Computational design of diverse nuclear factor erythroid 2 activators with cellular antioxidative activity

Area of Science:

• Biochemistry
• Molecular Biology
• Computational Biology

Background:

• Oxidative stress is implicated in chronic diseases.
• The KEAP1-NRF2 pathway is crucial for cellular antioxidant response.
• Existing synthetic antioxidants have limitations in activating this pathway.

Purpose of the Study:

• To design novel miniproteins using deep learning to activate the NRF2 pathway.
• To develop potent and selective activators of cellular antioxidant defenses.

Main Methods:

• Application of deep learning algorithms for miniprotein design.
• Biochemical assays to determine binding affinities to KEAP1.
• Cell-based assays to assess NRF2 activation and antioxidant gene expression.

Main Results:

• Five designed miniprotein binders demonstrated potent interference with the KEAP1-NRF2 complex (4.4 nM to 53.3 nM affinity for KEAP1).
• Two miniproteins, designed via motif scaffolding, successfully activated NRF2 in eukaryotic cells.
• Activated NRF2 led to a 3.8-fold increase in antioxidant gene expression and enhanced cell survival under oxidative stress.

Conclusions:

• Deep learning-integrated models can create stable miniproteins with diverse structures and properties.
• These miniproteins offer a promising therapeutic strategy against oxidative damage and related diseases.
• The designed miniproteins support adaptive homeostasis in cellular antioxidative systems.