Using human pluripotent stem cell-derived cardiomyocytes to test drug cardiac toxicity

Unexpected cardiotoxicity underlies high rates of attrition during drug development, posing a multi-billion dollar burden on the pharmaceutical industry. In the last decade of the 20^th century, eight non-cardiovascular drugs were withdrawn from clinical use because they prolonged the QT interval, resulting in ventricular arrhythmias and potentially sudden death. Over-reliance on the use of animals and materials derived from animals in preclinical assays to predict the cardiotoxic effect of new drugs in humans has contributed to this problem. The relationship between drug responses from animal-derived primary cardiomyocytes for in vitro assays or from tissue samples for ex vivo assays (e.g. guinea pig, mouse or rat ventricular myocytes) and humans is not always clear. The development of more predictive human in vitro models is key to addressing this. However, the very limited availability of human cardiac tissue/cells on which to investigate disease mechanisms and cardiotoxic effects of drugs remains a major challenge in the cardiac field. Human stem cell-derived cardiomyocytes (CM) may offer a potential solution.

Dr Tamed Mohammed and his team have successfully established a technology for the high-throughput production of human induced pluripotent stem cell (hiPSC)-derived CMs. These have been optimized for use in various assays routinely used by industry to screen for drug cardiotoxicity such as the MTT cell viability assay, ATP depletion, intracellular calcium and apoptosis assays e.g. caspase 3/7 activity, annexin V and Tunnel staining. The team has validated the hiPSC-CMs against a small panel of compounds, but they need to extend this compound set to better understand the predictive potential of their model in identifying cardiotoxic drugs.

Through CRACK IT Solutions they have secured support from the Fraunhofer IME Screening Port, who has an immediate need for their hiPSC-CM assay to refine their drug libraries. The Fraunhofer IME provided 200 known cardiotoxicants with which to validate their model. Following successful validation of their assays under CRACK IT Solutions funding, the Fraunhofer IME are currently using Dr Mohammed's CMs to refine their compound library of over 250,000 chemicals from those compounds that have a potential cardiotoxicity. This information would allow the early prioritisation of compounds for further study and development.

Full details about this CRACK IT Solution can be found on the CRACK IT website.

Using this established high-throughput, high-content cellular analysis confocal microscopy platform, Dr Mohammed investigated if their hiPSC-CMs could predict the cardiotoxic potential of 105 compounds (80 known cardiotoxins and 25 control compounds) provided by their industrial partner the Fraunhofer IME Screening Port. Using this approach he was able to assess various cell viability parameters such as: cellular ATP levels, imaging of nuclear condensation, calcium imaging, mitochondrial membrane potential, membrane permeability, and apoptosis. Dr Mohammed and his team also explored whether the current gold standard cellular models could identify the cardiotoxicants in the same compound library. These included neonatal rat cardiomyocytes, H9C2 cells and two cancer cell lines - A549 and ACHN.

The team tried various time points and found that 48 hours exposure to the compounds was sufficient to produce cardiotoxic effects. For high-throughput analysis of the data, they designed a scoring system from 1-5 to each compound where compounds with an EC₅₀ in the nanomolar or millimolar concentration range score one or five, respectively. Their results showed that the iPS-CMs were the best at predicting cardiotoxicity, identifying known cardiotoxic compounds with an accuracy of 88.5% compared to 75.5% for neonatal rat cardiomyocytes, 69.3% for H9C2, 53.3% for the cancer cell lines.

As a result of this study, these combinations of assays in their iPS-CMs are now employed as the standard method for testing cardiotoxicity in all compounds within the clinical application pipelines at the Fraunhofer IME screening port. This has significantly reduced the animal use in getting neonatal rat cardiomyocytes for cardiotoxicity testing as well as terminating the potentially cardiotoxic compounds from further drug development and regulatory in vivo testing. This work has also resulted in the formation of the TOXICON consortium between 13 different SMEs and academic partners from Europe and the USA to apply for the PHC-33-2015 call from Horizon 2020. The focus of the call is to establish a reliable in silico predictor for drug-mediated cardiac and neuronal toxicities based on high-throughput targeted and genome-scale in vitro assays on human iPSC derived cardiomyocytes and neurons.