Small-Molecule Inhibitors of TIPE3 Protein Identified through Deep Learning Suppress Cancer Cell Growth In Vitro.

Area of Science:

• Oncology
• Computational Chemistry
• Drug Discovery

Background:

• Tumor necrosis factor-α-induced protein 8-like 3 (TIPE3) is upregulated in various human cancers, promoting tumor proliferation, migration, invasion, and inhibiting apoptosis.
• Targeting TIPE3 is a potential anti-cancer strategy.
• Artificial intelligence (AI) is accelerating anti-cancer drug development.

Purpose of the Study:

• To identify novel inhibitors of TIPE3 using a combination of computational methods.
• To evaluate the in vitro anti-tumor activity of identified compounds.
• To investigate the selectivity and molecular interactions of potential drug candidates.

Main Methods:

• Utilized deep learning (DFCNN, DeepBindBC), molecular docking (Autodock Vina), and molecular dynamics (MD, metadynamics) for virtual screening.
• Screened a ZINC compound dataset against TIPE3.
• Selected six candidates for experimental validation, focusing on three promising compounds.

Main Results:

• Three small-molecule compounds (K784-8160, E745-0011, 7238-1516) exhibited significant in vitro anti-tumor activity, reducing viability, proliferation, and migration while enhancing apoptosis.
• Compounds E745-0011 and 7238-1516 demonstrated selective cytotoxicity against TIPE3-high expressing tumor cells, sparing normal cells.
• Molecular docking confirmed interactions between the inhibitors and TIPE3, revealing key hydrophobic interactions and residues.

Conclusions:

• AI-driven drug discovery, integrating deep learning and MD simulations, is effective for identifying TIPE3 inhibitors.
• The identified compounds show significant potential for developing new anti-cancer therapeutics targeting TIPE3.
• Further research into these selective inhibitors could lead to novel cancer treatments.