Preprint Models How Oxidative Stress and IKr Block Interact in Human Ventricular Cells

A Computational Link Between Mitochondria and Electrical Activity

A University of California, Davis team developed the Zukowski excitation–contraction–mitochondrial–ROS model, a computational representation of a human ventricular myocyte. The preprint combines the O’Hara–Rudy ventricular electrophysiology model with mitochondrial calcium, energy metabolism and reactive oxygen species (ROS) modules. It asks how oxidative stress interacts with calcium cycling, pacing rate and ion-channel behavior during doxorubicin-associated cardiac stress.

Cytosolic calcium connects electrical activity to mitochondrial uptake and metabolism. Mitochondrial ROS then modifies ryanodine-receptor release, SERCA uptake, L-type calcium current and late sodium current. These feedback relationships allow a perturbation in one component to affect both energy-related variables and the action potential.

Experimental Comparisons Constrain Different Components

The model draws on published experiments for component calibration and comparison. Mitochondrial calcium relationships include rat ventricular measurements; ROS-dependent current modulation uses guinea-pig and rabbit cardiac data. Calcium and action-potential comparisons also differ in species, pacing and experimental preparation. Those distinctions limit the quantitative equivalence of an individual simulated condition to a human experiment.

A separate human comparator used ventricular slices from nine young male donors, exposed to 0.5–10 μM doxorubicin for 24 hours and paced at 1 Hz. The model’s maximum stress setting predicted 15.1% prolongation of action-potential duration at 80% repolarization (APD80); the published slice response ranged from 2% to 13% across concentrations. Because the modeled doxorubicin effect is dimensionless, the comparison was made with the experimental response range rather than a matched drug concentration.

The authors also note that experiments applying extracellular hydrogen peroxide informed relationships driven by intracellular superoxide in the model. These species cannot be quantitatively mapped concentration for concentration. Some channel relationships were constrained by a single maximal-effect measurement, leaving their full concentration dependence uncertain.

Pacing Rate Changes the Predicted ROS–Calcium Feedback

Increasing modeled cytosolic calcium raised mitochondrial activity and ROS production nonlinearly. Faster pacing amplified the effect, as the interval available for recovery shortened. Moderate ROS increased calcium release, whereas greater oxidative stress depleted sarcoplasmic-reticulum stores. The direction and size of a calcium response therefore depended on the stress condition.

Doxorubicin stress was represented by a parameter D from 0 to 1, acting through increased electron-transport-chain ROS leakage and reduced scavenging. Direct drug effects on repolarizing potassium currents were omitted to isolate the oxidative mechanism. The leakage increase also had a modeling assumption because its magnitude at clinically relevant intracellular drug concentrations was not experimentally constrained.

Across the stress-parameter scan, mitochondrial calcium and ROS changed at settings where action-potential duration changed little. This result supports a hypothesis that mitochondrial measurements could detect a perturbation before a large electrical phenotype becomes apparent. It is a comparison across modeled stress levels, rather than a longitudinal demonstration of clinical injury onset.

Single and Combined Perturbations Give Different Population Results

The investigators generated one population of 1,000 virtual ventricular cells, varying 12 conductance factors independently within ±20% of baseline. The same parameter sets were used for all conditions, preserving a paired comparison. Cells received 10 beats at 2 Hz, a one-second pause and a final recorded beat.

Full doxorubicin stress alone produced no early afterdepolarizations. A 40% block of the rapid delayed rectifier potassium current, IKr, also produced none when applied alone. Adding 30% IKr block to full doxorubicin stress prolonged action potentials without producing early afterdepolarizations. At 40% block with full stress, a subset developed secondary voltage upstrokes during repolarization.

The combined 40% condition increased both mean action-potential duration and between-cell variation. These virtual outcomes describe susceptibility within the specified parameter population and pause protocol. They identify a possible interaction between oxidative dysfunction and reduced repolarization reserve for experimental testing.

The authors identify human cardiac mitochondrial kinetics and full concentration-response measurements for ROS-dependent currents as priorities for refining the parameters. Planned extensions include direct doxorubicin effects on IKs. The current model isolates mitochondrial/ROS feedback, so those additional mechanisms may change the thresholds and combinations predicted to impair repolarization.