Machine Learning Enabled Structure-Based Drug Repurposing Approach to Identify Potential CYP1B1 Inhibitors

Area of Science:

• Pharmacology and Medicinal Chemistry
• Computational Drug Discovery
• Oncology

Background:

• Drug-metabolizing enzyme (DME)-mediated resistance is a key factor in cancer treatment failure.
• CYP1B1 inactivates several vital anticancer drugs, including docetaxel, tamoxifen, imatinib, cisplatin, and paclitaxel.
• No current drugs target CYP1B1-mediated inactivation, necessitating novel therapeutic strategies.

Purpose of the Study:

• To identify potential inhibitors of CYP1B1 through a drug repurposing strategy.
• To leverage machine learning (ML) and structure-based virtual screening (SB-VS) for discovering new therapeutic agents.
• To address CYP1B1-mediated drug resistance in cancer treatment.

Main Methods:

• Development and application of three machine learning models: support vector machines (SVMs), random forest (RF), and artificial neural network (ANN).
• Virtual screening of the selleckchem database using the best-performing ML model to identify potential CYP1B1 inhibitors.
• In silico analysis of docking scores, binding affinities, and binding modes, followed by in vitro validation of selected compounds.

Main Results:

• Four compounds were selected based on computational analysis and progressed to in vitro testing.
• Chlorprothixene, nadifloxacin, and ticagrelor demonstrated significant CYP1B1 inhibitory activity.
• These compounds exhibited IC50 values in the micromolar range (0.07–3.00 μM).

Conclusions:

• Drug repurposing combined with ML and SB-VS is an effective strategy for identifying CYP1B1 inhibitors.
• Chlorprothixene, nadifloxacin, and ticagrelor represent promising chemical scaffolds for combating CYP1B1-mediated drug resistance.
• These identified compounds hold potential as adjuvant therapies to enhance existing cancer treatments.