Tissue-Specific Toxicokinetics of 6PPD Quinone Associated with Tyre Wear Particles Using a Genetic Algorithm-Calibrated PBTK Framework.

We evaluated multi-objective genetic algorithm (GA) calibration of a zebrafish physiologically based toxicokinetic model for 6PPD quinone (6PPDQ). Gill, liver and brain data from a 96-h study at 100µg/L nominal exposure informed calibration (1, 6 and 24h) and retrospective same-study temporal evaluation (48, 72 and 96h). Four parameter-bound cases revealed tissue-specific trade-offs. Compared with an uncalibrated quantitative structure-activity relationship reference, the Case 3 compromise reduced mean tissue natural-log mean squared error from 6.088 to 0.257 during calibration and from 6.568 to 0.744 during evaluation. Negative evaluation Nash-Sutcliffe efficiencies in all tissues indicated unresolved later dynamics. Global sensitivity analyses and conditional confidence intervals assessed parameter influence and uncertainty. Because aqueous concentrations were not measured, we examined alternative exposure scenarios. Improvement over the reference persisted, but parameter estimates varied with assumed exposure. GA improved study-specific concentration agreement and supported explicit selection of a compromise across tissues.