Scholarly article on topic 'Hexokinase—A limiting factor in lipid production from fructose in Yarrowia lipolytica'

Hexokinase—A limiting factor in lipid production from fructose in Yarrowia lipolytica Academic research paper on "Biological sciences"

CC BY-NC-ND
0
0
Share paper
Academic journal
Metabolic Engineering
OECD Field of science
Keywords
{Fructose / Sucrose / Hexokinase / "Lipid accumulation" / "Single-cell oil" / Yeasts}

Abstract of research paper on Biological sciences, author of scientific article — Zbigniew Lazar, Thierry Dulermo, Cécile Neuvéglise, Anne-Marie Crutz-Le Coq, Jean-Marc Nicaud

Abstract Microbial biolipid production has become an important part of making biofuel production economically feasible. Genetic engineering has been used to improve the ability of Yarrowia lipolytica, an oleaginous yeast, to produce lipids using glucose-based media. However, few studies have examined lipid accumulation by Y. lipolytica׳s ability to utilize other hexose sugars, and as of yet, the rate-limiting steps in this process are unidentified. In this study, we investigated the de novo accumulation of lipids by Y. lipolytica when grown in glucose, fructose, and sucrose. Three Y. lipolytica wild-type (WT) strains of varied origin differed significantly in their lipid production, growth, and fructose utilization. Hexokinase (ylHXK1p) activity partially explained these differences. Overexpression of the ylHXK1 gene led to increased hexokinase activity (6.5–12 times higher) in the mutants versus the WT strains; a pronounced reduction in cell filamentation in mutants grown in fructose-based media; and improved biomass production, particularly in the mutant whose parent had shown the lowest growth capacity in fructose (French strain W29). All mutants showed improved lipid yield and production when grown on fructose, although the effect was strain dependent (23–55% improvement). Finally, we overexpressed ylHXK1 in a highly modified strain of Y. lipolytica W29 engineered to optimize oil production. This modification was combined with Saccharomyces cerevisiae invertase gene expression to evaluate the resulting mutant׳s ability to produce lipids using cheap industrial substrates, namely sucrose (a major component of molasses). Sucrose turned out to be a better substrate than either of its building blocks, glucose or fructose. Over its 96h of growth in the bioreactors, this highly modified strain produced 9.15gL−1 of lipids, yielding 0.262gg−1 of biomass.

Academic research paper on topic "Hexokinase—A limiting factor in lipid production from fructose in Yarrowia lipolytica"

Contents lists available at ScienceDirect

Metabolic Engineering

journal homepage: www.elsevier.com/locate/ymben

Hexokinase—A limiting factor in lipid production from fructose in Yarrowia lipolytica

Zbigniew Lazara,b, Thierry Dulermoa, Cécile Neuvéglisea, Anne-Marie Crutz-Le Coqa, Jean-Marc Nicaud a,n1

a INRA, AgroParisTech, UMR1319 Micalis, F-78350 Jouy-en-Josas, France

b Department of Biotechnology and Food Microbiology, Wroclaw University of Environmental and Life Sciences, Chelmonskiego 37/41, 51-630 Wroclaw, Poland

CrossMark

ARTICLE INFO

Article history:

Received 27 May 2014

Received in revised form

27 August 2014

Accepted 12 September 2014

Available online 13 October 2014

Keywords:

Fructose

Sucrose

Hexokinase

Lipid accumulation

Single-cell oil

Yeasts

ABSTRACT

Microbial biolipid production has become an important part of making biofuel production economically feasible. Genetic engineering has been used to improve the ability of Yarrowia lipolytica, an oleaginous yeast, to produce lipids using glucose-based media. However, few studies have examined lipid accumulation by Y. lipolytica's ability to utilize other hexose sugars, and as of yet, the rate-limiting steps in this process are unidentified. In this study, we investigated the de novo accumulation of lipids by Y. lipolytica when grown in glucose, fructose, and sucrose. Three Y. lipolytica wild-type (WT) strains of varied origin differed significantly in their lipid production, growth, and fructose utilization. Hexokinase (ylHXK1p) activity partially explained these differences. Overexpression of the ylHXK1 gene led to increased hexokinase activity (6.5-12 times higher) in the mutants versus the WT strains; a pronounced reduction in cell filamentation in mutants grown in fructose-based media; and improved biomass production, particularly in the mutant whose parent had shown the lowest growth capacity in fructose (French strain W29). All mutants showed improved lipid yield and production when grown on fructose, although the effect was strain dependent (23-55% improvement). Finally, we overexpressed ylHXK1 in a highly modified strain of Y. lipolytica W29 engineered to optimize oil production. This modification was combined with Saccharomyces cerevisiae invertase gene expression to evaluate the resulting mutant's ability to produce lipids using cheap industrial substrates, namely sucrose (a major component of molasses). Sucrose turned out to be a better substrate than either of its building blocks, glucose or fructose. Over its 96 h of growth in the bioreactors, this highly modified strain produced 9.15 g L_1 of lipids, yielding 0.262 g g~1 of biomass.

© 2014 International Metabolic Engineering Society Published by Elsevier Inc. On behalf of International Metabolic Engineering Society. This is an open access article under the CC BY-NC-ND license

(http://creativecommons.org/licenses/by-nc-nd/3.0/).

1. Introduction

Yarrowia lipolytica is one of the most widely studied species of "nonconventional" oleaginous yeasts (Barth and Gaillardin, 1996; Nicaud, 2012) because of its status as a 'GRAS' (Generally Recognized as Safe) organism and its promising industrial potential (Groenewald et al., 2014). This potential is currently being explored in numerous studies that are attempting to enhance Y. lipoly-tica's ability to produce various molecules of interest, such as organic acids (e.g., citric acid; Rywinska and Rymowicz, 2010), sugar-alcohols or polyols (e.g., erythritol; Mironczuk et al., 2014),

* Corresponding author.

1 Present address: AgroParisTech, INRA, UMR1319, F78850 Thiverval-Grignon, France.

and fatty acids that can be used as biofuels (Blazeck et al., 2014). Indeed, the fact that Y lipolytica is able to synthesize and store amounts of fatty acids that are up to 40% of its dry cell weight (Beopoulos et al., 2009) has led this yeast to be regarded as a promising microbial producer of biofuels. In contrast, the standard reference yeast Saccharomyces cerevisiae is not oleaginous and can only accumulate lipids in amounts of up to 15% over its own biomass (Dyer et al., 2002). Thus, as a result of its phylogenetic and metabolic divergence from S. cerevisiae, Y. lipolytica has earned a place as a model organism in studies of lipid accumulation. In these studies, the fully sequenced Y. lipolytica genome (Dujon et al., 2004) has served as a valuable tool. It has enabled the improvement of some aspects of lipid metabolism through the manipulation of several genes involved in the bioconversion, synthesis, and mobilization of lipids (Beopoulos et al., 2008, 2012; Dulermo and Nicaud, 2011; Tai and Stephanopoulos, 2013; Blazeck et al., 2014). However,

http://dx.doi.org/10.1016/j.ymben.2014.09.008

1096-7176/© 2014 International Metabolic Engineering Society Published by Elsevier Inc. On behalf of International Metabolic Engineering Society. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/3.0/).

despite the increasing amount of information available on the biosynthesis of triacylglycerols and steryl esters, many questions regarding lipid homeostasis remain unanswered.

Y lipolytica begins to accumulate lipids when nitrogen in the medium is limiting and carbon resources are in excess. The process depends on temperature and pH and is also competitive with the production of citric acid, an immediate precursor of lipid accumulation. The C/N ratio of the medium affects various metabolic parameters, such as growth, organic acid production, and lipid biosynthesis (Beopoulos et al., 2009). When carbon resources are exhausted, cells remobilize stored lipids. The efficiency of carbon source utilization is therefore an important factor in biomass production and lipid accumulation.

Glucose and fructose, widespread in nature and easy to produce industrially, are relatively cheap raw materials for the production of intracellular lipids. Both monosaccharides are also components of the disaccharide sucrose (table sugar), a readily available compound that has already been successfully used in citric acid production by genetically modified strains of Y lipolytica (Lazar et al., 2011, 2013; Moeller et al., 2012). However, this process has revealed some issues related to the use of fructose: it appears that glucose is preferentially consumed over fructose and, therefore, fructose is only used after any available glucose has been completely consumed (Lazar et al., 2011, 2013). Fructose is thus utilized late in the production process and may not be completely consumed before cell growth is inhibited, partially due to citric acid production (Lazar et al., 2011). A similar situation occurs during ethanol fermentation of grape must by S. cerevisiae and can lead to fermentation defects (Liccioli et al., 2011). In both species, strains with different fructose utilization capacities have been characterized (Guillaume et al., 2007; Lazar et al., 2011; Liccioli et al., 2011).

Phosphorylation of hexoses, e.g., glucose and fructose, is one of the key steps in sugar metabolism. This process is carried out by specific kinases in the hexokinase gene family, namely, glucokinase, which is specialized for glucose phosphorylation, and hexokinase, which is involved in the phosphorylation of other hexoses, including fructose. These two enzymes have been experimentally identified in Y lipolytica (Petit and Gancedo, 1999) and are encoded by YAL10E15488g (ylGLKl ) and YAL10B22308g (ylHXKl ), respectively. In Y lipolytica, both glucokinase and hexokinase can use glucose as a substrate at a KM of 0.17 mM and 0.38 mM, respectively, whereas only hexokinase has an affinity for fructose, at a KM of 3.56 mM. 1n vivo glucose phosphorylation is mostly carried out by glucoki-nase, while fructose activation is performed by hexokinase. 1ndeed, Petit and Gancedo (1999) observed that deletion of the ylHXK1 gene increased doubling time by 15% (relative to wild type) in strains growing on glucose, but they saw no growth at all on fructose. In addition, Y. lipolytica hexokinase has been shown to be the functional equivalent of S. cerevisiae hexokinase 11 (scHXK2p, YGL253W), which is involved in glucose catabolite repression (Petit and Gancedo, 1999); likewise, ylHXK1 is suspected to be involved in glucose repression of the LIP2 gene, which encodes extracellular lipase in Y. lipolytica (Fickers et al., 2005a).

1n this study, we compared glucose and fructose utilization in three wild-type strains of Y. lipolytica obtained from different sources and examined the effect of hexokinase overexpression on growth, cell morphology, and lipid accumulation. Hexokinase overexpression led to a strong reduction in cell filamentation on fructose-based media and improved biomass production, particularly in the strain with the lowest growth capacity on fructose-based substrates - French strain W29. Lipid yield and production were also improved for all strains, particularly when grown on fructose, although the effect was strain

JMY1233 Apox1-6

PUTtgl4 (JME1364)

JMY2179 Apox1-6 Atgl4::URA3ex

pUB4-CRE 1

JMY3122 Apox1-6 Atgl4

pTEF-DGA2-LEU2ex (JMP1822)

JMY3373 Apox1-6 Atgl4 pTEF-DGA2-LEU2ex

pTEF-GPD1-URA3 (JMP1128)

JMY3501 Apox1-6 Atgl4 pTEF-DGA2-LEU2ex pTEF-GPD1-URA3ex

pUB4-CRE 1

JMY3820 Apox1-6 Atgl4 pTEF-DGA2 pTEF-GPD1

pTEF-YlHXK1-URA3ex (JMP2103) pTEF-SUC2-LEU2ex (JMP2347) +URA3ex fragment

JMY4059 Apox1-6 Atgl4 pTEF-DGA2 pTEF-GPD1 pTEF-YlHXK1-URA3ex

pTEF-SUC2-LEU2ex (JMP2347)

JMY4086 Apox1-6 Atgl4 pTEF-DGA2 pTEF-GPD1 Apox1-6 Atgl4 pTEF-DGA2 pTEF-YlHXK1-URA3ex pTEF-SUC2-LEU2ex JMY4498 pTEF-GPD1 pTEF-SUC2-LEU2ex

Fig. 1. Schematic representation of strain construction.

The JMY3501 strain was derived fromJMY1233 (Beopoulos et al. 2008). (i) TGL4 was inactivated by introducing the disruption cassette tgl4::URA3ex from JMP1364 (Dulermo et al., 2013), which generated JMY2179. (ii) An excisable auxotrophic marker, URA3ex, was then excised fromJMY2179 using JMP547 (Fickers et al., 2003), which generated JMY3122. (iii) JMY3501 was then obtained by successively introducing pTEF-DGA2-LEU2ex, from JMP1822, and pTEF-GPD1-URA3ex, from JMP1128 (Dulermo and Nicaud, 2011), into JMY3122. JMP1822 was obtained by replacing the URA3ex marker fromJMP1132 (Beopoulos et al. 2008) with LEU2ex.

The JMY4086 strain was generated by successively introducing pTEF-ylHXK1-URA3ex, from JMP2103, and pTEF-SUC2-LEU2ex, from JMP2347, into JMY3820. JMY3820 corresponds to JMY3501, but is different in that the URA3ex and LEU2ex markers in the former have been rescued, as previously described (Fickers et al., 2003). The JMY4498 strain was generated by introducing pTEF-SUC2-LEU2ex, from JMP2347, and I-Scel fragment of URA3ex from JMP2103, into JMY3820.

dependent. To further investigate lipid production on cheap industrial substrates, hexokinase and invertase were overexpressed in strains with improved lipid metabolisms and cultivated in bioreactors using sucrose as the carbon source.

2. Materials and methods

2.1. Yeast strains and plasmids

The plasmids and strains used in this study are listed in Table S1. Three Y. lipolytica wild-type (WT) strains were used (country of origin in parentheses): W29 (France), A-101 (Poland), and H222 (Germany) (Wojtatowicz and Rymowicz, 1991; Barth and Gaillardin, 1996). The following auxotrophic strains had previously been derived from these WT strains and were also used in this study: PO1d (Ura~Leu~) from W29 (Barth and Gaillardin 1996), A-101-A18 (Ura") from A-101 (Walczak and Robak, 2009), and Y322 (Ura") from H222 (Mauersberger et al., 2001). The other strains used in this study were strains Y3573, Y3812, and Y3850, which contained an expression cassette that included the Y. lipolytica HXK1 gene from W29 (ylHXKl, YALI0B22308g) under the control of the constitutive TEF promoter (Müller et al., 1998), and strain Y3572, which contained an expression cassette carrying the S. cerevisiae hexokinase gene HXK2 (scHXK2, YGL253W). Transformation of Y. lipolytica was performed with the lithium acetate procedure (Xuan et al., 1990), using NotI digested fragments for random chromosomal integration (Mauersberger et al., 2001).

To recover prototrophy, strains Y3572 and Y3573 were transformed with a purified SalI fragment of the plasmid pINA62 that contained the LEU2 gene (Gaillardin and Ribet, 1987). Construction of the Y4086 strain, which was modified for lipid production, is depicted in detail in Fig. 1.

2.2. Growth media

Media and growth conditions for Escherichia coli were identical to those in previous studies (Sambrook and Russell, 2001), as were those of Y. lipolytica (Barth and Gaillardin, 1996). Rich (YPD) medium was prepared using 20 g LBacto™ Peptone (Difco, Paris, France), 10 g L_ 1 yeast extract (Difco), and 20 g L_1 glucose (Merck, Fontenay-sous-Bois, France). Minimal (YNB) medium was prepared using 1.7 g L_1 yeast nitrogen base (without amino acids and ammonium sulfate, Difco), 10gL_ 1 glucose (Merck), 5gL_ 1 NH4Cl, and 50 mM phosphate buffer (pH 6.8). To complement auxotrophic processes, 0.1 g L_ 1 uracil or leucine (Difco, Paris, France) was added as necessary.

2.3. Growth in microtiter plates

Precultures were obtained from frozen stocks, inoculated into tubes containing 5 mL YPD medium, and cultured overnight (170 rpm, 28 °C). Precultures were then centrifuged and washed with sterile distilled water; cell suspensions were standardized to an OD600 of 0.1. Yeast strains were grown in 96-well plates in 200 ml of minimal YNB medium (see above) containing 10 g L_ 1 of either glucose or fructose. The culture was performed thrice, with 2-3 replicates for each condition. Cultures were maintained at 28 °C under constant agitation with a Biotek Synergy MX micro-titer plate reader (Biotek Instruments, Colmar, France); each culture's optical density at 600 nm was measured every 20 min for 72 h.

2.4. Media and growth for lipid biosynthesis experiments

For lipid biosynthesis in minimal media, cultures were prepared as follows: an initial preculture was established by inoculating 50 mL of YPD medium in 250-mL Erlenmeyer flasks; this was followed by an overnight shaking step at 28 °C and 170 rpm. The resulting cell suspension was washed thrice with sterile distilled water and used to inoculate 50 mL of YNB medium containing 15, 30, 60, 90, or 120 gL_ 1 of fructose (corresponding to a carbon/nitrogen (C/N) ratio of 15, 30, 60, 90, and 120, respectively). Each culture was incubated, with shaking, in non-baffled 250-mL Erlenmeyer flasks, at 28 °C and 170 rpm for 168 h, or until all available sugar had been consumed. We also evaluated lipid biosynthesis in several other types of media, including a glucose-only (60 g L_ 1, C/N = 60) control medium, and a sucrose-containing (60gL_ 1, C/N=60) medium. The preculture and growth conditions for each experiment were as described above. Media were prepared with distilled water except for experiments with Y. lipolytica overacumulating strains Y3501 and derivatives, in which tap water was used instead, as a cheaper ingredient and potential additional source of microelements.

2.5. Bioreactor studies

Lipid biosynthesis was also evaluated in batch cultures (BC) that were maintained for 96 h in 5-L stirred-tank BIO-STAT B-PLUS reactors (Sartorius, Frankfurt, Germany) under the following conditions: 2-L working volume, 28 °C, 800 rpm, and 4-Lmin~1 aeration rate. The production media contained 150 g sucrose, 1.7 g YNB, 3.75 g NH4Cl, 0.7 g KH2PO4, and 1.0 g MgSO4 x 7H2O, all in 1 L of tap water. Culture acidity was automatically maintained at pH 6.8 using a 40% (w/v) NaOH solution. Culture inocula were grown in 0.1 L of YNB medium with 30 g L~1 glucose in 0.5-L flasks on a rotary shaker kept at 170 rpm and 28 °C for 48 h; inocula were added to the bioreactor cultures in a volume equal to 10% of the total working volume.

To analyze lipid production in the bioreactor cultures, a 15-mL sample was taken from each culture 10 min after inoculation (Time=0); subsequent sampling was conducted every 12 h. Each sample was centrifuged for 10 min at 5000 rpm; supernatants and cell pellets were collected and used for further analyses.

2.6. General genetic techniques and plasmid construction

Standard molecular genetic techniques were used throughout this study following Sambrook and Russell (2001). Restriction enzymes were obtained from New England Biolabs (Ipswich, England). Genomic DNA from yeast transformants was prepared as described by Querol et al. (1992). PCR amplification was performed using an Eppendorf 2720 thermal cycler and employing both Taq (Promega, Madison, WI) and Pfu (Stratagene, LaJolla, CA) DNA polymerases. PCR fragments were then purified with a QIAgen Purification Kit (Qiagen, Hilden, Germany), and DNA fragments were recovered from agarose gels using a QIAquick Gel Extraction Kit (Qiagen, Hilden, Germany). The Staden software package was used for gene sequence analysis (Dear and Staden, 1991). To quantify hexokinase gene expression, genes were amplified with the primer pairs ylHXK1-fwd and ylHXK1-rev (GAGAAGATC-TATGGTTCATCTTGGTCCCCGAAAACCC and GCGCCCTAGGCTAAA-TATCGTACTTGACACCGGGCTTG, respectively), and scHXK2-fwd and scHXK2-rev (GCGCGGATCCATGGTTCATTTAGGTCCAAAAAAACC and GCGCCCTAGGTTAAGCACCGATGATACCAACG, respectively), all of which contained BamHI(BglII)-AvrII restriction sites. These restriction sites enabled the genes to be cloned into JME1128 plasmids that had been digested with BamHI-AvrII, as previously described (Beopoulos et al. 2008; Dulermo et al., 2013). To delete the genes of

J 0.100 в

•S 0.010 0.001

40 Time (h)

J 0.100 -| в

•S 0.0100.001

40 Time (h)

s 0.100-

•S 0.010

40 Time (h)

1.000 -

s 0.100-

•S 0.010

40 Time (h)

Fig. 2. Growth curves of different Y lipolytica WT strains (A,B) and ylHXKl- overexpression transformants (C,D) grown in YNB medium with 10 gL 1 glucose (A,C) or 10 g L_ 1 fructose (B,D). WT strains were W29 (■••), A-101 (-), and H222 (- -); growth was analyzed using a Biotek apparatus.

interest, the disruption cassettes were produced in accordance with the protocol of Fickers et al. (2003)). Auxotrophies were restored via excision using the Cre-lox recombinase system following transformation with the replicative plasmid pUB4-Cre1 (JME547) (Fickers et al., 2003).

2.7. Fluorescence microscopy

Images were obtained using a Zeiss Axio Imager M2 microscope (Zeiss, Le Pecq, France) with a 100 x objective lens and Zeiss filter sets 45 and 46 for fluorescence microscopy. Axiovision 4.8 software (Zeiss, Le Pecq, France) was used for image acquisition. To make the lipid bodies (LBs) visible, BodiPy® Lipid Probe (2.5mgmL_1 in ethanol; Invitrogen) was added to the cell suspension (OD600=5) and the samples were incubated for 10 min at room temperature.

2.8. Lipid determination

Fatty acids (FAs) in 15-mg aliquots of freeze-dried cells were converted into methyl esters using the method described in Browse et al. (1986) and were analyzed using a gas chromatograph (GC). GC analysis of FA methyl esters was performed using a Varian 3900 instrument equipped with a flame ionization detector and a Varian FactorFour vf-23 ms column, for which the bleed specification at 260 °C was 3 pA (30 m, 0.25 mm, 0.25 ^m). FAs were identified by comparing their GC patterns to those of commercial FA methyl ester standards (FAME32; Supelco) and quantified using the internal standard method, which involved the addition of 50 mg of commercial C17:0 (Sigma).

Total lipid extractions were obtained from 100-mg samples (cell dry weight (CDW)) in accordance with the method described by Folch et al. (1957). Briefly, Y. lipolytica cells were spun down, washed with water, freeze dried, and then resuspended in a 2:1 chloroform/methanol solution and vortexed with glass beads for 20 min. The organic solution was extracted and washed with 0.4 mL of 0.9% NaCl solution before being dried at 60 °C overnight and weighed to quantify lipid production.

2.9. Sugar and citric acid measurement

Citric acid (CA), glucose, fructose, and sucrose were identified and quantified by HPLC (UltiMate 3000, Dionex-Thermo Fisher Scientific, UK) using an Aminex HPX87H column coupled to UV (210 nm) and RI detectors. The column was eluted with 0.01 N H2SO4 at room temperature and a flow rate of 0.6 mL min_1. Identification and quantification were achieved via comparisons to standards. Before being subject to HPLC analysis, samples were filtered on 0.45-^m pore-size membranes.

2.10. Dry biomass determination

To determine dry biomass, the cell pellets from 15-mL culture samples were washed twice with distilled water, filtered on 0.45-^m pore-size membranes, and dried at 105 °C using a WPS 110S weight dryer (Radwag, Poznan, Poland) until a constant mass was reached.

2.11. Measurement of hexokinase activity

Total hexokinase activity was determined using whole cell extracts and a Hexokinase Colorimetric Assay Kit (Sigma-Aldrich, Saint Louis, MO, USA) in accordance with the manufacturer's instructions. The reaction was performed at 24 °C in 96-well plates using a Biotek Synergy MX microtiter plate reader and was monitored by measuring absorbance at 450 nm. One unit of hexokinase was defined as the amount of enzyme that generated 1.0 ^mole of NADH per minute at pH 8.0 at room temperature.

2.12. Reverse transcription and qRT-PCR

RNA extraction was performed using TRIzol® reagent (Invitrogen, Carlsbad, CA, USA) in accordance with the manufacturer's instructions. Nucleic acids amounts were measured using a Biochrom WPA Biowave II spectrophotometer (Biochrom Ltd., Cambridge, UK) equipped with a TrayCell (HelmaAnalytics, Müllheim, Germany). Following the manufacturer's instructions, cDNA was prepared using Maxima First Strand cDNA Synthesis Kits for RT-qPCR (Thermo-Scientific, Waltham, MA, USA).

Real-time PCR was performed using the DyNAmo Flash SYBR Green qPCR Kit (ThermoScientific, Waltham, MA, USA) with 0.5 ^M forward and reverse primers and 1 ^g of cDNA template in a final reaction volume of 10 ^L. Thermocycling was performed in the Eco Real-Time PCR System (Illumina, San Diego, CA, USA) with the following cycling parameters: 5 min incubation at 95 °C, followed by 40 cycles of 10 s at 95 °C, 10 s at 60 °C, and 8 s at 72 °C. Fluorescence data were acquired during each elongation step, and at the end of each run, specificity was controlled by melting curve analysis. Hexokinase expression was detected using the primers ylHXK1-qPCR-fwd (TCTCCCAGCTTGAAACCATC) and ylHXK1-qPCR-rev (CTTGACAACTCGCAGGTTGG). The results were normalized to actin gene expression (Lazar et al., 2011) and then analyzed using the ddCT method (Schmittgen and Livak, 2008).

All experiments in this paper were performed at least thrice.

3. Results and discussion

Y. lipolytica is a strictly aerobic microorganism that grows on hydrophobic substrates like n-alkanes, fatty acids, and oils (Fickers et al., 2005b). This yeast can metabolize a few different types of sugars, namely glucose, fructose, and mannose (Coelho et al., 2010; Michely et al., 2013) and preferentially consumes glucose over fructose (Wojtatowicz et al., 1997; Lazar et al., 2011).

3.1. Variable fructose utilization by Y. lipolytica strains of different origins

The ability of the three WT strains of different origins to grow in media containing either glucose or fructose was compared. The strains had similar growth kinetics in YNB containing 10 g L_ 1 glucose (m=0.355 h~1; Fig. 2A); however, their growth kinetics differed significantly in YNB containing 10 g L_ 1 fructose (Fig. 2B). In the fructose medium, H222 had a constant growth rate (0.282 h _ 1), whereas both A-101 and W29 showed reduced, varied growth rates. From about 8-20 h, A-101 exhibited a low growth rate (0.131 h_ 1), and W29's growth rate was even slower. In the subsequent phase, growth rates increased but remained lower than that of H222 (0.203 h^1 and 0.182 h"1 for A101 and W29, respectively). Taken together, these results clearly show that the WT strains differed in their growth in the fructose medium and that the French line, W29, had the most difficulty exploiting this substrate.

3.2. Overexpression of the ylHXKl gene enhances hexokinase activity, growth, and fructose uptake in Y. lipolytica

As Hxk1p is crucial for fructose assimilation in Y. lipolytica, we reasoned that interstrain variation in its activity could be responsible for the diverse growth patterns observed when the strains were grown in the fructose medium. We thus compared the sequences of the hexokinase gene and its promoter region among

the three strains (Neuveglise, unpublished data). No polymorphisms were found in the sequence encoding the hexokinase protein, and only a few changes were identified in the different strains' promoter regions (data not shown).

To increase hexokinase activity, an additional copy of ylHXKl under the control of a strong constitutive TEF promoter (Müller et al., 1998) was introduced. Overexpression of ylHXKl increased both HXK1 transcript abundance (at least 23 fold) and hexokinase activity (at least 6 fold) in yeast grown in both glucose and fructose media (Table 1). In the fructose medium, all three ylHXKl -over-expressing strains exhibited similar hexokinase activity (around 1700 U gcDw). This result confirmed that HXK1 had been successfully overexpressed.

Next, we compared the growth patterns of the ylHXKl -over-expressing strains and the wT strains. The growth profiles of the two groups were similar in the glucose medium (overexpressing strains: ¿u=0.367 h" 1; Fig. 2C). However, in contrast to the WT strains, all three overexpressing strains exhibited the same growth kinetics in the fructose medium (¿u=0.363 h" 1; Fig. 2D) as in the glucose medium. This finding indicates that hexokinase activity can limit W29's and A-101's growth in fructose. Interestingly, overexpression of hexokinase II in S. cerevisiae did not stimulate that species' growth in a fructose medium (Ernandes et al., 1998), suggesting that there are fundamental differences between S. cerevisiae and Y. lipolytica in the regulation of fructose metabolism.

Finally, to investigate the effects of hexokinase overexpression on glucose and fructose assimilation, we analyzed the uptake of these sugars during growth (Fig. S1). Both the WT strains and the ylHXKl -overexpressing strains consumed glucose at the same rate (0.65, 0.56, and 0.54 g L"1 h"1 for the W29, A-101, and H222 transformants, respectively; Fig. S1: A,C,E). In contrast, the overexpressing strains consumed fructose faster (0.64, 0.56, and 0.55 g L"1 h"1 for the W29, A-101, and H222 transformants, respectively) than did the WT strains (0.36, 0.38, and 0.54gL"1h"1 for W29, A-101, and H222, respectively; Fig. S1: B,D,F). It is worth noting that the overexpressing strains consumed fructose and glucose at the same rate. When its natural, limited hexokinase activity was boosted, W29 was better at exploiting glucose and fructose than the other two strains.

3.3. Overexpression of hexokinase inhibits Y. lipolytica filamentation

When Y. lipolytica was grown in the glucose medium, the growth of yeast-form cells was favored; however, when grown in the fructose medium, all three WT strains showed filamentous-form growth (Fig. 3: A,C,E). HXK1 overexpression strongly decreased filamentous-form growth in the fructose medium (Fig. 3: B,D,F), and even after 5 days of culture, cells remained in the yeast form.

In Y. lipolytica and other well-studied yeasts such as S. cerevi-siae, filamentation is known to be triggered by the consumption of non-glucose carbon sources: N-acetyl-glucosamine by Y. lipolytica (Herrero et al., 1999; Hurtado and Rachubinski, 1999); and mannose,

Table 1

Activity and mRNA fold change of hexokinase extracted from Y. lipolytica WT and ylHXKl mutants growing in YNB medium with 100 g 1 glucose or 100 g 1 fructose analyzed at 24 h of the culture.

Strain Glucose Fructose

Activity (U gcdw) Fold change Fold change in transcript level Activity (U gcdw) Fold change Fold change in transcript level

W29-HXK1 A-101

A-WÎ-HXK1 H222

H222-HXK1

193.5 7 23 1490.4 7 186

42.4 7 3 1148.6 7 87 33.1 7 4 1122.67 100

7.70 27.10 33.92

28.227 1.7 40.21 7 3.0 55.847 3.6

145.37 12 1766.27 118 155.67 10 1670.27 151 256.57 21 1653.67 181

12.15 10.73 6.45

96.11 7 4.8 55.397 2.8 23.61 7 1.2

ylHXKl

DIC (Nomarski) С s- ' -

f (■' , E

(, / Д, Ш

¿А/ }

Vf — 1 /

Fig. 3. Cell morphology of Y lipolytica WT and ylHXKl-overexpression transformants. Images are of the WT French line W29 (A), Polish line A-101 (C), and German line H222 (E), as well as of their respective overexpression transformants (B, D, E, respectively). Images were taken after 120 h of growth in flasks in YNB fructose medium (carbon source 100 g 1).

maltose, maltotriose, or sucrose by S. cerevisiae (da Silva et al., 2007; Van de Velde and Thevelein, 2008). Interestingly, in S. cerevisiae, both the consumption of fructose (da Silva et al., 2007) and the absence of hexokinase activity result in filamentous-form growth, as does the deletion of the gene encoding hexokinase II, when the species is grown in a glucose medium (Van de Velde and Thevelein, 2008).

3.4. Hexokinase overexpression increases biomass and lipid biosynthesis

We compared lipid production by WT and ylHXKl -overexpres-sing strains grown in YNB glucose and fructose media (100 g L_ 1 C/N molar ratio of 100). Production of dry biomass, fatty acids, and citric acid, as well as sugar consumption, were followed over the 120-h culturing period (Table 2).

In the glucose medium, W29 and its ylHXKl transformant produced the greatest amount of dry biomass: ~ 21.5 g L_ \ However, ylHXKl overexpression did not lead any strain to significantly increase biomass production when grown in the glucose medium. In contrast, in the fructose medium, ylHXKl overexpression greatly increased biomass production by the W29 transformant (around 4 g L_1, or 24% more than WT W29); the improvement shown by the A-101 and H222 transformants was smaller (increase of 0.6 to 5%). The strains differed less when it came to biomass yielded per

unit of substrate consumed (YX/S; Table 2). Nevertheless, ylHXKl overexpression slightly increased YX/S for the H222 transformant grown in the glucose and fructose media as well as for the W29 transformant grown in the fructose medium. On fructose, the W29 transformant had the highest levels of biomass production and bioconversion; similarly to its WT of origin, it also outperformed the other strains when grown in the glucose medium.

Lipid accumulation was evaluated by measuring one parameter— the amount of total fatty acids (FAs)—and calculating two yields—YFA/ X per unit biomass and YFA/S per unit substrate (Table 2). Among the WT strains, A-101 produced the greatest amount of total FAs in both the glucose (3.03 g L_ 1) and fructose (2.19 g L_ 1) media. H222 showed the weakest production (43% lower FA production than A-101 in both media). Although all three strains had lower FA production in fructose than in glucose, W29 exhibited the greatest difference (1 g L_1 less total FAs in fructose than in glucose), perhaps because this strain shows disparate biomass production between the two media. YFA/X and YFA/S showed the same patterns as FA production and were used to compare the different strains independently of growth: yield was highest and lowest for A-101 and H222, respectively. These results also indicate that all three strains more efficiently produce FAs from glucose than from fructose.

Hexokinase overexpression allowed all the strains to increase FA production in both media. Nevertheless, the improvement was

Parameters of fatty acids, biomass and citric acid production by different origin WT and ylHXK1 transformants of Y. lipolytica growing 120 h in YNB medium with glucose or fructose (carbon source 100 g L_1, C/N 100).

Parameters Glucose Fructose

W29 A-101 H222 W29 A-101 H222

WT ylHXK1 WT ylHXK1 WT ylHXK1 WT ylHXK1 WT ylHXK1 WT ylHXK1

X gL-1 21.4 21.9 16.8 17.3 15.4 16.2 17.4 21.6 16.9 17.8 15.7 15.8

yx/s gg 1 0.27 0.28 0.25 0.26 0.23 0.26 0.25 0.28 0.27 0.27 0.23 0.25

FA gL-1 2.57 3.28 3.03 3.12 1.69 1.95 1.56 3.02 2.19 2.85 1.26 1.90

yfa/x gg -1 0.12 0.15 0.18 0.18 0.11 0.12 0.09 0.14 0.13 0.16 0.08 0.12

yfa/s gg -1 0.032 0.044 0.045 0.046 0.025 0.032 0.022 0.039 0.035 0.043 0.019 0.029

CA gL"1 0.54 2.21 4.89 8.76 1.04 0.51 0.33 1.16 2.47 3.65 0.26 0.00

Symbols: X - dry biomass, FA - fatty acids, CA - citric acid, YX/S - yield of biomass from consumed substrate, YFA/S - yield of fatty acids from consumed substrate, YFA/X - yield of fatty acids from dry biomass; SD of all analyzed parameters did not exceed 7%.

Composition of FA produced by Y lipolytica WT and ylHXQ transformants growing 120 h in YNB glucose or fructose medium (carbon source 100 g L_1, C/N 100).

Fatty acid Glucose Fructose

W29 A-101 H222 W29 A-101 H222

WT ylHXK1 WT ylHXK1 WT ylHXK1 WT ylHXK1 WT ylHXK1 WT ylHXK1

C16:0 17.4 18.4 11.9 11.3 18.8 17.2 15.5 18.7 12.2 12.0 17.3 16.8

C16:1(n- 7) 7.1 6.2 7.6 8.7 5.7 6.1 7.0 6.7 7.9 8.6 7.2 7.5

C18:0 11.9 13.7 8.2 7.5 15.8 14.0 9.5 13.5 7.6 7.6 12.1 11.5

C18:1(n - 9) 52.5 47.1 61.1 62.9 47.9 50.2 54.2 47.4 60.3 61.6 48.9 49.0

C18:2(n - 6) 7.3 8.7 7.6 6.4 7.3 7.6 9.6 8.7 8.3 6.8 10.2 10.3

Others 4.0 5.9 3.6 3.3 4.5 4.8 4.2 5.0 3.7 3.4 4.3 4.9

most evident when the transformants were grown in the fructose medium. In contrast to the results for the WT strains, the W29 transformant produced the greatest amount of FAs in both the glucose (3.28 g L_and fructose (3.02 g L_1) media, partly because of its high level of biomass production. For all strains in both media, YFA/X and YFA/S increased as a result of hexokinase overexpression, except in the case of the A-101 transformant grown in the glucose medium: its WT already had a high level of production. As with FA production, FA yields were highest when the transformants were grown in the fructose medium. The W29 transformant showed the greatest improvement: its YFA/S demonstrated increases of 56% and 25% in fructose and glucose, respectively, compared to its WT.

Citric acid (CA) production was also examined since it is an expected consequence of nitrogen limitation in Y. lipolytica. In this experiment, strains varied little in CA production, and relatively small amounts of CA were produced. A-101 and its ylHXK1 transformant secreted the most CA (Table 2). This result is not surprising since A-101 has been selected for this trait (Wojtatowicz and Rymowicz, 1991). Indeed, when grown in batch cultures and under conditions optimized for CA production, WT A-101 can yield 0.45 g of CA per g of glucose (Rywinska et al., 2010); it produced a maximum of 0.13 g of CA per g of glucose under our experimental conditions (data not shown). Hexokinase overexpression led the W29 and A-101 transformants to produce more CA than their WTs, in both glucose and fructose media. In contrast, the opposite effect was observed for H222: the H222 transformant was better able to convert sugars into lipids with reduced CA production. The fact that the W29 and A-101 transformants had slightly increased CA production suggests lipid synthesis may become limiting in the WT strains.

We also examined the FA composition of the WT and ylHXK1 overexpressing strains. The three Y. lipolytica WT strains differed in their FA profiles (Table 3). They contained somewhat high levels of C18:1 (47.1-62.9%), medium levels of the saturated fatty acids C16:0 and C18:0 (11.3-18.8% and 7.5-15.8%, respectively), and lower levels of C18:2 (6.4-10.3%). A-101 had higher levels of

C18:1 and lower levels of C18:0 and C16:0, which suggests that FA elongation and desaturation were more efficient in this strain than in the other two strains, perhaps due to its higher activity of the A9-desaturase and elongase enzymes. Hexokinase overexpression most clearly affected the FA composition of W29, in both glucose and fructose media: the W29 transformant had lower levels of C18:1 and higher levels of saturated C16:0 and C18:0, which suggests that faster FA synthesis might have resulted in A9-desaturase becoming saturated, thus reducing the conversion of C18:0 to C18:1. In the other two strains, hexokinase overexpression did not significantly change FA composition. Indeed, in Y. lipolytica, the number of desaturases gene copies was shown to be important for efficient lipid production (Xue et al., 2013).

3.5. Impact of different hexokinase genes and C/N ratios on fatty acid production

Given the improved performance shown by our W29 transformant and the fact that many studies on lipid biosynthesis have been conducted using other W29 derivatives such as PO1d and PO1f (Dulermo and Nicaud, 2011; Beopoulos et al., 2012; Blazeck et al., 2014), we chose to conduct further analyses using W29.

Y lipolytica hexokinase is unique because it is very sensitive to trehalose-6-phosphate inhibition, much more so than Hxk2p in S. cerevisiae (Petit and Gancedo, 1999). Therefore, we examined FA production following overexpression of either scHXK2 or ylHXK1 in the W29 background (strain PO1d). Overexpression of both genes improved lipid production after yeast was cultured for 72 h in the fructose medium. However, there was a difference in magnitude: overexpression of ylHXK1 resulted in an 82% increase, while overexpression of scHXK2 lead to a 49% increase (Fig. S2).

The effect of C/N molar ratios on lipid production and profiles for strains grown in the fructose medium was then investigated (Fig. 4). C/N ratios significantly positively affected FA yield; the WT and ylHXK1 -overexpressing strains responded similarly. FA yield

was most improved when C/N=90. However, when C/N was above 60, lipid quantities remained similar, unawares high residual fructose was observed. After 120 h of culture, residual fructose was 10, 35, and 62 g Lfor C/N ratios of 60, 90, and 120, respectively (Fig. S3). In addition, we observed that levels of C16:0 and C18:0 increased as levels of C16:1 and C18:1 decreased (Table S2). This result again suggests that desaturase activity may limit efficient lipid production.

3.6. Effects of ylHXKl overexpression in a strain optimized for fatty acid accumulation

Finally, we investigated the impact of ylHXKl overexpression in a W29 transformant engineered to optimize lipid production. Experimental strains were constructed as described in Fig. 1. First, the genes encoding acyl-coenzyme A oxidases (POXl-6 genes) were deleted, blocking ß-oxidation (Beopoulos et al., 2008). In Y. lipoly-tica, lipids accumulate in specialized organelles called lipid bodies, mainly in the form of triacylglycerols (TAGs) (Daum et al., 1998; Mlickova et al., 2004; Athenstaedt et al., 2006). To inhibit TAG remobilization, the gene encoding triglyceride lipase, ylTGL4 (Dulermo et al., 2013), was deleted. In addition, to push and pull TAG biosynthesis, ylDGA2, which encodes the major acyl-CoA: diacylglycerol acyltransferase (Beopoulos et al., 2012), and ylGPDl, which encodes glycerol-3-phosphate dehydrogenase (Dulermo and Nicaud, 2011), were overexpressed; this process produced strain

Î" 14 -Я

g, 10 -■ö

« 8 -

f 6-ь

4 -2 -0

15 30 60

C/N ratio

Fig. 4. Fatty acid production by Y. lipolytica W29 (□) and its ylHXKl -overexpression transformant (■) in YNB fructose medium with different C/N molar ratios. In red: the improvement in FA production (%; ratio of ylHXKl to WT). Lipid content was analyzed after 120 h of culture or after complete fructose consumption. Different C/N ratios were obtained by increasing fructose concentration. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)

Y3501. Then, the modified cassette for the efficient expression of the S. cerevisiae invertase gene (Lazar et al., 2013) was introduced, resulting in strain Y4498. The corresponding strain that also over-expressed ylHXKl was named Y4086 (Fig. 1; Table S1).

Batch cultures of Y3501 (Suc +), Y4498 (Suc ++), and Y4086 (Suc + +, ylHXKl) were grown in non-baffled Erlenmeyer flasks in YNB media that contained 60 g L_ 1 of glucose, fructose, or sucrose as a carbon source; the C/N ratio for all media was 60. Strains Y4498 and Y4086 produced 14.83 and 14.65 g L"1 of dry biomass, respectively, in the sucrose medium; these were the highest concentrations of biomass obtained in this experiment (Table 4). These two strains also had the highest levels of biomass yield when grown in the sucrose medium (0.39 and 0.36 gg-1 for Y4498 and Y4086, respectively); however, when grown in the glucose or fructose media, they had lower yields of total biomass (39-50% lower) and were less efficient at converting substrate into biomass (24-38% less efficient). This pattern was probably due to the lower osmotic pressure in the sucrose medium, which allows cells to better adapt to culture conditions (Lazar et al., 2011). When double-transformant Y4086 (overexpressing both ylHXKl and scSUC2) was grown in the sucrose medium, it had the best FA yield per g of biomass (0.42 g g_1) and the most efficient conversion of substrate into FAs (0.15 g g_ 1) observed in this experiment. Interestingly, ylHXKl overexpression increased FA production, which reached up to 6.15 g Lwhen transformants were grown in the sucrose medium, while concomitantly decreasing CA production.

Strains Y3501, Y4498 and Y4086, which were optimized for lipid accumulation, produced more C16:0 than the WT W29 (~30% versus ~ 16.5%); however, they displayed decreased levels of unsaturated C16 and C18. Nonetheless, oleic acid remained the major FA stored (42-45%). This finding suggests that, in these strains, A9 desaturase and (more likely) C16-C18 elongase activity become limiting factors. However, neither ylHXKl nor invertase overexpression significantly affected FA profiles (Table 5).

3.7. Lipid production by engineered strains grown in a sucrose medium in a bioreactor

To evaluate lipid production by the optimized strain Y3501 and the synergistic impact of scSUC2 and ylHXKl overexpression in Y4086, strains were grown in a sucrose medium in a 5-L fermenter (Fig. 5). Within 25 h, Y4086 was hydrolyzing sucrose at a rate of 5.28 g L_ 1 h~\ resulting in the rapid accumulation of glucose and fructose in the medium. In addition, fructose began to be consumed only when the glucose present was almost exhausted. Lipid content increased over time, reaching 25% of CDWat 96 h of fermentation. In contrast, up until 60 h, Y3501 hydrolyzed sucrose very slowly, at a rate of 0.35 g L_ 1 h~1; neither glucose nor fructose accumulated in

Parameters of FA, biomass and CA production of 120 h flask culture using Y lipolytica Y3501, Y4498 and Y4086 strains growing in YNB medium with glucose, fructose or sucrose (carbon source 60 g L_1, C/N 60).

Parameters

Glucose

Fructose

Sucrose

X gL-1 12.20 12.30 12.55 13.01 12.68 13.78 11.54 14.83 14.65

Yx/S g g -1 0.25 0.26 0.26 0.28 0.28 0.29 0.33 0.39 0.36

FA gL"1 4.88 4.92 5.27 4.55 4.18 5.51 3.69 5.49 6.15

Yfa/x g g"1 0.40 0.40 0.42 0.35 0.33 0.40 0.32 0.37 0.42

Yfa/s g g "1 0.10 0.10 0.11 0.10 0.09 0.12 0.10 0.13 0.15

CA gL"1 3.89 4.47 3.32 2.69 3.92 2.30 1.83 6.32 5.67

Symbols: X - dry biomass, FA - fatty acids, CA - citric acid, YX/S - yield of biomass from consumed substrate, YFA/S - yield of fatty acids from consumed substrate, YFA/X - yield of fatty acids from dry biomass. SD of all analysis did not exceed 5%.

Table 5

Fatty acid composition in Y. lipolytica strains growing 120 h in YNB medium with glucose, fructose or sucrose (carbon source 60 g L- 1, C/N 60).

Parameters Glucose Fructose Sucrose

Y3501 Y4498 Y4086 Y3501 Y4498 Y4086 Y3501 Y4498 Y4086

C16:0 29.8 30.0 30.1 29.5 30.1 29.4 29.6 29.1 29.1

C16:1(n-7) 4.1 3.9 3.8 5.7 5.7 5.1 5.4 4.6 4.4

C18:0 14.2 14.9 15.5 11.1 11.2 12.0 11.1 12.9 13.6

C18:1(n-9) 43.0 41.8 41.9 44.4 43.5 45.4 44.3 44.7 45.0

C18:2(n-6) 6.0 6.1 5.5 6.3 6.4 5.3 6.6 5.6 5.0

Others 2.9 3.3 3.2 3.0 3.1 2.8 3.3 3.1 2.9

150 -| 125 -100 -75 -50 -25 -0

Time (h)

I- 25 SS

- 10 - 5

150 -, 125 100 75 50 25 0

Time (h)

30 g 5»

25 £ 20

15 1 10 I ;

5 •§

H- 0 100

Fig. 5. Sucrose (♦), glucose (■), fructose (▲), CA (•), dry biomass ( x ), and FA (o) concentration during Y. lipolytica Y4086 (A) and Y3501 (B) growth in YNB medium with sucrose over the 96 h of culture in the bioreactor. SD of all analysis did not exceed 5%.

the medium. Then, following the increased expression of the XPR2 promoter, the hydrolysis rate climbed to 2.16 g L- 1 h-1 and fructose and glucose (to a lesser extent) began to accumulate (Fig. 5A). Y3501's hydrolysis rate was slightly slower than that of JMY2529, a W29 transformant containing the XPR2-SUC2 cassette (2.16 versus 2.50 g L-1 h-1; Lazar et al., 2013). Similarly, Y4086's hydrolysis rate was slower than that of JMY2531, a W29 transformant containing the pTEF-SUC2 cassette (5.28 versus 7.63gL-1h-1; Lazar et al., 2013). These results indicate that improved lipid biosynthesis and accumulation may affect invertase production.

Y4086 had a faster growth and reached the stationary phase at about 60 h, while Y3501 grew slowly and continuesly until the end of fermentation (Fig. 5). This finding shows that invertase production was a limiting factor. However, both strains had a similar final biomass of 34 g L-1 after 96 h. Y4086 began to secrete CA into the medium at a rate of 1.06 g L-1 h-1 after 36 h of culture, while Y3501 started to secrete CA at a rate of 0.77 g L-1 h-1 after 72 h of culture (Fig. 5). A similar pattern was observed for lipid accumulation: it increased as CA secretion increased. Y4086 produced significantly greater amounts of lipids (9.15 g L-1) than Y3501: total lipid production, FA production, and FA yield increased by 60% (Table S3). Altogether, these results indicate that, by optimizing fermentation conditions, CA could be redirected to lipid accumulation, thus further increasing lipid levels.

In addition, as noted earlier, cell morphology plays an important role in lipid accumulation (Fig. S4). The reduced lipid yield obtained from Y3501 is consistent with the observation that this strain demonstrated both yeast-form and filamentous-form growth; in contrast, Y4086 demonstrated only yeast-form growth. The overexpression of hexokinase in the latter strain inhibited filamentous-form growth and also led to the development of larger lipid bodies inside its cells.

The two strains differed in their biomass yields and their per-unit conversion of substrates into lipids. Y4086 was better than Y3501 at converting sugars into lipids (YFA/S of 0.063 g g-1 and 0.057 g g-1, respectively). However, Y4086 was less efficient than

Y3501 at converting substrates into biomass (YX/S of 0.24 gg-1 and 0.36 gg-1, respectively; Table S3). For instance, using a Polg derivative strain overexpressing ACC1 and DGA1, Tai and Step-hanopoulos (2013) were able to produce 28.5 g L-1 of biomass and a lipid yield of 0.195 gg-1 using 90 gL-1 of glucose. Similarly, Blazeck et al. (2014) obtained 20 g L-1 of biomass and a lipid yield of 0.25 g g-1 using 80 g.L-1 of glucose, however, lipid accumulation level in the fructose medium was 35% lower than in the glucose medium. These results support the idea that hexokinase activity is limiting since PO1f is derived from W29. Even if lipid accumulation levels and yields from sucrose were lower in our bioreactor experiment than in our flask experiments, this finding indicates that they could be improved by optimizing fermentation conditions.

4. Conclusions

As part of efforts to develop alternative methods of biofuel production, Y. lipolytica's lipid metabolism has become the target of many studies in recent years. Much of this research work seeks to decipher the de novo biosynthesis and accumulation of lipids. Efforts to optimize this species' biolipid production have also made it clear that expanding the yeast's substrate range is of great importance (Abghari and Chen, 2014). From an economic point of view, these substrates must be cheap and widely available. One such substrate is sucrose, which is a major component of molasses. Overexpression of S. cerevisiae invertase in Y. lipolytica allows the latter to efficiently utilize sucrose (Lazar et al., 2013). In the present study, we investigated another problem linked to sucrose utilization: namely differences in fructose utilization among Y. lipolytica strains. We determined that impaired fructose assimilation can be successfully eliminated by overexpressing hexokinase. This modification improved growth and fructose uptake as well as fructose-based lipid production in Y. lipolytica. In addition, the strains overexpressing hexokinase remained in yeast form throughout the

culturing period. Hexokinase overexpression combined with genetic modifications of the lipid metabolism system increased FA accumulation from fructose by up to 40% of CDW, a level similar to that seen in strains grown on glucose. However, the highly engineered strain Y4086 encountered additional limiting factors; one of these may have been the activity of C16:C18 elongase, as suggested by the high levels of C16:0. In addition, higher levels of lipid accumulation resulted when sucrose was used as a carbon source instead of its constituents (glucose and fructose); bioreactor cultures grown in a sucrose medium generated 9 g L_ 1 of lipids. However, Y. lipolytica's preferential consumption of glucose over fructose is a limiting factor that must be addressed in order to increase lipid production.

Acknowledgments

This work was funded by the Agence Nationale de la Recherche (Investissements d'avenir program; reference ANR-11-BTBR-0003). Z. Lazar received financial support from the European Union in the form of an AgreenSkills Fellowship (Grant agreement no. 267196; MarieCurie FP7 COFUND People Program). We thank Claude Gaillardin for critical comments on the manuscript. We would also like to thank Jessica Pearce and Lindsay Higgins for their language editing services.

Appendix A. Supporting information

Supplementary data associated with this article can be found in the online version at http://dx.doi.org/10.1016/j.ymben.2014.09.008.

References

Abghari, A., Chen, S., 2014. Yarrowia lipolytica as an oleaginous cell factory platform for production of fatty acid-based biofuel and bioproducts. Font. Energy Res. 2, 1-21.

Athenstaedt, K., Jolivet, P., Boulard, C., Zivy, M., Negroni, L., Nicaud, J.M., Chardot, T.,

2006. Lipid particle composition of the yeast Yarrowia lipolytica depends on the carbon source. Proteomics 6, 1450-1459.

Barth, G., Gaillardin, C., 1996. Yarrowia lipolytica. In: Wolf, K. (Ed.), Non Conventional Yeasts in Biotechnology, a Handbook. Springer-Verlag, Berlin, Germany, pp. 313-388.

Beopoulos, A., Mrozova, Z., Thevenieau, F., Dall, M.T., Hapala, I., Papanikolaou, S., Chardot, T., Nicaud, J.M., 2008. Control of lipid accumulation in the yeast Yarrowia lipolytica. Appl. Environ. Microbiol. 74, 7779-7789.

Beopoulos, A., Cescut, J., Haddouche, R., Uribelarrea, J.L., Molina-Jouve, C., Nicaud, J.M., 2009. Yarrowia lipolytica as a model for bio-oil production. Prog. Lipid Res. 48, 375-387.

Beopoulos, A., Haddouche, R., Kabran, P., Dulermo, T., Chardot, T., Nicaud, J.M., 2012. Identification and characterization of DGA2, an acyltransferase of the DGAT1 acyl-CoA:diacylglycerol acyltransferase family in the oleaginous yeast Yarrowia lipolytica. New insights into the storage lipid metabolism of oleaginous yeasts. Appl. Microbiol. Biotechnol. 93 (4), 1523-1537.

Browse, J., McCourt, P.J., Somerville, C.R., 1986. Fatty acid composition of leaf lipids determined after combined digestion and fatty acid methyl ester formation from fresh tissue. Anal. Biochem. 152, 141-145.

Blazeck, J., Hill, A., Liu, L., Knight, R., Miller, J., Pan, A., Otoupal, P., Alper, H.S., 2014. Harnessing Yarrowia lipolytica lipogenesis to create a platform for lipid and biofuel production. Nat. Commun., http://dx.doi.org/10.1038/ncomms4131.

Coelho, M.A.Z., Amaral, P.F.F., Belo, I., 2010. Yarrowia lipolytica: an industrial workhorse. (Book Series no. 2). In: Méndez-Villas, A (Ed.), Current Research, Technology and Education Topics in Applied Microbiology and Microbial BiotechnologyFormatex, Badajoz, p. 930.

Daum, G., Lees, N.D., Bard, M., Dickson, R., 1998. Biochemistry, cell biology and molecular biology of lipids of Saccharomyces cerevisiae. Yeast 14, 1471-1510.

da Silva, P.C., Horii, J., Santos Miranda, V., Gallera Brunetto, H., Ceccato-Antonini, S.R.,

2007. Characterization of industrial strains of Saccharomyces cerevisiae exhibiting filamentous growth induced by alcohols and nutrient deprivation. World J. Micro-biol. Biotechnol. 23, 697-704.

Dear, S., Staden, R., 1991. A sequence assembly and editing program for efficient management of large projects. Nucleic Acids Res. 19 (14), 3907-3911.

Dujon, B., Sherman, D., Fischer, G., Durrens, P., Casaregola, S., Lafontaine, I., De Montigny, J., Marck, C., Neuveglise, C., Talla, E., Goffard, N., Frangeul, L., Aigle, M., Anthouard, V., Babour, A., Barbe, V., Barnay, S., Blanchin, S., Beckerich, J.M., Beyne, E., Bleykasten, C., Boisrame, A., Boyer, J., Cattolico, L., Confanioleri, F., De Daruvar, A., Despons, L., Fabre, E., Fairhead, C., Ferry-Dumazet, H., Groppi, A., Hantraye, F., Hennequin, C., Jauniaux, N., Joyet, P., Kachouri, R., Kerrest, A., Koszul, R., Lemaire, M., Lesur, I., Ma, L., Muller, H.,

Nicaud, G.F., Nikolski, M., Oztas, S., Ozier-Kalogeropoulos, O., Pellenz, S., Potier, S., Richard, G.F., Straub, M.L., Suleau, A., Swennen, D., Tekaia, F., Wesolowski-Louvel, M., Westhof, E., Wirth, B., Zeniou-Meyer, M., Zivanovic, I., Bolotin-Fukuhara, M., Thierry, A., Bouchier, C., Caudron, B., Scarpelli, C., Gaillardin, C., Weissenbach, J., Wincker, P., Souciet, J.L., 2004. Genome evolution in yeasts. Nature 430 (6995), 35-44.

Dulermo, T., Nicaud, J.M., 2011. Involvement of the G3P shuttle and p-oxidation pathway in the control of TAG synthesis and lipid accumulation in Yarrowia lipolytica. Metab. Eng. 13 (5), 482-491.

Dulermo, T., Tréton, B., Beopoulos, A., Kabran Gnankon, A.P., Haddouche, R., Nicaud, J.M., 2013. Characterization of the two intracellular lipases of Y. lipolytica encoded by TGL3 and TGL4 genes: new insights into the role of intracellular lipases and lipid body organisation. Biochim. Biophys. Acta 1831 (9), 1486-1495.

Dyer, J.M., Chapital, D.C., Kuan, J.W., Mullen, R.T., Pepperman, A.B., 2002. Metabolic engineering of Saccharomyces cerevisiae for production of novel lipid compounds. Appl. Microbiol. Biotechnol. 59 (2-3), 224-230.

Ernandes, J.R., De Meirsman, C., Rolland, F., Winderickx, J., de Winde, J., Brandâo, R.L., Thevelein, J.M., 1998. During the initiation of fermentation overexpression of hexokinase PII in yeast transiently causes a similar deregulation of glycolisys as deletion of Tps1. Yeast 14 (3), 255-269.

Fickers, P., Le Dall, M.T., Gaillardin, C., Thonart, P., Nicaud, J.M., 2003. New disruption cassettes for rapid gene disruption and marker rescue in the yeast Yarrowia lipolytica. J. Microbiol. Methods 55, 727-737.

Fickers, P., Nicaud, J.M., Destain, J., Thonart, P., 2005a. Involvement of hexokinase Hxk1 in glucose catabolite repression of L1P2 encoding extracellular lipase in the yeast Yarrowia lipolytica. Curr. Microbiol. 50 (3), 133-137.

Fickers, P., Benetti, P.H., Waché, Y., Marty, A., Mauersberger, S., Smit, M.S., Nicaud, J.M., 2005b. Hydrophobic substrate utilisation by the yeast Yarrowia lipolytica, and its potential applications. FEMS Yeast Res. 5, 527-543.

Folch, J., Lees, M., Sloane-Stanley, G.H., 1957. A simple method for the isolation and purification of total lipides from animal tissues. J. Biol. Chem. 226, 497-509.

Gaillardin, C., Ribet, A.M., 1987. LEU2 directed expression of p-galactosidase activity and phleomycin resistance in Yarrowia lipolytica. Curr. Genet. 11, 369-375.

Groenewald, M., Boekhout, T., Neuvéglise, C., Gaillardin, C., van Dijck, P.W.M., Wyss, M., 2014. Yarrowia lipolytica: safety assessment of an oleaginous yeast with a great industrial potential. Crit. Rev. Microbiol. 40 (3), 187-206.

Guillaume, C., Delobel, P., Sablayrolles, J.M., Blondin, B., 2007. Molecular basis of fructose utilization by the wine yeast Saccharomyces cerevisiae: a mutated HXT3 allele enhances fructose fermentation. Appl. Environ. Microbiol. 73 (8), 2432-2439.

Herrero, A.B., López, M.C., Fernández-Lago, L., Domínguez, A., 1999. Candida albicans and Yarrowia lipolytica as aternative models for analysing budding patterns and germ tube formation in dimorphic fungi. Microbiology 145, 2727-2737.

Hurtado, C.A., Rachubinski, R.A., 1999. MHY1 encodes a C2H2-type zinc finger protein that promotes dimorphic transition in the yeast Yarrowia lipolytica. J. Bacteriol. 181, 3051-3057.

Lazar, Z., Walczak, E., Robak, M., 2011. Simultaneous production of citric acid and invertase by Yarrowia lipolytica SUC + transformants. Bioresour. Technol. 102 (13), 6982-6989.

Lazar, Z., Rossignol, T., Verbeke, J., Crutz-Le Coq, A.M., Nicaud, J.M., Robak, M., 2013. Optimized invertase expression and secretion cassette for improving Yarrowia lipolytica growth on sucrose for industrial applications. J. Ind. Microbiol. Biotechnol. 40 (11), 1273-1283.

Liccioli, T., Chambers, P.J., Jiranek, V., 2011. A novel methodology independent of fermentation rate for assessment of the fructophilic character of wine yeast strains. J. 1nd. Microbiol. Biotechnol. 38 (7), 833-843.

Mauersberger, S., Wang, H.J., Gaillardin, C., Barth, G., Nicaud, J.M., 2001. 1nsertional mutagenesis in the n-alkane-assimilating yeast Yarrowia lipolytica: generation of tagged mutations in genes involved in hydrophobic substrate utilization. J. Bacteriol. 183, 5102-5109.

Michely, S., Gaillardin, C., Nicaud, J.M., Neuvéglise, C., 2013. Comparative physiology of oleaginous species from the Yarrowia clade. PLoS One 8 (5), 1-10 (e63356).

Mironczuk, A.M., Furgala, J., Rakicka, M., Rymowicz, W., 2014. Enhanced production of erythritol by Yarrowia lipolytica on glycerol in repeated batch cultures. J. Ind. Microbiol. Biotechnol. 41 (1), 57-64.

Mlickova, K., Roux, E., Athenstaedt, K., d'Andrea, S., Daum, G., Chardot, T., Nicaud, J.M., 2004. Lipid accumulation, lipid body formation, and acyl coenzyme a oxidases of the yeast Yarrowia lipolytica. Appl. Environ. Microbiol. 70, 3918-3924.

Moeller, L., Zehnsdorf, A., Aurich, A., Barth, G., Bley, T., Strehlitz, B., 2012. Citric acid production from sucrose by recombinant Yarrowia lipolytica using semicontinuous fermentation. Eng. Life Sci. 13 (2), 163-171.

Müller, S., Sandal, T., Kamp-Hansen, P., Dalboge, H., 1998. Comparison of expression systems in the yeasts Saccharomyces cerevisiae, Hansenula polymorpha, Kluy-veromyces lactis, Schizosaccharomyces pombe and Yarrowia lipolytica. Cloning of two novel promoters from Yarrowia lipolytica. Yeast 14 (14), 1267-1283.

Nicaud, J.M., 2012. Yarrowia lipolytica. Yeast 29 (10), 409-418.

Petit, T., Gancedo, C., 1999. Molecular cloning and characterization of the Gene HXK1 encoding the hexokinase from Yarrowia lipolytica. Yeast 15, 1573-1584.

Rywinska, A., Rymowicz, W., 2010. High-yield production of citric acid by Yarrowia lipolytica on glycerol in repeated-batch bioreactors. J. Ind. Microbiol. Biotechnol. 37 (5), 431-435.

Rywinska, A., Rymowicz, W., Zarowska, B., Skrzypinski, A., 2010. Comparison of citric acid production from glycerol and glucose by different strains of Yarrowia lipolytica. World J. Microbiol. Biotechnol. 26, 1217-1224.

Sambrook, J., Russell, D.W., 2001. Molecular Cloning: a Laboratory Manual, 3rd ed. Cold Spring Harbor Laboratory Press, New York.

Schmittgen, T.D., Livak, K.J., 2008. Analyzing real-time PCR data by the comparative CT method. Nat. Protoc. 3 (6), 1101-1108.

Tai, M., Stephanopoulos, G., 2013. Engineering the push and pull of lipid biosynthesis in oleaginous yeast Yarrowia lipolytica for biofuel production. Metab. Eng. 15, 1-9.

Querol, A., Barrio, E., Huerta, T., Ramon, D., 1992. Molecular monitoring of wine fermentations conducted by active dry yeast strains. Appl. Environ. Microbiol. 58, 2948-2953.

Van de Velde, S., Thevelein, J.M., 2008. Cyclic AMP-protein kinase A and Snf1 signaling mechanisms underlie the superior potency of sucrose for induction of filamentation in Saccharomyces cerevisiae. Eucaryotic Cell 7 (2), 286-293.

Walczak, E., Robak, M., 2009. Growth on sucrose of Yarrowia lipolytica yeasts clones with invertase gene from Saccharomyces cerevisiae [in Polish]. Acta Sci. Pol., Biotechnol. 8 (4), 25-36.

Wojtatowicz, M., Rymowicz, W., 1991. Effect of inoculum on kinetics and yield of citric acid production on glucose media by Yarrowia lipolytica A-101. Acta Aliment. Pol. 41 (2), 137-143.

Wojtatowicz, M., Rymowicz, W., Robak, M., Zarowska, B., Nicaud, J.M., 1997. Kinetics of cell growth and citric acid production by Yarrowia lipolytica Suc+ transformants in sucrose media. Pol. J. Food Nutr. Sci. 6/47 (4), 49-54.

Xuan, J.W., Fournier, P., Declerck, N., Chasles, M., Gaillardin, C., 1990. Overlapping reading frames at the LYS5 locus in the yeast Yarrowia lipolytica. Mol. Cell. Biol. 10, 4795-4806.

Xue, Z., Sharpe, P.L., Hong, S.P., Yadav, N.S., Xie, D., Short, D.R., Damude, H.G., Rupert, R.A., Seip, J.E., Wang, J., Pollak, D.W., Bostick, M.W., Bosak, M.D., Macool, D.J., Hollerbach, D.H., Zhang, H., Arcilla, D.M., Bledsoe, S.A., Croker, K., McCord, E.F., Tyreus, B.D., Jackson, E.N., Zhu, Q., 2013. Production of omega-3 eicosapentae-noic acid by metabolic engineering of Yarrowia lipolytica. Nat. Biotechnol. 31, 734-740.