Anja Lehmann (1,2), Ina Geburek (3), Anja These (3), Xiaojing Yang (4), Stefanie Hessel-Pras (5), Charlotte Kloft (6,2), Christoph Hethey (1,2)
(1) Junior Research Group Toxicokinetic Modelling, Dept. Exposure, German Federal Institute for Risk Assessment (BfR); (2) Graduate Research Training Program PharMetrX, Berlin/Potsdam; (3) Unit Contaminants, Dept. Safety in the Food Chain, German Federal Institute for Risk Assessment (BfR); (4) Wuya College of Innovation, Shenyang Pharmaceutical University, P. R. China; (5) Unit Effect-based Analytics and Toxicogenomics, Dept. Food Safety, German Federal Institute for Risk Assessment (BfR); (6) Institute of Pharmacy, Dept. Clinical Pharmacy & Biochemistry, Freie Universitaet Berlin
Introduction/Objective: Pyrrolizidine alkaloids (PAs) are a class of secondary metabolites in plants of which some are highly hepatotoxic, genotoxic, and carcinogenic [1]. Humans are exposed to PAs via intake of herbal supplements or medicines, and via contamination of foodstuffs such as tea, honey, and spices [2]. The combination of PAs with cytochrome P450 enzyme-inducing compounds, e.g. the drug phenobarbital, has been shown to dramatically increase PA toxicity [3]. Our aim was to develop a specifically tailored physiologically-based toxicokinetic (PBTK) model for PAs that includes the necessary detailed representation of PA metabolism to predict hepatic interactions.
Methods: In contrast to most pharmaceutical compounds, PAs are not well characterized in terms of their physico-chemical/biochemical properties. Kinetic data is sparse due to general efforts towards reduction of animal testing in chemical risk assessment. To overcome this sparse data situation, we used a combined in silico, in vitro, and in vivo approach for PBTK model development. We performed in vitro assays to determine lipophilicity, transcellular permeability (Caco-2), and metabolic clearance in mouse, rat, and human liver microsomes. In vivo metabolic clearance was predicted via in vitro to in vivo extrapolation. In silico methods were applied to predict ionization (SPARC v2018), plasma protein binding (GastroPlus v9.5), and tissue distribution [4]. We implemented the PBTK model in R based on the RxODE (v0.7.2-1) package [5] and inferred parameters via Maximum Likelihood Estimation and the Delayed Rejection Adaptive Metropolis algorithm [6].
Results: Retrorsine was identified as suitable PA for model development, since mouse and rat in vivo data (i.p. or i.v. administration) are available. The data include measurements of retrorsine and selected metabolites (protein adducts, DNA adducts, glutathione conjugates) in plasma, liver, urine or bile [7-10]. With regard to PBTK model development, we have specifically tailored the liver compartment by adapting the extended clearance model [11]. The adaption includes a representation of basic PA toxification and detoxification pathways. This allows to discriminate between metabolic and transport-related clearance, and makes the PBTK model well-suited to predict hepatic interactions of PAs. Determination of retrorsine metabolic clearance in liver microsomes revealed inter-species differences: metabolic clearance was more than 3-fold higher in rat compared to mouse and human. For all species, retrorsine depletion in liver microsomes followed a biexponential pattern. We explained and modelled this pattern mechanistically via end-product inhibition of the metabolic activity. Transcellular permeability of retrorsine in Caco-2 cells was determined to be 5.52·10-6 cm·s-1, which we use to predict oral absorption profiles.
Conclusions: Our research underpins highly relevant interactions between medical and foodborne compounds that should be considered in both drug dosing and food safety risk assessment, respectively. On the example of PA hepatic interactions, we demonstrate that translational toxicology is an efficient tool for the development of PBTK models, especially in sparse data situations. Next steps include the extrapolation of the PBTK model to humans and simulating real-life consumer exposure scenarios.
References:
[1] B. Dusemund, N. Nowak, C. Sommerfeld, O. Lindtner, B. Schäfer, and A. Lampen. Risk assessment of pyrrolizidine alkaloids in food of plant and animal origin. Food and Chemical Toxicology. 115: 63–72 (2018).
[2] P. Mulder, P. López Sanchez, A. These, A. Preiss-Weigert, and M. Castellari. Occurrence of Pyrrolizidine Alkaloids in food. EFSA supporting publication. EN-859: 1-116 (2015).
[3] J. R. Allen, C. F. Chesney, and W. J. Frazee. Modifications of pyrrolizidine alkaloid intoxication resulting from altered hepatic microsomal enzymes. Toxicol. Appl. Pharmacol. 23: 470-479 (1972).
[4] W. Schmitt. General approach for the calculation of tissue to plasma partition coefficients. Toxicology in Vitro. 22: 457-467 (2008).
[5] R: A language and environment for statistical computing. R Foundation for Statistical Computing. Vienna, Austria. Version 3.2.4.
[6] H. Haario, M. Laine, A. Mira, E. Saksman. DRAM: Efficient adaptive MCMC. Statistics and Computing. 16:339-354 (2006).
[7] X. Yang, W. Li, Y. Sun, X. Guo, W. Huang, Y. Peng, and J. Zheng. Comparative Study of Hepatotoxicity of Pyrrolizidine Alkaloids Retrorsine and Monocrotaline. Chem. Res. Toxicol. 30: 532–539 (2017).
[8] L. Zhu, J. Xue, Q. Xia, P. P. Fu, and G. Lin. The long persistence of pyrrolizidine alkaloid-derived DNA adducts in vivo: kinetic study following single and multiple exposures in male ICR mice. Arch. Toxicol., 91: 949–965 (2017).
[9] P.-S. Chu and H.J. Segall. Species difference in the urinary excretion of isatinecic acid from the pyrrolizidine alkaloid retrorsine. Camp. Biochem. Physiol. 100C: 683-686 (1991).
[10] I.N.H. White. Excretion of pyrrolic metabolites in the bile of rats given the pyrrolizidine alkaloid retrorsine or the bis-N-ethylcarbamate of synthanecine A. Chem.-Biol. Interactions, 16: 169-180 (1977).
[11] C. J. Endres, M. G. Endres, and J. D. Unadkat. Interplay of Drug Metabolism and Transport: A Real Phenomenon or an Artifact of the Site of Measurement? Mol. Pharm. 6: 1756–1765 (2009).
Reference: PAGE 28 (2019) Abstr 9154 [www.page-meeting.org/?abstract=9154]
Poster: Drug/Disease Modelling - Absorption & PBPK