Discovery, design, and engineering of enzymes based on molecular retrobiosynthesis

Abstract

Biosynthesis-a process utilizing biological systems to synthesize chemical compounds-has emerged as a revolutionary solution to 21st-century challenges due to its environmental sustainability, scalability, and high stereoselectivity and regioselectivity. Recent advancements in artificial intelligence (AI) are accelerating biosynthesis by enabling intelligent design, construction, and optimization of enzymatic reactions and biological systems. We first introduce the molecular retrosynthesis route planning in biochemical pathway design, including single-step retrosynthesis algorithms and AI-based chemical retrosynthesis route design tools. We highlight the advantages and challenges of large language models in addressing the sparsity of chemical data. Furthermore, we review enzyme discovery methods based on sequence and structure alignment techniques. Breakthroughs in AI-based structural prediction methods are expected to significantly improve the accuracy of enzyme discovery. We also summarize methods for de novo enzyme generation for nonnatural or orphan reactions, focusing on AI-based enzyme functional annotation and enzyme discovery techniques based on reaction or small molecule similarity. Turning to enzyme engineering, we discuss strategies to improve enzyme thermostability, solubility, and activity, as well as the applications of AI in these fields. The shift from traditional experiment-driven models to data-driven and computationally driven intelligent models is already underway. Finally, we present potential challenges and provide a perspective on future research directions. We envision expanded applications of biocatalysis in drug development, green chemistry, and complex molecule synthesis.

Keywords: artificial intelligence; enzyme design; enzyme discovery; enzyme engineering; molecular retrosynthesis planning.

Figure 1

Enzyme design and engineering based on retrosynthetic route planning. (A) Biochemical pathway discovery. The AI‐based retrobiosynthetic route prediction algorithm significantly enhances the discovery efficiency of new reaction pathways. (B) AI‐ and computation‐based multi‐route strategies can increase enzyme discovery efficiency. (C) Fully rational bioelement design methods based on computation aid in the efficient enhancement of element performance.

Figure 2

Synthetic pathway of cis‐Octahydropyrrolo[3,4‐b]pyridine via sequential single‐step retrosynthesis synthetic routes of the target molecule. In each retrosynthetic step, the product is listed first, followed by the substrate. These steps are connected sequentially until simpler, known, or commercially available precursor molecules are identified.

Figure 3

Enzyme discovery via multiple strategies for enhanced success rates. Enzyme discovery is performed using multiple approaches, including sequence search, structure search, EC (Enzyme Commission) number prediction and retrieval, and reverse virtual screening. The first two methods use the initial enzyme as a starting point, while the latter two begin with the reaction or molecule. These strategies can be used individually, in combination, or iteratively to increase the likelihood of successful enzyme discovery.

Figure 4

Enzyme discovery based on protein–ligand interactions using traditional and AI‐driven approaches. (A) Traditional bioinformatics methods using PointSite and zPoseScore for applications in drug design and enzyme discovery. (B) Direct structure prediction methods, such as AlphaFold3, leveraging AI to predict protein structures from amino acid sequences.

Figure 5

De novo enzyme generation and optimization based on reaction pathways. Starting from the chemical reaction, the sequential steps involved in enzyme generation and optimization: predicting the EC number using EC‐blast; generating the enzyme with tools like ZymCtrl; redesigning the catalytic pocket for enhanced small molecule binding using PocketGen and similar tools; predicting protein conformation with Protenix and analogous methods; and assessing enzyme activity via DLKcat and other predictive approaches. If necessary, the catalytic pocket is regenerated and re‐evaluated iteratively until the desired performance is achieved.

Figure 6

Typical enzyme engineering pathways. Protein sequence analyses, such as multiple sequence alignment (or multiple structure alignment, MSA), coevolution, position‐specific scoring matrix (PSSM), and conservation calculations, can be used to help select mutation sites and targets. Molecular Dynamics (MD) can be used to provide mutation guidance, such as catalytic mechanism, near‐attack state, and motion correlation between residues. AI‐based models can be used to rapidly and accurately assess properties of mutations, including thermal stability, solubility, enzyme activity, binding affinity, and interaction evaluations.