Small Molecules Targeting the Structural Dynamics of AR-V7 Partially Disordered Proteins Using Deep Ensemble Docking.

Area of Science:

• Computational Biology
• Drug Discovery
• Structural Biology

Background:

• Partially disordered proteins present significant challenges for traditional drug discovery due to their dynamic nature and numerous transient binding sites.
• The AR-V7 splicing variant is implicated in prostate cancer progression, making it a critical therapeutic target.

Purpose of the Study:

• To develop and validate a deep ensemble docking pipeline for accelerated drug screening against partially disordered proteins.
• To identify functional binding sites on the AR-V7 protein and discover small molecule binders that modulate its activity.

Main Methods:

• A deep ensemble docking pipeline was employed to analyze the conformational ensemble of AR-V7.
• Dimension reduction techniques were used to identify functionally relevant binding sites.
• Physics-based molecular docking combined with machine learning models screened for small molecule binders.
• Atomistic molecular dynamics simulations assessed the effect of identified binders on AR-V7 dynamics.

Main Results:

• The pipeline identified functional binding sites on AR-V7 at phase separation-prone regions, reducing binding site dimensionality by 90%.
• The multibinding site hit rate was increased by a factor of 17 compared to naive docking.
• A selected compound, ChEMBL22003, was found to reduce AR-V7 conformational entropy and modulate its phase separation-prone regions.

Conclusions:

• The developed deep ensemble docking pipeline effectively accelerates the discovery of binders for partially disordered proteins.
• ChEMBL22003 shows potential as an AR-V7 phase separation modulator, offering a new therapeutic strategy for prostate cancer.