SEARCH

SEARCH BY CITATION

Keywords:

  • Heart;
  • Confocal and light microscopy;
  • Mitosis;
  • Chimerism;
  • Bone marrow cell transdifferentiationEnhanced green fluorescent protein autofluorescence

Abstract

  1. Top of page
  2. Abstract
  3. Introduction
  4. Possibility 1: The Heart Is a Terminally Differentiated Postmitotic Organ
  5. Possibility 2: The Heart Is Not a Terminally Differentiated Postmitotic Organ
  6. Myocyte Division
  7. Cardiac Chimerism and Myocyte Formation
  8. Hematopoietic Stem Cells and Myocyte Formation
  9. Cardiac Stem Cells and Myocyte Formation
  10. Disclosures
  11. Acknowledgements
  12. References
  13. Supporting Information

This review discusses the current controversy about the role that endogenous and exogenous progenitor cells have in cardiac homeostasis and myocardial regeneration following injury. Although great enthusiasm was created by the possibility of reconstituting the damaged heart, the opponents of this new concept of cardiac biology have interpreted most of the findings supporting this possibility as the product of technical artifacts. This article challenges this established, static view of cardiac growth and favors the notion that the mammalian heart has the inherent ability to replace its cardiomyocytes through the activation of a pool of resident primitive cells or the administration of hematopoietic stem cells.


Introduction

  1. Top of page
  2. Abstract
  3. Introduction
  4. Possibility 1: The Heart Is a Terminally Differentiated Postmitotic Organ
  5. Possibility 2: The Heart Is Not a Terminally Differentiated Postmitotic Organ
  6. Myocyte Division
  7. Cardiac Chimerism and Myocyte Formation
  8. Hematopoietic Stem Cells and Myocyte Formation
  9. Cardiac Stem Cells and Myocyte Formation
  10. Disclosures
  11. Acknowledgements
  12. References
  13. Supporting Information

This article discusses the fundamental results that have contributed to create the view that the postnatal and adult heart cannot renew its parenchymal cells and that the number of myocytes that are present at birth dictates the destiny of this organ throughout life [1, 2]. According to this premise, myocyte turnover does not occur, and the cells originally formed during embryonic and fetal cardiac development are responsible for the preservation of myocardial performance in the young, adult, old, and senescent heart. Coronary vascular endothelial cells (ECs) and smooth muscle cells (SMCs) divide rarely, have a very long cell cycle, and are replaced slowly by division of resident ECs and SMCs within the wall, without changing the interaction between the established, irreplaceable myocytes and the modestly more dynamic coronary vessels. Importantly, circulating progenitor cells or primitive cells mobilized from the bone marrow do not transdifferentiate and therefore are incapable of acquiring the myocyte, EC, and SMC lineages [3]. By necessity, myocyte apoptosis has to be an occasional event of no or, at most, trivial consequence; if not, the heart would rapidly lose a significant number of parenchymal cells, and the small group of myocytes left would be unable to sustain ventricular performance. Together, these concepts project a rather static biological view of cardiac homeostasis and pathology. In addition, this perception of the heart imposes severe limitations on the development of therapeutic strategies that can be implemented for the management of human disease [4].

We will review the work that has promoted a shift in paradigm of the heart from a terminally differentiated postmitotic organ to a self-renewing organ. This dramatic change in understanding of cardiac behavior and function was originated by observations that were made in the 1940s and that have continued to accumulate over the years (for reviews, see [5, 6]). However, the disagreement that persists among the members of the scientific community today was present then and has persisted for nearly 60 years. The reason for the controversy is unclear, but it seems to reflect unshakable positions based on preconceived beliefs more than on careful analysis and proper consideration of published results [7]. In this article, we focus on methodological issues because, in the course of the controversy, recurrent statements have been made to undermine the technical protocols used in studies supporting the existence of cardiac regeneration. Consistently, the data in favor of myocardial regeneration are claimed to be the product of methodological artifacts [1, [2], [3]4, 8, [9], [10]11]. To properly evaluate the validity of these criticisms or lack thereof, methodologies need to be compared to provide readers with a better understanding of the substrate that governs the debate.

The old dogma has profoundly conditioned basic and clinical research in cardiology for the last three decades [7, 12]. Based on this paradigm, cardiomyocytes undergo cellular hypertrophy but cannot be replaced either by the entry into the cell cycle of a subpopulation of nonterminally differentiated myocytes or by the activation of a pool of primitive cells that become committed to the myocyte lineage. However, the efforts made to introduce a highly dynamic perspective of the heart have led to the identification and characterization of a resident pool of stem cells that can generate myocyte, and ECs and SMCs organized in coronary vessels [13]. This discovery has created a new, heated debate concerning the implementation of adult cardiac stem cells in the treatment of heart failure of ischemic and nonischemic origin.

Possibility 1: The Heart Is a Terminally Differentiated Postmitotic Organ

  1. Top of page
  2. Abstract
  3. Introduction
  4. Possibility 1: The Heart Is a Terminally Differentiated Postmitotic Organ
  5. Possibility 2: The Heart Is Not a Terminally Differentiated Postmitotic Organ
  6. Myocyte Division
  7. Cardiac Chimerism and Myocyte Formation
  8. Hematopoietic Stem Cells and Myocyte Formation
  9. Cardiac Stem Cells and Myocyte Formation
  10. Disclosures
  11. Acknowledgements
  12. References
  13. Supporting Information

The majority of organs contain dividing and nondividing cells. However, nondividing cells can rest in the G0 phase and reenter the cell cycle following growth activation or become terminally differentiated and die without undergoing further division [14]. The latter category of cells is not dormant in G0 and cannot reenter the cell cycle. This is typically seen at the tip of the villi of the small intestine [15], in the auditory hair cells of the ear [16], and in the upper layer of the epidermis [17]. Hepatocytes in vivo are in a G0 state, and after partial hepatectomy or severe injury, liver regeneration is accomplished by proliferation of hepatocytes, biliary epithelial cells, and fenestrated endothelial cells [18]. Because the liver contains a population of primitive cells, the oval cells, the growth and maturation of this cell category may contribute to the reconstitution of liver mass [19]. Probably both cell mechanisms are involved in the repair of the organ, although controversy exists in the field [20]. Multipotential cells have also been identified in the central nervous system, but questions have been raised concerning their ability to grow and generate Purkinje neurons in the adult brain [21, [22], [23]24]. Conversely, the regeneration of the skin [25], bone marrow [26], and skeletal muscle [27] is regulated by activation of skin stem cells, hematopoietic stem cells (HSCs), and satellite cells, respectively. These mechanisms of cellular growth have not been considered possible in the myocardium, and the notion has been proposed that ventricular myocytes cannot be replaced once cell division ceases immediately after birth in the mammalian heart [1, 2, 9, 10].

Thus, the dogma was established that the postnatal heart is composed of a fixed number of myocytes and that, if myocytes die, they are permanently lost and the myocardium must maintain its vital role with a reduced number of cells. The remaining myocytes are not in G0 and cannot be triggered into the replicating phase [28]; they continue to perform their physiological function, undergo cellular hypertrophy, and ultimately die [29]. Based on this paradigm, the age of myocytes, organ, and organism was assumed to coincide, implying that myocytes in humans may have a lifespan that exceeds 100 years [30]. For several decades, no effort was made to reexamine this rather unusual view of the biology of the heart and cardiac homeostasis. Remarkably, there is not a single piece of evidence that demonstrates the inability of the heart to replace its dying myocytes. It seems rather extravagant that cardiomyocytes can contract 70 times per minute over 100 years and continue to be functional. During this period, they would have contracted 3.7 billion times and still be operative. If this were to be the case, adult myocytes would be essentially immortal cells.

Possibility 2: The Heart Is Not a Terminally Differentiated Postmitotic Organ

  1. Top of page
  2. Abstract
  3. Introduction
  4. Possibility 1: The Heart Is a Terminally Differentiated Postmitotic Organ
  5. Possibility 2: The Heart Is Not a Terminally Differentiated Postmitotic Organ
  6. Myocyte Division
  7. Cardiac Chimerism and Myocyte Formation
  8. Hematopoietic Stem Cells and Myocyte Formation
  9. Cardiac Stem Cells and Myocyte Formation
  10. Disclosures
  11. Acknowledgements
  12. References
  13. Supporting Information

Historically, the foundations for the view of the heart as a terminally differentiated postmitotic organ incapable of regeneration were established in the mid-1920s [31]. This conclusion was dictated by the difficulty in identifying mitotic figures in myocyte nuclei of the human heart with the unsophisticated histologic preparations available 80–90 years ago. This notion gained support from autoradiographic studies of tritiated thymidine incorporation in the myocardium during postnatal development and pathological overloads [32, [33], [34]35], conducted in the mid- and late 1960s. DNA synthesis in myocyte nuclei was negligible, and this observation, together with the inability to recognize myocyte mitotic figures, led to the belief that the adult heart is composed of parenchymal cells that are permanently withdrawn from the cell cycle. Myocytes can increase in volume but not in number [36]. Again, this work used standard histologic preparations in combination with photographic emulsions, which imposed severe limitations in terms of microscopic resolution [37]. These factors made it highly problematic to distinguish whether DNA replication or mitosis was present exclusively in interstitial cells or, at times, involved cardiomyocytes as well. In fact, the possibility of myocyte division was acknowledged in two of the early reports [34, 35].

The qualitative results discussed so far were in sharp contrast with quantitative measurements of myocyte volume and number performed in human hearts obtained from patients who died because of decompensated cardiac hypertrophy and congestive heart failure [5]. In the late 1940s and early 1950s, Linzbach documented that, in the presence of a cardiac weight equal to or greater than 500 g, myocyte proliferation represented the predominant mechanism of increased muscle mass [38, 39]. An interesting aspect of this work is that these determinations were all based on two very simple but critical parameters: myocyte length, assessed by the distance between two nuclei of largely mononucleated myocytes, and myocyte diameter across the nucleus. These variables are easily recognizable in standard histologic sections and do not require high resolution and an impeccable preparation. Based on a cylindrical shape configuration, myocyte cell volume was computed. Moreover, the aggregate volume of the myocyte compartment of the myocardium was obtained, and the quotient of this quantity and myocyte cell volume yielded the total number of ventricular myocytes. Because of the simplicity of the approach, the original results were confirmed several years later in other laboratories in which more refined techniques were applied [40, 41]. In all cases, hearts weighing 500 g or more were characterized by a striking increase in myocyte number that was more prominent than cellular hypertrophy; this adaptation involved both the left and right ventricles [38, [39], [40], [41], [42]43].

In the mid- and late 1990s, new studies of the human heart examined the distribution of mononucleated and binucleated myocytes in 72 normal and 176 diseased hearts [43]. Aging, cardiac hypertrophy, and ischemic cardiomyopathy were characterized by the lack of changes in the relative proportion of mononucleated and multinucleated myocytes in the ventricular myocardium, confirming that the early and more recent measurements of myocyte proliferation were valid and did not represent and erroneous interpretation dictated by nuclear hyperplasia in the absence of myocyte division [6, 44]. Mononucleated cells constitute ∼75% of myocytes in the human heart, differing significantly from mice [45], rats [46], dogs [47, 48], and pigs [49].

Mononucleated cells are smaller than binucleated cells, and this cellular property may influence the ability of myocyte to divide [50]. Human myocytes larger than 30,000 μm3 cannot reenter the cell cycle and proliferate [50]. In fact, cycling and mitotic myocytes are predominantly small and mononucleated, and this feature may account for the massive cardiac hypertrophy that can be achieved in humans (Fig. 1). It would be inefficient for large myocytes to divide once or at most twice to expand the cardiac mass. Heart weight in humans can increase nearly threefold, reaching values of 1,000 g or larger [5, 38, 41, 43]. Heart failure typically shows increases in myocyte number that vary from 20%–100% or more [5, 38, 41, 43, 51, [52]53]. This phenomenon is not affected by age; in an analysis of 7,112 human hearts, from birth to 110 years of age, Linzbach and Akuamoa-Boateng have shown that extreme forms of organ hypertrophy are detectable up to the ninth decade of life, and heart weights of 500 and 600 g are present in patients at 100 years of age and older [54].

thumbnail image

Figure Figure 1.. Myocyte proliferation and ischemic cardiomyopathy in humans. Small dividing left ventricular myocytes (α-sarcomeric actin; red) are identified by Ki67 labeling (A) (white, arrow), MCM5 expression (B) (yellow, arrow), and metaphase chromosomes (C) (arrow). The boundaries of myocytes and interstitial cells are defined by laminin (green). Nuclei are stained by 4,6-diamidino-2-phenylindole (blue).

Download figure to PowerPoint

Importantly, scar formation, foci of replacement fibrosis scattered throughout the ventricular wall, diffuse interstitial fibrosis, and ongoing myocyte death by apoptosis and necrosis are typically present in the hypertrophied failing heart [44, 55, 56]. These factors markedly affect the number of cardiomyocytes and indicate that the extent of myocyte formation determined by the morphometric methodologies summarized above constitutes an underestimation of the actual magnitude of myocyte regeneration in the pathological heart. In addition, there is no alternative methodology able to demonstrate unequivocally myocyte proliferation. By definition, myocyte replication corresponds to an increase in the number of myocytes; in the absence of this documentation, biochemical, molecular, physiological, autoradiographic, and structural results can only be suggestive of myocyte regeneration but in themselves cannot provide the critical information, that is, an increased number of cells. Quantitative morphology is the only technique that can identify myocyte formation. Contrary to expectations, the observations in favor of myocyte regeneration were not received with enthusiasm by the scientific community [57, [58], [59], [60], [61], [62]63]. With some small variations, the morphometric results in the human heart were questioned, and the possibility of myocyte formation in the adult myocardium was first dismissed and subsequently attacked. This new idea challenged the comfortable paradigm of myocyte hypertrophy that had dominated the field of molecular cardiology and cell signaling for the previous 25 years. Heart failure was viewed as the consequence of a defective myocyte hypertrophy in which alterations of effector pathways regulating myocyte growth and contractility were considered responsible for the depression in organ, tissue, and cell function [64, 65].

However, the quantitative data collected in the human heart suggested that a reexamination of the mechanisms of myocyte growth needed to be considered and that additional studies needed to be performed to clarify the apparently contradictory results. For this purpose, experiments were conducted in animal models of cardiac failure and in human diseased hearts to provide relevant information in support of or against the notion of the regeneration potential of the adult myocardium. New methodologies were introduced to determine whether a subset of cardiomyocytes had the ability to reenter the cell cycle and divide and whether this process involved the activation of cell cycle-related genes, cyclins, and cyclin-dependent kinases and the expression of proteins modulating karyokinesis and cytokinesis [45, 50, 55, 66, [67], [68], [69], [70], [71], [71], [72], [73], [74], [75], [76], [77], [78], [79]80]. Rounds of cell division result in progressive telomere attrition, and critical telomeric shortening triggers irreversible growth arrest [81, [82]83]. Thus, the telomere-telomerase axis, together with the telomere-binding proteins, is a fundamental component of the multiple mechanisms that regulate the regeneration of non-postmitotic organs; because of this function, this system was evaluated and found to be operative in the adult heart of animals and humans [50, 74, [75], [76]77, 79, 84]. Unfortunately, these multiple studies left the controversy unresolved.

Myocyte Division

  1. Top of page
  2. Abstract
  3. Introduction
  4. Possibility 1: The Heart Is a Terminally Differentiated Postmitotic Organ
  5. Possibility 2: The Heart Is Not a Terminally Differentiated Postmitotic Organ
  6. Myocyte Division
  7. Cardiac Chimerism and Myocyte Formation
  8. Hematopoietic Stem Cells and Myocyte Formation
  9. Cardiac Stem Cells and Myocyte Formation
  10. Disclosures
  11. Acknowledgements
  12. References
  13. Supporting Information

In 1925, a highly significant paper was published in which the claim was made that there are no detectable mitotic figures in myocytes of the human heart, and therefore myocyte regeneration does not occur in mammals [31]. This work challenged the view proposed in numerous studies published from 1850 to 1911 in which the general belief was that cardiac hypertrophy is the consequence of hyperplasia and hypertrophy of existing myocytes [6]. The 1925 study can be considered the one that introduced the concept that the heart is a terminally differentiated postmitotic organ. The impact of this research was enormous. To the best of our knowledge, for several decades, many generations of pathologists and cardiovascular scientists have accepted the view that mitotic myocytes are not to be found in the adult myocardium, and only a few rare exceptions have been published, mostly restricted to the fetal-neonatal heart [85, [86], [87]88].

In the last 10–12 years, however, mitotic images in human cardiomyocytes have been documented in acute and chronic ischemic cardiomyopathy [72, 89, 90], idiopathic dilated cardiomyopathy [72], chronic aortic stenosis with moderate ventricular dysfunction [50], myocardial aging [84], acromegaly (Fig. 2A), and diabetes (Fig. 2B). Using light microscopic examination of routine histologic sections, an extensive search was required for the recognition of occasional mitotic figures in myocytes. Because in the early successful observations, dividing myocytes were seen only in the fetal and diseased failing heart [89], the assumption was made that replicating myocytes are essentially undetectable in normal adult myocardium. This hypothesis, in fact, suggested a rather minor role of myocyte regeneration in cardiac homeostasis and pathology in humans. However, in spite of the caution exercised in the interpretation of these results, this work was considered to reflect the incorrect interpretation of proliferating interstitial fibroblasts as dividing myocytes [1, 2]. Surprisingly, this conclusion was reached based on electron microscopic images that illustrated nondividing fibroblasts in proximity to differentiated nonreplicating cardiomyocytes [1]. This represents the normal organization of the myocardium and makes it difficult to understand how this well-established morphological characteristic of cardiac structure can be used as an argument against myocyte division and a case in favor of its erroneous identification. So Artifact 1 was introduced in the literature as the only logical explanation for these unexpected results.

thumbnail image

Figure Figure 2.. Myocyte generation in the diseased human heart. Metaphase chromosomes (blue, 4,6-diamidino-2-phenylindole [DAPI]; arrows) are apparent in two small dividing cardiomyocytes (red, α-sarcomeric actin) in patients affected by acromegaly (A) or diabetes (B). The organization of tubulin in the mitotic spindle (C) (white, arrows), the accumulation of actin in the contractile ring (D) (red, arrows), and cytokinesis (E) (arrows) in dividing cardiomyocytes are apparent. In all cases, chromosomes are labeled by DAPI (blue) and the myocyte cytoplasm by α-sarcomeric actin (red). Green indicates laminin.

Download figure to PowerPoint

An important point that necessitates discussion is the belief by some that electron microscopy is the only tool that can unequivocally prove or disprove the presence of myocyte mitotic images in the myocardium. In invited reviews questioning the existence of myocyte regeneration, interstitial fibroblasts have been described as extremely mobile cells capable of penetrating by several microns the thickness of myocytes, positioning themselves in the central portion of these cells [1, 2]. By this mechanism, the myocyte nucleus was apparently displaced from its original location, which was occupied by the “penetrating fibroblast.” Schemes were created to strengthen this unlikely possibility [1]. However, this highly imaginative understanding of cardiac morphology fell short on several accounts: the examples of “penetrating” interstitial cells did not show dividing cells, the myocyte nucleus was at its expected site in the middle of the myocyte cytoplasm, and the electron microscopic preparation was of poor quality. Myocytes were hypercontracted and had sarcomeres of ∼1.4–1.5 μm in length. Normally, diastolic sarcomere length is 2.1–2.2 μm, and during systolic contraction, it becomes 1.8–1.9 μm [91]. Hearts with sarcomeres ∼1.4 μm in length are dead, since no contraction can be generated under this condition. Contraction bands produced by the inappropriate fixation of the myocardium were analyzed to reinterpret myocardial ultrastructure. Surprisingly, these artifacts were accepted and introduced in the literature as facts.

The frequency of mitosis in myocytes of normal human, mouse, or rat heart, as assessed by confocal microscopy, is low and averages 14/106 myocytes [72], 37/106 myocytes [45], and 85/106 myocytes (J.K., unpublished data), respectively. The values for dividing interstitial fibroblasts are similar to those of cardiomyocytes [89]. Based on these data, ∼350, 140, and 60 mm2 of tissue need to be examined to identify a dividing myocyte in the human, mouse, and rat heart, respectively. This is why the suggestion to implement electron microscopy for this analysis is not realistic. The area of myocardium that is sampled in each thin section by electron microscopy is ∼0.2 mm2, and the area of the microscopic field at a magnification of approximately ×2,000 is 0.0054 mm2. Thus, the numbers of micrographs that would have to be collected to identify a single myocyte mitotic image are 66,000 pictures for humans, 25,000 for mice, and 11,000 for rats. If one wishes to use electron microscopy to challenge the existence of myocyte division and to demonstrate that proliferating fibroblasts were improperly interpreted as mitotic myocytes, that is what needs to be invested to obtain at least one example in support of or against this thesis [92].

Major advances in understanding myocyte replication in the adult human heart have been made by the use of confocal microscopy and immunolabeling in the analysis of the normal and diseased heart. With this approach, the identification of mitotic images in myocytes became easier and more accurate. These new results strengthened the notion that parenchymal cells are formed continuously in the normal myocardium and myocyte regeneration is markedly potentiated in the pathological failing heart [6, 50, 72, 84, 90]. This technology was also applied to the study of animal models of human diseases, and myocyte replication was unequivocally demonstrated in myocardial sections and isolated cell preparations [6, 45, 73, 93]. The latter observation conclusively refuted the claim that dividing myocytes represent proliferating fibroblasts inserted within or attached to terminally differentiated myocytes. The formation of the mitotic spindle by the arrangement of tubulin and the generation of the contractile ring by the accumulation of actin in the region of cytoplasmic separation were clearly documented, together with karyokinesis and cytokinesis (Fig. 2C–2E). However, these results did not change the position of certain groups, who continued to reject the regenerative ability of the adult mammalian myocardium [2, 4, 8, 9, 28]. These groups introduced the possibility of a new source of artifacts, Artifact 2, related to confocal microscopy; they stated that traditional light microscopy is greatly superior to confocal microscopy and, with this personal viewpoint, tried to invalidate the results obtained with the latter technique.

An important issue that requires discussion relates to the notion that the analysis of histochemical preparations by conventional light microscopy is superior to fluorescence labeling and confocal microscopy [94, 95]. The resolution of the micrographs provided by light microscopy is markedly inferior to that obtained by confocal microscopy, ∼0.62 versus 0.29 μm [37, 96]. In addition, only the very superficial layer of a tissue section can be examined by light microscopy (Fig. 3). This inherent problem with light microscopy was recognized immediately when confocal technology became available [97]. Clear examples were published in the Journal of Cell Biology in 1987 demonstrating the impossibility of identifying microtubules, mitotic spindle, cytoplasmic proteins, and chromosomal structures by light microscopy in cultured cells. Conversely, these morphological details were apparent when the same cells were examined by confocal microscopy [98]. The difference between epifluorescence light microscopy and confocal microscopy is even greater in tissue sections. When confocal microscopy was not available to us, we estimated the myocyte mitotic index in chronic heart failure to be 11 myocytes per million [89]. When comparable hearts were evaluated by confocal microscopy, the mitotic index reached a value of 152 cells per million [72]. Even more striking is the difference between the mitotic indices obtained by light and confocal microscopy in the border zone of acute infarcts in humans: light microscopy yielded 3.3 mitotic myocytes per million [89] and confocal microscopy 775 mitotic cells per million [90]. In all cases, the measurements of myocytes in mitosis were made in the same laboratory.

thumbnail image

Figure Figure 3.. Light epifluorescence microscopy of human myocardial tissue sections. Panels show different focal planes. (A): Superficial layer of the section. (B): Focal plane located 1 μm below the surface. (C): Focal plane located 2 μm below the surface. Note that sarcomere striations (green, α-sarcomeric actin) and details of nuclear structure (red, propidium iodide) are well-defined only in (A) and are not clearly visible in (B) and (C). The yellowish area adjacent to the nucleus reflects the accumulation of lipofuscin.

Download figure to PowerPoint

Cardiac Chimerism and Myocyte Formation

  1. Top of page
  2. Abstract
  3. Introduction
  4. Possibility 1: The Heart Is a Terminally Differentiated Postmitotic Organ
  5. Possibility 2: The Heart Is Not a Terminally Differentiated Postmitotic Organ
  6. Myocyte Division
  7. Cardiac Chimerism and Myocyte Formation
  8. Hematopoietic Stem Cells and Myocyte Formation
  9. Cardiac Stem Cells and Myocyte Formation
  10. Disclosures
  11. Acknowledgements
  12. References
  13. Supporting Information

The controversy concerning the recognition of dividing myocytes in the adult heart intensified when studies on cardiac chimerism were conducted in an attempt to determine whether myocardial regeneration could be accounted for by the activation of undifferentiated cells of recipient origin [99]. Although studies of the chimerism of the heart and other organs following sex-mismatched cardiac or bone marrow transplantation have provided consistent results concerning the migration of primitive cells from the host to the graft [94, 95, 99, [100], [101], [102], [103], [104], [105], [106], [107], [108], [109], [110], [111], [112], [113]114], findings from different laboratories have been inconsistent in terms of the extent of this process, and this variability has promoted a heated debate in the scientific community. After homing, host progenitor cells undergo replication and differentiation generating mature parenchymal cells and vascular structures in the transplanted organ. There is little disagreement among authors on the occurrence of this phenomenon in sex-mismatched heart transplant, but the magnitude of cardiac chimerism varies significantly in different reports [94, 95, 99, [100], [101]102, 104, 109, 110]. This discrepancy involves mostly ventricular myocytes and, to a much lesser extent, newly formed coronary vessels. For myocytes, the published values range from as high as 18% to as low as 0.02% or absent [99, 100, 104]. A plethora of intermediate results has also been obtained.

The surprising outcome of these findings was that the low extent of myocyte chimerism detected by some groups was used to question myocyte formation and to support the traditional view that the heart is a terminally differentiated postmitotic organ incapable of undergoing meaningful regeneration of any kind [2, 4, 8, [9]10]. However, to accomplish this task, a rational explanation was necessary to account for the data pointing to a high number of male myocytes within female hearts transplanted in male recipients [99]. In fact, these observations indicated that a significant proportion of male myocytes was present in the female heart as a result of activation and commitment of the host male progenitor cells. Interestingly, the discrepancy in the degree of cardiac chimerism was immediately adopted to initiate another controversy and introduced in the literature Artifact 3. The promoters of this new source of technical error reached this conclusion by assuming that only the absence or minimal levels of chimerism were correct, whereas high levels of chimerism had to be necessarily wrong [94, 95, 104, 115]. The critics decided to take the prerogative of reinterpreting data published by other groups, pointing to a number of crass morphological errors. Surprisingly, these negative comments were published and, once again, these questionable criticisms became facts.

The most obvious technical differences among the studies published so far is the use of conventional light microscopy [94, 95, 100, [101]102, 104] versus confocal microscopy [99, 109, 110]. The probes used for the detection of the Y chromosome in female hearts transplanted in male recipients were not always the same, and the procedures used for the recognition of the Y chromosome differed as well. These factors introduced critical variables in the acquisition of the data that, together with the method of analysis, have been a major cause for the discrepancy.

A serious technical and conceptual concern is that similar high values of Y-chromosome-positive myocyte nuclei have been obtained in control male human hearts, averaging 50%. However, the values reported in control male human hearts by conventional fluorescence microscopy and confocal microscopy in the absence of optical sectioning of the samples are unrealistic and therefore cast doubts on this series of studies. By these two methods of analysis, the fraction of Y-chromosome-positive myocyte nuclei is low and rarely exceeds 15% (Fig. 4A). This is due to the limited focal depth of the objective, ∼0.5 μm, [37] and the extent by which a nucleus has to be embedded within the section to be visible (i.e., the penetration factor) [44]. Moreover, with conventional fluorescence microscopy, the out-of-focus fluorescence blurs the image, complicating the identification of the Y-chromosome signal in nuclei (Fig. 4B). When histologic sections of male myocardium are examined by confocal microscopy, the superficial layer of the section, which corresponds to 0.6 μm, shows 12% Y-chromosome-positive myocyte nuclei (Fig. 4C). Importantly, the evaluation of the entire thickness of the section by the analysis of consecutive optical planes markedly increases the detection of Y-chromosome-labeled myocyte nuclei, reaching a cumulative value of 55% (Fig. 4D). Therefore, in several reports in which conventional fluorescence microscopy or confocal microscopy restricted to a single plane was used [94, 101, 102, 104], the finding of the Y chromosome in a large number of myocyte nuclei of control male hearts is at variance with reality. Because of this unacceptable premise, the extremely low number or absence of Y-chromosome-positive myocyte nuclei found in the transplanted female hearts can be seriously questioned.

thumbnail image

Figure Figure 4.. Detection of Y-chr. by fluorescence in situ hybridization. (A): One myocyte nucleus positive for the Y-chr. (white dot, arrow) is visible by conventional epifluorescence microscopy of 6-μm-thick section of human male myocardium. Seven nuclei pertaining to myocytes (α-sarcomeric actin, red, arrowheads) are present in the field. (B): Often, out-of-focus fluorescence makes it difficult the recognition of Y-chr. labeling. The arrowhead points to a nuclear dot that may correspond to the Y-chr. signal. (C, D): Confocal microscopy of a 6-μm-thick tissue section of human male myocardium; the two panels illustrate the same field. (C): Superficial layer of the section. (D): Combination of 10 optical projections of 0.6 μm each of the entire section thickness. Arrowheads indicate Y-chr.-positive myocyte nuclei in (C) (2 of 15) and (D) (9 of 16). Laminin (green) defines the periphery of the cells. (E): High-magnification confocal image of a myocyte nucleus positive for the Y-chr. The area occupied by the nucleus is 176 μm2, whereas the area occupied by the Y-chr. is 2.0 μm2, accounting for 1.1% of the nuclear cross-sectional area. Abbreviation: Y-chr., Y chromosome.

Download figure to PowerPoint

The volume of a myocyte nucleus is ∼250 μm3, and the signal for the Y chromosome has a volume of ∼0.5 μm3, occupying 0.2% of the entire nucleus. When the Y chromosome is present in a mid-section of a nucleus, the area of the Y-chromosome signal is ∼1 μm2, and the area of the nucleus profile is ∼75 μm2 [116]. Thus, the Y chromosome occupies ∼1.3% of the nuclear area (Fig. 4E). It follows that the probability of hitting the Y chromosome in myocyte nuclear profiles of a tissue section can be calculated. Importantly, the chance of a positive signal varies as a function of section thickness and is influenced by the orientation of the nuclei in myocytes. These variables can be computed according to the following equations [44, 116].

  • equation image

In these equations, P corresponds to the probability of finding a Y-chromosome signal with a diameter dY within a nucleus whose length is described by lN and width by wN. The parameter F represents the fraction of the nucleus embedded in the section. F is a function of the thickness of the section, tS, and the orientation and size of the nucleus, described by the variable k. The angle between the plane of the section and the long axis of the nucleus is given by α.

Supplemental online Table 1 lists the predicted minimal, maximal, and average values of Y-chromosome-positive myocyte nuclei in control male human hearts according to section thickness, nuclear orientation, and modality of analysis. Surprisingly, there is a dramatic inconsistency between the expected values of Y-chromosome labeling of myocyte nuclei in normal conditions and in published results. Because of the questionable validity of control values in these studies, the number of male cells identified in female-transplanted hearts suffers from lack of reliability and necessitates reevaluation. This issue is certainly more relevant than the claim concerning the difficulty of distinguishing myocyte nuclei from interstitial cell nuclei [1, 2, 94, 115]. As recently emphasized, an experienced pathologist can easily differentiate an interstitial cell from a cardiomyocyte or discriminate an interstitial cell nucleus from a myocyte nucleus, particularly by confocal microscopy [117]. This is confirmed by the clear recognition of the Y-chromosome signal and nuclear profiles by confocal microscopy [99, 109, 110] and the rather ill-defined Y-chromosome labeling and nuclear boundary by light microscopy [94, 95, 104]. In some cases, by light microscopy, the Y chromosome cannot be distinguished from the nucleolus [94] or properly identified because of little contrast, weak intensity of the signal, and diffuse pattern of staining [104].

Several other factors have to be considered when human myocardium is examined. The age of the patients, time from transplantation, level of rejection, immunosuppressive regimen, and, most importantly, the condition of the tissue to be studied may all contribute to the variability in the levels of cardiac chimerism reported in the last several years. Poor fixation and/or delayed fixation affect the integrity of the DNA and, thereby, the probability of detecting any nucleotide sequence [118]. DNA degradation precedes the alterations in protein antigenicity [119] so that Y-chromosome labeling of myocyte nuclei may constitute a significant underestimation of the actual degree of myocyte chimerism present in the organ [109, 117]. Likewise, prolonged preservation of the myocardium in formalin fixative produces excessive cross-linking of proteins [120], and this condition diminishes the accessibility of the Y-chromosome probe to the nuclear DNA [118, 121]. It is surprising that these crucial factors, together with elementary morphometric principles, have been ignored in the studies pointing to the absence or minimal levels of cardiac chimerism following sex-mismatched heart allograft or bone marrow transplantation.

An elegant protocol has been introduced in one study of cardiac chimerism following sex-mismatched bone marrow transplantation to correct, at least in part, this problem. The fluorescence in situ hybridization (FISH) assay was performed with two distinct probes, which detected, respectively, the Y chromosome and the X chromosome in myocyte nuclei [103, 109]. Nuclei negative for the X chromosome had severely altered DNA and had to be excluded from the analysis. When this approach was used, the extent of cardiomyocyte chimerism was found to be 30-fold higher than in other comparable studies [108]. Because questions have been raised concerning some of the values obtained in our study of cardiac chimerism following sex-mismatched heart transplantation [99], we have re-examined cases with the highest levels of Y-chromosome-positive myocyte nuclei using this dual-color FISH assay and evaluated the percentage of XX-positive and XY-positive myocyte nuclei (Fig. 5A, 5B). The original data were confirmed and strengthened by this more sophisticated approach. In addition, the measurements of the number of X and Y chromosomes in nuclei allowed us to exclude cell fusion as the mechanism of male myocyte formation in the female heart. Importantly, optical sectioning by confocal microscopy prevented any possible misinterpretation of interstitial cell nuclei as myocyte nuclei (Fig. 5C–5L).

thumbnail image

Figure Figure 5.. Cardiac chimerism in humans. Confocal images of a tissue section obtained from a female heart transplanted in a male recipient. (A, B): Nuclei (4,6-diamidino-2-phenylindole, blue) are positive for the X chromosome (magenta dots) and Y chromosome (white dots, arrowheads). The coexpression of X and Y chromosomes in the same nucleus indicates male diploid genotype. The majority of nuclei belong to myocytes (B) (α-sarcomeric actin, red). (B): Open arrow points to a male myocyte. The periphery of the cells is defined by laminin (green). (C–L): Y chromosome (white dots) and X chromosome (magenta dots) are detected in nine distinct optical sections 1 μm apart (D–L) of the area indicated by the inset in (C). The Y chromosome in the myocyte nucleus (C) (arrowhead) is not visible in (D) and (E) but is evident in (F–H). It is no longer present in (G–L). The X chromosome in the same myocyte nucleus is apparent only in (J) and (K). The nucleus of an interstitial cell is also visible in (C) (asterisk).

Download figure to PowerPoint

Together, the studies on cardiac chimerism in patients who had sex-mismatched bone marrow transplantation provide strong evidence that circulating progenitor cells mobilized from the bone marrow have the ability to repopulate the heart with cardiomyocytes and coronary vessels [103, 108, 109]. The generation of parenchymal cells and vascular structures mediated by engraftment and differentiation of primitive cells from the recipient is even greater following heart transplantation [99, 109]. However, in this case, the source of the cells that translocate from the host to the implanted heart remains uncertain [99]. The progenitor cells could have been of hematopoietic origin and migrated to the heart through the systemic circulation, or they could have come from the cardiac remnant preserved during surgery for the attachment of the donor heart [122]. These possibilities are not mutually exclusive.

The recognition that cardiac chimerism occurs rapidly and in a noticeable manner has had a dramatic impact on our understanding of the biology of the heart and mechanisms of myocardial repair. This work has formed the foundation for the concept of myocardial regeneration and has created the opportunity for testing the role that HSCs have in the restoration of the injured heart [123]. The notion that HSCs retain a significant degree of developmental plasticity and can acquire cell lineages distinct from the organ of origin has promoted the entire field of cell therapy of the failing heart [124]. Opponents of this emerging new paradigm in cardiology, however, continued to claim that these concepts are the product of artifacts [1, 2, 10, 94, 95, 108, 125]. A recurrent unqualified statement is that they have introduced in their work “rigorous” and “stringent” criteria, which implies that others do not adhere to the same high scientific standards. The old paradigm has to survive, and the notion that myocytes cannot be generated postnatally, in adulthood or senescence, must be defended. An example of this approach can be found in a recent report in which levels of EC chimerism of 25% and vascular SMC chimerism of 3.5% have been found in sex-mismatched heart transplants in the absence (0.0016%) of myocyte chimerism [95].

Hematopoietic Stem Cells and Myocyte Formation

  1. Top of page
  2. Abstract
  3. Introduction
  4. Possibility 1: The Heart Is a Terminally Differentiated Postmitotic Organ
  5. Possibility 2: The Heart Is Not a Terminally Differentiated Postmitotic Organ
  6. Myocyte Division
  7. Cardiac Chimerism and Myocyte Formation
  8. Hematopoietic Stem Cells and Myocyte Formation
  9. Cardiac Stem Cells and Myocyte Formation
  10. Disclosures
  11. Acknowledgements
  12. References
  13. Supporting Information

In the last decade, major discoveries have been made concerning the biology of adult HSCs. They can differentiate into cell lineages distinct from the organ in which they reside and into cells derived from a different germ layer [126]. These properties were considered to be restricted to embryonic stem cells, and because of their dramatic biological and clinical implications, they triggered a vigorous debate in the scientific community. The controversy reached unexpectedly high levels of intensity when HSCs were shown to be able to migrate to sites of injury repairing damage in various organs, in particular the brain and the heart [127, [128]129]. The brain and the heart have always been viewed as nonpermissive organs for exogenous and endogenous tissue regeneration. Studies presenting positive results [127, 129, [130], [131], [132], [133], [134]135] were immediately confronted by negative reports [136, [137], [138], [139]140], which challenged stem cell plasticity and the emerging new paradigm of organ and organism homeostasis and repair [141, [142]143]. Although the possibility that HSCs differentiate into Purkinje neurons was questioned based on data interpretation or inability to reproduce similar findings, a different reaction occurred in response to the documentation of the efficacy of adult HSCs in myocardial regeneration after infarction. Several laboratories have used genetically labeled HSCs from transgenic animals to promote the restoration of infarcted myocardium experimentally to identify the progeny formed [129, 133, [134]135]. HSCs expressing enhanced green fluorescence protein (EGFP) and sex-mismatched transplantation have been used frequently to have two independent markers for the recognition of newly generated myocardium: EGFP expression and Y-chromosome labeling [129, 134, 135]. 5-Bromo-2′-deoxyuridine (BrdU) was also administered chronically to animals after infarction and cell treatment to have a third marker of cell accumulation within the infarcted region (Fig. 6A–6F).

thumbnail image

Figure Figure 6.. Bone marrow cell transdifferentiation. Male c-kit-positive bone marrow cells were obtained from EGFP-transgenic mice and injected shortly after infarction in the female mouse heart. (A–C): The regenerated myocardium is composed of small myocytes positive for α-actinin (A, C) (red) that express EGFP (B, C) (green) and have a male genotype (C) (Y chromosome, white dots). (D–F): Again, the new myocytes express α-actinin (D, F) (red) and EGFP (E, F) (green). The vast majority of myocyte nuclei are positive for the cell cycle marker BrdU (F) (white). (G–I): The regenerated myocytes express α-actinin (G, I) (red) and EGFP (H, I) (green), and the nuclei are positive for MEF2C (I) (yellow). (J–L): A newly formed arteriole (J, L) (α-smooth muscle actin, red) expresses EGFP (K, L) (green); smooth muscle cells are positive for GATA-6 (L) (white). (M–O): The formed capillaries (M, O) (vWf, red) are EGFP-positive (N, O) (green), and endothelial cells express the transcription factor Ets1 (O) (white). Abbreviations: BrdU, 5-bromo-2′-deoxyuridine; EGFP, enhanced green fluorescence protein; vWf, von Willebrand factor.

Download figure to PowerPoint

Following the detection of EGFP-, Y-chromosome-, and BrdU-positive cells, transcription factors and cytoplasmic proteins specific to myocytes and vascular smooth muscle and endothelial cells were demonstrated (Fig. 6G–6O) so as to avoid potential questions on the validity and accuracy of the results [129, 135]. However, these observations were attacked on methodological grounds and were claimed to be the product of a new artifact [2, 9, 10, 138, 139, 144]. The characteristics of Artifact 4 were unexpected because a new twist was introduced in the “erroneous” interpretation of morphological images obtained by immunolabeling and confocal microscopy: autofluorescence. How these multiple markers of myocardial reconstitution by HSCs could all be the result of autofluorescence is an interesting issue, which unfortunately was never addressed. The autofluorescence was introduced by the same authors who raised the issues of Artifacts 1–3 and who, this time, joined forces with experimental hematologists opposing the concept of HSC plasticity [8]. So the never-documented autofluorescence artifact was invoked as the only logical explanation for the possibility that cardiac repair can be accomplished by exogenous cells.

Criticisms have focused on the need to use genetic markers to demonstrate whether HSCs can form heart muscle [2, [3]4, 9, 10]. Genetic markers have been used in the questioned positive studies [129, 135], as well as in the negative reports that have been claimed to be valid [138, [139]140]. Both positive and negative publications used immunostaining, but in the former case, autofluorescence was assumed to be a factor, whereas in the latter case, the technique used was considered immaculate. Unfortunately, the basis for this conclusion remains an enigma. Although EGFP labeling was applied in several of the papers that opposed HSC plasticity [139, 140, 145], one of these laboratories introduced the notion that green autofluorescence is commonly present in skeletal muscle samples and therefore the detection of EGFP-positive cells in the heart or skeletal muscle following implantation of EGFP-positive HSCs is by necessity an artifact [146]. This concept was adopted immediately by the International Society of Stem Cell Research [147] and emphasized in original studies and reviews [2, 10].

The rationale for Artifact 5 is incomprehensible. The green autofluorescence observed in histologic sections of skeletal muscle was largely due to the use of an inappropriate fixation protocol for immunolabeling studies. A mixture of paraformaldehyde and glutaraldehyde (Fig. 7A) was used rather than phosphate-buffered formalin (Fig. 7B), which carries minimal background autofluorescence. But most importantly, immunolabeling of glutaraldehyde-fixed tissue never works (Fig. 7C, 7D) because of the high level of cross-linking of proteins generated by glutaraldehyde [148]. Green autofluorescence can be clearly distinguished from EGFP-expressing cells by the use of specific green fluorescence protein antibodies as opposed to direct identification of the EGFP protein by epifluorescence light microscopy (Fig. 7E–7G). Therefore, immunolabeling can be easily optimized so that the emission signal is orders of magnitude greater than background fluorescence, or autofluorescence can be totally excluded by direct labeling of antibodies with quantum dots (Fig. 7H–7K).

thumbnail image

Figure Figure 7.. Autofluorescence of tissue sections and detection of EGFP in skeletal muscle. Autofluorescence (green) of tissue sections obtained from skeletal muscle after fixation with paraformaldehyde and glutaraldehyde (A) or formalin (B), both in the absence of immunolabeling. Nuclei are stained with 4,6-diamidino-2-phenylindole (blue). (C, D): After fixation with paraformaldehyde and glutaraldehyde, skeletal muscle tissue sections were labeled with α-skeletal actin (C) and EGFP (D) antibodies. The lack of sarcomere striation indicates that only unspecific labeling of these two antigens was obtained with this protocol. The signals in (C) and (D) correspond to autofluorescence. In this regard, compare (D) with (A). After fixation with formalin only, skeletal muscle tissue sections were labeled with α-skeletal actin (E, G) (red) and EGFP (F, G) (green) antibodies. The presence of striations in the cell cytoplasm indicates that specific labeling was obtained by this approach. Note the coexistence of EGFP-positive and EGFP-negative fibers in the same muscle. (H–J): After fixation with formalin, skeletal muscle tissue sections were labeled with α-skeletal actin (H, J) (red) and EGFP (I, J) (green) antibodies conjugated with quantum dots. The presence of striations in the cell cytoplasm indicates the specificity of labeling. With this approach, the autofluorescence of a tissue section can be totally avoided: the confocal image illustrated in (K) was acquired with the excitation and emission wavelengths used for quantum dot 655. Note the complete absence of autofluorescence. Importantly, the excitation and emission wavelengths used for quantum dot analysis are distinct from the autofluorescence wavelength of tissue sections. Abbreviation: EGFP, enhanced green fluorescence protein.

Download figure to PowerPoint

What remains unclear is the ultimate goal of these unqualified criticisms. They add to the already enormous amount of confusion and uncertainty present in the field of regenerative medicine. The critics of myocardial and skeletal muscle repair by exogenous or endogenous stem cells continue to implement in their work the same methodologies used by the investigators being criticized. The bias lies in the preconceived notion that immunolabeling and confocal microscopy are good for some but bad for others. We have examined two recent issues each of Nature, Nature Medicine, Science, and Proceedings of the National Academy of Sciences of the United States of America and found that 7 of 17, 13 of 25, 3 of 11, and 14 of 22 publications in biomedical sciences, respectively, have used immunostaining and confocal microscopy. Remarkably, in none of these 37 studies was the issue of autofluorescence discussed as a source of artifacts.

Cardiac Stem Cells and Myocyte Formation

  1. Top of page
  2. Abstract
  3. Introduction
  4. Possibility 1: The Heart Is a Terminally Differentiated Postmitotic Organ
  5. Possibility 2: The Heart Is Not a Terminally Differentiated Postmitotic Organ
  6. Myocyte Division
  7. Cardiac Chimerism and Myocyte Formation
  8. Hematopoietic Stem Cells and Myocyte Formation
  9. Cardiac Stem Cells and Myocyte Formation
  10. Disclosures
  11. Acknowledgements
  12. References
  13. Supporting Information

In the last 3 years, several laboratories have identified progenitor cells in the adult heart of small and large mammals, including humans. Cardiac progenitor cells are multipotent in vitro and give rise to cardiomyocytes and coronary vessels in vivo, thus possessing the fundamental properties of stem cells [13, 48, 149, [150], [151]153]. The recognition that stem cells reside in the myocardium had a dramatic impact on the field of cardiology from a biological and clinical perspective. This discovery has provided the missing link between the documentation of small dividing myocytes [6, 7, 12] and the uncertainty concerning the origin of these repopulating cells, and it has laid the groundwork for the possibility of introducing the use of these cells in the treatment of the failing human heart [12, 154]. A number of laboratories have invested resources and staff in exploring this novel and radically different perspective of cardiac pathophysiology [151, 152, 154]. Because of these efforts, major advances have been made, and preclinical studies are in progress with the objective of rapidly implementing this technology in humans.

The discovery that cardiac stem cells (CSCs) are present in the adult myocardium led to the proposal of a new artifact. The formulation of Artifact 6 required vivid imagination and unusual logic; since myocytes cannot be formed [1, 2], the preparations of CSCs must be a product of an artifact associated with the in vitro isolation [10]. The critics then go on to suggest that this artifact may, however, have some positive outcomes. But to question the presence and role of CSCs in the adult heart, Artifact 6 could not be restricted to the in vitro work and had to be extended to the demonstration that CSCs regenerate infarcted myocardium in vivo [13]. This task was accomplished by stating that cells that expressed cardiac myosin heavy chain, connexin 43, and N-cadherin and responded to electrical stimulation by shortening and relengthening were in reality fibroblasts [10].

In summary, the controversy concerning stem cell therapy for the damaged heart has its foundations in the debate that originated nearly 80 years ago when it was concluded that myocardial regeneration ceases at birth and cannot occur in adulthood or senescence [31]. The biochemical characterization of cardiac hypertrophy performed in the late 1960s and 1970s has contributed to reaffirming the notion that myocardial growth postnatally can be accomplished only by enlargement of the pre-existing myocytes [36], strengthening the notion of the heart as an organ formed by a predetermined number of myocytes incapable of reentering the cell cycle and dividing. Molecular cardiology was built on this inherited inviolable paradigm [57], which has been strongly supported by traditional cardiovascular pathologists reluctant to introduce modern technology in the analysis of the structure of the normal and diseased heart. Surprisingly, myocyte death and myocyte formation, which are the two critical variables that control cardiac cell number, are rarely evaluated concurrently to obtain a dynamic view of the heart in its various phases of life [45, 75, 76, 155]. By this simple approach, it would become clear whether the heart can function with the same cells throughout the lifespan of the organ and organism or whether an alternative possibility has to be considered.

References

  1. Top of page
  2. Abstract
  3. Introduction
  4. Possibility 1: The Heart Is a Terminally Differentiated Postmitotic Organ
  5. Possibility 2: The Heart Is Not a Terminally Differentiated Postmitotic Organ
  6. Myocyte Division
  7. Cardiac Chimerism and Myocyte Formation
  8. Hematopoietic Stem Cells and Myocyte Formation
  9. Cardiac Stem Cells and Myocyte Formation
  10. Disclosures
  11. Acknowledgements
  12. References
  13. Supporting Information

Supporting Information

  1. Top of page
  2. Abstract
  3. Introduction
  4. Possibility 1: The Heart Is a Terminally Differentiated Postmitotic Organ
  5. Possibility 2: The Heart Is Not a Terminally Differentiated Postmitotic Organ
  6. Myocyte Division
  7. Cardiac Chimerism and Myocyte Formation
  8. Hematopoietic Stem Cells and Myocyte Formation
  9. Cardiac Stem Cells and Myocyte Formation
  10. Disclosures
  11. Acknowledgements
  12. References
  13. Supporting Information
FilenameFormatSizeDescription
Supp_Table.pdf125KSupplemental Table

Please note: Wiley Blackwell is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.