Multifunctional Biocatalyst Structures Using Reactive Center Grafting
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current methods struggle to design enzymes with multiple active sites or reaction centers for industrial applications, requiring sub-Å level accuracy in AI-based structure prediction models, which is challenging due to the lack of experimental data on reactive center grafting across proteins.
Innovation Solution
The RC-Hydrolase database and CADSEEK 3D shape search engine identify structurally similar reactive centers across different enzyme classes, enabling the design of multifunctional enzymes by minimal mutations using ProteinMPNN, allowing for fusion enzymes with multiple catalytic activities.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If AI-based structure prediction models are used to design enzymes with multiple active sites, then enzyme versatility is improved, but manufacturing precision deteriorates due to lack of experimental data on reactive center grafting
Solution Approach 1:
The patent copies the reactive center structure from one enzyme to another by identifying structurally similar reactive centers using the CADSEEK engine. This allows the grafting of catalytic functionality without requiring de novo design, thereby improving precision while maintaining versatility.
Solution Approach 2:
The patent changes the parameter of structural similarity threshold in CADSEEK to identify appropriate reactive centers for grafting. By adjusting this parameter, the system optimizes the balance between finding suitable candidates and ensuring precise grafting outcomes.
2Productivity
If multiple enzymes are used for multi-step reactions, then catalytic functionality is improved, but device complexity increases due to fusion design requirements
Solution Approach 1:
The patent merges multiple enzyme functionalities into a single fusion enzyme by combining their reactive centers. This approach maintains high catalytic functionality while reducing the number of separate components needed in the system.
Solution Approach 2:
The patent creates universal fusion enzymes that can perform multiple catalytic functions simultaneously. By designing enzymes with multiple reactive centers, a single enzyme molecule can catalyze different reactions in a multi-step process.
3Manufacturing precision
If sub-Å level accuracy is required in structure prediction, then reactive center design precision is improved, but loss of information increases due to paucity of experimental data
Solution Approach 1:
The patent introduces CADSEEK as an intermediary tool that bridges the gap between limited experimental data and the need for high-precision structure prediction. This shape search engine uses structural similarity metrics to compensate for the lack of training data in AI models.
Solution Approach 2:
The patent focuses computational resources on achieving high precision specifically at the reactive center level rather than requiring uniform high accuracy across the entire enzyme structure. This localized approach reduces information loss while maintaining design precision where it matters most.
Applied Scientific Principles
This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.
Function Achieved in This Case
This approach enhances enzyme catalytic efficiency and versatility, reducing production costs and simplifying purification processes by creating enzymes that can perform multiple functions simultaneously, such as protease and nuclease activities, suitable for industrial applications like detergent manufacturing and biofuel production.
Implementation Method 1
CADSEEK 3D shape search engine identify structurally similar reactive centers across different enzyme classes
Implementation Method 2
enabling the design of multifunctional enzymes by minimal mutations using ProteinMPNN
Implementation Method 3
creating enzymes that can perform multiple functions simultaneously, such as protease and nuclease activities
Data Source
AI summary
A systematic pipeline is used to extract catalytically active pockets of the most diverse enzyme class—hydrolases, from the PDB database. A process extracts the 38029 hydrolase reactive centers (RC) and collates them into a publicly accessible active site collection (actiome; RC-Hydrolase). The process includes 128M pairwise shape comparisons across RC-Hydrolase using CADSEEK 3D Shape Search Engine to end up with 155,329 instances presented in a available, visually interactive dataset. Allowing comparisons of enzyme reactive centers across functional spaces (EC classification numbers) enables identification of enzyme backbones which can be minimally mutated to accommodate more than one type of catalytic activity to aid rational design of multifunctional enzymes. Such versatile enzyme backbones is leveraged by latest diffusion-based protein design models to design a library of structurally stable multifunctional enzyme pockets. Design of a bifunctional protease-nuclease shown as an example opens up a novel computational recipe for enzyme engineering.