Increasing lipid yield in Yarrowia lipolytica through phosphoketolase and phosphotransacetylase expression in a phosphofructokinase deletion strain

Abstract

Background: Lipids are important precursors in the biofuel and oleochemical industries. Yarrowia lipolytica is among the most extensively studied oleaginous microorganisms and has been a focus of metabolic engineering to improve lipid production. Yield improvement, through rewiring of the central carbon metabolism of Y. lipolytica from glucose to the lipid precursor acetyl-CoA, is a key strategy for achieving commercial success in this organism.

Results: Building on YB-392, a Y. lipolytica isolate known for stable non-hyphal growth and low citrate production with demonstrated potential for high lipid accumulation, we assembled a heterologous pathway that redirects carbon flux from glucose through the pentose phosphate pathway (PPP) to acetyl-CoA. We used phosphofructokinase (Pfk) deletion to block glycolysis and expressed two non-native enzymes, phosphoketolase (Xpk) and phosphotransacetylase (Pta), to convert PPP-produced xylulose-5-P to acetyl-CoA. Introduction of the pathway in a pfk deletion strain that is unable to grow and accumulate lipid from glucose in defined media ensured maximal redirection of carbon flux through Xpk/Pta. Expression of Xpk and Pta restored growth and lipid production from glucose. In 1-L bioreactors, the engineered strains recorded improved lipid yield and cell-specific productivity by up to 19 and 78%, respectively.

Conclusions: Yields and cell-specific productivities are important bioprocess parameters for large-scale lipid fermentations. Improving these parameters by engineering the Xpk/Pta pathway is an important step towards developing Y. lipolytica as an industrially preferred microbial biocatalyst for lipid production.

Keywords: Cell-specific lipid productivity; Central carbon metabolism; Lipid yield; Phosphoketolase; Phosphotransacetylase; Yarrowia lipolytica.

Fig. 1

Lipid synthesis pathway for triolein production in Y. lipolytica. The native glucose to lipid synthesis pathway is shown in black, the pentose phosphate pathway is shown in green, the heterologous Xpk/Pta pathway is shown in blue and PFK1 deletion in red. List of enzymes: ACL ATP:citrate lyase, CIT citrate synthase, ELO elongase, FAS fatty acid synthase, FBP fructose 1,6-bisphosphatase, OLE – Δ9 fatty acid desaturase, PDH pyruvate dehydrogenase, PFK phosphofructokinase, PGI phosphoglucoisomerase, PTA phosphotransacetylase, XPK phosphoketolase. DHAP dihydroxyacetone phosphate, Ga3P glyceraldehyde 3-phosphate, Ma-CoA malonyl-CoA

Fig. 2

Heterologous Pta activity in Y. lipolytica. PTA genes from seven different source organisms codon-optimized to S. cerevisiae (GeneArt) were expressed in YB-392 under the control of the Y. lipolytica EXP1 promoter using linear integrating expression cassettes. Cell-free extracts (CFE) from four transformants per test gene were analyzed for Pta activity using a DTNB assay. Data are presented as fold change of the measured specific activity over the averaged specific activity of the parent strain YB-392 (dashed line), which is included as control

Fig. 3

Heterologous Xpk activity in Y. lipolytica. XPK genes from four source organisms codon-optimized to Y. lipolytica (ATGme [59]) were expressed in YB-392 under the control of the Y. lipolytica TEF1 promoter on replicating plasmids. Cell-free extracts from 4 transformants per test gene were analyzed by ferric hydroxamate assay to measure Xpk activity on ribose 5-phosphate (R5P, black bars) and fructose 6-phosphate (F6P, grey bars). Absorbance was measured at 540 nm and normalized to total protein in the crude cell-free extract after a reaction time of 30 min. Data are presented as fold change of the normalized absorbance over the averaged normalized absorbance of the parent strain YB-392 (dashed line), which is included as control

Fig. 4

Y. lipolytica growth on glucose. Strains YB-392 (wild-type) and NS1047 (Δpfk1) were streaked on A) Yeast Nitrogen Base (YNB) media containing 2% glucose for 4 days, or B) YPD for 2 days as labeled in C

Fig. 5

Engineering the Xpk/Pta pathway in a Δpfk1 Y. lipolytica strain. a Pta and Xpk activity. Pta activity in all strains shown was measured using the DTNB assay (black bars). Xpk activity in the control strain YB-392 and all the strains that obtained a copy of CaXPK through transformation (NS1281, NS1292, NS1322, NS1457 and NS1475) was measured using the ferric hydroxamate assay with ribose 5-phosphate as the substrate (grey bars). NS1047, NS1341, NS1352 and NS1420 were excluded from this assay. b Growth and lipid accumulation assays. OD₆₀₀ was measured after 2 days of growth in lipid production media (black bars). Lipid accumulation was measured as fluorescence/OD after overnight growth in glycerol followed by seven days of culture in modified Verduyn media (grey bars). Modified Verduyn media contained glucose as the only carbon source and no nitrogen to induce lipid production

Fig. 6

Characterization of Xpk/Pta/Δpfk1 strain NS1475 and YB-392 in 1-L glucose batch fermentation. a Time-course profiles of glucose consumption, lipid free dry cell weight (LFDCW) and lipid titers of YB-392 and NS1475 over a 5-day fermentation. b Lipid content. c Total lipid yield. d Cell-specific lipid productivity (day 2 – day 5). Data are mean ± standard deviation for two replicate runs

Fig. 7

Rebuilding Xpk/Pta/Δpfk1 in Y. lipolytica with TsPTA(v2) and CaXPK(v2). a Strain construction flowchart. b Time-course profiles of glucose consumption, lipid free dry cell weight (LFDCW), lipid titers of Xpk/Pta/Δpfk1 strains NS1656-57 and YB-392 in 1-L glucose batch fermentations over a 5-day period. C Total lipid yield. d Cell-specific lipid productivity (day 2–day 5). e Lipid content. Data are mean ± standard deviation for two replicate runs