Bioactivation of Isoxazole-Containing Bromodomain and Extra-Terminal Domain (BET) Inhibitors

Area of Science:

• Medicinal Chemistry
• Computational Toxicology
• Drug Discovery

Background:

• The 3,5-dimethylisoxazole motif is frequently used in bromodomain and extra-terminal (BET) inhibitors as an acetyl-lysine mimic.
• This motif may lead to bioactivation into toxic reactive metabolites, posing a safety concern for drug development.

Purpose of the Study:

• To develop and validate a computational approach for predicting bioactivation pathways of BET inhibitors.
• To assess the potential for toxic reactive metabolite formation in isoxazole-containing BET inhibitors.

Main Methods:

• Coupled deep neural models for predicting quinone formation, metabolite structures, and biomolecule reactivity.
• Prediction of bioactivation pathways for 32 BET inhibitors.
• Experimental validation of bioactivation for selected inhibitors (OXFBD02, OXFBD04, I-BET151).

Main Results:

• The computational model predicted that BET inhibitors are more likely to undergo bioactivation than non-bioactivated isoxazole-containing molecules.
• OXFBD inhibitors showed bioactivation into both traditional quinones and novel extended quinone-methides, with traditional quinones being dominant experimentally.
• Modeled bioactivations for I-BET151 predicted extended quinone-methides, which were not experimentally verified, indicating areas for model improvement.

Conclusions:

• The coupled modeling approach successfully predicted BET inhibitor bioactivations, including novel pathways.
• Experimental validation confirmed predicted bioactivation pathways, highlighting potential toxicity concerns for drug development.
• The study underscores the need for improved computational models to accurately predict bioactivation and ensure the safety of new drug leads.