Combined metabolic engineering and lipid droplets degradation to increase vitamin A production in Saccharomyces cerevisiae.

Abstract

Background: In microbial cell factories, substrate accessibility to enzyme is a key factor affecting the biosynthesis of natural products. As a robust chassis cells for biofuels and bioproducts, Saccharomyces cerevisiae also encounters the challenge since different enzymes and precursors are typically compartmentalized in different organelles. Such spatial separation could largely limit the efficiency of enzymatic reactions. In this study, the production of the hydrophobic product (vitamin A) was highly improved by metabolic engineering combined with degrading lipid droplets (the primary organelle storing β-carotene) to achieve efficient contact between β-carotene and 15, 15'-β-carotene monooxygenases in Saccharomyces cerevisiae.

Results: To efficiently produce vitamin A in Saccharomyces cerevisiae, ten 15, 15'-β-carotene monooxygenases (BCMOs) were firstly evaluated. The strain carrying marine bacterium 66A03 (Mb. BCMO) achieved the highest vitamin A titer. Co-adding 10% dodecane and 1% dibutylhydroxytoluene increased vitamin A titer to 19.03 mg/L in two-phase fermentation. Since most β-carotene is stored in LDs while BCMO is located in the cytosol, we developed a strategy to release β-carotene from LDs to better contact with BCMO. By overexpressing TGL3 and TGL4 using an ion-responsive promoter after high accumulation of β-carotene in LDs, LDs were sequentially degraded, which dramatically improved vitamin A production. Finally, by overexpressing tHMG1, ERG20, and CrtI and introducing Vitreoscilla hemoglobin, vitamin A titer reached 219.27 mg/L, which was a 10.52-folds increase over the original strain in shake flasks, and finally reached 1100.83 mg/L in fed-batch fermentation. The effectiveness of LDs degradation on promoting the formation of β-carotene cleaved product has also been verified in β-ionone synthesis with 44.07% increased yield.

Conclusions: Overall, our results highlighted the significance of sequential degrading LDs on vitamin A overproduction in recombinant yeast, and verified that combining metabolic and LDs engineering is an efficient strategy to improve vitamin A production. This integrated strategy can be applied to the overproduction of other hydrophobic compounds with similar characteristics.