Workflow for predictive risk assessments of UVCBs: cheminformatics library design, QSAR, and read-across approaches applied to complex mixtures of metal naphthenates

Abstract

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 one such category of UVCBs that are the reaction products of naphthenic acids with metals to form complex mixtures. Metal naphthenates are often found or used in household and industrial materials with potential for human exposure, but very few of these materials have been evaluated for causing human health hazards. Herein, we evaluate metal naphthenates using predictions derived from read-across and quantitative structure-activity/property relationship (QSAR/QSPR) models. Accordingly, we first built a computational chemistry library by enumerating the structures of naphthenic acids and derived 11,850 QSAR-acceptable structures; then, we used open and commercial in silico tools on these structures to predict a set of physicochemical properties and toxicity endpoints. We then compared the QSAR/QSPR predictions with available experimental data on naphthenic acids to provide a more complete picture of the contributions of the components to the toxicity profiles of metal naphthenate mixtures. The available systematic acute oral toxicity values (LD₅₀) and QSAR LD₅₀ predictions of all the naphthenic acid components indicated low concern for toxicity. The point of departure predictions for chronic repeated dose toxicity for the naphthenic acid components using QSAR models developed from studies on rats ranged from 25 to 50 mg/kg/day. These values are in good agreement with findings from studies on copper and zinc naphthenates, which had no observed adverse effect levels of 30 and 118 mg/kg/day, respectively. Hence, this study demonstrates how published in silico approaches can be used to identify the potential components of metal naphthenates for further testing, inform groupings of UVCBs such as naphthenates, as well as fill the data gaps using read-across and QSAR models to inform risk assessment.

Keywords: cheminformatics; metal naphthenates; quantitative structure–activity relationships; quantitative structure–property relationships; read-across; risk assessment; unknown or variable composition complex reaction products and biological materials (UVCBs); virtual chemical library.

FIGURE 1

Sample naphthenic acid structures, where M+2 is the metal cation, R is an alkyl chain, Z is the hydrogen deficiency, and n is the number of CH₂ units. A copper naphthenate example structure is shown in the top-right box. More negative Z structures, such as tricyclic compounds, are described in the supporting information.

FIGURE 2

Overall methodology for predictive risk assessment of a UVCB. Step 1: The scaffolds and R-group library are designed based on UVCB characterization. Step 2: Cheminformatics tools are used to enumerate possible structures. Steps 3, 4, 5, and 6: Further in silico technologies (read-across, ADME, LD₅₀, and repeat-dose toxicity predictions) are applied for the thousands of component structures.

FIGURE 3

Creation of a virtual chemical library of naphthenic acids. Through two enumeration steps, 26 alkyl chains (middle) and 15 alkyl chains with carboxylic acids (bottom) were attached to the eight core rings. The attachment points are indicated by asterisks. The possible combinations of R-groups and ring positions resulted in the derivation of 11,850 unique naphthenic acid structures.

FIGURE 4

Distribution of molecular weights for the naphthenic acids in the virtual library. The gray box shows the experimentally derived range of mean molecular weights for crude naphthenic acids.

FIGURE 5

Generalized read-across predictions for the 14 naphthenic acids identified using the EPA CompTox Chemicals dashboard. The structures are shown along with their CAS numbers. The read-across prediction for each toxicity outcome is shown on a scale of red to green (red: 1.7–250; orange: 250–600; yellow: 600–1,200; green: 1,200–4,531 mg/kg/day).

FIGURE 6

(A) Scatter plot of the predicted octanol/water partitions and water solubilities (in log units) for all naphthenic acids in the virtual chemical library (each point represents one chemical). (B) Scatter plot of predicted human hepatic intrinsic clearance (µL/min/10⁶ cells) versus predicted human unbound plasma fraction. All panels have the same numbers of points, and the molecules are colored according to cyclization degrees of the structures ranging from blue (low) to red (high). (C) Scatter plot of the predicted median lethal oral dose (LD₅₀) in rats versus molecular weight for all naphthenic acids in the virtual chemical library. (D) Pie chart of the predicted EPA acute oral hazard categories for the naphthenic acids, with category 3 being “low toxicity” and category 4 being “very low toxicity.” No chemicals were predicted in the more toxic categories 1 and 2.

FIGURE 7

(A) Histogram of the predicted rat developmental POD values for all naphthenic acids in the virtual chemical library. The antilog-transformed values are equivalent to 79–95 mg/kg/day. (B) Histogram of the predicted rat chronic POD values. The antilog equivalents are between 25 and 50 mg/kg/day. (C) Density plot showing the naphthenic acid population with their predicted median lethal oral doses in rats (LD₅₀) and predicted rat chronic POD values.