You are here:
Updated models for predicting biotransformation impacts on chemical bioconcentration in rainbow trout
Citation:
Arnot, J., M. Embry, J. Armitage, AND J. Nichols. Updated models for predicting biotransformation impacts on chemical bioconcentration in rainbow trout. SETAC North America, Sacramento, CA, November 04 - 08, 2018.
Impact/Purpose:
In vitro-in vivo extrapolation methods have been developed as a means of incorporating measured rates of in vitro biotransformation into predictive models for chemical bioaccumulation in fish. To date, however, these methods have employed the QSAR-based bioaccumulation model (1-Co-QSAR) given previously by Arnot and Gobas (2003, QSAR Comb. Chem., 22:337). In this poster, we examine hypothetical effects of in vitro biotransformation on bioconcentration factors (BCFs) predicted using both the 1-Co-QSAR model and a 1-compartment physiologically based model (1-Co-PBPK) given by Arnot et al. (2008, Environ. Toxicol. Chem., 27:341). Model performance was generally similar; however, compared to the 1-Co-QSAR model, the impact of biotransformation predicted by the 1-Co-PBPK model at a fixed rate of activity was greater. This finding may be important, since the 1-Co-QSAR model has been reported to underestimate biotransformation impacts on bioaccumulation. The 1-Co-PBPK explicitly accounts for differences in temperature and fish size. As such, this model is better suited to address questions related to possible differences within and among fish species. In time, we anticipate that the 1-Co-PBPK model will be adopted as a replacement for the 1-Co-QSAR model for chemical bioaccumulation assessments.
Description:
Biotransformation may substantially reduce chemical accumulation in fish and is a recognized source of uncertainty in many bioaccumulation assessments. Here we describe and compare two one-compartment toxicokinetic models that can be used to predict the effect of hepatic biotransformation on bioconcentration factors (BCFs) in rainbow trout (Oncorhynchus mykiss). Both models employ established methods to extrapolate in vitrointrinsic clearance rates measured using liver S9 fractions or isolated hepatocytes to the whole animal; however, one model relies of QSARs to predict rates of chemical uptake and elimination, while the other describes these processes in physiological terms. Compared to the QSAR-based model, the physiological model more explicitly addresses how difference in fish size and water temperature affect rate constants for chemical uptake and elimination (e.g., branchial uptake rate) in the fish. In addition, the physiological model provides a more direct link to previously published estimates of whole-body biotransformation rates (kM), derived fromin vivoBCF and dietary bioaccumulation tests in various fish species. BCFs predicted by the physiological model in the absence of biotransformation are higher than those generated by the QSAR-based model, due to differences in predicted chemical partitioning and fecal egestion. However, the physiological model predicts lower BCFs for compounds that undergo biotransformation at all but very low levels of activity. This difference in predicted BCFs varies with log KOWand the rate of activity, and in some instances, exceeds 1.5-fold (BCFQSAR/BCFPhys). The utility of the physiological model was demonstrated by calculating kMvalues for a suite of polycyclic aromatic hydrocarbons using measured in vitrobiotransformation rate data. The resulting kMvalues exhibit remarkably good agreement with in vivoand in silico(QSAR) estimates. Finally, uncertainty in model selection for BCF calculation was put into context by highlighting variability and uncertainty in measuredin vivoBCFs.