Comparative Structure-Based Virtual Screening Utilizing Optimized AlphaFold Model Identifies Selective HDAC11...

Area of Science:

• Biochemistry
• Structural Biology
• Drug Discovery

Background:

• Histone deacetylase 11 (HDAC11) is a Class IV histone deacylase lacking a reported crystal structure.
• Low sequence homology of HDAC11's catalytic domain hinders traditional homology modeling.
• AlphaFold offers accurate protein structure prediction, even for novel structures, but requires optimization for drug design applications.

Purpose of the Study:

• To implement a structure-based virtual screening approach using an optimized HDAC11 AlphaFold model.
• To identify novel and selective HDAC11 inhibitors.
• To validate the utility of optimized AlphaFold models in drug discovery.

Main Methods:

• Optimization of an HDAC11 AlphaFold model by incorporating the catalytic zinc ion and performing minimization with known inhibitors.
• Comparative structure-based virtual screening utilizing the optimized model.
• In vitro enzymatic assays to evaluate identified hit compounds.
• Molecular dynamics simulations to confirm binding hypotheses.

Main Results:

• A virtual screening approach successfully identified a novel hit compound.
• The hit compound demonstrated an IC50 of 3.5 µM against HDAC11.
• The compound selectively inhibited HDAC11 over other HDAC subtypes at 10 µM.
• Molecular dynamics simulations supported the docking-derived binding hypothesis.

Conclusions:

• The study validates an optimized AlphaFold model approach for identifying selective HDAC11 inhibitors.
• AlphaFold models, when appropriately optimized, are applicable tools for novel inhibitor discovery.
• This work reinforces the potential of AI-driven structural biology in advancing drug discovery efforts.