A functional biologically motivated model for thyroid axis disturbances in the pregnant rat
On this page:
Disruption of the thyroid endocrine axis during development may result in irreversible adverse neurodevelopmental outcomes. New in vitro and in silico methods to evaluate chemicals for developmental toxicology potential are proposed for future regulatory use. Our aim is to build on the historical computational work with the thyroid axis and develop strategies for predicting in utero thyroid hormone disruption based on a combined in vitro, in vivo, and in silico workflow. This methodology may be used for (a) screening and prioritization (b) interpretation of existing and emerging animal studies (c) informing design of studies needed to improve in silico models (d) in vitro to in vivo extrapolation (IVIVE) and (e) in vivo animal to human extrapolation. Using perchlorate as a case study, a biologically based dose response (BBDR) model for the pregnant rat was constructed to predict maternal thyroid hormone disruption at the sodium iodide symporter (NIS) protein. Production of steady state concentrations of thyroid stimulating hormone (TSH), thyroxine and triiodothyronine plasma concentrations in the dam was based on a non-linear negative feedback loop. TSH production is increased under conditions of low thyroxine, these increases in TSH activating thyroxine production to maintain adequate levels in circulation. Mathematical representation of the complexity of these interactive control mechanisms across lifestage is challenging. Perchlorate’s action to block the NIS protein reduces thyroxine production. The reduction in thyroxine in turn serves to increase TSH, driving an upregulation in the production of NIS protein in the thyroid gland. Under low concentrations of perchlorate, the upregulation of NIS appears to be sufficient to maintain thyroxine levels, but fail in the presence of higher concentrations. Initial simulations with the calibrated BBDR model resulted in predictions of maternal Gestation Day 20 (GD20) plasma concentrations of TSH and thyroxine/triiodothyronine that were within 0.5 to 3-fold of observations. The perchlorate drinking water exposures ranged from 0.01 to 82 mg/kg/day. These interactive control mechanisms are present in the mother and the fetus, with more data available for the dam than the fetus. The BBDR model predicted maternal thyroxine concentrations were then linked to published GD20 fetal biomarkers for thyroid disruption (e.g., serum thyroxine or brain concentrations of thyroxine and triiodothyronine). Unfortunately lack of data hampered the creation of a TSH-T4 negative feedback loop in the rat fetus. However, once this model is established for the pregnant rat, chemicals exhibiting other modes of action (Molecular Initiating Events, MIEs) can be addressed with this BBDR HPT axis framework. From this work in the rodent model, construction of a BBDR model for the TSH-T4 negative feedback loop in the human fetus and mother may be possible using available published literature. Does not reflect EPA policy