Predicting the oral uptake efficiency of chemicals in mammals: combining the hydrophilic and lipophilic range

Abstract

Environmental risk assessment requires models for estimating the bioaccumulation of untested compounds. So far, bioaccumulation models have focused on lipophilic compounds, and only a few have included hydrophilic compounds. Our aim was to extend an existing bioaccumulation model to estimate the oral uptake efficiency of pollutants in mammals for compounds over a wide K(ow) range with an emphasis on hydrophilic compounds, i.e. compounds in the lower K(ow) range. Usually, most models use octanol as a single surrogate for the membrane and thus neglect the bilayer structure of the membrane. However, compounds with polar groups can have different affinities for the different membrane regions. Therefore, an existing bioaccumulation model was extended by dividing the diffusion resistance through the membrane into an outer and inner membrane resistance, where the solvents octanol and heptane were used as surrogates for these membrane regions, respectively. The model was calibrated with uptake efficiencies of environmental pollutants measured in different mammals during feeding studies combined with human oral uptake efficiencies of pharmaceuticals. The new model estimated the uptake efficiency of neutral (RMSE=14.6) and dissociating (RMSE=19.5) compounds with logK(ow) ranging from -10 to +8. The inclusion of the K(hw) improved uptake estimation for 33% of the hydrophilic compounds (logK(ow)<0) (r(2)=0.51, RMSE=22.8) compared with the model based on K(ow) only (r(2)=0.05, RMSE=34.9), while hydrophobic compounds (logK(ow)>0) were estimated equally by both model versions with RMSE=15.2 (K(ow)&K(hw)) and RMSE=15.7 (K(ow) only). The model can be used to estimate the oral uptake efficiency for both hydrophilic and hydrophobic compounds.