Engineering of Yarrowia lipolytica for producing pyruvate from glycerol

Abstract

The present study aims to increase pyruvate production by engineering Yarrowia lipolytica through modifying the glycerol metabolic pathway. Results: Wild-type Yarrowia lipolytica (Po1d) was engineered to produce six different strains, namely ZS099 (by over-expressing PYK1), ZS100 (by deleting DGA2), ZS101 (by over-expressing DAK1, DAK2, and GCY1), ZS102 (by over-expressing GUT1 and GUT2), ZS103 (by over-expressing GUT1) and ZSGP (by over-expressing POS5 and deleting GPD2). Production of pyruvate from engineered and control strains was determined using high-performance liquid chromatography (HPLC). Subsequently, the fermentation conditions for producing pyruvate were optimized, including the amount of initial inoculation, the addition of calcium carbonate (CaCO₃), thiamine and glycerol. Finally, for scaled-up purposes, a 20-L fermentor was used. It was observed that pyruvate production increased by 136% (8.55 g/L) in ZSGP strain compared to control (3.62 g/L). Furthermore, pyruvate production by ZSGP reached up to 110.4 g/L in 96 h in the scaled-up process. We conclude that ZSGP strain of Y. lipolytica can be effectively used for pyruvate production at the industrial level.

Supplementary Information

The online version contains supplementary material available at 10.1007/s13205-022-03158-7.

Introduction

Pyruvate (PA) is the final product of the glycolysis pathway and is an important intermediate molecule in protein, fat and carbohydrates metabolism (McCommis and Finck 2015b). PA is a three-carbon molecule that functions as a critical step in cell metabolism in aerobic and anaerobic conditions (Vander Heiden et al. 2009). The two PA molecules produced from glycolysis can be converted into various products like carbohydrates (through gluconeogenesis), alanine, or ethanol (through fermentation), fatty acids and energy (through Krebs cycle) (McCommis and Finck 2015b). Inside the cells, various metabolic pathways are connected by PA (McCommis and Finck 2015a). PA is a transparent, water-miscible liquid having an acetic acid-like aroma under normal conditions. PA and its derivatives are widely employed in the pharmaceutical, cosmetic, food, and chemical sectors to synthesis other important goods (Li et al. 2001; Soma et al. 2014). PA is also found in sports nutrition supplements, which help athletes to improve their physical condition and maintain body mass control (Jager et al. 2008). PA calcium salts have been demonstrated to speed up fatty acid metabolism while also helping to reduce serum cholesterol levels (Han et al. 2016). PA also possesses antioxidant effects and could be used as a nutraceutical to treat diabetes II (Ju et al. 2012; Plotnikov et al. 2019).

At the industrial level, pyruvate is widely used as a starting material for the production of L-tyrosine, L-tryptophan, N-acetylneuraminic acid (sialic acid), 3-Dihydroxyphenyl (DOPA), levodopa drug class, etc. (Cybulski et al. 2019; Maleki and Eiteman 2017). Pyruvate production mainly uses a chemical synthesis method, which increases its cost and greatly limits its application. Current research shows that the microorganisms that ferment to produce pyruvate are mainly Escherichia coli and yeast. In Escherichia coli, lactate dehydrogenase (ldhA), which is responsible for converting pyruvate into lactic acid, is inactivated, and the highest pyruvate production is 110.0 g/L with the mass yield of 0.87 g/g (Zelić et al. 2004). In yeast, pyruvate decarboxylase-negative Saccharomyces cerevisiae, through directed evolution, has obtained a high-yield pyruvate strain that is C2-independent and glucose-tolerant, with a maximum yield of 135 g/L. The pyruvate/glucose conversion rate was 0.54 g/g (Van Maris et al. 2004). Using 70 g/L NaCl as the selection criterion, the NaCl-resistant mutant Torulopsis glabrata RS23 was screened out through continuous culture with pH control. The yield of pyruvate reached 94.3 g/l, and the mass yield was 0.635 g/g (Liu et al. 2007). Most of the above-mentioned strains with high pyruvate production use glucose as a substrate, and achieve high production of pyruvate by reducing the consumption of pyruvate. However, the Yarrowia lipolytica yeast, which uses glycerol as the sole carbon source, has been more researched by screening wild bacteria with high pyruvate production(Krzysztof et al. 2018b; Zeng et al. 2017a), or directed evolution to high-yield pyruvate strains that are resistant to high osmotic pressure (Yuan et al. 2020). This study analyzed a number of key enzymes in the glycerol consumption pathway (glycerol 3-phosphate pathway and dihydroxyacetone pathway), and finally achieved high pyruvate production. Such previous information provided a lot of data for studying the decomposition of glycerol in yeast.

Materials and methods

Culture conditions

For plasmid propagation, Escherichia coli cells (DH5α strain) were grown at 37 °C in Luria–Bertani (LB) medium (10 g/L tryptone OXOID™, 5 g/L yeast extract OXOID™, 10 g/L NaCl) supplemented with antibiotic (ampicillin 100 mg/mL) at 220–230 rpm. Y. lipolytica and its engineered strains were grown at 30 °C in YPD medium (10 g/L yeast extract OXOID™, 20 g/L tryptone OXOID™, 20 g/L glucose OXOID™) at 180–200 rpm. The other yeast media used in the present study include Yeast nitrogen glucose (YND) (20 g/L glucose OXOID™, 50 g/L(NH₄)₂SO₄, 1.7 g/L YNB OXOID™) and Yeast nitrogen glycerol (YNG) (2 g/L glycerol, 0.2 g/L yeast extract, 0.4 g/L tryptone, 0.24 g/L KH₂PO₄, 1.7 g/L K₂HPO₄·3H₂O).

For evaluating the pyruvate production and growth rate in flasks, Y. lipolytica strains were grown for 108 h in YNG medium. For fermentation purposes, pH 5.5 was maintained followed by addition of CaCO₃ (40 g/L). For scale-up purposes, Y. lipolytica strains were firstly grown in YPD medium, followed by shifting to a 20-L fermentor containing 1.4 g/L MgSO₄·7H₂O, 10 g/L (NH₄)₂SO₄, 0.8 g/L CaCl₂, 2 g/L KH₂PO₄, 60 g/L glycerol, 1 μg/L thiamine and 0.5 g/L NaCl. Addition of glycerol was performed periodically, as it was the only source of carbon. Conditions maintained during fermentation were 40% dissolved O₂, pH 4.0 and 30 °C.

Plasmid construction

Target genes of Y. lipolytica were amplified using specific primers (as shown in Table S1) and were cloned in respective plasmids (as shown in Table S2). Recombinant plasmid E11 (GUT1-pEXP1-GUT2 cassette in JMP121 plasmid) and E12 (GUT1 cassette in JMP121 plasmid) were constructed using ClonExpress Ultra One Step Cloning Kit (Vazyme, China).

For construction of E11, the fragments GUT1 and GUT2 were amplified by PCR using genomic DNA of H. polymorpha as a template and inserted in to the pEXP1 vector. The large fragment GUT1-pEXP1-GUT2 was further obtained by overlap PCR using GUT1-pEXP1-GUT2 as template. The linearization of plasmid vector JMP121 was performed by digested with the restriction endonuclease Quick-Cut BamH I. Finally, the obtained GUT1-pEXP1-GUT2 was inserted to JMP121 plasmid using ClonExpress Ultra One Step Cloning Kit, resulting the plasmid E11. Similarly for construction of E12, the fragment GUT1 was amplified by PCR using genomic DNA of H. polymorpha as a template. The linearization of plasmid vector JMP121 was performed by digested with the restriction endonuclease Quick-Cut BamH I. The obtained GUT1 was inserted to JMP121 plasmid using ClonExpress Ultra One Step Cloning Kit, resulting the plasmid E12. For specific steps, please visit www.vazyme.com.

Transformation and Screening of engineered strains

Transformation of recombinant plasmids in to Y. lipolytica was performed using Yeast Transformation Kit (Sigma Aldrich). The Po1d strain of Y. lipolytica was used as a wild type host to produce engineered forms (as shown in Table 1). Strategy for the generation of Y. lipolytica engineered strains is shown in Figs. 1 and 2. The engineered Y. lipolytica produced in the current study includes ZS099 (by over-expressing PYK1), ZS100 (by deleting DGA2), ZS101 (by over-expressing DAK1, DAK2, and GCY1), ZS102 (by over-expressing GUT1 and GUT2), ZS103 (by over-expressing GUT1) and ZSGP (by over-expressing POS5 and deleting GPD2). All strain genotypes were confirmed by colony PCR. All Y. lipolytica colonies were first screened on YND medium containing 2% agar. All positive transformants with integrative cassettes were cultivated in YNG medium for 5 days to monitor glycerol production by using high-performance liquid chromatography (HPLC). The LoxP-Cre recombination system was used for marker rescue and to ensure the multistep insertion of the target genes. A Ura3 selection marker was recycled on YPD containing 50 µg/ml hygromycin by transforming F7 plasmid which contains Cre recombinase.

Table 1.

Production of pyruvate by microbial fermentation

Strain	Substrate	Pyruvate (g/L)	Yield (g/g)	References
Escherichia coli CGSC6162 Deltappc	Glucose	35 g/l	0.78	Tomar et al. (2003)
Escherichia coli TC44	Glucose	65.96	0.75	Hollmann and Deckwer (2004)
Escherichia coli ALS929	Glucose/acetate	90	0.68	Zhu et al. (2008)
E. coli YYC202 ldhA: Kan	Glucose	110.0	0.87	Zelic et al. (2004)
Bacillus megaterium	Glucose	28.7	0.38	Hollmann and Deckwer (2004)
Trichosporon cutaneum PD70	Glucose	34.6	0.429	Wang et al. (2002)
Blastobotrys adeninivorans VKM Y-2677	Glucose	43.2	0.77	Kamzolova and Morgunov (2016)
Torulopsis glabrata	Glucose	94.3	0.635	Liu et al. (2007)
Saccharomyces cerevisiae	Glucose	135	0.54	van Maris et al. (2004)
Yarrowia lipolytica WSH-Z06	Glycerol	39.3	0.71	Zeng et al. (2017b)
Yarrowia lipolytica	Glycerol	48.1	0.48	Krzysztof et al. (2018a)
Yarrowia lipolytica 374/4	Glycerol	61.3	0.71	Morgunov et al. (2004)
Yarrowia lipolytica	Glycerol	97.2	0.795	Yuan et al. (2020)

Fig. 1.

Metabolic strategies for redirecting glycerol to pyruvate in Yarrowia lipolytic. In Y. lipolytica, glycerol produces dihydroxyacetone phosphate through the 3-phosphoglycerol and the dihydroxyacetone pathways. The former is composed of glycerol kinase (GUT1) and 3-phosphoglycerol dehydrogenase (GUT2), and the latter is composed of glycerol dehydrogenase (GCY1) and dihydroxyacetone kinase (DAK) composition. Dihydroxyacetone phosphate can be catalyzed by diacylglycerol acyltransferase (DAG2) to generate triacylglycerol, or enter the glycolytic pathway, and finally generate pyruvate under the catalysis of pyruvate kinase (PYK1)

Fig. 2.

Flowchart for the construction of yeast strains. Each arrow represents one manipulation step for the strain. Where two arrows are depicted, the top arrow represents URA3 rescue and the bottom arrow represents transformation with the DNA fragments listed to the right. +  + represents over-expression, while triangle ▲ represents deletion

Analytical methods

To determine cell density, serial dilution of fermented broth was performed, followed by measuring the absorbance (OD₆₀₀) using a spectrophotometer (Lengguang Technology Co, Shanghai, China). To assay PA concentration, the fermentation broth was centrifuged at 10,000 × g for 10 min. The PA concentration was determined by HPLC (Agilent 1100 series, Santa Clara, CA, USA) with a Stable Bond C18 column (Shiseido Co. Ltd, Tokyo, Japan). The flow rate of the mobile phase (0.1% phosphoric acid aqueous solution) was 1 mL/min, the column temperature was 28℃, the injection volume was 10 μL, and detection was in the ultraviolet range of 210 nm. Values are means of at least three independent experiments.

Statistical analysis

SPSS 19.0 software was used to carry out statistical analysis. Experimental values were given as mean and standard error of mean. (P < 0.05 as statistically significant).

Results and discussion

Enhancing production of pyruvate by reducing TAG synthesis and over-expressing PYK1

Earlier, we have observed that glycerol was the best carbon source for producing pyruvate from Y. lipolytica compared to glucose and sucrose (Yuan et al. 2020). In the current study, Y. lipolytica cells were cultured using glycerol with different periods. With time (from 0 to 84 h), the pyruvate concentration increased, finally reached the maximum at 84 h (3.62 g/L) as shown in Fig. 3. In culture, Y. lipolytica continuously grows with time and accumulates more and more pyruvate till 84 h. However, pyruvate production began to decrease after 84 h and can be mainly due to the consumption of pyruvate by other metabolic pathways (Fig. 3B). Our results clearly showed that the best time for producing maximum pyruvate yield from Y. lipolytica is 84 h. We, therefore, chose 84 h as a standard period for pyruvate production from Y. lipolytica and its strains.

Fig. 3.

Over-expressing PYK1 and reducing TAG synthesis. Figure A strategy for over-expression and deletion of PYK1 and DGA2 respectively. Figure B pyruvate accumulation curve of Po1d in flask. Data represent the mean ± standard deviation (n = 3). Figure C effect of over-expression of PYK1 and deletion of DGA2 on pyruvate production. Data represent the mean ± standard deviation (n = 3)

PYK1 is the rate-limiting enzyme in glycolysis (Burke et al. 1983; Pearce et al. 2001). To enhance production of pyruvate, PYK1 (YALI0F09185p) from Y. lipolytica was over-expressed under the strong constitutive promoter TEF. Notably, pyruvate production increased by 22.5% in strain ZS099 (Fig. 3A, 3C). These results indicated that PYK1 is an important enzyme for the accumulation of pyruvate in Y. lipolytica. DGA2 enzyme makes a significant contribution in triacylglycerols synthesis via an acyl-CoA-dependent mechanism (Beopoulos et al. 2012). Deleting DGA2 (YALI0D07986g) impairs triacylglycerols synthesis and lipid production, ultimately resulting in the accumulation of DHAP and Gly-3-p in Y. lipolytica (Friedlander et al. 2016). In this study, DGA2 was deleted in ZS099 to yield the ZS100 strain. As shown in Fig. 3C, the pyruvate yield was further increased by 28.8% (reaching 6.03 ± 0.06 g/L).

Enhancing production of pyruvate by reducing synthesis of glycerol

There are two glycerol decomposition pathways in Saccharomyces cerevisiae, namely dihydroxyacetone pathway (involving GCY1, DAK1, DAK2) (Aßkamp et al. 2019) and glycerol-3-phosphate pathway (involving GUT1 and GUT2) (Pavlik et al. 1993; Ronnow and Kiellandbrandt 1993). Over-expression of GCY1 and DAK1 in S. cerevisiae enhances glycerol decomposition (Zhang et al. 2013). Similarly, over-expression of GUT1 and GUT2 has been reported to enable effective glycerol assimilation toward the synthesis of desired products in Y. lypolytica. To test whether enhancing glycerol decomposition pathways can increase the accumulation of pyruvate in Y. lypolytica strains. The GCY1 (YALI0B07117p), DAK1 (YALI0F09273p) and DAK2 (YALI0E20691p) from Y. lypolytica were over-expressed in ZS100 to generate ZS101 (Fig. 4A). Interestingly, the pyruvate concentration did not increase in ZS101, compared with ZS100 (Fig. 4B). Similarly, the GUT1 (YALI0F00484g) and GUT2 (YALI0B02948g) were over-expressed in ZS100 to generate ZS102 (Fig. 4A). Considering that GUT2 participates in the glycerol-3-phosphate shuttle, which is required for S. cerevisiae to maintain redox balance between NAD and NADH under aerobic conditions, The ZS103 was constructed by only over-expressing GUT1. As shown in Fig. 4B, there was no increase of pyruvate accumulation in ZS102 and ZS103, compared to ZS100. These data indicated that enhancing decomposition pathways in ZS100 did not push glycerol assimilation further toward pyruvate synthesis. From the obtained results, we can assume that there may be a third glycerol decomposition pathway existing in Y. lypolytica. Alternatively, enhancing glycerol decomposition pathways might not be the bottleneck for the accumulation of pyruvate in Y. lypolytica.

Fig. 4.

Reducing glycerol synthesis enhances pyruvate production. Figure A shows strategy for constructing ZS101 (over-expression of DAK1, DAK2, GCY1), ZS102 (over-expression of GUT1, GUT2), ZS103 (over-expression of GUT1), ZSGP (over-expression of POS5 and deletion of GPD2). Figure B production of pyruvate for different strains. Data represent the mean ± standard deviation (n = 3)

As shown in Fig. 1, glycerol is synthesized by reducing DHAP with glycerol-3-phosphate dehydrogenase (Fig. 1). GPD2 deletion markedly decreases glycerol accumulation in S. cerevisiae (Nissen et al. 2000; Papapetridis et al. 2017). Furthermore, GPD2 deletion allows slower growth of cells, as it disrupts NADH oxidation into NAD⁺. POS5 plays an important role in maintaining the stability of mitochondria and oxidative stress response. Over-expression of POS5 has been reported to increase the stability of mitochondria and oxidative stress response (Tomàs-Gamisans et al. 2020; Zhao et al. 2015). In this study, GPD2 (YALI0B13970p) was deleted and POS5 (YALI0E17963p) was over-expressed in ZS100 to yield the ZSGP strain. As shown in Fig. 4B, the pyruvate concentration was increased by 41.8%, reaching 8.55 ± 0.09 g/L in ZSGP. These results indicate that reducing synthesis of glycerol would help produce pyruvate. It is well reported that MPC1 and MPC2 are essential in transporting pyruvate into mitochondria (Bricker et al. 2012). Lower intracellular pyruvate content significantly enhances pyruvate production by over-expression of MPC1 and MPC2 in Candida glabrata (Luo et al. 2018). Interestingly, over-expression of MPC1 (YALI0F00264p) and MPC2 (YALI0C03223p) in ZSGP did not affect pyruvate production (data not shown).

Optimization of fermentation conditions for the production of pyruvate

An initial inoculation amount of yeast has proved to be very important for production of pyruvate (Gao et al. 2021; Jeon et al. 2010). Therefore, the optimal initial inoculation to produce pyruvate using glycerol in flask culture was evaluated for ZSGP. The pyruvate production obtained at different initial inoculation amount can be observed in Fig. 5A. An initial inoculation amount of 0.5 at OD₆₀₀ yielded the highest production of pyruvate (8.49 ± 0.3 g/L). A lower initial inoculation amount resulted in low cell density, which in turn limited the production of pyruvate. Similarly, a higher initial inoculation amount did not provide enough time for cells to accumulate pyruvate and, therefore, a major pyruvate was generated in stationary phase.

Fig. 5.

Optimization of fermentation conditions for ZSGP in shaking flask. Figure A effect of initial inoculation amount on pyruvate production. X1, X2, X3, X4, X5 and X6 represents OD₆₀₀ as 0.1, 0.3, 0.5, 1.0, 1.5, and 2.0, respectively. Figure B effect of concentration of thiamine on pyruvate production. Figure C effect of CaCO₃ concentration on pyruvate production. Similarly, Fig. D effect of glycerol concentration on pyruvate production

Thiamine has been found to affect pyruvate production directly. A low concentration of thiamine inhibits growth rate via inhibition of thiamine-dependent enzymes like pyruvate dehydrogenase (Cybulski et al. 2019). We, therefore, evaluated the optimal concentration of thiamine for ZSGP growing in culturing flask. Figure 5B shows pyruvate production to different concentration of thiamine. We observed that thiamine at 1.5 μg/L concentration allows highest yield of pyruvate (9.51 ± 0.35 g/L). A lower concentration of thiamine suppresses cell growth due to the thiamine auxotrophy of Y. lipolytica. Similarly, a high concentration of thiamine allows less accumulation of pyruvate due to degradation by pyruvate dehydrogenase.

In the current study, CaCO₃ was used as a buffering agent to produce pyruvate (Liu et al. 2003; Zahoor et al. 2019). We, therefore, evaluated the optimal concentration of CaCO₃ for ZSGP growing in culturing flask as shown in Fig. 5C. We observed that CaCO₃ concentration at 50 g/L yielded the highest yield for pyruvate (10.92 ± 0.23 g/L). The osmostress response of yeast is carbon source dependent (Babazadeh et al. 2017). Therefore, we evaluated the optimal glycerol concentration for ZSGP growing in culturing flask as shown in Fig. 5D. We observed that glycerol concentration at 70 g/L yielded the pyruvate (12.25 ± 0.25 g/L), means a yield of 0.175 g/g. Cellular adaptation to osmotic stress is an important characteristics of living cells like yeast (Hohmann 2002; Saito and Posas 2012). A higher concentration of glycerol suppresses the growth of cells less due to osmostress.

Scale-up fermentation in a 20-L bioreactor

We scaled-up the ZSGP fermentation to a 20-L fermentor to obtain maximum pyruvate yield for industrial use. The optimum concentration of thiamine (1.5 μg/L) and glycerol (70 g/L) was used, after 50 and 64 h of fermentation, more glycerol (30 g/L) was added. During the whole process of fermentation, a pH of 4.0 was maintained using 6 M NaOH and after 120 h of fermentation, the highest pyruvate yield (110.4 g/L) at 96 h was obtained from ZSGP strain (Fig. 6).

Fig. 6.

Pyruvate yield in a 20-L fermentor. As shown in the figure, a maximum yield of pyruvate was obtained after 96 h from ZSGP strain using optimum concentration of substrates

Conclusion

We have designed a new method for efficient pyruvate production from Y. lipolytica. We conclude that ZSGP strain of Yarrowia lipolytica can be effectively used for pyruvate production at the industrial level.

Supplementary Information

Below is the link to the electronic supplementary material.

Abbreviations

DAK

Dihydroxyacetone kinases

GUT1

Glycerol kinase

GUT2

3-Phosphoglycerol dehydrogenase

GPD

Glyceraldehyde-3-phosphate dehydrogenase

GPP

Glycerol-1-phosphatase

GCY1

Glycerol dehydrogenase

Gly-3-p

Glycerol-3- phosphate

DHA

1, 3-Dihydroxyacetone

PYK1

Pyruvate kinase

POS5

NADH kinase POS5

TAG

Triacylglycerol

DGA

Diacylglycerol acyltransferase

DHAP

Dihydroxyacetone phosphate

MPC

Mitochondrial pyruvate carrier

GAP

Glyceraldehyde-3-phosphate

Declarations

Conflict of interest

Please check the following as appropriate: (a) all authors have participated in (a) conception and design, or analysis and interpretation of the data; (b) drafting the article or revising it critically for important intellectual content; and (c) approval of the final version. (b) This manuscript has not been submitted to, nor is under review at another journal or other publishing venue. (c) The authors have no affiliation with any organization with a direct or indirect financial interest in the subject matter discussed in the manuscript. Moral statement: (a) We have no conflict of interest with the editors, reviewers and other personnel of three biotech journals. (b) The subject of our research is microorganisms, which does not involve any human or animal subjects, and does not violate any ethical standards.