Fragment-based diffusion modeling and molecular dynamics simulation validation for the discovery of PD-L1...

Area of Science:

• Drug Discovery and Development
• Computational Chemistry
• Cancer Immunotherapy

Background:

• The programmed cell death-1/programmed cell death-ligand 1 (PD-1/PD-L1) pathway is crucial in cancer immunotherapy.
• Monoclonal antibodies targeting PD-1/PD-L1 are effective but have limitations like poor tissue penetration and high costs.
• Small-molecule inhibitors offer a complementary approach, but designing them for the large PD-1/PD-L1 interface is challenging.

Purpose of the Study:

• To develop an efficient strategy for discovering small-molecule inhibitors of the PD-1/PD-L1 pathway.
• To overcome the limitations of traditional antibody-based immunotherapies.
• To computationally prioritize novel drug candidates for experimental validation.

Main Methods:

• Integration of fragment-based drug design (FBDD) with conditional diffusion modeling.
• Utilizing core scaffolds from fragment analysis as conditional inputs for molecular generation.
• Employing molecular docking and molecular dynamics (MD) simulations for screening and evaluation.
• Using MM-PBSA with IE correction to calculate binding free energies.

Main Results:

• Generation and screening of structurally plausible candidate inhibitors using diffusion modeling and docking.
• Identification of four promising candidates (bo1-bo4) after MD simulations.
• Three candidates (bo1, bo2, bo3) exhibited predicted binding free energies below -40 kcal/mol.
• Demonstrated potential of candidates to stabilize the PD-L1 dimer interface in silico.

Conclusions:

• Fragment-based diffusion modeling is an efficient and interpretable strategy for discovering PD-L1 small-molecule inhibitors.
• This approach provides a promising framework for targeting challenging protein-protein interactions in cancer immunotherapy.
• The identified candidates are computationally prioritized for further experimental investigation of PD-1/PD-L1 blockade.