Application of toxicology in silico methods for prediction of acute toxicity (LD50) for Novichoks

Abstract

Novichoks represent the fourth generation of chemical warfare agents with paralytic and convulsive effects, produced clandestinely during the Cold War by the Soviet Union. This novel class of organophosphate compounds is characterised by severe toxicity, which, for example, we have already experienced three times (Salisbury, Amesbury, and Navalny's case) as a society. Then the public debate about the true nature of Novichoks began, realising the importance of examining the properties, especially the toxicological aspects of these compounds. The updated Chemical Warfare Agents list registers over 10,000 compounds as candidate structures for Novichoks. Consequently, conducting experimental research for each of them would be a huge challenge. Additionally, due to the enormous risk of contact with hazardous Novichoks, in silico assessments were applied to estimate their toxicity safely. In silico toxicology provides a means of identifying hazards of compounds before synthesis, helping to fill gaps and guide risk minimisation strategies. A new approach to toxicology testing first considers the prediction of toxicological parameters, eliminating unnecessary animal studies. This new generation risk assessment (NGRA) can meet the modern requirements of toxicological research. The present study explains, using QSAR models, the acute toxicity of the Novichoks studied (n = 17). The results indicate that the toxicity of Novichoks varies. The deadliest turned out to be A-232, followed by A-230 and A-234. On the other hand, the "Iranian" Novichok and C01-A038 compounds turned out to be the least toxic. Developing reliable in silico methods to predict various parameters is essential to prepare for the upcoming use of Novichoks.

The prediction of acute toxicity (LD50) for organophosphorus-based chemical warfare agents (V-series) using toxicology in silico methods

“…The advantages and disadvantages of the methods incorporated in TEST are described in a previous publication (Noga et al 2023 ). Calculation options (end point: oral rat LD 50 , method: Consensus and FDA, relax fragment constraint: disabled, chemical transformation simulator: disabled).…”

Underestimations in the In Silico-Predicted Toxicities of V-Agents

“…The program allows the prediction of the acute rat oral LD50 for the query compound with three different approaches: consensus, hierarchical clustering, nearest neighbor. Further details on TEST were recently published [ 9 ]. Also, it allows the prediction of various physical properties including density, surface tension, viscosity and vapor pressure using the following approaches: consensus, hierarchical clustering, nearest neighbor, group contribution and single model (additional model for viscosity).…”

“…Thus, in silico toxicology is an important field for the assessment of potential chemical warfare agents. Previously, the TEST software was used to predict the acute oral toxicity in rats of Novichok nerve agents [ 9 ]. The ProTox-II program was developed as an alternative to TEST and is an online platform for the prediction of acute oral rat toxicity [ 10 ].…”

A-agents, misleadingly known as “Novichoks”: a narrative review

“…The trend of the hydrolysis rate between A-230 > A-232 > A-234 corresponded with the lipophilicity of molecule A-234 > A-232 > A-230 Noga et al ( 2023a ) A-230, A-232, A-234, A-242, A-262 Hydrolysis and biodegradation Evaluation of hydrolysis estimates showed extremely rapid degradation of compounds A-230 and A-242 in contrast to A-232, A-234, and A-262 ChemSpider ( b , 2018c , 2018a ) A-230, A-232, A-234 Density, boiling point, vapor pressure, enthalpy, flash point, etc. Predicted by ACD/Labs ChemAxon Means of modeling: Franca et al ( 2019 )—chemicalize.com; Carlsen et al ( 2019 )—QSAR modeling; Lyagin and Efremenko ( 2019 )—molecular docking; Bhakhoa et al ( 2019 )—DFT; Tan et al ( 2019 )—DFT; De Farias ( 2019 )—SE method; Nakano et al ( 2019 )—DFT; Imrit et al ( 2020 )—DFT; Motlagh et al ( 2020 )—DFT; Yar et al ( 2021 )—DFT; Otsuka and Miyaguchi ( 2021 )—DFT; Vieira et al ( 2021 )—DFT, QSAR modeling; Chernicharo et al ( 2021 )—DFT; Sajid et al ( 2021 )—DFT; Eskandari et al ( 2022 )—DFT; Jeong et al ( 2022a )—DFT, QSAR modeling; Kim et al ( 2022 )—DFT; Jeong et al ( 2022b )—DFT; Rashid et al ( 2023 )—DFT; Noga et al ( 2023a , b )—QSAR modeling DFT density functional theory, SE semiempirical, QSAR quantitative structure–activity relationship, n/a not available …”

“… Means of modeling: Franca et al ( 2019 )—chemicalize.com; Carlsen et al ( 2019 )—QSAR modeling; Lyagin and Efremenko ( 2019 )—molecular docking; Bhakhoa et al ( 2019 )—DFT; Tan et al ( 2019 )—DFT; De Farias ( 2019 )—SE method; Nakano et al ( 2019 )—DFT; Imrit et al ( 2020 )—DFT; Motlagh et al ( 2020 )—DFT; Yar et al ( 2021 )—DFT; Otsuka and Miyaguchi ( 2021 )—DFT; Vieira et al ( 2021 )—DFT, QSAR modeling; Chernicharo et al ( 2021 )—DFT; Sajid et al ( 2021 )—DFT; Eskandari et al ( 2022 )—DFT; Jeong et al ( 2022a )—DFT, QSAR modeling; Kim et al ( 2022 )—DFT; Jeong et al ( 2022b )—DFT; Rashid et al ( 2023 )—DFT; Noga et al ( 2023a , b )—QSAR modeling DFT density functional theory, SE semiempirical, QSAR quantitative structure–activity relationship, n/a not available …”