Anthracycline Cardiomyopathy Is Mediated by Depletion of the Cardiac Stem Cell Pool and Is Rescued by Restoration of Progenitor Cell Function
Background— Anthracyclines are the most effective drugs available in the treatment of neoplastic diseases; however, they have profound consequences on the structure and function of the heart, which over time cause a cardiomyopathy that leads to congestive heart failure.
Methods and Results— Administration of doxorubicin in rats led to a dilated myopathy, heart failure, and death. To test whether the effects of doxorubicin on cardiac anatomy and function were mediated by alterations in cardiac progenitor cells (CPCs), these cells were exposed to the anthracycline, which increased the formation of reactive oxygen species and caused increases in DNA damage, expression of p53, telomere attrition, and apoptosis. Additionally, doxorubicin resulted in cell-cycle arrest at the G2/M transition, which led to a significant decrease in CPC growth. Doxorubicin elicited multiple molecular adaptations; the massive apoptotic death that occurred in CPCs in the presence of anthracycline imposed on the surviving CPC pool the activation of several pathways aimed at preservation of the primitive state, cell division, lineage differentiation, and repair of damaged DNA. To establish whether delivery of syngeneic progenitor cells opposed the progression of doxorubicin cardiotoxicity, enhanced green fluorescent protein–labeled CPCs were injected in the failing myocardium; this treatment promoted regeneration of cardiomyocytes and vascular structures, which improved ventricular performance and rate of animal survival.
Conclusions— Our results raise the possibility that autologous CPCs can be obtained before antineoplastic drugs are given to cancer patients and subsequently administered to individuals who are particularly sensitive to the cardiotoxicity of these agents for prevention or management of heart failure.
Received July 19, 2009; accepted November 2, 2009.
Anthracyclines are some of the most effective drugs currently available in the treatment of neoplastic diseases1; however, anthracyclines have profound consequences on the structure and function of the heart that with time cause a cardiomyopathy that leads to intractable congestive heart failure.2 The cardiotoxicity of anthracyclines is dose dependent, and this limits the clinical implementation of anthracyclines at optimal antitumor efficacy. Doxorubicin is the most powerful and widely used anthracycline, and considerable effort has been made to elucidate the cause of doxorubicin-induced cardiotoxicity to prevent the mechanisms implicated in the initiation and dramatic evolution of ventricular dysfunction.3 The generation of reactive oxygen species is a critical mediator of myocardial damage,4 but the target cell actually responsible for the deterioration of cardiac performance remains to be determined.
Clinical Perspective on p 292
The recognition that the adult heart in animals and humans contains a pool of resident primitive cells that are self-renewing, clonogenic, and multipotent in vitro and that regenerate myocytes and coronary vessels in vivo5–8 raises the question of whether the effects of doxorubicin on cardiac homeostasis and repair are directed primarily to the stem cell compartment, partially ablating the reserve of functionally competent cardiac progenitor cells (CPCs). CPCs are particularly sensitive to oxidative stress and die rapidly by apoptosis. Myocytes are more resistant to reactive oxygen species formation than are CPCs, which strengthens the possibility that loss of CPCs together with the attenuated generation of myocyte progeny may be critical in the development of doxorubicin-mediated cardiomyopathy. Theoretically, CPCs can be isolated from biopsy samples, and after their expansion in vitro, they can be implanted locally within regions of damage, where they reconstitute the injured myocardium.5–8 This strategy may allow aggressive chemotherapy followed by CPC repopulation of the depleted myocardium, which may rescue the cardiomyopathic heart. These hypotheses have been tested in the present study to determine whether doxorubicin-induced cardiomyopathy can be viewed as a stem cell disease and whether CPC therapy reverses heart failure in an animal model. Here, we report that intramyocardial injection of syngeneic CPCs positively interferes with anthracycline cardiotoxicity, largely restoring the structural and functional integrity of the diseased heart.
CPCs and Doxorubicin
Clonogenic c-kit–positive CPCs were infected with a retrovirus carrying enhanced green fluorescent protein (EGFP). CPCs were treated for 12, 24, and 48 hours with 0.1-, 0.5-, and 1-μmol/L concentrations of doxorubicin. CPC apoptosis and proliferation were determined.
Telomerase activity was measured by quantitative polymerase chain reaction, and telomere length was measured by Q-FISH (quantitative fluorescent in situ hybridization).
Reverse-Transcription–Polymerase Chain Reaction Array
The transcriptional profile of CPCs in the absence and presence of doxorubicin was assessed by quantitative reverse-transcription–polymerase chain reaction array.
Fischer 344 rats with doxorubicin-induced cardiomyopathy were treated with CPCs. A total of 5×104 EGFP-labeled CPCs were injected at 4 sites in the left ventricular (LV) myocardium. This dose was selected on the basis of previous results in which the delivery of progenitors ranging from 10 000 to 100 000 to 200 000 produced similar positive effects on myocardial regeneration.
Data Analysis and Statistics
Results are presented as mean±SD. For additional information, see the Methods section of the online-only Data Supplement.
Doxorubicin and CPC Death and Growth
To establish the effects of doxorubicin on clonogenic c-kit–positive CPCs,5 these cells were exposed to doxorubicin 0.1, 0.5, and 1 μmol/L for 12, 24, and 48 hours. Cell viability was assessed by a colorimetric MTT assay. In the presence of doxorubicin 0.1 μmol/L, CPC survival was not affected; however, doxorubicin at 0.5 and 1 μmol/L reduced CPC viability by 24% and 33%, respectively, at 24 hours and by 66% and 90%, respectively, at 48 hours (Figure 1⇓A). Additionally, apoptosis measured by terminal deoxynucleotidyl transferase assay, DNA laddering, and caspase-3 activity increased with time and the dose of doxorubicin. These 3 indicators of apoptosis peaked after 48 hours of treatment with doxorubicin 1 μmol/L (Figures 1B through 1⇓D). The terminal deoxynucleotidyl transferase assay was restricted to adherent cells, and after 48 hours of exposure to doxorubicin 1 μmol/L, the number of adherent CPCs was reduced by ≈90%, which indicates that this drug promoted apoptosis in almost all cells.
The impact of doxorubicin on CPC division was determined by bromodeoxyuridine (BrdU) and phospho-H3 labeling. The number of BrdU-positive CPCs and the mitotic index decreased with increasing concentration of doxorubicin and time (Figures 1E and 1⇑F). Moreover, the molecular regulators of G1, G1/S transition, and G2/M transition were measured. Cyclin D1, which drives cells from G1 to S, is activated by the cyclin-dependent kinase (cdk) cdk4, and this complex phosphorylates retinoblastoma (Rb), which inhibits its repressive function on cell-cycle progression. During G2, the cyclin B1–cdc2 (cell division control 2) complex is inactivated by phosphorylation. At the end of G2, the cdc25 phosphatase dephosphorylates this complex, and cells enter mitosis. Cyclin D1, cdk4, and phosphorylated Rb were decreased in CPCs exposed to doxorubicin in a dose- and time-dependent manner. The increase in cyclin B1 and cdc2 phosphorylation may reflect the arrest of the cell cycle at the G2/M transition (Figure 2⇓A; online-only Data Supplement Figure IA). These data are consistent with the delay in the decrease of BrdU labeling in CPCs with respect to phospho-H3 (Figures 1E and 1⇑F).
Subsequently, the protein level of the cyclin-dependent kinase inhibitors p21Cip, p27Kip1, and p16INK4a was determined in CPCs. Doxorubicin resulted in a transient increase in p21Cip and a persistent increase in p16INK4a (Figure 2⇑B; online-only Data Supplement Figure IB); expression of p27Kip1 in CPCs was not affected by doxorubicin. The early upregulation of p21Cip may represent an attempt by CPCs to repair DNA damage, whereas the persistently high quantity of p16INK4a indicates irreversible growth arrest and cellular senescence.
Doxorubicin and Oxidative Stress in CPCs
There is general consensus that the generation of reactive oxygen species plays a relevant role in the development of anthracycline-induced cardiomyopathy.2,4 To determine whether a similar process was operative in CPCs, the presence of 8-OH-deoxyguanosine (8-OHdG) was measured in nuclei by immunocytochemistry and confocal microscopy. Doxorubicin treatment was characterized by a striking increase in the number of 8-OHdG–positive CPCs (Figure 2⇑C). Moreover, expression of the antioxidant enzymes manganese superoxide dismutase (Mn SOD), copper-zinc superoxide dismutase (Cu/Zn SOD), and catalase did not change, whereas the activity of these enzymes decreased markedly at 48 hours, failing to counteract reactive oxygen species–mediated DNA damage (Figure 2⇑D; online-only Data Supplement Figure IIA). Doxorubicin resulted in an average 30% shortening of telomeres in CPCs and a shift to the left in the distribution curve of telomere lengths. Additionally, the percentage of CPCs with telomeres <8 kilobase pairs (kbp) increased 4-fold with doxorubicin (Figure 3⇓A). Telomere attrition occurred in spite of the preservation of telomerase activity in doxorubicin-treated CPCs (Figure 3⇓B).
Dysfunctional telomeres trigger a DNA damage response in which the major determinant is the transcription factor p53. The ataxia-telangiectasia mutated (ATM) protein kinase is required for phosphorylation of p53 at serine 15; ATM kinase and phospho-p53 at serine 15 and 20 were upregulated in doxorubicin-treated CPCs (Figure 3⇑C; online-only Data Supplement Figure IIB). ATM kinase expression peaked at 12 hours, whereas phospho-p53 at serine 15 and 20 increased primarily at 12 and 24 hours and remained elevated at 48 hours. Phosphorylation at serine 15 activates a cascade of posttranslational modifications of p53 that result in transcription of p53 target genes followed by activation of apoptosis or cellular senescence.9 In the present study, p53 phosphorylation at serine 15 was accompanied by enhanced but transient expression of p21Cip1 (see above), possibly in an attempt to promote DNA repair. Also, the proapoptotic proteins Bax and Bad increased in doxorubicin-treated CPCs (Figures 1E and 3⇑⇑⇑D). The prolonged upregulation of p16INK4a in CPCs is consistent with the role of this protein in the modulation of irreversible growth arrest and cellular senescence. p16INK4a rarely colocalizes with DNA double-strand breaks and represents a delayed response10 that follows the induction of p53 and p21Cip1.
Thus, anthracyclines promote oxidative stress and the activation of p53, which together inhibit the growth and survival of CPCs, and this supports the notion that defects in progenitor cell function may condition the development of cardiac myopathy in vivo. Additionally, these in vitro observations raise the possibility that CPC death may represent the primary event responsible for impaired myocyte turnover, accumulation of senescent cells, apoptosis, and the onset of ventricular dysfunction, which are unrecognized aspects of doxorubicin-mediated cardiotoxicity. The in vivo experiments discussed in the subsequent sections were performed as an attempt to document whether alterations at the level of the controlling cell, the CPC, dictate the dramatic outcome of doxorubicin treatment in patients with neoplastic diseases.
Doxorubicin and Cardiac Anatomy and Function
To evaluate the effects of anthracyclines in vivo, Fischer 344 rats were injected intraperitoneally over a period of 14 days with 6 doses of doxorubicin 11 (online-only Data Supplement Figure III). One week after the last administration, there was a significant impairment of LV function characterized by a decrease in ejection fraction, which decreased further at 6 weeks (Figure 4⇓⇓A). The question was then whether the abnormalities detected echocardiographically were due to the prolonged presence of doxorubicin in the organism or whether the anthracycline had an acute toxic effect that persisted with time that depressed myocyte mechanical behavior. Because doxorubicin has a half-life of ≈30 hours and its direct action on cells is no longer detectable after 1 to 2 days,12 myocyte contractility and Ca2+ transients were determined in LV myocytes isolated from animals at 3 weeks. Sarcomere shortening and Ca2+ transients in myocytes were decreased with doxorubicin (Figure 4⇓⇓B). Both the time constant (τ) of Ca2+ decay and the time to 90% relaxation of myocytes were longer in these cells.
To establish whether doxorubicin activated cell death, cardiomyocyte apoptosis was determined. Compared with control hearts, doxorubicin treatment resulted in a 7-fold and 4-fold increase in myocyte apoptosis at 3 and 6 weeks, respectively (Figure 4⇑⇑C). Importantly, the corresponding increases in the fraction of cardiomyocytes that expressed the senescence-associated protein p16INK4a were 2-fold and 3-fold (Figure 4⇑⇑D). More than 70% of LV myocytes were p16INK4a positive at 6 weeks. Conversely, myocyte formation measured by the expression of Ki67 decreased 95% and 65% at 3 and 6 weeks, respectively (Figure 4⇑⇑E). Therefore, myocyte loss was not counteracted by an adequate formation of new cells, which led to a significant decrease in the aggregate number of parenchymal cells in the LV myocardium. This reduction in myocyte number was more pronounced at 6 weeks than at 3 weeks. Additionally, myocyte cell volume increased with time, which reflects the inadequate level of myocyte regeneration seen in the presence of doxorubicin (Figure 4⇑⇑F). Collectively, these observations suggest that doxorubicin led to a cardiac myopathy in which myocyte death predominated and contributed together with the depression in cell mechanics to the deterioration of ventricular function in this animal model.
Doxorubicin and CPC Transcriptional Profile
To establish whether doxorubicin treatment influences CPC fate, the molecular identity of these cells was defined by analyzing their transcriptional profile after exposure to the anthracycline. We used a quantitative reverse-transcription–polymerase chain reaction array and examined a restricted set of genes linked to the undifferentiated state of the cells and their specification to cardiovascular lineages. Additionally, genes involved in cell proliferation, survival, death, and senescence were studied (online-only Data Supplement Figure IV).
Doxorubicin induced profound changes in global gene expression of CPCs: 103 and 21 genes were upregulated and downregulated, respectively. Doxorubicin resulted in a 9-fold increase in the expression of the ATP-binding cassette ABC transporter Abcg2/Mdr1, which is implicated in drug efflux and cell protection from toxic agents.13 Although c-kit receptor mRNA was similar in untreated and treated CPCs, transcripts for the downstream effectors MITF and Snail-homolog 2 (Slug) increased in the presence of the anthracycline (Figure 5; online-only Data Supplement Figure IV).
Genes involved in self-renewal and progenitor cell expansion,14,15 including fibroblast growth factor 8 (FGF8) and 10 (FGF10), the catalytic subunit of telomerase (TERT), and the histone acetyltransferases Myst1 and Myst2, were more abundant in doxorubicin-treated than untreated CPCs. Similarly, Numb and Prospero-related protein (Prox1), which modulate asymmetrical division,16 were more abundant with doxorubicin (Figure 5). Importantly, transcripts for Klf4, Klf5, Oct4, and c-Myc were significantly increased in CPCs exposed to the anthracycline. Growth differentiation factor-3 (GDF3) and Nanog were enhanced with doxorubicin, whereas Sox2 was decreased, but these changes in gene expression were not significant (online-only Data Supplement Figure IV). Klf4, Sox2, c-Myc, and Oct4 are the 4 genes that promote reprogramming of fibroblasts into inducible pluripotent stem cells.17 The core Klf circuitry, composed of Klf2, Klf4, and Klf5, is critical for the preservation of the undifferentiated state of embryonic stem cells.17 Together with GDF3, these genes integrate into the Nanog transcriptional network that specifies the “stemness” of various progenitors.18 Additionally, several cell-cycle regulators comprising cyclins D1, E, and A2 and the cyclin-dependent kinase cdc2 were more abundant in doxorubicin-treated CPCs.
The mechanisms that control cardiomyogenesis in the adult heart are largely unknown; however, the differentiation of CPCs into myocytes reiterates in part the molecular programs of cardiac development. The majority of cardiac regulatory transcription factors were upregulated in doxorubicin-treated CPCs. These included GATA4, GATA5, MEF2A, Tbx1, Tbx3, Tbx20, and Hand2. Consistently, the downstream targets BNP (brain natriuretic peptide), α-sarcomeric actin, myosin light chain-4, and β-myosin heavy chain were more highly expressed in these cells. Notch1 receptor is a critical determinant of the transition of CPCs to amplifying myocytes.19 Although Notch1 expression was decreased, transcripts of the Notch pathway, including the delta-like 3 and the Jagged1 ligands, the mastermind-like 1 cofactor, and the Hes1 effector, were more abundant in doxorubicin-treated CPCs (Figure 5; online-only Data Supplement Figure IV).
The positive effect of doxorubicin on CPC commitment was not restricted to myocyte lineage. The expression of several vascular-specific genes increased in CPCs in response to doxorubicin. This molecular adaptation involved primarily endothelial cell–related genes, including Vezf1, Flk1, Flt1, Tie2, PECAM, multimerin, selectin, and von Willebrand factor. Together with the enhanced expression of Flk1, the upregulation of GATA1, CD34, and Tal1 indicated that the anthracycline triggered the activation of the molecular program that controls the formation of hemangioblasts.20 For the acquisition of smooth muscle cell lineage, only transforming growth factor-β receptor 1 and smooth muscle myosin heavy chain were upregulated in doxorubicin-treated CPCs (Figure 5; online-only Data Supplement Figure IV). Similarly, a group of p53-related genes implicated in cell death, DNA damage response, and growth arrest were expressed more highly in these cells. These included ATM kinase, Rad50, Mre11, Bax, p21Cip1, Gadd45a, and Mdm2 (Figure 5; online-only Data Supplement Figure IV).
Collectively, these findings at the transcriptional level indicate that doxorubicin triggers multiple biological adaptations in CPCs. The massive apoptotic death that occurs in CPCs in the presence of the anthracycline implies that the surviving CPC pool activates several pathways aimed at the preservation of the primitive state, cell division, lineage differentiation, and repair of damaged DNA.
Doxorubicin and CPC Death and Growth In Vivo
The data above raised the possibility that one of the major consequences of doxorubicin in cardiomyocyte death, hypertrophy, and dysfunction in vivo was mediated by defects at the level of the progenitor cell compartment. Therefore, these variables of CPC function were evaluated quantitatively in the LV myocardium. Compared with control hearts, doxorubicin produced a 5-fold and 8-fold increase in CPC apoptosis at 3 and 6 weeks, respectively (Figure 6A). Additionally, the fraction of p16INK4a-positive CPCs that reached irreversible growth arrest10 was increased in these hearts dramatically (Figure 6B). In contrast, the percentage of Ki67-positive CPCs was severely reduced with doxorubicin treatment (Figure 6C). These findings were consonant with the enhanced oxidative stress and DNA damage promoted by doxorubicin, as documented by the generation of 8-OHdG in CPC nuclei (Figure 6D). Collectively, the impact of doxorubicin on CPC apoptosis and senescence decreased by 79% and 94% the compartment of functionally competent CPCs in the LV myocardium at 3 and 6 weeks, respectively (Figure 6E). Thus, anthracyclines have negative effects on cell viability and growth, depleting the CPC pool available for cardiac homeostasis and repair.
CPC Repopulation of the Myocardium
If the detrimental consequences of anthracyclines on the heart were dependent on the loss of CPCs, exogenously administered immunocompatible CPCs would be expected to partially restore cardiac function and structure, improving the outcome of the dilated myopathy and animal survival. Therefore, doxorubicin-treated rats were divided into 2 groups at 3 weeks (online-only Data Supplement Figure V). The first group received intramyocardial injections of syngeneic CPCs (DOXO-CPC), and the second received vehicle only (DOXO-vehicle). CPCs were genetically tagged with EGFP for identification of their progeny. All animals were euthanized 3 weeks later, ie, 6 weeks after the onset of doxorubicin administration and 3 weeks after CPC or vehicle delivery.
Shortly after cell implantation, preliminary studies were performed to document by immunocytochemistry the presence of EGFP-positive CPCs within the myocardium. Additionally, the expression of Ki67 in EGFP-positive CPCs was demonstrated to prove that these cells, at least in part, successfully engrafted and continued to grow within the recipient myocardium (online-only Data Supplement Figure VI). After treatment, animals were exposed continuously to BrdU to label newly formed structures within the damaged decompensated heart. Therefore, regenerated myocytes and coronary vessels were expected to be both EGFP and BrdU positive in DOXO-CPC hearts. Previous results at 2 days after delivery of a comparable number of cells were ≈20%; however, this value is the product of 2 variables, death of the nonengrafted cells and proliferation of engrafted cells.21
Three weeks after CPC therapy, there was an amelioration of the conditions of the animals; they were less lethargic and had modest or no abdominal enlargement. The amount of fluid in the abdomen was ≈6-fold lower in DOXO-CPC rats (5.3±2.8 mL) than in DOXO-vehicle rats (30±14 mL; P<0.001). Most importantly, the mortality rate was dramatically reduced after CPC injection (Figure 7A). At 3 weeks, before treatment, the mortality rate averaged 45%; however, from 3 to 6 weeks, the animal mortality rate was decreased by 66% with CPC implantation. In the animals that survived, cardiac function was largely restored by CPC administration. With respect to DOXO-vehicle rats, LV developed pressure and +dP/dt and −dP/dt were markedly increased in DOXO-CPC hearts, reaching hemodynamic values similar to those in control animals (Figure 7B). Similarly, ejection fraction was essentially restored by CPC delivery (online-only Data Supplement Figure VII). The decrease in ventricular mass and wall thickness and the increase in chamber diameter and volume with the doxorubicin-induced myopathy were partially reversed by cell therapy (Figure 7C), which suggests that CPCs promoted myocardial regeneration, which contributed to the recovery of structure and function of the damaged heart.
Large clusters of newly formed cardiomyocytes were detected throughout the LV wall. These cells were EGFP and BrdU positive and expressed the contractile protein α-sarcomeric actin (Figures 8A through 8⇓C). Areas of myocardial regeneration were identified in all CPC-treated animals and varied in size from 0.05 to 2.5 mm2. Connexin 43 and N-cadherin were detected between new myocytes and preexisting and regenerated myocytes, which demonstrated that formed cells expressed the junctional proteins responsible for electrical and mechanical coupling (online-only Data Supplement Figure VIII). Seeding of EGFP-positive cells was also shown by polymerase chain reaction, which confirmed the morphological results (online-only Data Supplement Figure IX).
These foci of cardiac tissue consisted of cardiomyocytes and resistance arterioles and capillaries distributed throughout the regenerated myocardium (online-only Data Supplement Figure X). New myocytes retained a fetal-neonatal phenotype that varied in volume from 300 to 4300 μm3 (online-only Data Supplement Figure XI). Similarly, the number of capillaries and arterioles (Figure 8⇑D) reflected those of a developing heart.8 Myocardial regeneration decreased by 52±12% the extent of doxorubicin-induced tissue damage. After cell therapy, replacement fibrosis was 34% and 53% lower than in untreated animals at 3 and 6 weeks, respectively (Figures 8E through 8⇑G; online-only Data Supplement Figure XII), which indicates that CPC differentiation partially restored the structural integrity of the cardiomyopathic heart. However, activation of resident CPCs or the recruitment of circulating progenitors may have contributed to cardiac repair.
The cause of doxorubicin-induced cardiomyopathy remains unclear, although several mechanisms, including DNA damage, attenuation in protein synthesis, enhanced release of catecholamines, alterations in the adrenergic system, and defects in Ca2+ homeostasis, have been identified, together with the remarkable increase in oxidative stress and lipid peroxidation.1,2,4 These detrimental effects, however, result in a rather unspecific cardiac pathology1–3 in which cytoplasmic vacuolization, loss of myofibrils, and the relatively modest increase in interstitial and replacement fibrosis in the ventricular wall are not consistent with the severity of the disease and the dramatic manifestations of heart failure in animal models and humans.2,3 Understanding of anthracycline-mediated cardiomyopathy is further complicated by the timing of appearance of cardiotoxicity. Ventricular dilation, wall thinning, and depressed cardiac function may become evident during the course of therapy or later after the completion of treatment. The possibility of developing heart failure persists throughout life in patients who successfully survive cancer and its management with this antineoplastic agent.3
The results of the present study offer an alternative mechanism underlying the pathophysiology of doxorubicin-induced cardiomyopathy. The present in vitro and in vivo findings indicate that the determining event responsible for the initiation and evolution of the myopathy occurs at the level of the CPC compartment. Cardiotoxicity of the anthracycline is not restricted to cardiomyocytes but affects resident CPCs even more dramatically. Over a period of 6 weeks, doxorubicin led to an almost complete depletion of the CPC pool within the myocardium. Inhibition of CPC division in combination with accumulation of oxidative DNA damage, growth arrest, cellular senescence, and apoptosis decreased by 94% the number of functionally competent progenitors in the failing heart. This devastating consequence of doxorubicin for CPCs interfered with the physiological turnover of cardiomyocytes and their regeneration in the presence of diffuse cell death. The number of ventricular myocytes was reduced by 48%, and the majority of cells expressed the senescence-associated protein p16INK4a. Hypertrophied p16INK4a-positive myocytes show depressed mechanics and Ca2+ transients.22 Collectively, the lack of activation of CPCs and formation of a myocyte progeny, myocyte loss, and impaired contractile behavior of spared, enlarged myocytes appear to be critical variables of the development of heart failure with doxorubicin administration. Thus, doxorubicin cardiomyopathy is primarily a stem cell disease that conditions the pathophysiology of cardiomyocytes and ventricular hemodynamics.
Treatment of CPCs with doxorubicin induced dramatic changes of the cellular transcriptome, which highlights the molecular mechanisms implicated in the functional response of CPCs to anticancer drugs. Doxorubicin evoked multiple and apparently contradictory pathways in CPCs by simultaneously activating transcriptional regulators that promote multipotency and trigger differentiation.15,18 By necessity, gene-expression profiling was obtained in the subset of cells that survived doxorubicin toxicity and attempted to restore the homeostatic balance within the CPC compartment. To counteract ongoing apoptotic death, spared CPCs enhanced the expression of the complex network of transcription factors required for the maintenance of stemness. These genes integrate into the chromatin regulatory loop, which involves the Nanog promoter and is hierarchically supervised by Oct4.18 Klf4, Klf5, Oct4, c-myc, GDF3, and Nanog were upregulated in CPCs. This cluster of genes was initially believed to be restricted to embryonic stem cells18; however, the same transcriptional system appears to control the undifferentiated state of CPCs and other adult progenitor cells.23
Additionally, doxorubicin upregulated Flk1, GATA1, CD34, and Tal1, which control the formation of embryonic and adult hemangioblasts, in CPCs.20 Hemangioblasts are common precursors for endothelial and hematopoietic cell lineages, but as shown above, the presence of several endothelial cell transcripts in the absence of bone marrow markers indicated that CPCs preferentially acquired the endothelial cell phenotype. The attempt by DOXO-CPCs to differentiate into vascular cells was coupled with the activation of transcription factors that modulate cell fate and expression of contractile proteins during cardiac morphogenesis. These regulatory genes were upregulated in DOXO-CPCs and comprised the cardiac-restricted members of the GATA family, the Mef2 transcription factors, the T-box-family of transcription factors, and genes of the secondary heart field.
To strengthen the possibility that targeted therapies for cancer patients may be implemented in a manner that maximizes their antitumor effects24 without creating another equally devastating disease such as chronic heart failure,25 we have successfully documented that immunocompatible CPCs may be used to repopulate the cardiomyopathic heart with new cardiomyocytes and coronary vessels. Myocardial regeneration mediated by delivery and differentiation of CPCs restored cardiac hemodynamics in large part and most importantly decreased the animal mortality rate dramatically. The present findings have provided the first experimental documentation that negative ventricular remodeling characterized by cavitary dilation, wall thinning, severely impaired function, and ascites, typically present in chronic heart failure,26 can be reversed by CPC therapy. Although extreme caution must be exercised in the translation of these animal studies to human beings, the possibility is raised that myocardial biopsy samples may be obtained before antineoplastic drugs are given to cancer patients. Autologous CPCs can be isolated and expanded from these myocardial samples8 for the treatment of heart failure in individuals who are particularly sensitive to the cardiotoxicity of these chemotherapeutic agents.
A difficult question to answer concerns the variability in the response to doxorubicin in patients and the rather common observation that anthracycline-induced cardiomyopathy may develop years after the administration of the anticancer drug. Two important factors may account for these data in humans, which appear to be in contrast with the present experimental results. The first may be related to the age of the patient, which conditions the growth behavior and resistance to apoptosis of resident CPCs,27 and the second may reflect the intrinsic properties of CPCs, which are dictated by the patient’s medical history. The latter may have negatively influenced telomere length and telomerase, which together protect the pool size of CPCs and their ability to self-renew, divide asymmetrically, and survive. Finally, experimental findings cannot be translated to human beings easily.
Sources of Funding
This work was supported by grants from the Italian Ministry of Education (MIUR-PRIN 2007), the Italian Ministry of Health, and the National Institutes of Health (5R37HL081737-05, 5R01HL039902-18, 5R01AG017042-09, 5P01HL092868-02, 7P01AG023071-05, 5R01HL065573-07, 5R01HL065577-07, 5R01AG026107-04, 1R01HL091021-01A2, 5R21HL094894-02). J.F.-M. was supported by the Ministry of Science and Higher Education of Portugal.
van Dalen EC, Caron HN, Dickinson HO, Kremer LC. Cardioprotective interventions for cancer patients receiving anthracyclines. Cochrane Database Syst Rev. 2008; CD003917.
Pfister O, Mouquet F, Jain M, Summer R, Helmes M, Fine A, Colucci WS, Liao R. CD31- but not CD31+ cardiac side population cells exhibit functional cardiomyogenic differentiation. Circ Res. 2005; 97: 52–61.
Smith RR, Barile L, Cho HC, Leppo MK, Hare JM, Messina E, Giacomello A, Abraham MR, Marbán E. Regenerative potential of cardiosphere-derived cells expanded from percutaneous endomyocardial biopsy specimens. Circulation. 2007; 115: 896–908.
Bearzi C, Rota M, Hosoda T, Tillmanns J, Nascimbene A, De Angelis A, Yasuzawa-Amano S, Trofimova I, Siggins RW, Lecapitaine N, Cascapera S, Beltrami AP, D'Alessandro DA, 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: 14068–14073.
Siveski-Iliskovic N, Kaul N, Singal PK. Probucol promotes endogenous antioxidants and provides protection against adriamycin-induced cardiomyopathy in rats. Circulation. 1994; 89: 2829–2835.
Arola OJ, Saraste A, Pulkki K, Kallajoki M, Parvinen M, Voipio-Pulkki LM. Acute doxorubicin cardiotoxicity involves cardiomyocyte apoptosis. Cancer Res. 2000; 60: 1789–1792.
Sorrentino BP, Brandt SJ, Bodine D, Gottesman M, Pastan I, Cline A, Nienhuis AW. Selection of drug-resistant bone marrow cells in vivo after retroviral transfer of human MDR1. Science. 1992; 257: 99–103.
Tang H, Rompani SB, Atkins JB, Zhou Y, Osterwalder T, Zhong W. Numb proteins specify asymmetric cell fates via an endocytosis- and proteasome-independent pathway. Mol Cell Biol. 2005; 25: 2899–2909.
Gurdon JB. Melton DA. Nuclear reprogramming in cells. Science. 2008; 322: 1811–1815.
Levasseur DN, Wang J, Dorschner MO, Stamatoyannopoulos JA, Orkin SH. Oct4 dependence of chromatin structure within the extended Nanog locus in ES cells. Genes Dev. 2008; 22: 575–580.
Boni A, Urbanek K, Nascimbene A, Hosoda T, Zheng H, Delucchi F, Amano K, Gonzalez A, Vitale S, Ojaimi C, Rizzi R, Bolli R, Yutzey KE, Rota M, Kajstura J, Anversa P, Leri A. Notch1 regulates the fate of cardiac progenitor cells. Proc Natl Acad Sci U S A. 2008; 105: 15529–15534.
Rota M, Kajstura J, Hosoda T, Bearzi C, Vitale S, Esposito G, Iaffaldano G, Padin-Iruegas ME, Gonzalez A, Rizzi R, Small N, Muraski J, Alvarez R, Chen X, Urbanek K, Bolli R, Houser SR, Leri A, Sussman MA, Anversa P. Bone marrow cells adopt the cardiomyogenic fate in vivo. Proc Natl Acad Sci U S A. 2007; 104: 17783–17788.
Rota M, Hosoda T, De Angelis A, Arcarese ML, Esposito G, Rizzi R, Tillmanns J, Tugal D, Musso E, Rimoldi O, Bearzi C, Urbanek K, Anversa P, Leri A, Kajstura J. The young mouse heart is composed of myocytes heterogeneous in age and function. Circ Res. 2007; 101: 387–399.
Chen MH, Kerkelä R, Force T. Mechanisms of cardiac dysfunction associated with tyrosine kinase inhibitor cancer therapeutics. Circulation. 2008; 118: 84–95.
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 leads to premature cardiac aging and heart failure. Circ Res. 2003; 93: 604–613.
The mechanisms by which anthracyclines lead to the development of dilated cardiomyopathy in cancer patients are currently unknown. The work presented in this study raises the possibility that doxorubicin-induced cardiac failure is a stem cell disease mediated by a severe loss of resident cardiac progenitor cells (CPCs). Doxorubicin results in CPC apoptosis, which with time affects myocardial homeostasis and tissue repair. The dramatic reduction in the pool of CPCs conditions myocyte renewal, which promotes the accumulation of senescent, poorly functioning cardiomyocytes. In the long term, these abnormalities in cardiac structure and myocyte performance markedly impair ventricular hemodynamics and the likelihood of animal survival. Notably, repopulation assays with syngeneic CPCs rescue the cardiac phenotype and largely restore the functional properties of the diseased heart. Although extreme caution must be exercised in the translation of these animal studies to humans, the possibility is raised that myocardial biopsy samples could be obtained before antineoplastic drugs are given to cancer patients. Autologous CPCs could be isolated and expanded from these myocardial samples for the treatment of heart failure in individuals who are particularly sensitive to the cardiotoxicity of chemotherapeutic agents.
↵*Drs De Angelis and Piegari contributed equally to this work.
Guest Editor for this article was Buddhadeb Dawn, MD.
The online-only Data Supplement is available with this article at http://circ.ahajournals.org/cgi/content/full/CIRCULATIONAHA.109.895771/DC1.