Machine learning-driven drug discovery for the management of TNBC: focus on IDO1 and TDO targets

Area of Science:

• Biochemistry
• Pharmacology
• Computational Chemistry

Background:

• Tryptophan catabolism via the kynurenine pathway generates kynurenine, an oncometabolite linked to cancers like triple-negative breast cancer (TNBC).
• Indoleamine 2,3-dioxygenase (IDO1) and tryptophan 2,3-dioxygenase (TDO) enzymes drive this pathway, creating an immunosuppressive tumor microenvironment.
• Targeting IDO1 and TDO is a promising therapeutic strategy, but no approved drugs inhibit both simultaneously.

Purpose of the Study:

• To employ a machine learning (ML)-driven virtual screening pipeline to identify potent dual inhibitors of IDO1 and TDO.
• To discover novel therapeutic candidates for triple-negative breast cancer (TNBC) by targeting key enzymes in tryptophan catabolism.

Main Methods:

• Developed an in-house ML classification model (XGBRF classifier with 95% accuracy) using IC50 values from ChEMBL and BindingDB.
• Screened the MEGxp database using the trained ML model, followed by molecular docking, MM-GBSA calculations, rescoring, and ADMET profiling.
• Performed 100 ns molecular dynamics simulations to confirm the stability of protein-ligand complexes.

Main Results:

• Identified two promising dual IDO1 and TDO inhibitor candidates: NP000319 and NP003833.
• Both compounds demonstrated predicted anticancer potential against MDA-MB-231 TNBC cells.
• Confirmed the stability of the identified protein-ligand complexes through molecular dynamics simulations.

Conclusions:

• Machine learning-driven strategies are valuable for identifying dual-inhibitor drug candidates.
• The identified compounds NP000319 and NP003833 represent strong leads for further experimental validation and potential therapeutic development for TNBC.
• This study highlights a promising approach for targeting cancer-associated metabolic pathways.