Hirn-Derksen B. (1); Ryu C. (2); Hyunki C. (2); Kim J. Y. (2); Hempel G. (1)
(1) Institute of Pharmaceutical and Medicinal Chemistry, Clinical Pharmacy, Westfälische Wilhelms-Universität Münster, Corrensstr. 48, 48149 Münster, Germany. (2) Korea Institute of Science and Technology (KIST) Europe, Campus E7 1, 66123 Saarbrücken, Germany
Objectives: In the last two decades, the zebrafish and especially its embryos gained a lot of interest as animal models especially in toxicokinetics. Reasons are i.e. the genetic conformity with the human genome and the easy and cost-effective handling of the embryos [1]. However, to extrapolate a certain effect of a xenobiotic to the embryos the (toxico-)kinetic behaviour must be known. For this reason, we built a pharmacokinetic model to predict the concentration in zebrafish. As the main substance of interest we choose Dutasteride because of its nature to act as an endocrine disruptor in fish [2].
Methods: Based on a published model for zebrafish embryos we built a two-compartment model including the yolk sac and the embryo itself using GNU MCSim under R-Studio [3,4]. First estimations of different model parameters were done using Quantity Structure-Activity Relationship- (QSAR) and Virtual in vitro distribution (VIVD) model approaches [5,6]. Furthermore, we created an estimation method for absorption rate constant using a minimized design. We also integrated a plastic adsorption model to account for loss of chemical due to adsorption to wells [7], which also easily can be switched off to apply for scenarios were adsorption is negligible. Parameter calibration was done with measured zebrafish embryo concentration data set using the Marcov Chain Monte Carlo (MCMC) algorithm integrated in GNU MCSim. Convergence was checked using visual checks and scale reduction factor Rhat [8]. The performance of the calibrated model was checked against datasets of either 30 or 60 embryos per exposure vessel treated with different concentrations of Dutasteride. In this context, it was also checked if the model could predict concentration in zebrafish embryo when exposure vessel was changed from plastic well to glass beaker, to minimize adsorption. Control of the model performance was done using visual and computational predictive checks.
Results: The model is able to describe the pharmacokinetic time profile in zebrafish embryos for Dutasteride at different exposure concentrations. Mean prediction error (MPE) and mean absolute prediction error (MAPE) were -1.67 % and 12.18 % for the 500 nM experiment, -45 % and 45 % for the 100 nM experiment, 8.76 % and 21.02 % for the 50 nM experiment and -25.91 % and 41.62 % for 5 nM experiment. Performance of the model at 100 nM concentration experiments, which were performed in glass beaker instead of plastic wells were slightly worse than the other concentrations. This may also be due to other reasons such as some adsorption of dutasteride on the glass surface. In addition, we have also found a possible association between measured total protein concentration of embryos per experiment and internal concentration of Dutasteride in embryos.
Conclusion: It was possible to build a simple and dynamic model for zebrafish embryos which also accounts for decrease of medium concentration due to adsorption to plastic. However, the correlation between protein content and internal concentration still needs to be verified and could thus possibly improve the model prediction. We also plan to extend the model to other xenobiotics to check whether the pharmacokinetics of other substances can be predicted with our model. Because of its simple and generic nature, the model could maybe give an easily overview of the pharmacokinetic behavior of different xenobiotics in zebrafish embryos.
References:
[1] Cassar S, Adatto I, Freeman J, Gamse J, Iturria I, Lawrence C et al. Use of Zebrafish in Drug Discovery Toxicology. Chemical Research in Toxicology. 2019;33(1):95-118.
[2] Margiotta-Casaluci L, Hannah R, Sumpter J. Mode of action of human pharmaceuticals in fish: The effects of the 5-alpha-reductase inhibitor, dutasteride, on reproduction as a case study. Aquatic Toxicology. 2013;128-129:113-123.
[3] Siméon S, Brotzmann K, Fisher C, Gardner I, Silvester S, Maclennan R et al. Development of a generic zebrafish embryo PBPK model and application to the developmental toxicity assessment of valproic acid analogs. Reproductive Toxicology. 2020;93:219-229.
[4] Bois F. GNU MCSim: Bayesian statistical inference for SBML-coded systems biology models. Bioinformatics. 2009;25(11):1453-1454.
[5] Spacie A, Hamelink J. Alternative models for describing the bioconcentration of organics in fish. Environmental Toxicology and Chemistry. 1982;1(4):309-320.
[6] Fisher C, Siméon S, Jamei M, Gardner I, Bois Y. VIVD: Virtual in vitro distribution model for the mechanistic prediction of intracellular concentrations of chemicals in in vitro toxicity assays. Toxicology in Vitro. 2019;58:42-50.
[7] Fischer F, Cirpka O, Goss K, Henneberger L, Escher B. Application of Experimental Polystyrene Partition Constants and Diffusion Coefficients to Predict the Sorption of Neutral Organic Chemicals to Multiwell Plates in in Vivo and in Vitro Bioassays. Environmental Science & Technology. 2018;52(22):13511-13522.
[8] Gelman A, Rubin D. Inference from Iterative Simulation Using Multiple Sequences. Statistical Science. 1992;7(4).
Reference: PAGE 30 (2022) Abstr 9984 [www.page-meeting.org/?abstract=9984]
Poster: Drug/Disease Modelling - Endocrine