Application of in silico methods to predict the acute toxicity of bicyclic organophosphorus compounds as potential chemical weapon

Bicyclic organophosphorus compounds (BOPCs), including flame retardants and plasticisers, are widely used in industrial applications because of their thermal stability and resistance to degradation. However, their unique structural properties and mechanisms of toxicity raise concerns regarding their potential misuse. Unlike classical organophosphorus compounds that inhibit acetylcholinesterase, BOPCs exert toxicity by antagonising gamma-aminobutyric acid receptors, resulting in severe neurotoxic effects, including convulsions and seizures. This underscores the urgent need to prioritise predictive toxicity studies on these compounds as part of a national defence strategy. The present study represents the first extensive application of in silico toxicological approaches to investigate the acute toxicity of a BOPC dataset (n = 18) utilising advanced in silico tools, such as QSAR models and probabilistic software/platforms, to predict acute oral toxicity in rats. All the investigated BOPCs are highly acutely toxic, judging from LD50 values ranging for humans between < 1 mg and > 1.000 mg/kg bw, depending on the applied model. Noticeable variation between model predictions reminds us that present in silico approaches have significant limitations, at least when addressing chemically complex compounds, such as the BOPC class. This calls for wet-laboratory experimentation. Major toxicophoric groups, such as phosphate and phosphorothione moieties, have been identified as significant contributors to their toxicity. This study considers the need for high-level computational tools, well-founded experimental validation, targeted antidotes, and regulatory measures to reduce the risks from BOPCs and improve public health protection and chemical safety.