Redesigning myoglobin via functional site scaffolding for enhanced catalytic functions

Area of Science:

• Protein engineering and computational structural biology.
• Design of novel protein scaffolds with tailored functions.

Background:

• Protein redesign is often hindered by a lack of stable and adaptable protein scaffolds.
• Computational methods can explore new protein sequences, but replicating metalloprotein structural details is difficult.

Purpose of the Study:

• To develop a deep learning-based strategy for designing robust and functional protein scaffolds.
• To create a miniaturized myoglobin variant (bitMb) that maintains native heme-binding properties and O2 behavior.

Main Methods:

• Utilized a diffusion model for backbone generation and ProteinMPNN for sequence optimization.
• Employed AlphaFold and OmegaFold for structural consistency checks, focusing on the heme-binding site.
• Generated and computationally screened over 100,000 sequences to identify the bitMb scaffold.

Main Results:

• Successfully designed bitMb, a miniaturized myoglobin with preserved heme-binding and O2 behavior.
• Engineered bitMb variants showed increased peroxidase and carbene transferase activities.
• bitMb demonstrated enhanced thermal stability (5.5°C higher Tm) and remarkable stability in organic solvents (up to 96.7% methanol).

Conclusions:

• The functional site scaffolding strategy effectively generates robust and versatile protein scaffolds.
• Designed protein scaffolds can exhibit diverse catalytic functions and improved stability.
• This approach advances the field of de novo protein design for various applications.