Bioaccumulation and uptake kinetics of per- and polyfluoroalkyl substances in Pimephales promelas

Abstract

Due to the ubiquitous contamination of water resources by per- and polyfluoroalkyl substances (PFAS), there has been growing interest in understanding the disposition of PFAS in the environment. However, bioaccumulation dynamics remain poorly understood for many substances within this group of chemicals, and kinetic-based information is limited, particularly for fathead minnows (Pimephales promelas), a common ecotoxicology model that is not commonly utilized during bioaccumulation studies. Therefore, the aim of this study was to examine uptake of 19 PFAS by adult fathead minnows over 7 days at levels that did not elicit standardized adverse effects in larval fish. A PFAS mixture consisting of short and long chain perfluoroalkyl carboxylic acids (PFCAs) and perfluoroalkyl sulfonates (PFSAs), a sulfonamide, and fluorotelomer sulfonates was used with an exposure system simulating groundwater introduction to surface waters. Water, whole-body tissue, and plasma were collected at multiple timepoints, and samples were analyzed via liquid chromatography-tandem mass spectrometry. Uptake kinetics were estimated for each PFAS with nonlinear regressions for both tissue and plasma. Increasing PFAS concentrations were observed with increasing chain length in tissue and plasma for both PFCAs and PFSAs. Ratio-based bioaccumulation factors (BAFs) and blood-water partitioning coefficients (PBW) were estimated, with steady-state BAFs and PBWs up to 26.6 L/kg and 85.8, respectively, for perfluorooctanoic acid. We also estimated apparent volume of distribution (VD) for each chemical to examine the distribution of PFAS in fish. These VD estimates were typically below 1 L/kg, indicating PFAS were more distributed in plasma than tissue, and for both PFCAs and PFSAs, VD estimates generally decreased with increasing chain length. Our findings contribute to improving a predictive understanding of PFAS bioaccumulation in a common fish model.