Modeling Doxorubicin-Induced Cardiotoxicity in Human Pluripotent Stem Cell Derived-Cardiomyocytes.

Doxorubicin and other anthracyclines are highly efficacious anti-cancer medications and are used to treat a broad spectrum of adult and childhood malignancies, including Hodgkin’s and non-Hodgkin’s lymphoma, neuroblastoma, soft tissue sarcomas and breast cancer. Doxorubicin is also one of the most cardiotoxic medications in clinical use. Doxorubicin-induced cardiotoxicity (DIC) includes acute atrial and ventricular arrhythmia, as well as chronic cardiomyopathy and heart failure1,2. Despite decades of study, the precise mechanisms of DIC remain elusive and we currently lack the ability to predict or prevent this adverse drug reaction (ADR) in individual patients. DIC is dose-dependent and at cumulative doses of 450 to 500 mg/m2 the incidence of congestive heart failure is 4 to 5%, whereas at doses of 550 to 600 mg/m2 the incidence increases to 18%3. One major hypothesis regarding the mechanism of DIC is that doxorubicin causes increased production of reactive oxygen species (ROS), leading to damage to DNA, proteins and lipids, and ultimately causing death and dysfunction of cardiomyocytes4,5. Other proposed mechanisms include mitochondrial dysfunction and alterations of Ca2+ homeostasis6,7. More recently, studies in mice have suggested an important role for topoisomerase-II beta (Top2β) in the pathogenesis of DIC5. A major barrier to studying DIC has been the lack of appropriate model systems. This relates to the fact that human cardiac tissue, the primary site of DIC, is largely inaccessible and cannot be maintained in tissue culture. In addition, animal models may not accurately recapitulate DIC because of inter-species differences in both drug metabolism and cardiac structure and function. In particular, significant differences between mouse and human cardiac system in terms of electrophysiology and contractile features limit the extrapolation of findings from studies in murine systems to humans8,9,10,11,12. An ideal model system would possess both high human physiological relevance and potential for high throughput applications. Recently, human pluripotent stem cell-derived cardiomyocytes (hPSC-CMs) have emerged as a powerful tool to model cardiac toxicity in highly physiologically relevant human cells13,14,15,16. Human pluripotent stem cells (hPSC) are capable of self-renewal and, because of their capacity to differentiate into cell type derivatives of all three germ layers, are a powerful tool for disease modeling and drug screening. Highly efficient differentiation protocols for generating cardiomyocytes (CMs) from hPSCs have been described in the past decade17,18,19,20,21,22,23, and these cells express major human cardiac ion channels and sarcomeric proteins, suggesting that they may have high human physiological relevance13,24. Key advantages of hPSC-CMs for studying cardiac toxicity include a higher degree of homology with human compared to animal CMs, and potential for high throughput applications. The objective of this study was to investigate the characteristics and molecular mechanisms of DIC in a hPSC-CM model system.