In silico fragment-based discovery of CIB1-directed anti-tumor agents by FRASE-bot

Area of Science:

• Drug Discovery and Development
• Computational Biology
• Medicinal Chemistry

Background:

• Identifying chemical probes for novel therapeutic targets, especially hard-to-drug proteins, is a significant challenge in drug discovery.
• Unconventional targets require innovative approaches to expedite the identification of potential drug candidates.

Purpose of the Study:

• To introduce the FRASE-based hit-finding robot (FRASE-bot) for expediting drug discovery against unconventional therapeutic targets.
• To demonstrate the utility of FRASE-bot in identifying a novel ligand for Calcium and Integrin Binding protein 1 (CIB1).

Main Methods:

• FRASE-bot mines 3D structures of ligand-protein complexes to build a database of FRAgments in Structural Environments (FRASE).
• The database is screened for environments similar to the target protein, seeding the target structure with relevant ligand fragments.
• A neural network model prioritizes fragments likely to be native binders, informing ultra-large-scale virtual screening.

Main Results:

• FRASE-bot was applied to identify ligands for CIB1, a target implicated in triple-negative breast cancer.
• FRASE-based virtual screening successfully identified a small-molecule CIB1 ligand.
• The identified ligand demonstrated specific cell-killing activity in CIB1-dependent cancer cells, confirmed via TR-FRET assay.

Conclusions:

• FRASE-bot is an effective tool for expediting drug discovery, particularly for hard-to-drug and unconventional targets.
• The identified CIB1 ligand shows therapeutic potential for triple-negative breast cancer.
• This approach facilitates the translation of biological discoveries into novel therapeutic strategies.