Input of expressed hERG current into a complex neuron in silico cardiac atrial cell: Implications for cardiotoxicity screening.

Abstract

Listen

Translate article

NEURON is the leading in silico software for generating computational models of excitable cells (www.neuron.yale/edu/neuron/). We have combined our real-time, dynamic clamp system with offline NEURON models allowing for interfacing of living cells with NEURON models during electrophysiological recording. For these studies, we employed a complex in silico atrial cardiomyocyte (Courtemanche et al., 1998; Jacobson, 1998). In Proof of Concept studies, the cardiac atrial cardiomyocyte IKr model was replaced with the FDA ORD HERG model (Li et al., 2017) via dynamic clamp. The in silico atrial cardiomyocyte AP displayed a spike and dome morphology with a resting membrane potential of −82±0.004 mV, an amplitude of 104±2 mV, a dV/dt of 183±8 V/s and an APD90 of 303±2 ms (n=8). The NEURON in silico cardiomyocyte IKr was replaced with the FDA ORD HERG model and current magnitude scaled such that the APD90 value was ∼300 ms. The AP also displayed a spike and dome morphology with a resting membrane potential of −82±0.005 mV, an amplitude of 99±1 mV, a dV/dt of 151±7 V/s and an APD90 of 305±2 ms (n=8). Virtual drug block with 10 nM dofetilide significantly prolonged the AP90 to 350±3 ms (p<0.001). The effect of blocking real HERG current expressed in HEK cells on in silico atrial cardiomyocyte AP morphology will be examined. Future studies will measure the effect of blocking real HERG current on AP morphology using the FDA CiPA's O’Hara-Rudy in silico ventricular cardiomyocyte. This assay will have utility in screening drugs for proarrhythmic activity where the measured output of HERG block is AP behavior. Importantly, this assay can readily be implemented on automated patch clamp instrumentation with our dynamic clamp system.