Terpenoids are valuable natural products that are widely used in medicine, agriculture, energy, and food. Traditional production by plant extraction or chemical synthesis is inefficient, costly, and polluting. Microbial fermentation via synthetic biology offers a greener alternative but faces challenges such as metabolic flux competition, cofactor imbalance, and product toxicity that limit yields. Genome-scale metabolic models (GSMMs), as essential tools in systems biology, can provide computational guidance for the rational design. This paper systematically reviews the progress of GSMMs in four typical terpenoid-producing microorganisms: the model microorganisms Escherichia coli and Saccharomyces cerevisiae, as well as the nonmodel microorganisms cyanobacteria and Yarrowia lipolytica. It focuses on their applications in fermentation process optimization and metabolic engineering strategies. Furthermore, future development directions, such as multiconstraint models and the integration of machine learning with synthetic biology, are discussed, aiming to provide a theoretical reference for the intelligent design and efficient construction of terpenoid cell factories.
Keywords: genome-scale metabolic models; machine learning; multiomics constraints; synthetic biology; terpenoids.