitope of MSCs and angiogenesis (314, 382), CD29, CD44, and CD90 have been proposed to be important determinants of the mesenchymal phenotype (382, 383) while STRO-1 may identify an immature state (175). The immunophenotype of MSCs has only been partially determined, although the SH2, SH3, CD29, CD44, CD71, CD90, CD106, CD120a, and CD124 epitopes have consistently been found in MSCs (85, 383, 456). Moreover, MSCs are negative for markers of the hematopoietic lineage, including CD34 and the common leukocyte antigen CD45 (43, 383). However, Prockop and others (105, 423, 439) support the notion that MSCs cannot be distinguished solely by antigen expression but necessitate functional assays demonstrating their multipotent growth and differentiation behavior. Thus a definitive consensus on the properties of MSCs has not been reached yet.
Putative MSCs have been identified in embryonic, fetal, and postnatal organs (84, 389). During prenatal life, mesenchymal progenitors accumulate in sites harboring HSCs (307). MSCs appear in the aorta-gonad-mesonephros region and colocalize with HSCs (117). Additionally, MSCs are found in the embryonic circulation including the cord blood and amniotic fluid (213, 214, 400, 489). In adulthood, the bone marrow and the systemic circulation constitute the main sources of MSCs (389, 402), although they have also been detected in tissues distant from the bone marrow. The oval cells of the liver, prostatic stem cells, metanephric mesenchymal cells, precursors of the Leydig cells in the testis, primitive osteoprogenitors, preadipocytes, and satellite cells of the skeletal muscle have been classified as MSCs (13, 114, 399, 488, 543, 552). Because not all studies agree that MSCs reside in these mesenchymal tissues, the possibility has been raised that a long-distance traffic of MSCs occurs between the bone marrow and distant sites through the bloodstream. Alternatively, embryonic primordia of MSCs may be stored in organs of mesodermal origin and, in response to injury, may participate in tissue repair (383).
Characteristically, MSCs adhere quickly to the culture dish and grow, forming colonies that become visible 1 wk after plating (314, 383, 389). Only 0.001–0.01% of the initial unfractionated bone marrow cell population consists of MSCs (383, 389). Some preparations of MSCs can be expanded over 15 cell doublings, while others cease replicating after 4 cell doublings (389). This difference may depend on several determinants, including the procedure used to harvest the marrow (314), the low frequency of MSCs in marrow harvests (84), and the age or condition of the donor from which MSCs were prepared (450). This limited growth potential is associated with preservation of the karyotype, telomerase activity, and telomere length (381, 382). The constant level of telomerase activity and the absence of telomeric shortening have been interpreted as unique properties of MSCs, which were considered capable of escaping replicative senescence. However, MSCs from young donors experience 20 population doublings while MSCs from old donors are compromised in their proliferative capacity (381, 450). Extensive subcultivation impairs cell function, and evident signs of senescence (119) and/or apoptosis (314) become apparent. Both young and old MSCs undergo senescence-associated growth arrest (314); they become telomerase incompetent and have short telomeres. The senescent phenotype can be reversed by forced expression of human telomerase (2).
During the phase of amplification, MSCs do not differentiate spontaneously but do so in the presence of growth factors and cytokines (72, 85, 456). They can acquire multiple phenotypes including osteoblasts, chondrocytes, adipocytes, endothelial cells, and neuronal-like cells (99, 221, 286, 382, 383, 405, 412, 422, 538). Bone marrow MSCs differentiate into cardiomyocytes in the presence of 5-aza-cytidine (159, 185, 290); the morphology of the cells changes from a spindle-shaped to a ball-like form and, subsequently, a rod-shaped configuration. The cells then fuse in a syncitium-like structure that resembles a myotube (160). Although these characteristics mimic the organization of skeletal muscle cells, the committed progeny expresses markers commonly found in cardiomyocytes at different stages of fetal development (185). Transcription factors of the cardiac and myocyte lineage, GATA4, Nkx2.5, and HAND1/2, have been detected (159). Interestingly, a switch in MEF2 isoform occurs. From early to late passages, MEF2C is replaced by MEF2A and MEF2D. Additionally, the -isoform of cardiac myosin heavy chain is more abundant than the -isoform. Similarly, -skeletal actin predominates, and -cardiac actin is moderately present. Myosin light-chain 2v is also expressed. Functionally competent - and -adrenergic and muscarinic receptors have been found (185). Cells beat spontaneously and synchronously, and the rate of contraction increases with stimulation by isoproterenol. Conversely, contractile activity is inhibited by 1-blocking agents (290). Thus the differentiation process of these cardiomyogenic cells recapitulates in part the developmental program of gene expression of prenatal life.
An elusive stem cell that may or may not belong to the MSC category is the primitive cell identified by Verfaillie and co-workers (225). This human or mouse multipotent adult progenitor cell (MAPC) can be obtained by growing CD45 and glycoprotein A-depleted BMCs in selective culture conditions (Fig. 3). When this cell population is plated at low density in laminin-coated dishes with a low serum medium containing EGF and PDGF, MAPCs appear several months later after 20 population doublings. In this in vitro system, MAPCs differentiate into cells of the three germ layers including hepatocytes, endothelial cells, and neurons (398, 419). This multilineage differentiation occurs also in vivo when MAPCs are injected in the early blastocysts of mice. However, in adult sublethally irradiated immunodeficient mice, the distribution of systemically delivered MAPCs is much more restricted. MAPCs engraft only in hematopoietic tissues, lung, liver, and intestine but not in the heart (Fig. 3). Whether MAPCs injected locally in the myocardium acquire the myocyte lineage has not been tested. In summary, MAPCs could represent a rare population of MSCs or a by-product of long-term culture.
|
Autologous MSC-derived myogenic cells (321) and autologous undifferentiated MSCs (424) have been employed for cardiac repair after infarction or ischemia-reperfusion injury in pigs. Cells were injected directly in proximity or within the injured myocardium, acutely or chronically after the ischemic event, and the effects of these interventions were determined several weeks later. In all cases, myocardial regeneration was documented and the reconstitution of dead myocardium correlated with the improvement in ventricular function and reappearance of wall motion activity (321, 424). Moreover, cotransplantation of human MSCs and human fetal cardiomyocytes resulted in an amelioration of cardiac performance greater than with MSCs alone (311). Formation of myocytes and coronary vessels has also been observed in rats after cryoinjury and MSC implantation (355). However, the intracoronary delivery of MSCs after infarction led to the differentiation of MSCs into fibroblasts in the scarred region and to the regeneration of myocytes in the surviving, unaffected portion of the ventricular wall (506). Importantly, in the absence of injury, MSCs engraft into the myocardium but remain in a viable quiescent state and do not participate in the physiological turnover of myocytes as well as vascular smooth muscle and endothelial cells. Thus the microenvironment significantly conditions the developmental pathway of MSCs in vivo.
D. Controversy on Bone Marrow Cell Transdifferentiation and Cardiac Repair
Stem cell transdifferentiation in the adult organism is a highly questioned mechanism of growth (9, 10, 150, 178, 206, 253, 279, 384, 392, 438, 473, 487, 502, 507, 508, 518). The most versatile cell is the bone marrow progenitor cell. The documentation that adult BMCs are capable of generating mature cells beyond their own tissue boundaries has provided an unexpected and powerful new form of cell therapy. The wave of enthusiasm created by this discovery and the lack of effective new drugs for the treatment of heart failure has prompted cardiologists to the rapid implementation of BMCs in the management of the infarcted human heart. Currently, bone marrow cells or circulating bone marrow cells, including endothelial progenitor cells (EPCs), are utilized in patients, and several initial clinical trials have been performed (39, 69, 158, 163, 273, 372, 373, 415, 421, 448, 449, 452, 528). Because of the compelling need to seek new treatments for severely ill patients, clinicians are leading the field of cell therapy and myocardial regeneration. Conversely, statements of caution have been made concerning the necessity to acquire a better understanding of the mechanisms of recovery of cardiac function before these protocols are introduced in the treatment of the diseased human heart. The fundamental question to be addressed is whether this more conservative position is justified because strong criticisms have been made not only against the notion of stem cell plasticity (45, 95, 279, 324, 474, 507, 508, 518) but also against observations that have challenged the perennial view of the heart and the brain as postmitotic organs incapable of regenerating after birth (508). In fact, the revolutionary work on the brain and the heart that has led to the identification of neural stem cells (139, 500) and cardiac stem cells (51, 201, 293, 296, 338) has been attacked as inconclusive, methodologically incorrect (508), and, more recently, a collection of artifacts (45, 95, 324, 495).
The early studies on BMCs and cardiac repair were followed by numerous reports in which different types of BMCs and modalities of interventions were employed for the treatment of the damaged heart (24, 26, 34). These cells of bone marrow origin included several subsets making it difficult to dissect the impact that each cell type had on the function and structure of the diseased heart. For example, lineage-negative c-kit-positive cells (228, 258, 356), lineage-negative c-kit-positive Sca-1-positive cells, c-kit-positive-Thy1.1-low-lineage-negative Sca-1-positive, long-term reconstituting HSCs (45), CD34-positive cells (545), mesenchymal-like progenitor cells (424, 434), EPCs (281, 282), mononuclear BMCs, and clonally expanded human multipotent stem cells (539) have been tested. Moreover, BMCs with the potential of restoring the dead myocardium have been mobilized from the bone marrow into the systemic circulation utilizing several cytokines; SCF, G-CSF, and SDF-1 have been administered alone or in combination in various animal species (1, 38, 162, 188, 334, 357, 462). The effects of these various modalities of therapy ranged from improvement in regional cardiac function together with the restoration of the dead myocardium to a modest recovery of ventricular performance in the absence of tissue regeneration. The lack of beneficial consequences on the heart has also been reported. However, predominant observations have been positive supporting the feasibility of this form of cell therapy for the failing heart.
Although these experimental results and the therapeutic efficacy of BMCs in patients with ischemic and nonischemic heart failure were indicative of BMC transdifferentiation (Fig. 4), the plasticity of BMCs has been questioned. Recent publications in mice have emphasized negative results, criticizing the documentation that BMCs can convert into cardiac cell lineages. The ability of BMCs to regenerate dead myocardium in the mouse heart was questioned, and the claim was made that the fate of the injected cells was to acquire only the hematopoietic lineages (45, 324). An additional report suggested that the engraftment of BMCs in the damaged heart is transient and hematopoietic in nature. BMC-derived cardiomyocytes were observed at low frequency and only outside the infarcted myocardium (336).
|
The intramyocardial delivery of BMCs in the mouse heart is complex and faced with difficulties dictated by the extremely high heart rate and very thin LV wall. Laboratories with years of experience in inducing coronary artery occlusion, coronary artery narrowing, or ischemia-reperfusion injury in mice are well aware that the successful imposition of myocardial infarction is 80%, with 20% unsuccessful surgery (115, 228, 356, 357, 539). Additionally, in positive cases, infarct size can vary from 20 to 90% of the entire LV inclusive of the interventricular septum. Infarcts involving 50–60% of the LV carry a nearly 50% mortality within 24–48 h after surgery (33). Even more problematic is the recognition of a successful administration of BMCs within the LV myocardium, 40–50%, and the identification of the site of injection in proximity of the infarcted tissue (228, 356).
Because of these unpredictable factors, methodologies have been developed in which rhodamine-labeled polystyrene spheres are mixed with BMCs to have a reliable reference point for the recognition of the actual positioning of BMCs in the LV wall and with respect to the infarct. When this simple protocol was applied, myocardial regeneration was demonstrated in 100% of mice, which had rhodamine particles and BMCs implanted in the border zone of the infarcted myocardium (228). Therefore, differences in the protocols employed in the administration of BMCs could account for the differences observed experimentally between positive studies of myocardial regeneration (228, 356, 539) and reports challenging the ability of BMCs to form myocytes and coronary vessels (45, 324). Moreover, the negative studies claim that cell transplantation was 100% successful, that a mortality of 8% occurred with infarcts of 60%, and that a normal LV end-diastolic pressure was found with infarcts commonly incompatible with life in this model (45).
A critical issue in studies of myocardial regeneration mediated by transplantation of BMCs involves the morphological analysis of the heart. This is an important point because one of the reports questioning the ability of BMCs to form myocytes and coronary vessels utilized conventional light microscopy and immunoperoxidase staining to demonstrate the presence or absence of newly formed structures within the infarcted myocardium (324). This approach reflects the contention that conventional light microscopy is superior to fluorescence labeling and confocal microscopy (250, 467). The resolution in images obtained by light microscopy is markedly inferior to that provided by confocal microscopy (28). Additionally, only the very superficial layer of the section can be examined by light microscopy (Fig. 5), whereas the entire thick histological section can be viewed by confocal microscopy (391). At high-magnification light microscopy, the observer is faced with a blurred image with loss of detail and undefined cell boundaries. This inherent problem with light microscopy was recognized immediately when confocal technology became available (521). Clear examples were published documenting the impossibility of identifying microtubules, mitotic spindle, cytoplasmic proteins, and chromosomal structures by light microscopy of cultured cells. Conversely, these morphological details were apparent when the same cells were evaluated by confocal microscopy. The difference between epifluorescence microscopy or bright-field microscopy and confocal microscopy is magnified in tissue sections. In fact, focal depth in a 5- to 6-µm tissue section examined by epifluorescence microscopy or bright-field microscopy is only 8%, while the focal depth in a 0.5-µm optical section analyzed by confocal microscopy is 12.5-fold higher, i.e., 100% (29, 283; Fig. 5).
|
Paraffin-embedded tissue sections result, at most, in an underestimation of specific immunostaining not in an overestimation and, therefore, differences in the technical approach could explain why positive and negative results concerning BMC transdifferentiation have been obtained. Hearts studied immunocytochemically should be the same examined functionally and with routine histology. This was not the case in the negative studies (45, 324). Whether coronary ligation was unsuccessful or a small or large infarct was generated was not determined. This is critical, because of the complexity of the model and the difficulty of producing an infarct and a correct injection of BMCs. Moreover, the utilization of frozen sections may interfere with the recognition of small cells of the size of newly formed myocytes (115, 228, 258, 356). Double labeling for the identification of the transgene present in the administered BMCs and the detection of myocyte cytoplasmic proteins was never performed on the same section (324), and negative claims were based on only two animals (45) that were supposedly properly infarcted and injected with cells comparable to those used in the positive studies (356). This limited cohort calls the statistical validity of this type of experimentation into question, an issue that has recently been recognized in a large number of published reports in Nature and Nature Medicine (496). When these problems are dealt with, myocardial regeneration mediated by BMC transdifferentiation can also be documented by immunolabeling of frozen sections of infarcted myocardium (539).
The animal models of parabiosis with complete blood chimerism or with reconstituted bone marrow through the injection of a single EGFP positive HSC (45, 507, 508) have been introduced to question the ability of BMCs to acquire a cardiomyocyte lineage. The utilization of the irradiation protocol in studies aiming at the documentation of the transdifferentiation potential of adult stem cells has been criticized (4, 297). Irradiation represents a form of injury that complicates the interpretation of the results. Most importantly, it cannot be used as documentation of physiological trafficking of cells from the bone marrow to distant organs. In fact, bone marrow transplantation in mice affected by amyotropic lateral sclerosis typically shows cell engraftment in the central nervous system (CNS) shortly after the intravenous injection of BMCs. But this process decreases with time (110). Because the damage is not repaired and the CNS continues to send the same signals for eventual cell recruitment, the progressive decline in BMCs colonized to the CNS strongly suggests that homing of these cells occurred only during the phase of high concentration of repopulating cells in the bloodstream. The administered BMCs targeted the depleted bone marrow, and their circulating pool markedly decreased with time. After the initial engraftment, there was no longer translocation of cells from the bone marrow to the CNS (110). Thus positive and negative results obtained with this protocol are difficult to interpret.
The limitation of the therapeutic potential of circulating BMCs is not new. If circulating BMCs had the ability to spontaneously repair damaged organs, infarcts of the heart, brain, skin, kidney, and intestine would be easily reconstituted and the majority of current human diseases would not exist. These models are of little value to resolve the controversy at hand and can only add confusion to the confusion. The issue in need of resolution is whether BMCs injected directly in the infarct, in the border zone or systemically, differentiate into cardiac cell lineages and contribute to myocardial regeneration. The paradigms offered by the models of parabiosis and repopulated bone marrow with single cell injection are nonphysiological. More than 90% of the animals die, and the question is how relevant is an experimental procedure that represents the exception rather than the rule.
Intrinsic genetic markers have been proposed as today's gold standard for these types of studies (45, 95, 218, 324, 495). Although the detection of the Y chromosome and EGFP protein in the regenerated myocytes (228, 356) falls within this category, recent experiments were performed with a genetic marker consisting of a cardiomyocyte-restricted transgene. Donor c-kit-positive BMCs were obtained from male transgenic mice carrying a c-myc tagged nuclear Akt under the control of the cardiac specific -myosin heavy chain promoter (430). BMCs from these animals were injected acutely after infarction in wild-type female mice, and myocardial regeneration within the infarct was identified a few days later. The band of formed myocardium contained small myocytes that expressed -sarcomeric actin, troponin I, -actinin, connexin43, and N-cadherin. In all cases, the nuclei of these developing myocytes were positive for the c-myc tag (Fig. 6A). Vascular structures and CD45 labeled cells in the reconstituted tissue were negative for the c-myc tag. The Y chromosome was detected in endothelial and smooth muscle cell nuclei but not in CD45 positive cells. Therefore, BMCs injected in the dead myocardium do not adopt the hematopoietic fate (45) but assume the cardiac phenotype and repair the infarcted heart.
|
The emphasis placed on genetic markers for the "objective" recognition of newly formed structures is only in part correct. Any genetic marker requires its subsequent identification by histochemical (218, 324) or immunocytochemical (45, 218) procedures. If limitations exist in these protocols, then the powerful genetic markers lead to false collection of data and erroneous interpretations and conclusions. Collectively, the data available so far demonstrate that myocardial regeneration occurs and reestablish the plasticity of BMCs and their ability to acquire a phenotype different from the organ of origin. Whether the process of cardiac repair occurs independently from the formation of synkaryons and heterokaryons or requires fusion of the progenitor cell with a preexisting partner cell is discussed in section IV.B of this review.
|
III. CARDIAC PROGENITOR CELLS |
|---|
|
TopPrevious |
|---|
