You are here:
Refining High-Throughput In Vitro-In Vivo-Extrapolation Modeling through Incorporation of Intestinal Toxicokinetics
Citation:
Karol-Bexell, E., A. Jarnagin, M. Hughes, J. Wambaugh, AND B. Wetmore. Refining High-Throughput In Vitro-In Vivo-Extrapolation Modeling through Incorporation of Intestinal Toxicokinetics. SOT, Nashville, TN, March 19 - 23, 2023. https://doi.org/10.23645/epacomptox.22285516
Impact/Purpose:
N/A
Description:
The Toxic Substances Control Act (TSCA) sanctions the US Environment Protection Agency (EPA) to regulate commercially available chemicals that do not fall under the jurisdiction of other US federal agencies. As of February 2022, there were 86,631 chemicals on the TSCA inventory. Given this vast number, there is an escalating need for a high-throughput (HT) risk-based chemical prioritization and assessment. HT screening (HTS) for toxicity and toxicokinetic (TK) data are often used in combination with HT in vitro-in vivo extrapolation (IVIVE) modeling to allow the conversion of in vitro point of departure (POD) and steady state blood concentration (Css) values to an administrated equivalent dose (AED) in mg/kg/day. The current HT IVIVE extrapolation modeling system exhibits a gap between the predicted AED levels and the actual in vivo low effect levels (i.e., PODs) observed in animal studies. This is partially because the HT IVIVE model only considers plasma protein binding and hepatic clearance TK data, thus oversimplifying the whole-body metabolism contribution to chemical clearance. Comparing predicted AEDs to in vivo-derived PODs, the model proved to be on average 100-fold more conservative. This study attempts to narrow this gap by incorporating extrahepatic clearance data, namely intestinal clearance through consideration of CYP3A4 enzymatic contribution. So far, our new approach results in more predictive AED values CYP3A4 enzymatic kinetics are introduced into our IVIVE approach for our chemical test set. CYP3A4 enzymatic contribution to clearance was evaluated for 12 chemicals using human liver microsome clearance rates derived with or without the CYP3A4 inhibitor- CYP3cide. Since over 80 % of intestinal cytochrome-P450 activity is due to CYP3A4, incorporating CYP3A4 enzymatic contribution to clearance could serve as an efficient surrogate to predict intestinal involvement in chemical clearance rate. The addition of CYP3A4 enzyme kinetics altered the bioavailability (F), Css and fraction escaping gut metabolism (Fg) predictions made by the HT IVIVE extrapolation model. Css output predictions were between 1.2 to 2.7 fold lower after the incorporation of CYP3A4 contribution. In addition, Fg for the compounds for which data was already collected (diazinon, acetochlor, azoxystrobin and difenoconazole) drop from 1 to as low as 0.14 (difenoconazole) after CYP3A4 enzyme kinetics were incorporated in the calculations. To calculate AED values, we used bioactivity endpoints with the 5th percentile modeled activity concentration at cutoff (modlacc) values assigned as the in vitro POD (in vitro bioactivity data obtained from the EPA’s ToxCast HTS dataset) together with our Css values. The results showed a consistent increase in the predicted AEDs when CYP3A4 TK data was incorporated. Efforts are ongoing to continue evaluating CYP3A4 enzyme kinetics as a surrogate for intestinal clearance in IVIVE TK modeling. Human and rat intestinal microsome clearance rates as well as rat in vivo toxicokinetic data are being generated for in vitro-in vivo comparison. If incorporating intestinal TK data through CYP3A4 contribution proves successful, in silico CYP3A4 prediction could be employed in place of in vitro intestinal clearance data to refine the current HT-IVIVE approach. This will result in a more accurate risk-based prioritization of the existing TSCA chemical space. This abstract does not necessarily reflect the views of the U.S. EPA.
URLs/Downloads:
DOI: Refining High-Throughput In Vitro-In Vivo-Extrapolation Modeling through Incorporation of Intestinal Toxicokinetics
POSTER.PDF (PDF, NA pp, 608.34 KB, about PDF)