Abstract
Eight wine yeast strains of Saccharomyces sp. were tested for polygalacturonase (PGase) activity, after cultivation on various carbon sources. No strain showed any activity when grown on glucose, while five strains produced PGase in the presence of galactose and polygalacturonate. These data suggest that the PGase of wine strains is repressed by glucose and induced by galactose and polygalacturonate. The existence of the PGase gene in the wine strains and its similarity with that of the laboratory strains was proved by Southern hybridization and PCR amplification. The promoter region of the PGase gene in the wine strains was slightly different from that of the laboratory strains. This possibly explains the different pattern of gene expression in wine and laboratory strains. The PGase of wine strains produced di- or tri-galacturonic acid from polygalacturonic acid, different from the fungal PGase.
1. Introduction
Pectinolytic enzymes are widely used in the beverage industry to clarify fruit juices and wine. Endopolygalacturonase (PGase: EC 3.2.1.15), which degrades the pectic substrate (polygalacturonic acid) by splitting 1,4-α-glycosidic bonds, is a main component of industrial pectinolytic enzymes [1]. This enzyme plays important roles in the food industry, especially in wine making. Most strains of Saccharomyces cerevisiae, usually used in wine making, do not show the capacity to degrade pectic substrates. A few wild strains have been reported to have the ability to degrade pectin in wine fermentation [[2–[4]. Recently, by chemical mutagenesis a mutant producing PGases has been derived from a laboratory strain of S. cerevisiae[5], and the PGase has been characterized biochemically [6]. Cloning and sequencing of the PGase gene from the mutant and the parent revealed that both genes were completely identical in the PGase-coding region and 5′-upstream region, whereas the PGase gene was expressed in the mutant but not in the parent [[7,[8]. On the other hand, the expression of the PGase gene reported in the SCPP strains of Saccharomyces bayanus is upregulated by the presence of pectin and surprisingly is enhanced under conditions of filament formation [9]. These reports suggest that the regulation of the PGase gene in Saccharomyces is complicated.
It has been expected that pectinolytic wine yeasts can improve liquefaction, clarification and filterability of grape must, releasing more color and flavor compounds entrapped in the grape skins, thereby making a positive contribution to the wine bouquet [10]. Here, we have studied the pectinolytic activity of commercial wine yeasts and the influence of various carbon sources on the PGase production by the yeasts to elucidate the regulatory mechanism of PGase gene expression, and also have constructed mutants from some wine yeasts constitutively producing PGase to improve wine making.
2. Materials and methods
2.1. Media, strains and culture conditions
Yeast strains used are shown in Table 1. Media used were YPD (2% glucose, 1% yeast extract and 1% peptone), and YPGal (2% galactose, 1% yeast extract and 1% peptone). In some experiments, the 2% glucose in YPD was replaced with 1% glucose and 1% galactose in YP(D:Gal), 1% glucose and 1% sodium galacturonate in YP(D:MGA), and 1% glucose and 1% sodium polygalacturonate in YP(D:PGA). Halo-formation plates were prepared by upper-layering 1% sodium polygalacturonate with 1.5% agar on every plate medium. The yeast strains were cultured at 30 °C. The Escherichia coli strain used was DH-5α (Toyobo Biochemicals, Osaka, Japan).
Yeast strains used
| Strain | Property | Source |
| Saccharomyces cerevisiae | ||
| KW1 | Wine yeast | NRIBa |
| KW3 | Wine yeast | NRIB |
| KW4 | Wine yeast | NRIB |
| OC2 | Wine yeast | IBRCb |
| L2226 | Wine yeast | IBRC |
| UvaFerm | Wine yeast | IBRC |
| UvaFerm CEG | Wine yeast | IBRC |
| S288C (IFO 1136) | Laboratory strain | IFOc |
| DKD-5DH | Laboratory strain | [11] |
| SMF3 | PGase+ mutant from DKD-5DH | [6] |
| Saccharomycesi bayanus | ||
| EC1118 | Wine yeast | IBRC |
| Strain | Property | Source |
| Saccharomyces cerevisiae | ||
| KW1 | Wine yeast | NRIBa |
| KW3 | Wine yeast | NRIB |
| KW4 | Wine yeast | NRIB |
| OC2 | Wine yeast | IBRCb |
| L2226 | Wine yeast | IBRC |
| UvaFerm | Wine yeast | IBRC |
| UvaFerm CEG | Wine yeast | IBRC |
| S288C (IFO 1136) | Laboratory strain | IFOc |
| DKD-5DH | Laboratory strain | [11] |
| SMF3 | PGase+ mutant from DKD-5DH | [6] |
| Saccharomycesi bayanus | ||
| EC1118 | Wine yeast | IBRC |
National Research Institute of Brewing (Higashi Hiroshima, Japan).
Iwate Biotechnology Research Center (Kitakami, Japan).
Institute of Fermentation Osaka (Osaka, Japan).
Yeast strains used
| Strain | Property | Source |
| Saccharomyces cerevisiae | ||
| KW1 | Wine yeast | NRIBa |
| KW3 | Wine yeast | NRIB |
| KW4 | Wine yeast | NRIB |
| OC2 | Wine yeast | IBRCb |
| L2226 | Wine yeast | IBRC |
| UvaFerm | Wine yeast | IBRC |
| UvaFerm CEG | Wine yeast | IBRC |
| S288C (IFO 1136) | Laboratory strain | IFOc |
| DKD-5DH | Laboratory strain | [11] |
| SMF3 | PGase+ mutant from DKD-5DH | [6] |
| Saccharomycesi bayanus | ||
| EC1118 | Wine yeast | IBRC |
| Strain | Property | Source |
| Saccharomyces cerevisiae | ||
| KW1 | Wine yeast | NRIBa |
| KW3 | Wine yeast | NRIB |
| KW4 | Wine yeast | NRIB |
| OC2 | Wine yeast | IBRCb |
| L2226 | Wine yeast | IBRC |
| UvaFerm | Wine yeast | IBRC |
| UvaFerm CEG | Wine yeast | IBRC |
| S288C (IFO 1136) | Laboratory strain | IFOc |
| DKD-5DH | Laboratory strain | [11] |
| SMF3 | PGase+ mutant from DKD-5DH | [6] |
| Saccharomycesi bayanus | ||
| EC1118 | Wine yeast | IBRC |
National Research Institute of Brewing (Higashi Hiroshima, Japan).
Iwate Biotechnology Research Center (Kitakami, Japan).
Institute of Fermentation Osaka (Osaka, Japan).
2.2. Preparation of enzyme solution
Every medium was inoculated with 106 cells ml−1 preinoculum of every yeast strain. After cultivation for 72 h at 30 °C the supernatants were dialyzed overnight against 0.01-M sodium-acetate buffer (pH 5.0) to be used as crude enzymes.
2.3. Assay of pectinolytic activity
Pectin- and pectate-lyase activities were assayed using methyl-esterified polygalacturonate and polygalacturonic acid (free acid), respectively, as substrates, by measuring the increase of the optical density at 235 nm [12].
Two methods were employed for determining PGase activity. For a rapid assay the yeasts were cultivated on halo-formation plates for 3–4 days at 30 °C. PGase-positive strains were detected by showing a halo on the plate upon application of 6-N HCl.
PGase activity was assayed by measuring the amount of reducing sugar released, using the Somogyi–Nelson method [6], as follows. The reaction mixture was composed of 0.1% sodium polygalacturonate and crude enzyme solution (0.2% volume of the total reaction mixture) in 0.02-M sodium acetate buffer (pH 5.5). The reaction was performed at 37 °C for 60 min. One unit of PGase activity was defined as the activity that liberates reducing groups corresponding to 1 μmol of d-galacturonic acid under the above-mentioned conditions.
2.4. DNA techniques
Extraction of yeast genomic DNA was according to Wach et al. [13]. The polymerase chain reaction (PCR) was performed with a Takara Thermal Cycler (Takara Bio, Kusatsu, Japan) in 35 cycles of 60 s at 94 °C for denaturation, 30 s at 48 °C for annealing, and 90 s at 72 °C for DNA-polymerase reaction. Primers, synthesized from the nucleotide sequences based on the N-terminus, the C-terminus and the 1536-bases upstream from the initiation codon of the S. cerevisiae PGase gene, PSM1/PGU1, were, respectively, as follows: PGSM-N1, CCTAGATCTATGATTTCTGCTAATTCATTA; PGSM-C2, CTGCGGATCCTTAACAGCTTGCACCAGATC; PGSM-P3, ACAAGTCGACTTGTCCTGCC. The PCR products were purified with Purification kit (Amersham Pharmacia Biotech, Buckinghamshire, England) and cloned into pGEM-T Easy Vectors (Promega, Madison, WI, USA). Southern analysis against genomic DNA transferred onto Hybond-N+ membrane (Amersham Pharmacia Biotech) was done using ECL labeling and Detection kit (Amersham Pharmacia Biotech). The PSM1 gene [8] was used as the probe of Southern analysis. Protocols of kits were according to supplier. Other DNA techniques were done as described by Sambrook et al. [14].
2.5. Thin-layer chromatography
Degradation products from polygalacturonate by PGase from mutants were analyzed by thin-layer chromatography. An endo-polygalacturonase of Aspergillus kawachii IFO 4033 was used as control. The reaction mixture contained crude enzyme and 0.2% polygalacturonate in 0.02-M sodium acetate buffer (pH 5.0). Two samples, after 1-h and 3-h reaction time, were brought on the chromatography plate. The thin-layer plate used was Silica Gel 60 F254 (Merck, Darmstadt, Germany). The developing solvent consisted of 50:25:42 (v/v) ethyl acetate–glacial acetic acid–water. Saccharide spots were detected by spreading the plate with 25% sulfuric acid and then heating at 110 °C for 3–5 min.
2.6. Chemicals
Glucose, galactose and pectin were from Wako Pure Chemical Industry, Osaka, Japan. Methyl-polygalacturonate (esterified pectin), Na-polygalacturonic acid and Na-galacturonate were from Sigma Chemical Co., St. Louis, MO, USA.
3. Results
3.1. Effect of carbon source on the production of pectinolytic enzymes
The production of pectinolytic enzymes was tested in the media containing different carbon sources. None of the industrial wine yeasts tested produced a pectin- or pectate-lyase, since absorbance at 235 nm of the reaction mixture with crude enzymes was not increased using methyl-esterified polygalacturonate or free polygalacturonate as substrate.
In the YPD medium, none of the wine yeasts produced PGase; neither did the laboratory strains S288C and DKD-5DH (Table 2). However, PGase activity was recognized in crude enzymes of five strains, KW3, KW4, UvaFerm, UvaFerm CEG and EC1118, when cultured in the YPGal medium (Table 2). This suggested that the PGase production in these strains was regulated by the carbon source. To analyze for carbon catabolite repression in PGase production, KW3, KW4, UvaFerm, UvaFerm CEG and EC1118 were cultured on various carbon sources. KW3 and KW4 did show PGase activity upon cultivation on lactic acid as a carbon source, but none in the other carbon source media (data not shown). Further, the effect of glucose on the PGase production by these strains was tested by culturing in glucose-added galactose medium YP(D:Gal). The PGase production in YP(D:Gal) was at the same level as that in YPGal (Table 2). This indicated that the PGase genes in KW4, UvaFerm, UvaFerm CEG and EC1118 possibly undergo induction by galactose rather than repression by glucose.
PGase production of wine yeasts cultured on various carbon sources
| Strain | PGase activity (Uaml−1 culture) | ||||
| YPD | YPGal | YP(D:Gal) | YP(D:MGA) | YP(D:PGA) | |
| OC2 | <1 | <1 | <1 | <1 | <1 |
| L2226 | <1 | <1 | <1 | <1 | n.t. |
| KW1 | <1 | <1 | <1 | <1 | n.t. |
| KW3 | <1 | 110 | 85 | n.t. | n.t. |
| KW4 | <1 | 100 | 80 | 10 | 340 |
| UvaFerm | <1 | 50 | 40 | 12 | 330 |
| UvaFerm CEG | <1 | 42 | 48 | 5 | 290 |
| EC1118 | <1 | 77 | 57 | 5 | 176 |
| S288C | <1 | <1 | <1 | <1 | <1 |
| DKD-5DH | <1 | <1 | <1 | <1 | <1 |
| Strain | PGase activity (Uaml−1 culture) | ||||
| YPD | YPGal | YP(D:Gal) | YP(D:MGA) | YP(D:PGA) | |
| OC2 | <1 | <1 | <1 | <1 | <1 |
| L2226 | <1 | <1 | <1 | <1 | n.t. |
| KW1 | <1 | <1 | <1 | <1 | n.t. |
| KW3 | <1 | 110 | 85 | n.t. | n.t. |
| KW4 | <1 | 100 | 80 | 10 | 340 |
| UvaFerm | <1 | 50 | 40 | 12 | 330 |
| UvaFerm CEG | <1 | 42 | 48 | 5 | 290 |
| EC1118 | <1 | 77 | 57 | 5 | 176 |
| S288C | <1 | <1 | <1 | <1 | <1 |
| DKD-5DH | <1 | <1 | <1 | <1 | <1 |
n.t., Not tested.
1 U is the amount of enzyme to produce 1 μmol galacturonate in 1 min.
PGase production of wine yeasts cultured on various carbon sources
| Strain | PGase activity (Uaml−1 culture) | ||||
| YPD | YPGal | YP(D:Gal) | YP(D:MGA) | YP(D:PGA) | |
| OC2 | <1 | <1 | <1 | <1 | <1 |
| L2226 | <1 | <1 | <1 | <1 | n.t. |
| KW1 | <1 | <1 | <1 | <1 | n.t. |
| KW3 | <1 | 110 | 85 | n.t. | n.t. |
| KW4 | <1 | 100 | 80 | 10 | 340 |
| UvaFerm | <1 | 50 | 40 | 12 | 330 |
| UvaFerm CEG | <1 | 42 | 48 | 5 | 290 |
| EC1118 | <1 | 77 | 57 | 5 | 176 |
| S288C | <1 | <1 | <1 | <1 | <1 |
| DKD-5DH | <1 | <1 | <1 | <1 | <1 |
| Strain | PGase activity (Uaml−1 culture) | ||||
| YPD | YPGal | YP(D:Gal) | YP(D:MGA) | YP(D:PGA) | |
| OC2 | <1 | <1 | <1 | <1 | <1 |
| L2226 | <1 | <1 | <1 | <1 | n.t. |
| KW1 | <1 | <1 | <1 | <1 | n.t. |
| KW3 | <1 | 110 | 85 | n.t. | n.t. |
| KW4 | <1 | 100 | 80 | 10 | 340 |
| UvaFerm | <1 | 50 | 40 | 12 | 330 |
| UvaFerm CEG | <1 | 42 | 48 | 5 | 290 |
| EC1118 | <1 | 77 | 57 | 5 | 176 |
| S288C | <1 | <1 | <1 | <1 | <1 |
| DKD-5DH | <1 | <1 | <1 | <1 | <1 |
n.t., Not tested.
1 U is the amount of enzyme to produce 1 μmol galacturonate in 1 min.
Then, the inductive effects of galacturonate and polygalacturonate were examined. Since the wine yeasts could not grow in 2% galacturonate or 2% polygalacturonate medium, they were cultured in the glucose-supplemented media, YP(D:MGA) and YP(D:PGA). Growth of KW4, UvaFerm, and EC1118 in YP(D:MGA) were not significant, resulting in low level of PGase production. Different influence of polygalacturonate on the PGase production was shown. KW4, UvaFerm, UvaFerm CEG, and EC1118 showed good growth and high PGase production in the YP(D:PGA) medium. The level of PGase production in the presence of polygalacturonate was higher than that in the presence of galactose. (Table 2). The different effects of galacturonate and polygalacturonate are probably due to the endo-type PGase that does not degrade polygalacturonate to galacturonate.
The remaining wine yeast strains, KW1, L2226 and OC2, and the laboratory strains S288C and DKD-5DH developed no PGase activity in any of the cultivation media tested. This suggested the possibility that the PGase genes in these wine strains were always repressed, as in the laboratory strains, or were lost.
3.2. Presence of the PGase gene in wine yeasts
Southern-hybridization analysis, with the PSM1/PGU1 gene as a probe, was carried out against the chromosomal DNAs from KW4 and EC1118, wine yeasts showing PGase production in YPGal; OC2, a PGase non-producing wine yeast; and S288C, a PGase non-producing laboratory strain. Hybridization signals were detected on KW4, EC1118, and S288C DNAs, but not on OC2 DNA (Fig. 1). A PGase gene similar to PSM1/PGU1 may be absent in OC2. The existence of a PGase gene in wine yeasts was also examined by PCR. The primers PGSM-N1 and PGSM-C2 were used to amplify the PGase-coding region of the PSM1/PGU1 gene from the chromosomal DNA of wine yeast strains (Fig. 2(a)). The DNA fragment (whole size ∼1.1 kb) corresponding to PSM1/PGU1 was amplified for all except three, OC2, KW1, and L2226, which did not produce PGase in various media. These results show the possibility that OC2, KW1 and L2226 can not produce PGase because they do not contain the PSM1/PGU1 gene. In addition, the genes attached with promoter region were amplified using the primers PGSM-P3 and PGSM-C2 (Fig. 2(b)). KW3, UvaFerm, UvaFerm CEG and EC1118 did amplify in the corresponding segment with the PSM1/PGU1 gene, as did the laboratory strains DKD-5DH and S288C. However, KW4 did not amplify the promoter-containing PSM1/PGU1 gene. These results can be a proof of the presence of the promoter-containing PSM1/PGU1 gene in the laboratory strains, DKD-5DH and S288C as well as in the wine strains in KW3, UvaFerm, UvaFerm CEG and EC1118. Also these results indicate the possibility that the promoter region of the PSM1/PGU1 gene in KW4 was changed by mutation.
Southern analysis of various strains. Hybridization was done with labeled PSM1/PGU1 gene. Genome DNA was digested with HindIII. Lane 1: KW4; lane 2: OC2; lane 3: EC1118; lane 4: S288C.
PCR amplification of chromosomal DNA with PSM1/PGU1 primers. Agarose gel electrophoresis of PCR products using a primer set of PGSM-N1 and PGSM-C2 (a) and a set of PGSM-P3 and PGSM-C2, respectively (b). Lane 1 and 13: EcoT14I-digested λ DNA; lane 2: OC2; lane 3: KW1; lane 4: KW3; lane 5: KW4; lane 6: L2226; lane 7:UvaFerm; lane 8: UvaFermCEG; lane 9: EC1118; lane 10: S288C; lane 11: DKD-5DH.
3.3. Sequencing analysis of the PGase gene cloned from the wine yeast KW4
The amplified genes of KW3, KW4, UvaFerm, and EC1118 were cloned into a pGEM-T plasmid and a part of the DNA sequences were identified. The results indicated that the ORF region was identical to that of the PSM1/PGU1 gene recorded in the DNA database (data not shown). A part of 5′-upstream region of the amplified genes from KW3, UvaFerm and EC1118 was slightly different from that of the PSM1/PGU1 gene. The sequence similarities for the three strains in 600–700 bp upstream from the initiation codon of PSM1/PGU1, which contain the Gal4-binding sequence, were 92%, 93% and 92% to that of PSM1/PGU1, respectively. These results prove that the wine yeast strains contain the same ORF of the PSM1/PGU1 gene as the laboratory strain. They also indicate that the difference between the promoter regions in wine and laboratory strains may determine whether the PGase is induced in the galactose and polygalacturonate media or not.
3.4. End products of PGase activity from the wine yeasts
The PGases produced by KW3, KW4, UvaFerm, and EC1118 in YPGal culture medium were tested for their end products from polygalacturonic acid by thin-layer chromatography (Fig. 3). The controls were PGase SM, produced from the S. cerevisiae PGase-producing mutant SMF3 [6] and PGase B known as an endo-polygalacturonase of A. kawachii[15]. Chromatography of the hydrolysis products revealed the production of an endo-polygalacturonase. The position of the products of wine yeast PGases is different from that of mono-galacturonic acid which is produced by PGase B of A. kawachii. Considering the Rf value, the products of KW3 and Uva Ferm PGases seem to be a dimer just like the product of PGase SM. The products of KW4 and EC 1118 PGases seem to be a trimer.
![Thin-layer chromatogram of PGase reaction products. Authentic galacturonic acid (lane 1) and reaction products (lanes 2–7) were fractionated. Lane 2: SMF3-Pgase; lane 3: KW4-Pgase; lane 4: KW3-Pgase; lane 5: UvaFerm-Pgase; lane 6: EC1118-Pgase; lane 7: PGase B from A. kawachii[14].](https://oup.silverchair-cdn.com/oup/backfile/Content_public/Journal/femsyr/5/6-7/10.1016_j.femsyr.2004.09.006/1/m_FYR_663_f3.jpeg?Expires=1686539242&Signature=MG7AWxsvjukVPjuVLNnP82QHnQOFqdvfxuvswfwoyHClC722uYsJBjCLrFPF6eO9R3RnraHPMYLRK4vK0KebNIkfTj1PjHfacjiT8XG2VQKDWmiV1d-Xnj26S0QZvi4WsLyTvhDjzFyl3bdExSEaIZhaUgwoElAvaOSaQ-VNT41jGmfP2c9MU5vZXhdUTexNnD65lOE4JBYqpu3qOD7dCHqHq7D73ssuC8bNFo2lks8~60Z8WL0msYW-tDv5wEuUoKiuLooX5bOa1L6juSewslJtSbQ3G9STg5evH-Pi2aY2GI4zb34m6BVaXyUe0fdXTgjXkUSGvFX6ANB3iwZdoA__&Key-Pair-Id=APKAIE5G5CRDK6RD3PGA)
Thin-layer chromatogram of PGase reaction products. Authentic galacturonic acid (lane 1) and reaction products (lanes 2–7) were fractionated. Lane 2: SMF3-Pgase; lane 3: KW4-Pgase; lane 4: KW3-Pgase; lane 5: UvaFerm-Pgase; lane 6: EC1118-Pgase; lane 7: PGase B from A. kawachii[14].
4. Discussion
Pectinolytic wine yeasts are expected to make a positive contribution to the wine making [[10,[16]. A recombinant wine yeast strain which can express a pectate lyase gene (pelA) from Fusarium solani has been studied [17]. If S. cerevisiae strains have a potential ability to produce PGase, expression of this ability is more useful in wine making than the use of a recombinant strain. Recently, it has been reported that laboratory strains of S. cerevisiae possess the PGase gene without being expressed [5]. The reason why the PGase gene is not expressed has not been elucidated [[7,[8]. Industrial strains of wine yeast were not known to produce PGase in glucose-rich media such as grape must. This study has suggested the possible explanation why wine yeast strains do not produce PGase. One is that some strains like KW1, L2226 and OC2 are lacking in the original PGase gene, and another is that the PGase gene is regulated by carbon catabolite repression, as observed in KW3, KW4, UvaFerm and EC1118. The latter strains probably contain a promoter region different from that of the laboratory strain, whose PGase gene usually is repressed.
The major sugars in must are glucose and fructose, the first being consumed during the vinification. The level of glucose is up to 17% at the first step of the wine fermentation and less than 0.2% at the end. The catabolism of wine yeast strains is known to be repressed by glucose under wine-making conditions, and PGase production may also be repressed by glucose. It has been reported that the pectinase activity was completely abolished by the addition of 2% glucose in a few wine strains of S. cerevisiae[3]. There is a complex system of glucose repression of the synthesis of many enzymes involved in the catabolism of exogenously supplied sugars [[18,[19]. Our study suggests that also the PGase production in wine yeasts is regulated by glucose repression and by induction with galactose and polygalacturonate. The same regulation including glucose repression and galactose induction was reported for pectinase production in Neurospora crassa and Botrytis cinerea[[20,[21]. Such regulated gene expression is similar to that of GAL1, GAL7 and GAL10, involved in galactose metabolism of S. cerevisiae laboratory strains. The promoter regions of those genes contain regulatory elements (UAS and URS) that are controlled by glucose and galactose [[22,[23]. The PGase genes of wine strains are possibly regulated by a similar mechanism, since the putative cis-element bound to the GAL4 protein, the regulator of the galactose induction [24], is present 610 and 632 bp upstream from the initiation codon of the PGase gene, PSM1/PGU1. However, laboratory strains of S. cerevisiae can not produce PGase in the galactose culture. These data suggest that the wine strains producing the galactose-inducible PGase do not contain any negative regulators that are silencing the PSM1/PGU1 gene in the laboratory strains. The silencing factor of the PSM1/PGU1 gene in the laboratory strains of S. cerevisiae might be elucidated by analysis of the promoter region in wine strains.
Acknowledgements
We thank Dr. Nakazawa (Akita Prefectural University) and Dr. Ito (National Research Center of Brewing) for their kind gift of wine strains. A part of this work was supported by Grants-in-Aid for Scientific Research by the Ministry of Education, Science, and Culture of Japan (No. 10,760,057).
References
