You are here:
Estimating Likelihood of Fetal In Vivo Interactions Using In Vitro HTS Data (Teratology meeting)
Citation:
Sipes, N., J. Wambaugh, R. Pearce, D. Kapraun, B. Wetmore, M. Devito, S. Auerbach, AND S. Ferguson. Estimating Likelihood of Fetal In Vivo Interactions Using In Vitro HTS Data (Teratology meeting). Presented at Teratology Annual Meeting, San Antonio, TX, June 25 - 29, 2016. https://doi.org/10.23645/epacomptox.5189332
Impact/Purpose:
This is an abstract for a poster to be given by a NIEHS researcher at the Teratology 2016 meeting in San Antonio, TX.
Description:
Tox21/ToxCast efforts provide in vitro concentration-response data for thousands of compounds. Predicting whether chemical-biological interactions observed in vitro will occur in vivo is challenging. We hypothesize that using a modified model from the FDA guidance for drug interaction studies, Cmax/AC50 (i.e., maximal in vivo blood concentration over the half-maximal in in vitro activity concentration), will give a useful approximation for concentrations where in vivo interactions are likely. Further, for doses where maternal blood concentrations are likely to elicit an interaction (Cmax/AC50>0.1), where do the compounds accumulate in fetal tissues? In order to estimate these doses based on Tox21 data, in silico parameters of chemical fraction unbound in plasma and intrinsic hepatic clearance were estimated from ADMET predictor (Simulations-Plus Inc.) and used in the HTTK R-package to obtain Cmax values from a physiologically-based toxicokinetics model. In silico estimated Cmax values predicted in vivo human Cmax with median absolute error of 0.81 for 93 chemicals, giving confidence in the R-package and in silico estimates. A case example evaluating Cmax/AC50 values for peroxisome proliferator-activated receptor gamma (PPARγ) and glucocorticoid receptor revealed known compounds (glitazones and corticosteroids, respectively) highest on the list at pharmacological doses. Doses required to elicit likely interactions across all Tox21/ToxCast assays were compared to estimated daily exposures (Wambaugh et. al., 2014). 199 compounds were estimated to have likely interactions across 1-32 assays for the most conservative 95th % population at doses lower than estimated daily environmental exposures. The major chemical use-categories included pharmaceuticals, chemical intermediates and dyes. Maximum fetal tissue accumulation (2nd trimester-birth) ranged from the most to least accumulated tissue: rest of body, gut, kidney, lung, brain, and thyroid. 3,3',5,5'-Tetrabromobisphenol A and Triphenyltin acetate were among the top affecters across tissues (excluding thyroid) at concentrations disrupting nuclear receptors (PPARγ and retinoid X receptor, respectively). This approach can prioritize compounds and biological pathways quickly when no experimental data exists. Out of domain compounds, passive transport, and later developmental stage are present and need to be evaluated. This approach has shown promise toward estimating in vivo interaction concentrations for HTS data. This abstract does not reflect official NTP or EPA views.
URLs/Downloads:
DOI: Estimating Likelihood of Fetal In Vivo Interactions Using In Vitro HTS Data (Teratology meeting)
FIVIVE_NSS_POSTER06162016V2-SSFV3.PDF (PDF, NA pp, 2801.2 KB, about PDF)
TERATOLOGY2016_SIPESNS_WAMBAUGH_ABSTRACT_FETALMODEL.PDF (PDF, NA pp, 39.238 KB, about PDF)