Type III polyketide synthases: from catalytic plasticity to predictive engineering

Covering: 1998 to 2026Type III polyketide synthases (PKSs) are homodimeric enzymes that catalyze starter-substrate loading, iterative condensation, and cyclization to produce diverse natural products, including chalcones, stilbenes, cannabinoid precursors, and curcuminoids. Despite their conserved catalytic architecture, type III PKSs exhibit remarkable plasticity in starter-substrate utilization, chain extension, and cyclization chemistry. This versatility makes them attractive but challenging targets for enzyme engineering. This review examines three major strategies for improving type III PKSs: screening and genome mining, structure-guided engineering, and directed evolution coupled with high-throughput screening or selection. We discuss how these approaches can be used to improve catalytic activity, substrate specificity, and product selectivity, both in vitro and in heterologous production hosts. We also examine recent applications of artificial intelligence (AI) and machine learning (ML) to enzyme discovery, functional prediction, and variant prioritization, as well as the limitations imposed by small and heterogeneous datasets. Finally, we provide a practical framework for selecting engineering strategies and discuss future directions toward more predictable engineering of type III PKSs.