For nearly a century, the adult heart has been considered a

Transcription

1 Cardiomyogenesis in the Adult Human Heart Jan Kajstura, Konrad Urbanek, Shira Perl, Toru Hosoda, Hanqiao Zheng, Barbara Ogórek, João Ferreira-Martins, Polina Goichberg, Carlos Rondon, Domenico D Amario, Marcello Rota, Federica del Monte, Donald Orlic, John Tisdale, Annarosa Leri, Piero Anversa Rationale: The ability of the human heart to regenerate large quantities of myocytes remains controversial, and the extent of myocyte renewal claimed by different laboratories varies from none to nearly 20% per year. Objective: To address this issue, we examined the percentage of myocytes, endothelial cells, and fibroblasts labeled by iododeoxyuridine in postmortem samples obtained from cancer patients who received the thymidine analog for therapeutic purposes. Additionally, the potential contribution of DNA repair, polyploidy, and cell fusion to the measurement of myocyte regeneration was determined. Methods and Results: The fraction of myocytes labeled by iododeoxyuridine ranged from 2.5% to 46%, and similar values were found in fibroblasts and endothelial cells. An average 22%, 20%, and 13% new myocytes, fibroblasts, and endothelial cells were generated per year, suggesting that the lifespan of these cells was approximately 4.5, 5, and 8 years, respectively. The newly formed cardiac cells showed a fully differentiated adult phenotype and did not express the senescence-associated protein p16 INK4a. Moreover, measurements by confocal microscopy and flow cytometry documented that the human heart is composed predominantly of myocytes with 2n diploid DNA content and that tetraploid and octaploid nuclei constitute only a small fraction of the parenchymal cell pool. Importantly, DNA repair, ploidy formation, and cell fusion were not implicated in the assessment of myocyte regeneration. Conclusions: Our findings indicate that the human heart has a significant growth reserve and replaces its myocyte and nonmyocyte compartment several times during the course of life. (Circ Res. 2010;107:00-00.) Key Words: myocyte regeneration cell lifespan DNA repair ploidy cell fusion For nearly a century, the adult heart has been considered a postmitotic organ in which the number of parenchymal cells is established at birth and cardiomyocytes lost with age or as a result of cardiac diseases cannot be replaced by newly formed cells. The recent explosion of the field of stem cell biology, with the recognition that the possibility exists for extrinsic and intrinsic regeneration of myocytes and coronary vessels, 1 has imposed a reevaluation of cardiac homeostasis and pathology. Several laboratories have identified resident cardiac stem cells (CSCs) in the developing, postnatal, and adult heart of animals and humans, 2 4 suggesting that myocyte turnover and tissue regeneration may be more profound than previously predicted. The documentation that CSCs reside in the myocardium, are stored in discrete niche structures, and divide symmetrically and asymmetrically in vitro and in vivo 4 makes the heart a self-renewing organ. Cardiac cells continuously lost by wear and tear are constantly replaced by activation and commitment of CSCs. 5 Based on retrospective 14 C birth dating of cells, the claim has been made that throughout life, myocyte turnover in humans is restricted to a subset of 50% of cardiomyocytes. 6 Although the process of myocyte regeneration was confirmed, these data pointed to a modest degree of myocyte renewal, contrasting previous evaluations of myocyte regeneration in the normal, hypertrophied, and failing human heart. 7 9 To assess in a more precise manner the growth potential of the adult myocardium, the percentage of myocytes labeled by a thymidine analog was determined in hearts collected postmortem from 8 cancer patients who received infusion of the radiosensitizer iododeoxyuridine (IdU) for therapeutic purposes. 10 We have taken advantage of the fact that halogenated nucleotides are rapidly incorporated in cycling cells at delivery and are progressively diluted in the derived progeny, 11 providing a measurement of myocyte formation over time. This strategy mimics the pulse chase assay performed in animals. 5 Methods The presence of IdU in myocyte, endothelial cell, and fibroblast nuclei was determined by immunocytochemistry and confirmed by Original received April 28, 2010; revision received May 20, 2010; accepted May 21, In April 2010, the average time from submission to first decision for all original research papers submitted to Circulation Research was 15.2 days. From the Departments of Anesthesia and Medicine and the Cardiovascular Division, Brigham and Women s Hospital, Harvard Medical School, Boston, Mass; the National Heart, Lung and Blood Institute, National Institutes of Health, Bethesda, Md; and the Cardiovascular Institute, Beth Israel Deaconess Medical Center, Harvard Medical School, Boston, Mass. Correspondence to Jan Kajstura, PhD, or Piero Anversa, MD, Departments of Anesthesia and Medicine, and Cardiovascular Division, Brigham and Women s Hospital, Harvard Medical School, 75 Francis St, Boston, MA jkajstura@partners.org or panversa@partners.org 2010 American Heart Association, Inc. Circulation Research is available at DOI: /CIRCRESAHA

2 2 Circulation Research July 23, 2010 Non-standard Abbreviations and Acronyms IdU CSCs ECs NPC iododeoxyuridine cardiac stem cells endothelial cells nuclear pore complex spectral analysis and biochemical assay. Data in cardiac cells were measured quantitatively and their average lifespan evaluated. Protocols are described at Results Regeneration of Myocytes The extent of myocyte renewal reported by different laboratories varies dramatically,2,6,7,12 requiring an extremely cautious approach. For this purpose, human samples were coded at the National Institutes of Health (NIH), and all data were collected blindly at Brigham and Women s Hospital (BWH). Subsequently, 3 investigators from BWH (J.K., T.H., K.U.) went to Bethesda and repeated with our NIH coauthors IdU staining and analysis of the 8 hearts. Only NIH reagents were used. Although the results were in close agreement with those obtained at BWH, 2 of our NIH colleagues (S.P., J.T.) came to BWH and reevaluated IdU labeling of myocytes in 3 myocardial samples with the lowest, intermediate, and highest degree of positive cells. Again, essentially identical values were found. Stringent criteria were applied in the evaluation of IdUlabeled cardiac cells (Online Supporting Text I; Online Figure I through III). IdU-positive myocyte and nonmyocyte nuclei were identified in the left ventricle of all 8 cases (Figure 1A and Online Figure IV). Mitotic images labeled by IdU were also found (Figure 1B). With the exception of 1 sample in which the percentage of IdU-positive myocyte nuclei was 2.5%, in the other 7 patients, the fraction of labeled myocytes varied from 11% to 46% (Figure 1C). IdU staining was never detected in the left ventricle of 6 normal hearts not exposed to the halogenated nucleotide, which were Figure 1. IdU in human cardiomyocytes. A, Several myocyte nuclei (arrows) and nonmyocyte nuclei (arrowheads) are labeled by IdU (green). Myocytes are positive for -sarcomeric actin ( -SA, red). Laminin (white) defines the boundary of myocytes and interstitial cells. B, Metaphase chromosomes in a dividing myocyte nucleus (upper panel) are labeled by IdU (lower panel). C, Percentage of IdU-positive myocyte nuclei in each of the 8 patients injected with the thymidine analog. The interval between each IdU administration and patient s death is shown. IdU was not detected in control hearts. The average degree of myocyte formation in the 8 patients is also shown. D, Myocyte nuclei included in rectangles show punctuated IdU labeling (white). Diffuse IdU labeling of myocyte nuclei (arrows). Both punctuated and diffuse IdU labeling were negative for MCM5, indicating that DNA synthesis was no longer present at the time of sampling of the myocardium at patient s death. Positive controls for MCM5 (yellow) are shown in the 2 small lower panels.

3 Kajstura et al Cardiomyocyte Renewal in Humans 3 Figure 2. DNA repair and ploidy. A, Punctuated (rectangles) and diffuse (rectangles and arrows) localization of bromodeoxyuridine (BrdU) in myocyte nuclei from fresh samples of human myocardium perfused for 1 hour with 5 mol/l of the thymidine analog. Ki67 (yellow) is present only in uniformly BrdU-labeled myocyte nuclei. B, The increased intensity of the DNA dye propidium iodide (PI, red) is apparent in the enlarged myocyte nucleus (rectangle), reflecting an octaploid DNA content. Lower levels of PI fluorescence are present in the smaller diploid myocyte nuclei (circles). Myocytes: -SA, white. used as negative controls. In view of the difficulty to evaluate all the variables present in these patients regarding the dose and number of infusions of IdU, disease state, its duration, and antineoplastic therapy (Online Table I; Online Supporting Text II; Online Figure V), the 8 individual values were combined. An average 24% new myocytes were formed over an average period of 405 days, or 22% new myocytes were formed per year, or 0.06% new myocytes were formed per day. Because in the age range of these patients, the left ventricle contains myocytes, myocytes were generated per day or myocytes were generated in 405 days (Figure 1C). Three potential confounding factors had to be considered in the quantitative analysis of IdU labeling of myocyte nuclei: DNA repair, polyploidy, and cell fusion. These variables may lead to an overestimation of the degree of myocyte regeneration. DNA Repair DNA repair results in a punctuated pattern of IdU labeling,14 reflecting its integration at discrete sites of DNA damage. Conversely, uniform IdU incorporation identifies dividing cells in which the entire genome is replicated. Foci of IdU were occasionally seen in myocyte nuclei (Figure 1D), possibly caused by the low sensitivity of this methodology for the detection of DNA repair.15 Importantly, only nuclei with diffuse IdU labeling were measured (Figure 1A and 1B; Online Supporting Text I; Online Figure I through IV). To further validate our approach, large pieces of human myocardium from 2 explanted hearts were perfused with bromode-

4 4 Circulation Research July 23, 2010 Figure 3. Distribution of DNA content in myocyte nuclei. A and B: Frequency distribution of DNA content in IdU-positive (green bars) and IdU-negative (red bars) cardiomyocytes. Values are shown individually in each of the 8 patients (A) and as an average (B). The fraction of Ki67-positive myocyte nuclei is illustrated in black (B). Lymphocytes from human tonsils were used as control for 2n DNA content (yellow bars). C, Frequency distribution of DNA content in nuclei isolated from pure preparations of human myocytes. D, Three human myocyte nuclei show 2, 4, and 8 X-chromosomes (green dots). E, In these 3 sections of human myocardium, only 1 octaploid enlarged myocyte nucleus (PI, red) is present. Note the rather uniform size of the large majority of diploid myocyte nuclei. Myocytes: -SA, white. oxyuridine (BrdU). The colocalization of BrdU and Ki67 in myocyte nuclei was then determined because the cell cycle protein Ki67 is not implicated in DNA repair.16 By this protocol, we showed that the speckled appearance of BrdU in nuclei was negative for Ki67; this was not the case when the halogenated nucleotide was evenly distributed throughout the nucleus (Figure 2A). Ploidy and Cell Fusion Polyploidy is characterized by an exponential increase in DNA content dictated by the number of doublings of the entire genome from 2n to 4n to 8n. This exponential increase in DNA content is reflected by a significant increase in nuclear volume (Figure 2B). To evaluate the contribution of polyploidy, DNA content per nucleus was measured by confocal microscopy in IdU-positive and IdU-negative myocyte nuclei and compared with that of lymphocytes from human tonsils, which were used as control for diploid nuclei. In the majority of cases, 2n diploid DNA content was found in IdU-positive and IdU-negative myocyte nuclei. In a small number of IdU-positive and IdU-negative myocyte nuclei, DNA content was higher than 2n but less than 4n, suggesting

5 Kajstura et al Cardiomyocyte Renewal in Humans 5 Figure 4. IdU in human myocytes and nonmyocytes. A and B, IdU labeling (green) of fibroblast nuclei (A, procoll, yellow; arrows) and EC nuclei is shown (B, vwf, white; arrows). C, Percentage of IdU-positive fibroblasts and ECs in each case. Only the interval from the first IdU injection to patient s death is listed below each bar. Moreover, the percent difference in the degree of IdU labeling of myocytes, fibroblasts, and ECs is shown. *P vs myocytes; **P 0.05 vs fibroblasts. D, Chemiluminescent signals generated by IdU in myocardial samples (upper panel, black dots). Binding of the ss/ds DNA antibody to the membrane (central panel, black dots). Methylene blue staining of the DNA loaded and fixed onto the membrane is shown in the lower panel (blue dots). HEK293 cells cultured with, HEK( ), and without, HEK( ), the thymidine analog were used as positive and negative control, respectively. B16( ), B16 mouse melanoma cells cultured in the absence of the halogenated nucleotide. Buffer indicates buffer only. that these cells represented cycling, Ki67-positive, amplifying myocytes that had not reached terminal differentiation (Figure 3A and 3B). A few Ki67-labeled nuclei had 4n DNA content reflecting amplifying myocytes in G2. Additionally, DNA content was measured by flow cytometry in nuclei collected from pure preparations of myocytes isolated from normal and pathological hearts (n 8; Online Table II). Diploid, tetraploid, and octaploid nuclei represented 83%, 14%, and 3% of myocyte nuclei, respectively (Figure 3C). Rare examples of tetraploid and octaploid nuclei together with the majority of diploid nuclei were also found by Q-FISH in IdU-negative cells of cancer patients (Figure 3D). These observations are consistent with the relatively similar size of myocyte nuclei in sections of human myocardium (Figure 3E). To test the possibility of cell fusion, the number of sex chromosomes was determined in 614 myocyte nuclei from 4 cases with high levels of IdU-positive cells. Only 2 X-chromosomes or 1 X- and 1 Y-chromosome were identified in 591 nuclei (Online Figure VI). In 23 nuclei, more than 2 sex chromosomes were detected. However, dividing phospho-h3-positive myocytes with 2 sets of sex chromosomes were found, further decreasing the number of potential fusion events in the human myocardium. Therefore, the human heart is mostly composed of diploid cells, and cell fusion is at most an occasional phenomenon. Because the proportion of mononucleated and binucleated myocytes does not change in the human left ventricle with aging, cardiac hypertrophy, and ischemic injury, 17 measurements at the nuclear level are equivalent to measurements at the cellular level. Regeneration of Endothelial Cells and Fibroblasts Myocyte renewal was then compared with that of fibroblasts and endothelial cells (EC)s (Figure 4A through 4C). The percentage of IdU-positive fibroblasts was similar to that of myocytes, whereas the fraction of labeled ECs was 37% (P 0.01) and 33% (P 0.005) lower than myocytes and fibroblasts, respectively. By inference, the average lifespan of cardiomyocytes, fibroblasts, and ECs was estimated to be approximately 4.5, 5, and 8 years, respectively. A biochemical method was developed to provide an independent assessment of the presence of IdU in the human myocardium. Genomic DNA was extracted from tissue sections obtained blindly from 6 cases. DNA collected from HEK-293 cells cultured in the presence and absence of the thymidine analog was used as positive and negative control, respectively. DNA was blotted onto individual spots on nylon membranes and subjected to Western blotting. IdU was found in all cases (Figure 4D), confirming that it was incorporated in the ndna. Cardiac Cell Formation and Death To establish whether the similar rates of cardiac cell renewal were consistent with comparable levels of cell death, apopto-

6 6 Circulation Research July 23, 2010 Figure 5. Cell proliferation. A through C, Mitosis in a human myocyte (A), fibroblast (B), and EC (C) is recognized by phospho-h3 labeling (bright blue, arrows). Dividing cells are shown at higher magnification in the insets; DAPI, left insets; phospho-h3, right insets. D and E, Percentage of Ki67- (D) and phospho-h3-labeled (E) fibroblasts, ECs and myocytes in patients exposed to IdU. Values for myocytes in 6 control hearts are also shown. F, Comparison of the average values of Ki67 and phospho-h3 in patients (Pt) treated with IdU. *P 0.05 vs fibroblasts, **P 0.05 vs ECs. P 0.05 vs cardiomyocytes of patients. C indicates control hearts. sis of myocytes, ECs, and fibroblasts was determined. Also, the number of cycling cells and their mitotic index was evaluated by Ki67 and phospho-h3, respectively, to define the characteristics of the myocardium at patient s death. In the human heart, 9,18 cell apoptosis is invariably coupled with the expression of the aging-associated protein p16 INK4a that is a critical determinant of senescence and growth arrest of progenitor cells in various organs. 19 Thus, the interaction of p16 INK4a, growth inhibition, and apoptosis was tested to obtain a common denominator of the processes that regulate replication and death of cardiac cells. In the 8 hearts, the percentage of p16 INK4a -myocytes was higher than that of fibroblasts and ECs but, in all cases, the fraction of p16 INK4a -cells exceeded the fraction of cells undergoing apoptosis. Cell death was detected only in senescent fibroblasts, ECs, and myocytes. Small differences were found in the degree of apoptosis, but fibroblasts and myocytes were the most affected (Online Figure VII). The data on p16 INK4a /cell death were paralleled by similar results on cell proliferation. Ki67 and phospho-h3 labeling were comparable in myocytes and fibroblasts and lower in ECs (Figure 5A through 5F; Online Figure VIII). In each case, the magnitude of cell regeneration mimicked the extent of cell death, suggesting that a balance appeared to be operative between these 2 determinants of cardiac homeostasis. These observations challenge the notion that significant differences exist in the degree of cell renewal between myocytes and nonmyocytes, pointing to a relatively similar lifespan of cardiac cells in the adult human heart. The heart of the 8 cancer patients had pathological manifestations (Online Table I), which may have contributed to the extent of myocyte renewal. Importantly, IdU amplifies the effects of radiation therapy, having little or no impact on cell function alone. Thus, the fraction of cycling myocytes and their mitotic index were evaluated in 6 normal hearts (Online Table III). Both parameters were significantly lower than those measured in cancer patients (Figure 5F), confirming previous results in which cardiac pathology has been found to enhance myocyte regeneration. An additional possibility may involve the release of growth factors from cancer cells, which could have influenced the growth of CSCs and myocytes. Most importantly, the findings in cancer patients do not reflect only physiological cell turnover but a combination of this process with different disease states.

7 Kajstura et al Cardiomyocyte Renewal in Humans 7 Figure 6. Myocyte formation in the postnatal human heart. A through C, Metaphase (A and B) and telophase (C) chromosomes (green, arrows) in myocytes ( -SA, red) of a human heart at 2 years of age. Mitoses are labeled by phospho-h3 (insets, blue). Birth Dating of Cardiac Cells Comparable degrees of myocyte and nonmyocyte death and renewal were found in the present study. However, retrospective 14 C birth dating has claimed that the average lifespan of nonmyocytes is 4 years in young and old human beings, whereas the lifespan of myocytes is 18-fold and 40-fold longer at 25 and 75 years of age, respectively. 6 Several factors may explain these differences in the degree of myocyte formation in the human heart. 14 C birth dating of human myocytes was restricted to nuclei expressing the contractile protein troponin I (TnI), but whether this subset of nuclei was representative of the myocyte population was not determined. Moreover, the isolation of myocyte nuclei from specimens of frozen myocardium is problematic, and the presence of myocardial scarring in the pathological hearts included in the study further interfered with the isolation and sampling of myocyte nuclei. The mathematical model used assumes that the number of cardiomyocytes remains constant throughout life, although myocyte number increases with postnatal development 20 (Figure 6) and myocyte loss progressively occurs as a function of age 13 and cardiac disease. 21 In our 8 cases, the fraction of TnI-positive myocyte nuclei varied from 33% to 61%, and 98% of TnI-positive myocyte nuclei expressed p16 INK4a ; 1% expressed only TnI or p16 INK4a (Figure 7A and 7B). Additionally, 2% of TnIpositive p16 INK4a -positive myocyte nuclei were labeled by IdU but none by Ki67 or phospho-h3 (Online Figure IX), suggesting that the thymidine analog was incorporated before this small subset of cardiomyocytes reached cellular senescence. The expression of Nkx2.5 and GATA4 was tested to discriminate myocytes from nonmyocytes, 6 but these transcription factors are not uniformly present in myocyte nuclei 22 so that the lack of these nuclear proteins cannot be equated to a nonmyocyte phenotype. The consistent association of TnI and p16 INK4a in myocytes together with the absence of markers of cell cycle progression and mitosis strongly suggests that the localization of TnI in nuclei identified almost exclusively a category of senescent cells. The colocalization of TnI and p16 INK4a was also examined in 7 normal human hearts, 46 to 74 years old (Online Table IV), and found to be present in 90% of myocyte nuclei (Figure 7C). The fraction of TnI-positive nuclei varied from 17% to 60%, constituting an inconsistent marker for the isolation of a representative pool of myocytes. Further, 7 of the 8 pure preparations of isolated human myocytes (Online Table II) were used for the collection of nuclei and FACS analysis. The fraction of TnI-labeled myocyte nuclei ranged from 22% to 73%, and the coexpression of TnI and p16 INK4a involved 80% of nuclei (Figure 7D). Aging and Permeability of Myocyte Nuclei To identify the mechanism responsible for the expression of TnI in senescent cells, the nuclear properties of myocytes isolated from Fischer 344 rats at 3 and 24 months of age were determined. Nuclear pore complexes (NPCs) are assembled during mitosis, and their proteins do not turnover during interphase, 23 suggesting that alterations in NPC function may occur in long-lived cells. This hypothesis was tested in nuclei of

8 8 Circulation Research July 23, 2010 Figure 7. Nuclear TnI and myocyte senescence. A, Section of human myocardium in which TnI in myocyte nuclei (left panel, red, arrows) colocalizes with p16 INK4a (central panel, bright blue, arrows); nuclei are stained by DAPI (white). Right panel, Merge. B, Expression of TnI and p16 INK4a in myocyte nuclei of the 8 patients treated with IdU. C, Expression of TnI and p16 INK4a in myocyte nuclei of 7 normal hearts not exposed to IdU. D, Bivariate distribution of coexpression of TnI and p16 INK4a in pure preparations of myocyte nuclei. Q1 indicates nuclei negative for TnI and p16 INK4a ; Q2, nuclei positive for p16 INK4a only; Q3, nuclei positive for TnI only; and Q4, nuclei positive for TnI and p16 INK4a. Individual values are indicated by green bars and average values by blue bars. postmitotic cells of the rat brain cortex. Old cells showed defects in the permeability of NPCs with translocation to the nucleus of proteins restricted physiologically to the cytoplasm. 24 We applied similar protocols and used 70-kDa dextran to detect alterations in the permeability of myocyte nuclei. This molecule does not cross the nuclear membrane in the presence of intact NPCs. 24 Only 30% of nuclei were permeable to 70-kDa dextran in young myocytes, whereas this polysaccharide was found in 80% of old myocyte nuclei (Figure 8A and 8B); this may reflect the comparable levels of p16 INK4a present in young and old myocytes. 25 Importantly, TnI was detected in 95% of myocyte nuclei that were permeable to the 70-kDa dextran, and spectral analysis confirmed the specificity of the signals (Figure 8C through 8E). As previously shown in nuclei of postmitotic cells of rat brain cortex, 24 the deterioration of NPCs in old myocyte nuclei was coupled with a 2.2-fold downregulation of the scaffold nucleoporin Nup93 (Figure 8F). Thus, the translocation of TnI from the cytoplasm to the nucleus occurs only in leaky nuclei of senescent myocytes. Discussion In the past few years, several hypotheses have been raised to clarify the findings of dividing myocytes in the normal and diseased heart. They vary from the existence of a small class of myocytes that dedifferentiate and subsequently proliferate to a subset of cells that retain the ability to reenter the cell cycle and multiply. 2 Alternatively, progenitor cells from the bone marrow may migrate to the myocardium, where, after transdifferentiation, lead to the formation of a myocyte progeny. However, the search for the origin of dividing myocytes remained unclear until a resident stem cell com-

9 Kajstura et al Cardiomyocyte Renewal in Humans 9 Figure 8. TnI expression and lifespan of cardiac cells. A and B, Myocyte nuclei (DAPI, blue) exclude FITC-500 kda dextran (negative control, green). TRITC-70-kDa dextran (red) diffuses into a fraction of myocyte nuclei (arrows); this value is higher in nuclei from senescent hearts (B). C, Myocyte nuclei (DAPI, blue) positive for TRITC-70-kDa dextran (red) express TnI (white). Almost all dextran-permeable myocyte nuclei express TnI; unlabeled nuclei (circles). D, Fraction of dextran TnI-positive myocyte nuclei in young and old hearts; almost all dextran-permeable myocyte nuclei express TnI. E, Emission spectra from myocyte nuclei positive for TRITC- 70-kDa dextran and TnI (green lines). Blue lines correspond to autofluorescence of myocyte nuclei negative for TRITC 70-kDa dextran and TnI. F, Nup93 expression in myocyte nuclei. Loading conditions, lamin B1. Expression levels are shown by optical density (OD). *P partment was recognized in the adult heart. 2 4 This information has changed the interpretation of myocyte regeneration and has advanced our understanding of the regulatory processes involved in cardiac homeostasis and tissue repair. Replicating myocytes are small, mononucleated cells that represent the progeny of CSCs; they divide and hypertrophy, and these 2 cellular processes coexist until proliferation no longer occurs. Data in the current study indicate that a large number of myocytes is formed every year, suggesting that the entire heart is replaced several times during the course of life in humans. Our findings are in contrast with results obtained by the integration of atmospheric 14 C in the DNA of cardiomyocyte nuclei. 6 An inherent limitation in 14 C birth dating of cardiac cells was the analysis of only TnI-positive senescent myocyte nuclei, which was further confounded by the introduction of

10 10 Circulation Research July 23, 2010 an inappropriate mathematical model. The number of myocytes in the heart and their rate of turnover were considered constant. This model of invariant organ growth defines parenchyma in a steady state in which cell death is compensated by cell regeneration in young healthy individuals. This can hardly be applied to the biology of myocardial aging, hypertension, and acute and chronic myocardial infarction, conditions that were present in the small cohort of 12 patients included in the study. In fact, it is unrealistic to define the physiological mechanisms of myocardial aging with a sample size of 12 individuals of either sex, some of which are affected by severe cardiac pathologies. Additionally, the number of cardiomyocytes was postulated to be established at birth and to remain constant throughout life. Myocytes are formed postnatally, 20 and the decrease in the number of myocytes with age has been well defined. 13 Without basis, 14 C levels were analyzed and interpreted differently in young (19 to 42 years) and old (50 to 73 years) hearts to arbitrarily draw an exponential curve that indicated a progressive decline in cell generation as a function of age. 6 There is no gradual decrease in the rate of myocyte turnover but rather 2 distinct sets of data for young and old hearts. The runs test, used to evaluate the goodness of the fit, demonstrated a significant deviation (P ) from the exponential decay model used. If the data used to assess 14 C birth dating of cardiac cells are corrected for 5 known variables, that is, TnI-positive and TnI-negative myocyte nuclei, p16 INK4a -positive and p16 INK4a -negative myocyte nuclei, and the colocalization of TnI and p16 INK4a in myocyte nuclei, an 18% myocyte renewal per year can be calculated (Online Supporting Text III). The presence of CSCs in the human heart is apparently at variance with the small foci of tissue regeneration commonly found in acute and chronic infarcts or pressure overload hypertrophy. 7 9 CSCs are limited in their capacity to reconstitute necrotic myocardium, 9 and this phenomenon has been interpreted as proof of the inability of the heart to form large quantities of cardiomyocytes. 26 The current findings challenge this view. However, the growth reserve of the heart fails to restore the structural integrity of the myocardium after infarction and healing is associated with scar formation. A possible explanation for this inconsistency has been obtained in animal models of the human disease. 27 Stem cells are present throughout the infarct but die rapidly by apoptosis, following the destiny of myocytes. Myocyte regeneration is restricted to the viable portion of the heart. 7 A similar phenomenon occurs in solid and nonsolid selfrenewing organs including the skin, liver, intestine, and bone marrow. In all cases, occlusion of a supplying artery leads to scar formation mimicking cardiac pathology With polyarteritis nodosa and vasculitis, microinfarcts develop in the intestine and skin. Similarly, with sickle cell anemia, microinfarcts appear in the bone marrow, and resident stem cells do not repair the damaged organ. 31 The progenitor cell compartment may be properly equipped to modulate tissue homeostasis but does not respond effectively to ischemic injury or to aging and senescence of the organ and organism. 19 The level of myocyte renewal documented in the present study in the human heart is consistent with the activation and differentiation of resident CSCs. 4 A rapid turnover of cardiomyocytes regulated by a stem cell compartment has also been shown in rodents, 5,32 projecting a dynamic view of the mammalian myocardium. CSCs possess the fundamental properties of stem cells; they are self-renewing, clonogenic, and multipotent in vitro and in vivo. 4,5 Thus, the IdU-positive cardiomyocytes constituted the progeny of activation and commitment of resident CSCs. Transit amplifying myocytes divide and concurrently mature until the adult phenotype is reached and terminal differentiation is acquired. 2 The notion that cardiac homeostasis and repair in the adult organ are controlled by a pool of CSCs remains controversial. 26 Lineage tracing protocols that are currently considered the gold standard for the identification of progenitor cells 26 do not offer information on the self-renewing property and clonogenicity of primitive cells or clonal origin of daughter cells in vivo. 11 Conversely, the in vivo delivery of single cell derived clonogenic human CSCs results in the generation of cardiomyocytes and coronary vessels, providing strong evidence in favor of the multilineage differentiation of CSCs. 4 Recently, the self-renewal, clonogenicity, and multipotentiality of mouse and human CSCs in vivo was documented by viral gene tagging and the recognition of common sites of integration of the viral genome in CSCs, myocytes, and other cardiac cells. 5 Once more, the rate of myocyte turnover measured by genetic marking was 4 to 5 orders of magnitude higher than that reported more than a decade ago. 33 Sources of Funding This work was supported by National Institutes of Health grants. None. Disclosures References 1. Losordo D, Dimmeler S. Therapeutic angiogenesis and vasculogenesis for ischemic disease, part II: cell-based therapies. Circulation. 2004;109: Anversa P, Kajstura J, Leri A, Bolli R. Life and death of cardiac stem cells: a paradigm shift in cardiac biology. Circulation. 2006;113: Smith RR, Barile L, Cho HC, Leppo MK, Hare JM, Messina E, Giacomello A, Abraham HR, Marban E. Regenerative potential of cardiosphere-derived cells expanded from percutaneous endomyocardial biopsy specimens. Circulation. 2007;115: Bearzi C, Rota M, Hosoda T, Tillmanns J, Nascimbene A, De Angelis A, Yasuzawa-Amano S, Trofimova I, Siggins RW, Cascapera S, Beltrami AP, Zias E, Quaini F, Urbanek K, Michler RE, Bolli R, Kajstura J, Leri A, Anversa P. Human cardiac stem cells. Proc Natl Acad Sci U S A. 2007;104: Hosoda T, D Amario D, Cabral-Da-Silva MC, Zheng H, Padin-Iruegas ME, Ogorek B, Ferreira-Martins J, Yasuzawa-Amano S, Amano K, Ide-Iwata N, Cheng W, Rota M, Urbanek K, Kajstura J, Anversa P, Leri A. Clonality of mouse and human cardiomyogenesis in vivo. Proc Natl Acad Sci U S A. 2009;106: Bergmann O, Bhardwaj RD, Bernard S, Zdunek S, Barnabé-Heider F, Walsh S, Zupicich J, Alkass K, Buchholz BA, Druid H, Jovinge S, Frisén J. Evidence for cardiomyocyte renewal in humans. Science. 2009;324: Beltrami AP, Urbanek K, Kajstura J, Yan SM, Finato N, Bussani R, Nadal-Ginard B, Silvestri F, Leri A, Beltrami CA, Anversa P. Evidence that human cardiac myocytes divide after myocardial infarction. N Engl J Med. 2001;344: Urbanek K, Quaini F, Tasca G, Torella D, Castaldo C, Nadal-Ginard B, Leri A, Kajstura J, Quaini E, Anversa P. Intense myocyte formation from cardiac stem cells in human cardiac hypertrophy. Proc Natl Acad Sci USA. 2003;100:

11 Kajstura et al Cardiomyocyte Renewal in Humans Urbanek K, Torella D, Sheikh F, De Angelis A, Nurzynska D, Silvestri F, Beltrami CA, Bussani R, Beltrami AP, Quaini F, Bolli R, Leri A, Kajstura J, Anversa P. Myocardial regeneration by activation of multipotent cardiac stem cells in ischemic heart failure. Proc Natl Acad Sci U S A. 2005;12: Kinsella TJ. Radiosensitization and cell kinetics: clinical implications for S-phase-specific radiosensitizers. Semin Oncol. 1992;19: Suh H, Consiglio A, Ray J, Sawai T, D Amour KA, Gage FH. In vivo fate analysis reveals the multipotent and self-renewal capacities of Sox2 neural stem cells in the adult hippocampus. Cell Stem Cell. 2007;1: Narula J, Kolodgie FD, Virmani R. Apoptosis and cardiomyopathy. Curr Opin Cardiol. 2000;15: Olivetti G, Melissari M, Capasso JM, Anversa P. Cardiomyopathy of the aging human heart: myocyte loss and reactive cellular hypertrophy. Circ Res. 1991;68: Kao GD, McKenna WG, Yen TJ. Detection of repair activity during the DNA damage-induced G2 delay in human cancer cells. Oncogene. 2001; 20: Palmer TD, Willhoite AR, Gage FH. Vascular niche for adult hippocampal neurogenesis. J Comp Neurol. 2000;425: Scholzen T, Gerdes J. The Ki-67 protein: from the known and the unknown. J Cell Physiol. 2000;182: Olivetti G, Cigola E, Maestri R, Corradi D, Lagrasta C, Gambert SR, Anversa P. Aging, cardiac hypertrophy and ischemic cardiomyopathy do not affect the proportion of mononucleated and multinucleated myocytes in the human heart. J Mol Cell Cardiol. 1996;28: Chimenti C, Kajstura J, Torella D, Urbanek K, Heleniak H, Colussi C, Di Meglio F, Nadal-Ginard B, Frustaci A, Leri A, Maseri A, Anversa P. Senescence and death of primitive cells and myocytes lead to premature cardiac aging and heart failure. Circ Res. 2003;93: Sharpless NE, DePinho RA. How stem cells age and why this makes us grow old. Nat Rev Mol Cell Biol. 2007;8: Rakusan K. Cardiac growth, maturation, and aging. In: Zak R, ed. Growth of the Heart in Health and Disease. New York: Raven Press; 1984: Dorn GW II. Novel pharmacotherapies to abrogate postinfarction ventricular remodeling. Nat Rev Cardiol. 2009;6: Heineke J, Auger-Messier M, Xu J, Oka T, Sargent MA, York A, Klevitsky R, Vaikunth S, Duncan SA, Aronow BJ, Robbins J, Crombleholme TM, Molkentin JD. Cardiomyocyte GATA4 functions as a stress-responsive regulator of angiogenesis in the murine heart. J Clin Invest. 2007;117: D Angelo MA, Anderson DJ, Richard E, Hetzer MW. Nuclear pores form de novo from both sides of the nuclear envelope. Science. 2006;312: D Angelo MA, Raices M, Panowski SH, Hetzer MW. Age-dependent deterioration of nuclear pore complexes causes a loss of nuclear integrity in postmitotic cells. Cell. 2009;136: Kajstura J, Pertoldi B, Leri A, Beltrami CA, Deptala A, Darzynkiewicz Z, Anversa P. Telomere shortening is an in vivo marker of myocyte replication and aging. Am J Pathol. 2000;156: Parmacek MS, Epstein JA. Cardiomyocyte renewal. N Engl J Med. 2009;361: Urbanek K, Rota M, Cascapera S, Bearzi C, Nascimbene A, De Angelis A, Hosoda T, Chimenti S, Baker M, Limana F, Nurzynska D, Torella D, Rotatori F, Rastaldo R, Musso E, Quaini F, Leri A, Kajstura J, Anversa P. Cardiac stem cells possess growth factor-receptor systems that after activation regenerate the infarcted myocardium, improving ventricular function and long-term survival. Circ Res. 2005;97: Lopez LR, Schocket AL, Stanford RE, Claman HN, Kohler PF. Gastrointestinal involvement in leukocytoclastic vasculitis and polyarteritis nodosa. J Rheumatol. 1980;7: Saegusa M, Takano Y, Okudaira M. Human hepatic infarction: histopathological and postmortem angiological studies. Liver. 1993;13: Watanabe K, Abe H, Mishima T, Ogura G, Suzuki T. Polyangitis overlap syndrome: a fatal case combined with adult Henoch-Schönlein purpura and polyarteritis nodosa. Pathol Int. 2003;53: Dang NC, Johnson C, Eslami-Farsani M, Haywood LJ. Bone marrow embolism in sickle cell disease: a review. Am J Hematol. 2005;79: Gonzalez A, Rota M, Nurzynska D, Misao Y, Tillmanns J, Ojaimi C, Padin-Iruegas ME, Müller P, Esposito G, Bearzi C, Vitale S, Dawn B, Sanganalmath SK, Baker M, Hintze TH, Bolli R, Urbanek K, Hosoda T, Anversa P, Kajstura J, Leri A. Activation of cardiac progenitor cells reverses the failing heart senescent phenotype and prolongs lifespan. Circ Res. 2008;102: Soonpaa MH, Field LJ. Assessment of cardiomyocyte DNA synthesis in normal and injured adult mouse hearts. Am J Physiol. l 1997;272:H220 H226. Novelty and Significance What Is Known? Myocyte renewal in the adult human heart has been viewed as a modest phenomenon that comprises at most 50% of parenchymal cells. Fibroblasts and endothelial cell renewal has been claimed to be severalfold higher than that of cardiomyocytes. The expression of troponin I in myocyte nuclei has been considered a physiological process. What New Information Does This Article Contribute? The heart of cancer patients exposed to the radiosensitizer iododeoxyuridine shows a myocyte turnover rate of nearly 20% per year. The degree of fibroblast and endothelial cell turnover is comparable to that of myocytes. The expression of contractile proteins in myocyte nuclei is mediated by alterations in function of nuclear pore complexes characterized by downregulation of the scaffold nucleoporin Nup93. In the past decade, several studies have challenged the notion that the adult heart is a postmitotic organ in which the number of parenchymal cells is established at birth and cardiomyocytes lost with age or as a result of cardiac diseases cannot be replaced by newly formed cells. A pool of resident cardiac stem cells has been identified and characterized in the human heart and high levels of myocyte regeneration mediated by stem cell activation have been reported. However, based on retrospective 14 C birth dating of cells, it has been suggested that throughout life, myocyte replacement in humans is restricted to a subset of 50% of cardiomyocytes. The current work challenges this view by providing evidence to the contrary and documenting significant limitations in the approach and conclusions reached by 14 C dating of cardiac cells. Our data indicate that a large number of myocytes is formed every year, suggesting that the entire heart is replaced several times during the course of life in humans. These findings have important clinical implications because they emphasize the regeneration potential of the adult human heart.