2007 Progress Report: Development of a BBPK Model for the Thyroid Axis in the Pregnant Rat and Fetus for the Dose Response Analysis of Developmental Neurotoxicity

EPA Grant Number: R832134
Title: Development of a BBPK Model for the Thyroid Axis in the Pregnant Rat and Fetus for the Dose Response Analysis of Developmental Neurotoxicity
Investigators: Fisher, Jeffrey W. , Ferguson, Duncan , Wagner, John
Institution: University of Georgia
EPA Project Officer: Hahn, Intaek
Project Period: December 1, 2004 through November 30, 2008 (Extended to November 30, 2009)
Project Period Covered by this Report: December 1, 2006 through November 30,2007
Project Amount: $749,127
RFA: Development and Characterization of Biological Systems for Studying Low Dose Effects of Endocrine Disrupting Chemicals (2004) RFA Text |  Recipients Lists
Research Category: Endocrine Disruptors , Health Effects , Economics and Decision Sciences , Health , Safer Chemicals


The goal under this cooperative agreement is to develop biologically based pharmacokinetic (BBPK) maternal/neonatal models of the hypothamalic-pituitary-thyroid (HPT) axis in the rat for use in assessing developmental neurotoxicity mediated via the HPT axis. These BBPK HPT models will describe the highly non-linear relationship(s) between administered dose of thyroid active chemical, mode of action (MOA) specific perturbation in the HPT axis, and developmental neurotoxicity.


The biological based model for the thyroid will be created using literature derived information, unpublished studies, and collecting data sets to allow the successful development and validation of the thyroid model in the pregnant rat. Memory and learning studies in pups whose mother’s were exposed to perchlorate and low dietary iodide will conducted using slices of their brains and obtaining electrophysiological information.

Progress Summary:

A meeting occurred on March 14-16, 2007, at the University of Massachusetts to discuss research plans. UGA attendees included Dr. Ferguson, Matt Taylor, Dr. Fisher, Eva McLanahan, and Libby Myers. The decision to conduct experiment 3 at UGA was made at this meeting (see results below). Dr. Fisher invited Dr. Tom Zoeller to speak at UGA on Nov. 15th, 2007. He spent two days at UGA. We discussed research issues and outlined potential future work, if funding was obtained. Drs. Crofton and Gilbert joined in for a short conference call while Dr. Zoeller was at UGA.

In the third year of the grant, a third animal study was completed to resolve differences in results noted between the first and second studies. The study involved 36 timed pregnant dams distributed across four PTU doses and a control group, spanning the dose range covered by studies 1 and 2 (0, 0.3, 1, 3, and 10 ppm PTU in drinking water from GD2 until weaning). The overall study design was very similar to that of the second study, though target litter sizes were increased by two pups.

As before, the dams were significantly affected at high dose levels, though not as affected as their pups. Specifically, cortical Type 2 deiodinase (D2) was significantly increased at the highest dose, and serum T3 and T4 levels were decreased at that dose as well. In contrast to the first study, no alterations were found in the cortical T3 (or T4) levels of the dams.

As before, the PND14 pups were more affected by treatment than their PND21 littermates. This may be quite significant given that this timeframe is a key period for development of the hypothalamic-pituitary-thyroid axis. D2 activity is increased at 10, 3, and 1 ppm doses in this age group. Serum T3 declines at 10 ppm, and serum T4 declines at both 3 and 10 ppm. Cortical TH levels follow the same pattern, with cortical T3 declining at 10 ppm and cortical T4 at both 3 and 10 ppm.

The PND 21 pups were less affected than the first study as well. Cortical D2 activity was increased only at the highest dose, and serum T3 and T4 are decreased only at that dose as well. This contrasts with the first study where D2 and serum T3 and T4 were decreased at both 3 and 10 ppm dose levels. Also in contrast to the first study, cortical T3 (and T4) levels only approached significance (p=0.055). However, electrophysiology results from this age group parallel the results in the first study, with synaptic potential impaired at both the 3 and 10 ppm dose levels. Strangely, for reasons not clear except that the diminution of serum T4 concentration was not as severe at the same PTU dosage, the striking correlation from study 1 between D2 activity and synaptic potential was not evident in this study. As in the first study, synaptic activity (srmax) was impaired at 3 and 10 ppm. However, LTP results were somewhat contradictory in this study, with LTP impairments at the 30 minute timepoint (“weak” LTP) evident in the 1 and 3 ppm groups, but in the 1 and 0.3 ppm groups at 90 minutes.

Once again, pups weaned and allowed to grow under standard housing conditions for 2 months after weaning returned to normal in most respects. Interestingly, we found that serum T4 levels were increased at all in the face of normal TSH, but the highest dose level at the PND 90 mark. Additionally, cortical T3 levels were increased in the 10 ppm adults. As D2 activity had normalized, the possibilities exist that D3 activity was reduced or that TH transporters were persistently altered. However, neither of these parameters were measured in this study. In contrast to the first study, no lasting electrophysiological changes were detected. In PND21 pups from the second study, cortical D2 mRNA by in situ hybridization was found to increase by less than 2-fold while activity increased by over 7-fold in 3 ppm pups, indicating that post-translational regulation of D2 is the more important component of that response at the studied doses. Samples for brain in situ hybridization for the thyroid hormone-responsive markers, RC3 and neurogranin, D2 and MCT8 were sent to the Zoeller laboratory and results are still pending. We expect that these results may help evaluate the possibility of diversity of regulation of thyroid hormone action, metabolism and transport (in that order) in key brain loci. Nissl-stained histological sections of coronal sections of the brain in the region of the hippocampus are also currently being evaluated in studies 2 and 3 to establish whether cellular bilateral heterotopia in the corpus callosum is observed as in the studies of PTU by Gilbert, et al. ( Endocrinology 148:2593-2597, 2007 ). This heterotopia has been associated with GABAergic inhibitory neurons and is believed to be neurons which have arrested their migration to the hippocampus.

Overall, because the dosages of PTU spanned the widest range of oral PTU dosages (0, 0.3, 1, 3, 10) and developmental times yet, and seems to have captured for some parameters a NOAEL, but for others may indicate a small degree of hormesis at the lowest PTU dosage (0.3 ppm). Otherwise, this study confirmed the greater sensitivity of the fetus and neonate to chemically induced thyroxine deficiency. The neurophysiological changes in the hippocampus were also confirmed in this study, qualitatively, if not quantitatively, consistent with the results of the first PTU study. The dosage curve of dosage versus effect was shifted slightly to the right for some parameters and was as or more sensitive than study 1 and 3 for others. The most sensitive timepoint appeared to be the PND14 developmental point. In order to distinguish the study results of studies 2 and 3, and because we suspect possible dietary factors such as variability of isoflavonoid content, we plan on measuring genistein and daidzein in serum of PND14 pups.

The BBDR-HPT axis sub-models for the adult rat were constructed using simple model structure that allowed us to focus on an empirical ‘system based evaluation’ of key biochemical features of the HPT axis, such as the negative feedback loop. For example, the production of thyroid hormones is controlled, in part, by the model predicted serum TSH concentration, while the maximal rate of active sequestration of iodide into the thyroid is also controlled by the serum TSH concentration. Future modeling efforts focused on expanding the BBDR-HPT axis model to include other tissues of interest (e.g., brain for correlation of tissue concentrations to developmental effects) can be readily integrated into our model structure.

Models were coded using acslXtreme version (Aegis Technologies, Huntsville, Alabama) and solved with the Gear algorithm for stiff systems. Standardized units of nanomoles (nmol), liters (L), kilograms (kg), and hours (hr) were used in the sub-models. The approach taken for the development of the BBDR-HPT axis model was to first create simple and independent sub-model structures for radiolabeled –iodide, -TSH, -T4, and -T3 using radiotracer studies reported in literature for the adult rat. This provided some BBDR-HPT axis model parameter values, although sometimes preliminary, and helped to evaluate the adequacy of using the proposed structure for each sub-model. Then, literature derived datasets with endogenous information for the sub-models (iodide, TSH, T4, and T3) were gathered and the sub-models were linked as a system to simulate the HPT axis in the euthyroid adult rat. Key features of the BBDR-HPT axis model included the negative feedback loop, thyroid hormone production using available dietary iodide, and the metabolism of thyroid hormones with release of free iodide available for reuse in thyroid hormone production or urinary excretion. The euthyroid steady-state BBDR-HPT axis model relied on dietary iodide as the only exogenous input. Finally, the calibrated euthyroid, iodide sufficient adult rat BBDR-HPT axis model was tested for its ability to predict perturbations in the system under iodide deficient conditions. The modeling effort was successful.

Expected Results:

This project is expected to assist the US EPA in the development of dose-response assessment methodology for chemical induced hypothyroidism in rodents.

Future Activities:

Computational efforts to develop HPT axis model(s) for different reproductive states will be undertaken.

Journal Articles on this Report : 2 Displayed | Download in RIS Format

Other project views: All 15 publications 5 publications in selected types All 5 journal articles
Type Citation Project Document Sources
Journal Article McLanahan ED, Andersen ME, Fisher JW. A biologically based dose-response model for dietary iodide and the hypothalamic-pituitary-thyroid axis in the adult rat: evaluation of iodide deficiency. Toxicological Sciences 2008;102(2):241-253. R832134 (2007)
R832134 (2008)
  • Abstract from PubMed
  • Full-text: Oxford Journals-Full Text HTML
  • Abstract: Oxford Journals-Abstract
  • Other: Oxford Journals-PDF
  • Journal Article Taylor MA, Swant J, Wagner JJ, Fisher JW, Ferguson DC. Lower thyroid compensatory reserve of rat pups after maternal hypothyroidism: correlation of thyroid, hepatic, and cerebrocortical biomarkers with hippocampal neurophysiology. Endocrinology 2008;149(7):3521-3530. R832134 (2007)
    R832134 (2008)
  • Abstract from PubMed
  • Full-text: Endocrine Society-Full Text HTML & PDF Link
  • Supplemental Keywords:

    thyroid, perchlorate, biologically based model, pregnancy, learning, memory, iodide, rat, Modeling, thyroid, endocrine disruptors, hippocampal long-term potentiation, risk assessment, children, fetus, toxics, RFA, Health, Scientific Discipline, PHYSICAL ASPECTS, Environmental Chemistry, Health Risk Assessment, Endocrine Disruptors - Environmental Exposure & Risk, Risk Assessments, endocrine disruptors, Physical Processes, Biochemistry, Biology, Endocrine Disruptors - Human Health, neurotoxic, bioindicator, exposure, EDCs, thyroid toxicants, exposure studies, endocrine disrupting chemicals, sexual development, dose response, developmental biology, endocrine disrupting chemcials, human growth and development, toxicity, hormone production, assessment technology, ecological risk assessment model, human health risk

    Progress and Final Reports:

    Original Abstract
  • 2005 Progress Report
  • 2006 Progress Report
  • 2008 Progress Report
  • Final