Enzyme-constrained genome-scale model of Yarrowia lipolytica predicts growth-phase specific metabolic engineering targets.

The oleaginous yeast Yarrowia lipolytica has been gaining increasing importance as an industrial biotech platform, supported by several available metabolic engineering tools. Genome-scale models (GEMs) are relevant to the iterative improvement of this yeast, and their predictive ability is enhanced by enzymatic activity constraints (ecGEMs). Although the newest tools for ecGEM reconstruction use deep learning to expand the coverage of these constraints, this approach has not yet been applied to Y. lipolytica models. This paper describes the reconstruction of an ecGEM of Y. lipolytica (eciYali5-GEM) and its application in predicting metabolic engineering targets for enhanced lipid and carotenoid production, respectively, in this yeast. To achieve this, we constrained a manually curated Y. lipolytica model with physiological flux and mass-spectrometry proteomics data collected from distinct growth phases of two engineered strains (producing lipids and carotenoids, respectively) and their parental strain. We found that the enzymatic constraints enable the prediction of growth-phase-specific metabolic engineering targets, a feature not displayed by regular GEMs. Combining these targets with other ecGEM-based insights, we propose two strategies for further metabolic engineering, including the use of inducible promoters for precise, growth-phase-specific expression of targets such as phytoene dehydrogenase for carotenoid production. These targets included genes previously validated elsewhere, as well as novel genes awaiting experimental validation. This model, which is publicly available, can be similarly adapted and used by different metabolic engineering efforts, making it a versatile tool for the development of Y. lipolytica as a microbial cell factory. KEY POINTS: • eciYali5-GEM reconstruction used in silico and in vitro proteomics data. • Enzyme constraints enriched Y. lipolytica ecGEM predictions. • eciYali5-GEM improves metabolic engineering rational design.