Toward a biologically based dose-response model for developmental toxicity of 5-fluorouracil in the rat: acquisition of experimental data.

Abstract

Biologically based dose-response (BBDR) models represent an emerging approach to improving the current practice of human health-risk assessment. The concept of BBDR modeling is to incorporate mechanistic information about a chemical that is relevant to the expression of its toxicity into descriptive mathematical terms, thereby providing a quantitative model that will enhance the ability for low-dose and cross-species extrapolation. Construction of a BBDR model for developmental toxicity is particularly complicated by the multitude of possible mechanisms. Thus, a few model assumptions were made. The current study illustrates the processes involved in selecting the relevant information for BBDR modeling, using an established developmental toxicant, 5-fluorouracil (5-FU), as a prototypic example. The primary BBDR model for 5-FU is based on inhibition of thymidylate synthetase (TS) and resultant changes in nucleotide pools, DNA synthesis, cell-cycle progression, and somatic growth. A single subcutaneous injection of 5-FU at doses ranging from 1 to 40 mg/kg was given to pregnant Sprague-Dawley rats at gestational day 14; controls received saline. 5-FU was absorbed rapidly into the maternal circulation, and AUC estimates were linear with administered doses. We found metabolites of 5-FU directly incorporated into embryonic nucleic acids, although the levels of incorporation were low and lacked correlation with administered doses. On the other hand, 5-FU produced dose-dependent inhibition of thymidylate synthetase in the whole embryo, and recovery from enzyme inhibition was also related to the administered dose. As a consequence of TS inhibition, embryonic dTTP and dGTP were markedly reduced, while dCTP was profoundly elevated, perhaps due to feedback regulation of intracellular nucleotide pools. The total contents of embryonic macromolecules (DNA and protein) were also reduced, most notably at the high doses. Correspondingly, dose-related reductions of fetal weight were seen as early as GD 15, and these deficits persisted for the remainder of gestation. These detailed dose-response parameters involved in the expression of 5-FU developmental toxicity were incorporated into mathematical terms for BBDR modeling. Such quantitative models should be instrumental to the improvement of high-to-low dose and cross-species extrapolation in health-risk assessment.