Physics-based Modeling in the New Era of Enzyme Engineering

Zhongyue Yang,Christopher Jurich,Qianzhen Shao,Xinchun Ran

DOI: https://doi.org/10.26434/chemrxiv-2024-0z1gn

2024-09-10

Abstract:Enzyme engineering techniques optimize enzymes to synthesize value-added chemicals, degrade environmental pollutants, and improve therapeutics. The field is entering a new era characterized by the increasing integration of computational strategies. While bioinformatics and artificial intelligence (AI) have been extensively applied to accelerate the screening of function-enhancing mutants, physics-based modeling methods, such as molecular mechanics and quantum mechanics, serve as essential complements in engineering objectives where setting up high-throughput screening is difficult or where a deep understanding of unknown physical principles is crucial. In this perspective, we discuss the enormous, untapped potential of physics-based modeling in guiding the next step of computational enzyme engineering. We first explore the paradigm of physics-based design principles wherein insights from natural, efficient enzymes are applied to recommend beneficial mutations in silico. We examine current development of high-throughput molecular modeling workflows that aid enzyme engineering campaigns through large-scale virtual applications of design principles. We then emphasize how physics-based modeling empowers AI techniques through enriching data expressiveness and interpretability. Finally, we proposed unmet challenges for the next step advancement of computational tools for enzyme engineering.

Chemistry

What problem does this paper attempt to address?

The paper aims to address several key issues in the field of enzyme engineering:

1. **Limitations of High-Throughput Screening**: Although Directed Evolution (DE) technology has achieved significant success in enzyme engineering, it relies on high-throughput experimental screening, which has limitations for certain enzyme systems and engineering targets. For example, when side reactions cannot be ignored or when pure protein samples are required for detection, constructing screening protocols becomes difficult. Additionally, reliably reducing the size of enzymes through high-throughput deletion or truncation is challenging.

2. **Enzyme Engineering Under Extreme Conditions**: Industrial biosynthesis often requires enzymes to perform optimally under non-native, extreme conditions, which existing high-throughput screening methods struggle to meet. For instance, highly efficient extreme enzymes can significantly reduce costs and environmental impact, but engineering these systems remains difficult due to varying conditions such as temperature and pH.

3. **Limitations of Bacterial Expression Systems**: High-throughput screening typically uses bacterial expression systems, making it difficult to handle plant-based and mammalian enzymes, despite their many biosynthetic functions or lower immunogenicity.

4. **Lack of Quantitative Predictions**: Treating catalysis as a black-box process can lead to evolutionary dead ends, preventing further screening from improving enzyme efficiency. Therefore, complementary techniques are needed to overcome these limitations.

To address these issues, the paper emphasizes the importance of physics-based molecular modeling techniques. Quantum Mechanics (QM) and Molecular Mechanics (MM) can be applied to any system with atomic-resolution 3D structures, regardless of the enzyme's source or preferred environment. These methods not only help understand the origins of enzyme mechanisms and efficiency but also guide the formulation of design principles and enhance the data representation and interpretability of Machine Learning (ML) models. The ultimate goal is to achieve comprehensive optimization of any enzyme system using physics-based computational methods, including topology engineering, charge field analysis, and kinetic studies.