Andrew J. Prussia ^* Clement Welsh Tarah Somers Patricia Ruiz

• Agency for Toxic Substances and Disease Registry, Atlanta, United States

Substances of unknown or variable composition, complex reaction products, and biological materials (UVCBs) are commonly found in the environment. However, assessing their human toxicological risk is challenging due to their variable composition and many constituents. Metal naphthenate salts are a category of UVCBs, reaction products of naphthenic acids with metals which form complex mixtures. Metal naphthenates are typically found or used in household and industrial materials with potential for human exposure, but few have been evaluated for their human health hazards. Here, we evaluate metal naphthenates using predictions derived from read-across and quantitative structure-activity/property relationship (QSAR/QSPR) models. To do this, we first built a computational chemistry library by enumerating structures of naphthenic acids. For the 11,850 QSAR-acceptable structures derived, established open and commercial in silico tools were used to predict a selection of physicochemical properties and toxicity endpoints. Comparing the QSAR/QSPR predictions with available experimental data for the naphthenic acids provided a more complete picture of the contribution that the mixture components played in the toxicity profile of the metal naphthenate mixtures. Published systematic acute oral toxicity values (LD50) and QSAR LD50 predictions indicated a low concern of toxicity for all the naphthenic acid components. Point of departure (POD) predictions for chronic repeated dose toxicity using QSAR models developed from studies in rats ranged from 25 to 50 mg/kg/day for the components of naphthenic acids. These were in good agreement with available studies of copper and zinc naphthenates, which had No Observed Adverse Effect Levels (NOAELs) of 30 and 118 mg/kg/day, respectively. This study demonstrated how published in silico approaches can be used to identify potential components of metal naphthenates for further testing, inform groupings of UVCBs such as naphthenates, as well as fill data gaps using read-across and QSAR models to inform risk assessment.