Identifying potential protein targets for toluene using a molecular similarity search, in silico docking and in vitro validation

The toxicity of chemicals greatly depends on their interaction with macromolecular targets. The main goal of this study was to develop an approach for predicting protein targets for chemical toxins using a molecular similarity search of toxin–target information collected in the Toxin and Toxin-Target Database. The developed method was used to identify new targets for toluene which could predict potential cellular toxicity and to validate the approach with in vitro laboratory studies. We obtained 124 potential targets for toluene from a molecular similarity search. Results were further analysed using in silico molecular docking methods. The binding of toluene to two proteins, hemoglobin and serum albumin, was validated by the measurement of binding using microscale thermophoresis. The measured binding constant between toluene and hemoglobin was 1.9 μM, while albumin demonstrated toluene-induced aggregation. These results demonstrate the applicability of an exploratory in silico toxicity tool, based on a molecular similarity search and protein–ligand docking, for identification of potential targets for chemical toxins.