The construction and application of physiologically based toxicokinetic models for acrylamide, glycidamide and their biomarkers in rats and humans

Acrylamide (AA), a class 2A probable carcinogen to humans classified by the International Agency for Research on Cancer, has attracted extensive attention worldwide since it was widely used in industrial and domestic water treatment and detected in thermal processing foods. The metabolic adducts of AA and its primary metabolite glycidamide (GA) have been served as biomonitoring markers of AA intake, but the physiologically based toxicokinetics (PBTK) models to estimate internal dosimetry still remain unclear. An updated PBTK model for AA, GA and their metabolic biomarkers in rats and humans was developed and extended with time-course datasets from both literatures and our experiments. With adjustments to the model parameters, linear regression correlation coefficient ( R 2 ) between the fitting values and the validation datasets of rats and humans was greater than 0.76. The current model fits well with the experimental datasets of urinary N -acetyl- S -(2-carbamoylethyl)- l -cysteine (AAMA) and ( N -( R , S )-acetyl- S -(carbamoyl-2-hydroxyethyl)- l -cysteine) (GAMA) of rats exposed to AA from 0.1 to 50 mg/kg b.w. and humans exposed to AA from 0.0005 to 0.020 mg/kg b.w., indicating the robustness of the current models. Parameters for adduct of AA with N -terminal valine of hemoglobin (AAVal) were extended to humans and validated. Kinetic parameters for rats were assessed and validated based upon fit to the experimental datasets for liver N3-(2-carbamoyl-2-hydroxyethyl)-adenine (N3-GA-Ade) and N7-(2-carbamoyl-2-hydroxyethyl)-guanine (N7-GA-Gua) adducts. Compared with the previous model, the developed model included the correlation between AA intake and its mercapturic acid adducts, AAMA and GAMA, in a larger dose range with new experimental data, and parameters for AAVal, N3-GA-Ade and N7-GA-Gua were improved and verified. The current multi-component PBTK models provide a superior foundation for the estimation of short-term to medium and long-term intake levels of human exposure to AA. • PBTK models for acrylamide, glycidamide and their biomarkers in rats and humans were developed and improved. • The fitting range of AAMA and GAMA in urine after AA exposure was expanded. • The fitting results of rat AAVal parameters in human model fell within a reasonable range. • PBTK models for N3-GA-Ade and N7-GA-Gua adducts in rat liver were innovatively developed. • The multi-adduct model provided a superior foundation for the assessment of human exposure levels.