Discovery of a Novel DCAF1 Ligand Using a Drug-Target Interaction Prediction Model: Generalizing Machine Learning to...

Area of Science:

• Biochemistry
• Molecular Biology
• Drug Discovery

Background:

• DCAF1 acts as a crucial substrate recruitment subunit for CRL4 and EDVP E3 ubiquitin ligases.
• The DCAF1 WDR domain binds substrates and is targeted by lentiviruses for host factor degradation.
• DCAF1 is a promising target for developing chemical inhibitors and PROTACs for targeted protein degradation.

Purpose of the Study:

• To identify selective ligands for the DCAF1 WDR domain using AI-driven virtual screening.
• To validate predicted ligands through biophysical and crystallographic methods.
• To showcase the efficacy of AI in ligand discovery for previously unligated targets.

Main Methods:

• Proteome-scale drug-target interaction prediction using the MatchMaker model.
• Cheminformatics filtering and molecular docking for ligand identification.
• Biophysical screening (e.g., SPR, ITC) and X-ray crystallography for validation.

Main Results:

• Identification of selective ligands for the DCAF1 WDR domain.
• Experimental confirmation of predicted binders.
• X-ray crystallography revealed ligand binding within the central WDR domain cavity.
• Demonstrated successful application of AI virtual screening without prior ligand data.

Conclusions:

• AI-enabled virtual screening is effective for discovering ligands for novel targets like DCAF1.
• This approach accelerates the development of chemical inhibitors and PROTACs.
• The identified ligands provide a foundation for further therapeutic development targeting DCAF1.