Applied Microbiology and Biotechnology

, Volume 96, Issue 4, pp 981–991

Characterization of triglyceride lipase genes of fission yeast Schizosaccharomyces pombe

Authors

  • Hisashi Yazawa
    • Bioproduction Research InstituteNational Institute of Advanced Industrial Science and Technology
  • Hiromichi Kumagai
    • ASPEX DivisionAsahi Glass Co. Ltd.
    • Bioproduction Research InstituteNational Institute of Advanced Industrial Science and Technology
Applied genetics and molecular biotechnology

DOI: 10.1007/s00253-012-4151-8

Cite this article as:
Yazawa, H., Kumagai, H. & Uemura, H. Appl Microbiol Biotechnol (2012) 96: 981. doi:10.1007/s00253-012-4151-8

Abstract

Triglycerides (TG) are major storage lipids for eukaryotic cells. In this study, we characterized three genes of fission yeast Schizosaccharomyces pombe, SPCC1450.16c, SPAC1786.01c, and SPAC1A6.05c, that show high homology to Saccharomyces cerevisiae TG lipase genes, TGL3, TGL4, and TGL5. Deletion of each gene increased TG content by approximately 1.7-fold compared to the parental wild-type strain, and their triple deletion mutant further increased TG content to 2.7-fold of the wild-type strain, suggesting that all three genes encode TG lipase and are functioning in S. pombe. The triple deletion mutant showed no growth defect in rich and synthetic medium, but its growth was sensitive to cerulenin, an inhibitor of fatty acid synthesis. This growth defect by cerulenin was restored by adding oleic acid in media, suggesting that these genes were involved in the mobilization of TG in S. pombe. When ricinoleic acid was produced in the triple mutant by introducing CpFAH12 fatty acid hydroxylase gene from Claviceps purpurea, percent composition of ricinoleic acid increased by 1.1-fold compared to the wild-type strain, in addition to a 1.6-fold increase in total fatty acid content per dry cell weight (DCW). In total, the ricinoleic acid production per DCW increased by 1.8-fold in the triple deletion mutant.

Keywords

Schizosaccharomyces pombeYeastTriglyceride lipaseTriglycerideFatty acid

Introduction

We have been studying the heterologous production of polyunsaturated fatty acids (PUFAs) and fatty acid derivatives in model yeasts Saccharomyces cerevisiae and Schizosaccharomyces pombe. So far, we have produced linoleic (C18:2) and α-linolenic (C18:3n − 3) acids (Kainou et al. 2006), so-called essential fatty acids for human, and dihomo-gamma-linolenic acid (C20:3n − 6), a PUFA with anti-inflammatory effect, in S. cerevisiae (Yazawa et al. 2007). Recently, we are attempting to produce ricinoleic acid (12-hydroxy-octadeca-cis-9-enoic acid, C18:1–OH), an important natural product as a petrochemical replacement in a variety of industry, in S. pombe by introducing oleate Δ12-hydroxylase gene (FAH12). We chose S. pombe instead of S. cerevisiae because the amount of oleic acid (C18:1n − 9), a precursor of ricinoleic acid, is much higher in S. pombe and it reaches to around 75 % of its total fatty acids. Indeed we succeeded to produce more than 50 % ricinoleic acid content to total fatty acids in S. pombe (Holic et al. 2012). In all cases described earlier, the next step to take is to increase their lipid content to apply our knowledge to industrial use because they do not accumulate storage lipids and usually only 5 to 7 % of lipids are produced per dry cell weight (DCW). Since triglycerides (TGs) are the major storage lipids, we are focusing on the study of their metabolism in an effort to increase TG content per cell by removing TG lipase genes and reducing TG degradation in S. pombe.

In S. cerevisiae, a series of recent studies have identified three TG lipase genes (TGL3, TGL4, and TGL5). Whereas enzymes involved in TG synthesis in S. cerevisiae appear both structurally and functionally conserved to mammalian cells, the level of sequence conservation for TG-degrading enzymes is less pronounced. The first characterized S. cerevisiae Tgl1p displays more than 30 % sequence identity to mammalian acid lipases, but it turned out to function as a steryl ester hydrolase rather than a TG lipase in vivo (Jandrositz et al. 2005; Koffel et al. 2005). The first S. cerevisiae TG lipase identified was Tgl3p (Athenstaedt and Daum, 2003), but it lacks any significant structural homology to known lipases in higher eukaryotes. Subsequently, the second and the third TG lipase genes, TGL4 and TGL5, were identified based on the homology to Tgl3p (Athenstaedt and Daum, 2005).

In fission yeast S. pombe, however, studies on metabolism are less advanced than those in S. cerevisiae. S. pombe has been extensively used to study the molecular mechanisms of many aspects of cell physiology, including cell cycle and stress signaling, but little work has been done in the area of lipid cell biology. Its genome sequences predicted candidate genes based on sequence similarity, but no experimental evidence has been obtained so far. In this study, we constructed deletion mutants of candidate genes and examined their impact on TG content to experimentally identify TG lipase genes in S. pombe. We also provided evidence for their biological significance in TG mobilization.

Materials and methods

Strains, media, and growth

Plasmid manipulations were carried out using E. coli DH5α (F-endA1 hsdR17 supE44 thi-1 recA1 gyrA96 relA1 ∆(argF-lacZ)U169 ϕ80 ∆(lacZ)M15) grown in Luria–Bertani broth, supplemented with ampicillin. The S. pombe strain ARC010-1 (h-leu1-32 ura4-Δ18) (Idiris et al. 2006) was used for this study. S. pombe strains were grown in YES rich media and EMMS synthetic media or EMMS dropout media depending on the selective pressure required to maintain plasmids (Alfa et al.1993). EMMS was used for normal growth, and nitrogen-limited minimal medium (EMM-C/N3) (Holic et al. 2012) was used for the production of ricinoleic acid because our previous experiments indicated that the fatty acid content was higher in the cells grown in EMM-C/N3. EMM medium contained 0.5 % (w/v) ammonium chloride and 2 % (w/v) glucose, but in nitrogen-limited EMM medium (EMM-C/N3), the concentration of ammonium chloride was reduced to 0.1 % and the concentration of glucose was increased to 10 %.

Yeast cells were transformed by using Frozen-EZ Yeast Transformation II kit (Zymo Research, CA, USA) according to the protocol recommended by the manufacturer. For the assay of fatty acids, ARC010-1 strains were pre-cultured overnight at 30 °C in EMMS medium and the resultant cultures were inoculated into fresh EMMS at OD600 = 0.2 unless otherwise specified, and the cultures were shaken at 40 rpm. For the assay of ricinoleic acid production in the CpFAH12 transformant, EMM-C/N3-Leu was used instead of EMMS. Growth in liquid media was monitored by measuring the turbidity of the cells at OD600.

For automatic detection of growth, ARC010-1 disruptants grown in EMMS overnight at 30 °C were inoculated into fresh EMMS at OD630 = 0.05 and the growth curves in liquid media were obtained by monitoring the turbidity of the cells at 630 nm with an automatic detector (Bio-Plotter, Toyo-Sokki, Japan).

Plasmid construction

The standard techniques of DNA manipulation used in this study are described in Sambrook and Russell (2001). The FAH12 gene from Claviceps purpurea (CpFAH12, Meesapyodsuk and Qiu 2008) was cloned under the nmt1 promoter of S. pombe in pSL10 integration-type plasmid to construct pL2428-9 (Holic et al. 2012). pSL10 is designed to integrate between the leu1-32 and topo2+ loci of chromosome II (Idiris et al. 2010).

Disruption of lipase genes

The open reading frame (ORF) regions of ptl1 (SPCC1450.16c) (+10 bp to +1,658 bp), ptl2 (SPAC1786.01c) (+21 bp to +2,080 bp), and ptl3 (SPAC1A6.05c) (+10 bp to +1534 bp) were removed from the genome of ARC010-1 by the two-step pop-in and pop-out method of ura4 marker, that was named “latency to universal rescue system” or “Latour system” (Hirashima et al. 2006). Their proper disruptions were verified by colony PCR analysis (data not shown).

Fatty acid analysis

Total fatty acid contents were determined by gas chromatographic analysis as described (Kainou et al. 2006). Fatty acid analysis was performed by the application of 0.2-μl aliquots to a gas chromatograph (GC2010, Shimadzu, Japan) equipped with a TC-70 capillary column (30 × 0.25 mm i.d., GL Sciences, Japan) under temperature programming (160–234 °C at 4.5 °C/ min increments). Fatty acid composition was calculated based on the area of each peak, and the amount was determined by comparison with the methylheptadecanoate (C17:0) standard.

Thin-layer chromatography of lipids

Total lipids were extracted from cells by the procedure of Bligh and Dyer (1959). Cells (10 ml of culture each) of S. pombe grown to a stationary phase for 5 days were washed once with water, suspended in 2 ml of methanol, and disrupted by vortexing for 90 s (three times for 30 s each time) together with 0.5 g of glass beads (Brown ϕ = 0.4 to 0.5 mm). Then, 1 ml of chloroform was added to the extract and the tubes were kept shaken for 1 h at room temperature. Liquids were recovered by spinning tubes at 3,000 rpm for 5 min. Glass beads were washed twice with 1.5 ml of chloroform/methanol (1:2) mixture and the liquids were combined into the first solution. Lipids were extracted in chloroform by vortexing the solution with 2 ml each of chloroform and water, and the chloroform phase was separated by spinning them at 1,000 rpm for 5 min. The extracted chloroform was vaporized under nitrogen gas and the dried lipids were suspended in 200 ul of iso-propanol containing 1 % Triton-X100. Samples were kept stored at −80 °C until processed for lipid analysis.

Lipid compositions were analyzed by thin-layer chromatography (TLC). Lipids corresponding to 2.5 OD equivalent cells were separated by TLC (silica gel 60 plate, Merck) using the solvent systems chloroform/acetone/methanol/acetic acid/water (50:20:10:10:5 per volume; developed to one half of the total distance) first for the separation of phospho lipids and then hexane/diethylether/acetic acid (80:40:1 per volume; total distance) for the separation of neutral lipids. Separated lipids were visualized on TLC plates by post-chromatographic staining. Plates were dipped for 10 s into a developing reagent consisting of 4.65 g of MnCl24H2O, 450 ml of methanol, and 30 ml of concentrated sulphuric acid per liter, briefly dried, and heated to 120 °C for 30 min. Subsequently, separated lipids were quantified by densitometric scanning with LAS-1000 image analyzer (Fuji film).

Triglyceride analysis

Cellular total lipids were prepared as described earlier. The amount of triglyceride was determined spectrophotometrically by using the L-type WAKO TG-M kit (Wako, Japan) according to the protocol of the manufacturer. The amount of FA in TG was calculated based on the factor derived from triolein (glycerol trioleate) that was assayed with both gas chromatography and this kit. When the same amount of triolein was measured with the L-type WAKO TG-M kit (TG assay) and gas chromatography (FA assay), the value of TG measured by the TG assay kit was 1.18-fold higher than the value of FA measured by gas chromatography.

Fluorescence microscopy

The S. pombe strains grown to a stationary phase in EMMS at 30 °C were harvested by centrifugation, washed once with PBS buffer (0.137 mM NaCl, 2.68 μM KCl, 10.1 μM Na2HPO4, 1.76 μM KH2PO4), and stained in Nile Red solution (5 μg/ml in PBS) for 10 min in the dark. Nile Red fluorescence images of living cells were detected with an Olympus BX51 fluorescent microscope system in the combination of a 460–495-nm band pass exciter filter (BP460-495 IF) and a 510-nm-long pass barrier filter (BA510 IF) (WIB3 mirror unit). Pictures were taken with a DP30BW CCD camera (Olympus, Japan).

Results

Identification of TG lipase genes in S. pombe

The search of S. pombe genome sequence (Wood et al. 2002) identified three sequences [SPCC1450.16c (ptl1), SPAC1786.01c (ptl2), and SPAC1A6.05c (ptl3)], which encode proteins with the highest degree of homology to S. cerevisiae TG lipase genes (TGL3, TGL4, and TGL5), here designated, and subsequently confirmed, as ptl1, ptl2, and ptl3 for S. pombeTG lipase, respectively. Figure 1 shows their phylogenetic relationships with the S. cerevisiae TG lipase gene products (Tgl3p, Tgl4p, and Tgl5p). Among them, ptl2p and ptl3p were similar to S. cerevisiae Tgl4p and Tgl5p (Tgl4p and Tgl5p are 55 % homologous to each other).
https://static-content.springer.com/image/art%3A10.1007%2Fs00253-012-4151-8/MediaObjects/253_2012_4151_Fig1_HTML.gif
Fig. 1

Phylogenetic tree of putative lipase genes. Homology was analyzed by ClustalW (protein weight matrix, BLOSUM; clustering, neighbor joining; Thompson et al., 1994) and the tree was drawn by using “Tree view” (Page, 1996). The scale bar indicates genetic distance. Numbers at the branch points of the phylogenetic tree indicate bootstrap values. ptl1p, ptl2p, and ptl3p indicate gene products of SPCC1450.16c (ptl1), SPAC1786.01c (ptl2), and SPAC1A6.05c (ptl3) genes of S. pombe, respectively. Tgl3p, Tgl4p, and Tgl5p indicate proteins of S. cerevisiae triglyceride lipase genes TGL3, TGL4, and TGL5, respectively

SPCC1450.16c (ptl1), SPAC1786.01c (ptl2), and SPAC1A6.05c (ptl3) genes predict proteins of 545, 630, and 483 amino acids, respectively, whereas Tgl4p encodes a protein of 910 amino acids. Figure 2 shows the comparison of their deduced amino acid sequences with S. cerevisiae Tgl4p. The region corresponding to the patatin-like domain in Tgl4p (aa 282 to aa 483, underlined in Fig. 2) was highly conserved in them, and their identities (homologies) to Tgl4p in patatin-like domains are 34.9 % (46.2 %), 40.5 % (54.2 %), and 47.5 % (61.0 %), respectively. Since the patatin domains are conserved in all three S. cerevisiae TG lipases and are also characteristic for mammalian adipose triglyceride lipase (the major lipase acting on TG in mouse adipocytes) (Zimmermann et al. 2004) and its Drosophila ortholog Brummer lipase (Grönke et al. 2005), homology in this region strongly indicates that ptl1, ptl2, and ptl3 encode lipase. Furthermore, the consensus sequence characteristic for lipolytic enzymes (GXSXG) (Athenstaedt and Daum 2005) was completely conserved in ptl2p and ptl3p. Even the residues corresponding to X were identical between ptl3p and Tgl4p. However, the central serine residue, which was demonstrated to be essential for catalytic activity of Tgl4p (Kurat et al. 2006), was not conserved in ptl1p.
https://static-content.springer.com/image/art%3A10.1007%2Fs00253-012-4151-8/MediaObjects/253_2012_4151_Fig2_HTML.gif
Fig. 2

Comparison of the deduced amino acid sequences of S. pombe SPCC1450.16c (ptl1), SPAC1786.01c (ptl2), and SPAC1A6.05c (ptl3) gene products with S. cerevisiae Tgl4p. Numbers on the right indicate the positions of amino acid residues. Compared to the sequence of Tgl4p, identical amino acid residues are shaded black and homologous amino acid residues are shaded gray. The lipolytic motifs (GXSXG) characteristic for lipolytic enzymes (Athenstaedt and Daum, 2005) are highlighted with a box, and the region corresponding to the patatin-like domain of Tgl4p (aa 282–483) is underlined

Deletion of ptl1, ptl2, and ptl3 resulted in an increase in TG content

To experimentally address whether TG lipase candidates ptl1, ptl2, and ptl3 are actually involved in TG degradation in S. pombe, their deletion mutants were constructed on the same genetic background of ARC010-1 haploid strain by homologous recombination as described in “Materials and methods”. All mutants were viable at 30 °C on rich and EMMS synthetic media, and we were unable to observe any obvious morphological changes in the deletion mutants under a light microscope (data not shown). The deletion mutants were grown to a stationary phase in EMMS, and TG and FA contents were measured by TG assay kit and gas chromatography, respectively (Fig. 3). In order to combine TG and FA data in Fig. 3, the amount of FA in TG was re-calculated from the value of TG based on the factor obtained from triolein that was assayed by both gas chromatography (FA assay) and the L-type WAKO TG-M kit (TG assay).
https://static-content.springer.com/image/art%3A10.1007%2Fs00253-012-4151-8/MediaObjects/253_2012_4151_Fig3_HTML.gif
Fig. 3

Cellular triglycerides (TG) and fatty acids (FA) levels of the wild-type strain ARC010-1 and its deletion mutants of ptl1, ptl2, and ptl3. Cellular levels of TG (black bars) and FA (entire bars including black and white bars) in the wild-type ARC010-1 and its single, double, and triple deletion mutants of ptl1, ptl2, and ptl3 are indicated as the amount in terms of μg of FA/mg of DCW. The deletion mutants of Δptl1, Δptl2, and Δptl3, and the wild-type strain (WT) were grown in EMMS for 5 days at 30 °C. Total fatty acid contents were determined by gas chromatographic analysis. The amount of TG was determined spectrophotometrically by using the L-type WAKO TG-M kit. The amount of FA in TG was calculated from the value of TG based on the factor obtained from triolein that was assayed by both gas chromatography (assay of FA) and the L-type WAKO TG-M kit (assay of TG). Values are means of three independent experiments

TG contents in ptl1, ptl2, and ptl3 single deletion mutants were increased by 1.67, 1.67, and 1.70-fold, respectively. The increased levels of TG were similar in the three single deletion mutants, suggesting that all of them have similar TG lipase activities. Most of the FA increase was attributed to the increase in TG content because the increase in FA derived from other than TG was small in these mutants (from 39 μg/mg of DCW (WT) to 46, 42, and 47 μg/mg of DCW, respectively). TLC analysis further confirmed the specific increase in TG content in the deletion mutants (Fig. 4a). Densitometric measurement showed that the amounts of major phospholipid [phosphatidylethanolamine (PE) and phosphatidylcholine (PC)] and ergosterol (ES) were not affected and sterol ester (SE) was slightly affected, but the amount of TG was clearly increased in all deletion mutants (Fig. 4b). The increase in TG content in all three single deletion mutants of ptl1, ptl2, and ptl3, together with their high sequence similarity to S. cerevisiae TGL genes, strongly suggested that they encode TG lipases.
https://static-content.springer.com/image/art%3A10.1007%2Fs00253-012-4151-8/MediaObjects/253_2012_4151_Fig4_HTML.gif
Fig. 4

Analysis of the lipid patterns of the single deletion mutants of Δptl1, Δptl2, and Δptl3. a Analysis of the lipid patterns by TLC. The single deletion mutants of Δptl1, Δptl2, Δptl3 and the wild-type strain (WT) were grown to a stationary phase as described in the legend of Fig. 3. Lane 1: lipid extract of total cells of the parental wild-type (WT) strain ARC010-1. Lanes 2, 3, and 4: lipid extracts of total cells of Δptl1, Δptl2, and Δptl3 single deletion mutants, respectively. Positions of standards containing ergosteryl oleate (SE, sterol ester), triolein (TG), ergosterol (ES), phosphatidylethanolamine (PE), and phosphatidylcholine (PC) are indicated on the left. Solvent systems: chloroform/acetone/methanol/acetic acid/water (50:20:10:10:5 per volume; developed to one half of the total distance) first for the separation of phospholipids, and then hexane/diethylether/acetic acid (80:40:1 per volume; to the full distance) for the separation of neutral lipids. Post-chromatographic staining was performed as described in “Materials and methods”. b Densitometric measurement of each lipid component on the TLC plate. The relative amount of SE, TAG, ES, PE, and PC to the wild-type strain was indicated, respectively. Total means the summation of all components. Equal amounts of lipid extracts were applied on TLC based on the OD600 values of cultures

Characterization of double and triple deletion mutants of ptl1, ptl2, and ptl3

To determine whether the functions of ptl1p, ptl2p, and ptl3p are redundant or not, we constructed their double and triple deletions by using a repetitive homologous recombination of Latour system (Hirashima et al. 2006). Like single mutants, all double and triple mutants were viable at 30 °C, and no obvious morphological changes were observed under a light microscope (data not shown), indicating that even their triple deletions did not show synthetic lethality.

Their TG and FA contents are summarized in Fig. 3. As we had expected, the TG contents were further increased in the double deletion of ptl1 plus ptl2 and ptl2 plus ptl3 (2.85 and 2.59-fold), but not in the ptl1 plus ptl3 deletion mutant (1.70-fold, similar to the level of single mutants). The TG content level in the triple mutant (2.68-fold) was similar to that of the double deletion of ptl1 plus ptl2 and ptl2 plus ptl3. In these mutants, the percentage of FA derived from TG increased to around one third (34 to 38 %) of the total FA.

The fluorescence microscopic image of the triple deletion mutant (Fig. 5) agreed well with the result of the TG measurement, and both the number and intensity of the lipid particles in the triple deletion cells were significantly higher than those in the wild-type cells. TLC analysis further verified that the stronger Nile Red fluorescence observed in the triple mutant (Fig. 5) was due to the increase in TG, but not steryl esters (SE), other neutral lipids in lipid particles, and the amount of major phospholipids (PE and PC) was not affected (Fig. 6).
https://static-content.springer.com/image/art%3A10.1007%2Fs00253-012-4151-8/MediaObjects/253_2012_4151_Fig5_HTML.gif
Fig. 5

Lipid particle formation in the triple deletion mutant of ptl1, ptl2, and ptl3. The triple mutant of Δptl1, Δptl2, and Δptl3, and the wild-type strain (WT) were grown to a stationary phase as described in the legend of Fig. 3, and Nile Red fluorescent images (right panels) were taken with fluorescent microscopy. Differential interference contrast (DIC) image of the wild-type strain (a) and its Nile Red-stained image (b). DIC image of the triple mutant of Δptl1, Δptl2, and Δptl3 (c) and its Nile Red-stained image (d). Size bar, 10 μm

https://static-content.springer.com/image/art%3A10.1007%2Fs00253-012-4151-8/MediaObjects/253_2012_4151_Fig6_HTML.gif
Fig. 6

Analysis of the lipid pattern in the triple mutant. a Analysis of the lipid patterns by TLC. The triple mutant of Δptl1, Δptl2, and Δptl3 and the wild-type strain (WT) were grown to a stationary phase as described in the legend of Fig. 3. Lane 1: lipid extract of total cells of the wild-type strain. Lane 2: triple deletion mutant of Δptl1, Δptl2, and Δptl3. Positions of standards containing ergosteryl oleate (SE, sterol ester), triolein (TG), ergosterol (ES), phosphatidylethanolamine (PE), and phosphatidylcholine (PC) are indicated on the left. Solvent systems: chloroform/acetone/methanol/acetic acid/water (50:20:10:10:5 per volume; developed to one third of the total distance) first for the separation of phospholipids and then hexane/diethylether/acetic acid (80:40:1 per volume; to the full distance) for the separation of neutral lipids. b Densitometric measurement of each lipid component on the TLC plate. The relative amount of SE, TAG, ES, PE, and PC to the wild-type strain was indicated, respectively. Total means the summation of all components. Equal amounts of lipid extracts were applied on TLC based on the OD600 values of cultures

Biological significance of ptl1, ptl2, and ptl3 genes in TG metabolism

To examine whether the TG lipase activities of ptl1p, ptl2p, and ptl3p, predicted from the analysis of the triple deletion mutant, are relevant in vivo, we addressed the mobilization of cellular TG by examining the growth of the triple mutant in the presence of cerulenin, an inhibitor of fatty acid synthesis in yeast (Inokoshi et al. 1994). Whereas the triple mutant grew like the wild-type strain in the absence of celurenin, its growth was severely retarded in the presence of celurenin (Fig. 7a). In contrast to the triple mutant, most of the lipid particles disappeared in the wild-type strain after growing for 8 h in the presence of cerulenin (Fig. 8, lower panels). The difference was obvious even in the DIC images, and unlike the usual rough cell surface that reflects the formation of lipid particles, the wild-type cells showed a smooth cell surface (Fig. 8, see WT cells in the presence of cerulenin). Its hypersensitivity to cerulenin may be caused by the failure of this mutant to mobilize TG depots under the conditions of fatty acid depletion due to the blocking of de novo fatty acid synthesis by the inhibition of Fas2p function with cerulenin (see Fig. 9). Interestingly, the level of cerulenin sensitivity of the single and the double deletion mutants was correlated with the increase of their TG levels as shown in Fig. 3. The single mutants were less sensitive to cerulenin and the double mutant of ptl1 and ptl3, which showed almost the same level of TG increase as the single deletion mutant, also showed similar cerulenin sensitivity to the single deletion mutant. The other double mutants, Δptl1 plus Δptl2 and Δptl2 plus Δptl3, showed similar sensitivity as the triple mutant (data not shown).
https://static-content.springer.com/image/art%3A10.1007%2Fs00253-012-4151-8/MediaObjects/253_2012_4151_Fig7_HTML.gif
Fig. 7

Growth of the triple mutant in the presence of cerulenin. a The wild-type strain (circle) and the triple mutant (triangle) were grown in EMMS in the absence (closed symbols) or presence (open symbols) of 0.2 μg/ml of cerulenin at 30 °C. The turbidity of the cells was monitored at 630 nm by an automatic detector (Bio-Plotter, Toyo-Sokki, Japan). Arrow indicates the position where cells were sampled for the fluorescent microscope observations shown in Fig. 8. b In addition to the presence of celurenin (open circles and open triangles, respectively), the wild-type and triple deletion strains were grown in EMMS in the presence of oleic acid (0.125 mM) together with celurenin (0.2 μg/ml) (closed circles and closed triangles). Furthermore, restoration of growth was examined by adding oleic acid at t = 15 h (indicated by an arrow) to the triple deletion mutant grown in the presence of celurenin (closed squares)

https://static-content.springer.com/image/art%3A10.1007%2Fs00253-012-4151-8/MediaObjects/253_2012_4151_Fig8_HTML.gif
Fig. 8

Lipid particle formation in the wild-type strain and the triple deletion mutant grown in the presence of cerulenin. The triple mutant of Δptl1, Δptl2, and Δptl3 and the wild-type strain (WT) were grown with or without cerulenin for 8 h as described in the legend of Fig. 7a, and differential interference contrast (DIC, left panels) and Nile Red fluorescent images (right panels) were taken. Size bar, 10 μm

https://static-content.springer.com/image/art%3A10.1007%2Fs00253-012-4151-8/MediaObjects/253_2012_4151_Fig9_HTML.gif
Fig. 9

Predicted model for the TG lipase function in S. pombe. Lipid metabolism pathway was depicted in relation to glycolytic pathway and TCA cycle. TG lipases are predicted to function for mobilization of fatty acids. Cerulenin inhibits the function of Fas2p and prevents the de novo formation of fatty acids. G-3-P glycerol-3-phosphate, Lyso-PA lyso-phosphatidic acid, PA phosphatidic acid, DG diglyceride, TG triglyceride

To further confirm this point, the triple mutant was grown in the presence of cerulenin but with a supply of exogenous oleic acid at the same time. Under this condition, its growth inhibition was dramatically restored almost to the level of cells grown in the absence of cerulenin (Fig. 7b, see closed triangles). Furthermore, the hampered growth by cerulenin was partially restored by adding oleic acid at t = 15 h in the medium (Fig. 7b, see closed squares). Taken together, these results clearly demonstrated that ptl1, ptl2, and ptl3 genes were functioning for vital TG mobilization in S. pombe.

Expression of CpFAH12 gene in the triple deletion mutant

Recently, we attempted to produce ricinoleic acid in S. pombe by introducing an oleate Δ12-hydroxylase gene of C. purpurea (CpFAH12). Although we succeeded to produce more than 50 % ricinoleic acid content to total fatty acids in S. pombe (Holic et al. 2012), we need to increase its lipid content to apply our knowledge to industrial use because S. pombe does not accumulate storage lipids. Since TGs are the major storage lipids, we addressed if this triple mutant is suitable for the increased production of a heterologous gene product. The hydroxylase gene (CpFAH12), which converts oleic acid (C18:1n − 9) to ricinoleic acid (C18:1–OH), was integrated in the wild-type strain and the triple deletion mutant, and their ricinoleic acid production was compared. Table 1 shows the fatty acid composition of total FA (%) and FA content (μg/mg of DCW) in the triple mutant expressing CpFAH12 gene under the S. pombe nmt1 promoter. Compared to the wild-type strain, levels of C18:1 decreased by 7.71 %, and in contrast C18:1–OH and C18:2, a by-product of hydroxylase function, increased by 3.86 and 1.55 %, respectively, in the triple deletion mutant (compare lines 3 and 4). Thus, a combination of the increase in the ricinoleic acid composition (1.13-fold) and the increase in the total fatty acid content per cell (1.61-fold) in the triple mutant resulted in the 1.83-fold increase in ricinoleic acid production per DCW.
Table 1

Fatty acid composition of total fatty acids in the triple mutant expressing CpFAH12 gene

Line

Host

Expressed gene

Fatty acid composition in weight %

Total FA (μg)/mg of DCW

[second row: fatty acid contents (μg/mg of dry cell weight)]

C12:0

C14:0

C16:0

C16:1

C16:2

C18:0

C18:1

C18:2

C18:1–OH

1

WT

Vector

0.67 ± 0.06

0.44 ± 0.04

11.07 ± 0.57

3.21 ± 0.40

0.00

5.26 ± 0.42

77.94 ± 0.49

0.00

0.00

46.51 ± 9.17

[0.31 ± 0.03]

[0.20 ± 0.02]

[5.12 ± 0.75]

[1.48 ± 0.11]

[0.00]

[2.47 ± 0.68]

[36.27 ± 7.37]

[0.00]

[0.00]

2

Triple mutant

Vector

0.70 ± 0.04

0.47 ± 0.04

10.48 ± 0.25

3.24 ± 0.23

0.00

6.85 ± 0.23

76.80 ± 0.15

0.00

0.00

62.94 ± 3.97

[0.44 ± 0.06]

[0.30 ± 0.05]

[6.37 ± 0.58]

[1.98 ± 0.30]

[0.00]

[4.74 ± 0.16]

[48.16 ± 2.90]

[0.00]

[0.00]

3

WT

CpFAH12

0.55 ± 0.02

0.44 ± 0.02

10.60 ± 0.35

1.07 ± 0.09

0.13 ± 0.00

9.83 ± 0.14

40.67 ± 0.95

6.31 ± 0.34

29.06 ± 0.33

77.37 ± 2.20

[0.43 ± 0.01]

[0.34 ± 0.01]

[8.20 ± 0.04]

[0.82 ± 0.05]

[0.10 ± 0.00]

[7.61 ± 0.33]

[35.48 ± 1.63]

[4.88 ± 0.12]

[22.48 ± 0.39]

4

Triple mutant

CpFAH12

0.67 ± 0.01

0.43 ± 0.02

10.87 ± 0.14

1.10 ± 0.08

0.18 ± 0.02

11.40 ± 0.20

32.96 ± 1.14

7.86 ± 0.21

32.93 ± 1.11

124.56 ± 5.50

[0.84 ± 0.03]

[0.54 ± 0.04]

[13.53 ± 0.50]

[1.37 ± 0.15]

[0.23 ± 0.03]

[14.19 ± 0.50]

[41.03 ± 1.50]

[9.80 ± 0.63]

[41.04 ± 2.91]

The wild-type strain ARC010-1 and its triple deletion mutant expressing CpFAH12 gene were pre-cultured in EMM-C/N3-Leu in the absence of thiamine at 37 °C for 1 day. Then, the cells were inoculated in the same medium at OD600 = 2 and incubated at 25 °C for 5 days. ARC010-1 and its triple mutant integrating pL2428-9 (CpFAH12, leu1marker) were used to examine the fatty acid composition. An empty vector pSL10 (with the nmt1 promoter and a leu1 marker) was used as control. Fatty acid compositions in weight % are shown in the first row and fatty acid contents (mg/mg of DCW) are shown in the second row. Values are means of three to six experiments using the independently obtained transformants. Error range represents SD

Discussion

Studies on metabolism are rare in S. pombe compared to those in S. cerevisiae. To understand the metabolism of TGs, major storage lipids, and apply it to the production of useful FAs in S. pombe, we attempted to identify its TG lipase genes. The predicted lipase genes were selected from the genome database, and their lipase activities were addressed by constructing their deletion mutants. In S. cerevisiae, three genes have been identified to encode TG lipases (TGL3, TGL4, and TGL5) (Athenstaedt and Daum, 2003; Athenstaedt and Daum, 2005), and it was demonstrated that the single deletion of TGL3 or TGL4 increased TG content by 2.39-fold and 1.73-fold, respectively, but deletion of TGL5 did not (0.98-fold) (Athenstaedt and Daum, 2005).

In S. pombe, three genes also have high homology to S. cerevisiae TG lipase genes, and we showed that their disruptions affected the TG and fatty acid contents. Their single deletions showed a similar level of increase in TG content, suggesting that they are functioning equally. Their additive effects were observed in the double deletion mutants of ptl1 plus ptl2 and ptl2 plus ptl3 (2.85- and 2.59-fold increase, respectively), but not in ptl1 plus ptl3 deletion mutant (1.70-fold, similar to the level of single mutants). The TG level in the triple mutant (2.68-fold increase) was similar to that of Δptl1 plus Δptl2 and Δptl2 plus Δptl3, also supporting that the effect of ptl1 plus ptl3 deletions was not additive.

The reason for the non-additive effect of ptl1p plus ptl3p is not clear, but one possibility is the compensation of their expressions. For example, the expression of adh4, the minor alcohol dehydrogenase gene that is not normally produced, is induced when adh1 is disrupted, and it supports fermentative growth in the absence of Adh1p (Sakurai et al. 2004). Thus, the expression levels of ptl1, ptl2, and ptl3 were examined by RT-PCR and normalized based on the act1 expression level to address whether the deletion of one gene affects the expression of other genes to compensate each other (data not shown). The expression levels of ptl1, ptl2, and ptl3 were about 5, 60, and 40 % of act1 in the wild-type strain, and their expression levels were basically not affected by gene disruption of other lipase genes, indicating that their expressions were not compensating each other. Furthermore, in spite of the lower-level expression of ptl1 and a partial conservation of lipolytic motif in ptl1p, ptl1 is not considered to be a minor gene because the single deletion of ptl1 increased the TG level to a similar level as the deletion of another gene.

The other possibility is the formation of a complex between ptl1p and ptl3p. Since the homo-dimer formation of Tgl3p and Tgl4p was predicted based on the genome-wide two-hybrid experiments in S. cerevisiae (Ito et al. 2001), if the products of ptl1p and ptl3p work together by making a complex, the phenomenon could be explained. Furthermore, we should also consider if ptl1p, ptl2p, and ptl3p have additional functions in lipid metabolism because it has been reported recently that Tgl3p and Tgl5p of S. cerevisiae, which are homologs of ptl1p, ptl2p, and ptl3p, had a highly conserved H-(X)4-D aceyltransferase motif, and indeed they had acyltransferase activities (Rajakumari and Daum, 2010). The same motif (169HREQKD174) was found in ptl2p upstream of its lipolytic motif, but in contrast to the deletion of TGL3 or TGL5, no significant decrease in phospholipids was observed in the single deletion mutant of ptl1p, ptl2p, or ptl3p (Fig. 4). However, additional biochemical analyses would be required to examine their additional functions, including their protein level interactions.

To examine whether the TG lipase activity of ptl1p, ptl2p, and ptl3p, predicted from the analysis of the triple deletion mutant, has biological significance, we addressed the mobilization of cellular TG by examining the growth of the triple deletion mutant in the presence of cerulenin. The cerulenin-sensitive phenotype of the triple mutant and its growth recovery by supplementation of oleic acid strongly suggested that they are functioning in TG mobilization in S. pombe cells.

We are studying the possibility of using S. pombe as a host for the production of useful fatty acids and their derivatives because we consider that S. pombe can be a better microbe than S. cerevisiae, owing to its very high C18:1 content. The heterologous expression of CpFAH12 gene in the triple mutant demonstrated that this strain was useful and the production of ricinoleic acid per DCW was increased by 1.8-fold in the triple mutant, suggesting that the prevention of TG degradation by the disruption of TG lipase genes may be one of the strategies to increase lipid content. At the current moment, however, the disadvantage of S. pombe is that its fatty acid content is not sufficiently high to apply transgenic S. pombe strains to industrial applications. S. cerevisiae and S. pombe do not accumulate storage lipids and usually only 5 to 7 % of lipids are produced per DCW, whereas some oleaginous yeasts and fungi have 30 to 50 % lipids. Recently, we have established an S. cerevisiae strain with a total lipid content of 29 % by the DGA1 (diacylglycerol acyltransferase) overexpression in combination with SNF2 disruption and LEU2 overexpression (Kamisaka et al. 2007). Thus, the next step to take is to combine the prevention of TG degradation by TG lipase deletion and the enhancement of TG synthesis by the overexpression of diacylglycerol acyltransferases (Zhang et al. 2003) to assess whether we could further increase TG content in the cells.

Acknowledgments

A part of this work was supported by an A-STEP (Adaptable and Seamless Technology transfer program through target-driven R&D) research grant from Japan Science and Technology Agency (JST) to HK and HU. We also wish to thank Masakazu Yamaoka, Yasushi Kamisaka, and Kazuyoshi Kimura for continuous encouragement.

Copyright information

© Springer-Verlag 2012