Machine-Assisted Discovery of Chondroitinase ABC Complexes toward Sustained Neural Regeneration

Area of Science:

• Biomaterials Science
• Neuroscience
• Enzyme Engineering

Background:

• Chondroitin sulfate proteoglycans (CSPGs) inhibit neuronal regeneration.
• Chondroitinase ABC (ChABC) degrades CSPGs to promote regeneration.
• ChABC is thermally unstable at physiological temperatures, limiting its therapeutic use.

Purpose of the Study:

• To develop novel random copolymers for stabilizing ChABC at 37°C.
• To identify optimal copolymer designs using active machine learning.
• To enhance the therapeutic potential of ChABC for tissue regeneration.

Main Methods:

• Active machine learning paradigm for copolymer design.
• Iterative synthesis and testing of random copolymers.
• Automated PET-RAFT polymerization and retained enzyme activity (REA) assays.
• Gaussian process regression and Bayesian optimization for design selection.

Main Results:

• Significant improvements in ChABC thermostability and REA after 24h at 37°C.
• Identification of high-performing copolymers through iterative design.
• One copolymer maintained nearly 30% ChABC activity after one week at 37°C.
• Outperformed common ChABC stabilization methods.

Conclusions:

• Tailor-made random copolymers effectively stabilize ChABC at physiological temperatures.
• Active machine learning accelerates the discovery of effective enzyme stabilizers.
• This approach offers a promising strategy for sustained enzyme activity in regenerative medicine applications.