Predicted drug-induced bradycardia related cardio toxicity using a zebrafish in vivo model is highly correlated with results from in vitro tests

Abstract

Several in vivo and in vitro studies have assessed methods of evaluating the cardio toxicity of compounds during drug development due to its importance for predicting human toxicity. However, in vivo/in vitro relationships have not yet been reported using a zebrafish model. This study determined the bradycardia of 15 compounds by evaluating the change in heart beat rate (HBR) in zebrafish, hERG fluorescence polarization (hERG-FP), and ionic current change using a patch clamp (hERG-PC). In addition, a model for prediction of drug-induced bradycardia was established using in vivo and in vitro assays designed for high-throughput toxicological screening. The IC(50) values correlated well in two in vitro studies (R(2)=0.9). The change in HBR in zebrafish caused by the compounds could be estimated using the IC(50) from the hERG-FP assay [(i.e., % of HBR=19.5×log(IC(50), hERG-FP)] or hERG-PC assay [(i.e., % of HBR=19.6×log(IC(50), hERG-FP)]. To validate the predictive model, 10 unknown compounds were used and the percentages of the HBR were estimated using the model. The observed and predicted HBR% for the compounds in zebrafish were well-correlated (R(2)=0.948). Therefore, the proposed models were useful for prediction of drug-induced bradycardia related cardio toxicity.