Direct production of fatty alcohols from glucose using engineered strains of Yarrowia lipolytica

Abstract

Fatty alcohols are important industrial oleochemicals with broad applications and a growing market. Here, we sought to engineer Yarrowia lipolytica to serve as a renewable source of fatty alcohols (specifically hexadecanol, heptadecanol, octadecanol, and oleyl alcohol) directly from glucose. Through screening four fatty acyl-CoA reductase (FAR) enzyme variants across two engineered background strains, we identified that MhFAR enabled the highest production. Further strain engineering, fed-batch flask cultivation, and extractive fermentation improved the fatty alcohol titer to 1.5 g/L. Scale-up of this strain in a 2L bioreactor led to 5.8 g/L total fatty alcohols at an average yield of 36 mg/g glucose with a maximum productivity of 39 mg/L hr. Finally, we utilized this fatty alcohol reductase to generate a customized fatty alcohol, linolenyl alcohol, from α-linolenic acid. Overall, this work demonstrates Y. lipolytica is a robust chassis for diverse fatty alcohol production and highlights the capacity to obtain high titers and yields from a purely minimal media formulation directly from glucose without the need for complex additives.

Keywords: Extractive fermentation; Fatty acyl-CoA reductase; Fatty alcohols; Product yield; Yarrowia lipolytica.

Fig. 1

Clonal variation of FAR enzymes expressed in Y. lipolytica strains Po1fpmD and L36DGA1. Each data point represents an individual clone obtained from the random integration transformation of the FAR into (a) Po1fpmD or (b) L36DGA1. This level of variation is expected due to random integrations of the FAR expression cassette used in this approach.

Fig. 2

Total fatty alcohol titer evaluating the impact of second integration of MhFAR enzyme and dodecane overlay. Po1fpmD strains showed no significant differences in total fatty alcohol production across conditions. L36DGA1 based strains shows an improved fatty alcohol profile based on genetics and conditions. Aqueous phase represents the combined intracellular and extracellular fatty alcohols in the water phase. Error bars represent the standard deviation of biological triplicate experiments. Significance is denoted as: *p ​< ​0.05, ***p ​< ​0.005 (tested via two-way ANOVA).

Fig. 3

Fatty acid: Fatty alcohol ratio matrix for strain L36DGA1 2x MhFAR cultured with and without extractive fermentation. The fatty acid to fatty alcohol ratio is maintained between (a) normal fermentation conditions and (b) extractive fermentation with a dodecane overlay. Experimental data represents the average values of a biological triplicate experiment.

Fig. 4

Fatty alcohol yield with multiple integrations and extractive fermentation using YSC media. Yields were calculated directly from glucose due to the use of minimal media. Error bars represent the standard deviation of biological triplicate experiments. Yields do not significantly vary across groups as tested by a two-way ANOVA test.

Fig. 5

Fatty alcohol production and localization using dodecane extraction and/or a glucose pulse. A glucose pulse paired with extractive fermentation leads to highest fatty alcohol production in flasks. Data shown represents final time point tested for this flask fermentation (117 ​h). Aqueous phase represents the combined intracellular and extracellular fatty alcohols in the water phase. Error bars represent the standard deviation of biological triplicate experiments. Significance is denoted as: *p ​< ​0.05, **p ​< ​0.01 (tested via two-way ANOVA).

Fig. 6

Fed-batch bioreactor of strain L36DGA1 2x MhFAR with two glucose pulses, pH 5.0. Controlled bioreactor fermentation data enables 5.8 g/L of fatty alcohol production directly from minimal glucose media. (a) Growth parameters show biomass accumulation paired with a corresponding decrease in measured glucose concentration. Arrows represent the addition of 80 g glucose at 72 and 120 h. (b) Fatty alcohol titer and specific productivity demonstrate high level production over the course of the fermentation with a peak productivity at 5 days. Error bars represent the standard deviation of biological duplicate experiment.