Abstract
Oleaginous yeasts cultivation in low-cost substrates is an alternative for more sustainable production of lipids and oleochemicals. Lipomyces starkeyi accumulates high amounts of lipids from different carbon sources, such as glycerol, and glucose and xylose (lignocellulosic sugars). Systems metabolic engineering approaches can further enhance its capabilities for lipid production, but no genome-scale metabolic networks have been reconstructed and curated for L. starkeyi. Herein, we propose lista-GEM, a novel genome-scale metabolic model of L. starkeyi. We reconstructed the model using two high-quality models of oleaginous yeasts as templates and further curated the model to reflect the metabolism of L. starkeyi. We simulated phenotypes and predicted flux distributions in good accordance with experimental data. We also predicted targets to improve lipid production in glucose, xylose, and glycerol. The phase plane analysis indicated that the carbon availability affected lipid production more than oxygen availability. We found that the maximum lipid production in glucose and xylose required more oxygen than glycerol. Enzymes related to lipid synthesis in the endoplasmic reticulum were the main targets to improve lipid production: stearoyl-CoA desaturase, fatty-acyl-CoA synthase, diacylglycerol acyltransferase, and glycerol-3-phosphate acyltransferase. The glycolytic genes encoding pyruvate kinase, enolase, phosphoglycerate mutase, glyceraldehyde-3-phosphate dehydrogenase, and phosphoglycerate kinase were predicted as targets for overexpression. Pyruvate decarboxylase, acetaldehyde dehydrogenase, acetyl-CoA synthetase, adenylate kinase, inorganic diphosphatase, and triose-phosphate isomerase were predicted only when glycerol was the carbon source. Therefore, we demonstrated that lista-GEM provides multiple metabolic engineering targets to improve lipid production by L. starkeyi using carbon sources from agricultural and industrial wastes.