Notice: Wiley Online Library will be unavailable on Saturday 30th July 2016 from 08:00-11:00 BST / 03:00-06:00 EST / 15:00-18:00 SGT for essential maintenance. Apologies for the inconvenience.
If you can't find a tool you're looking for, please click the link at the top of the page to "Go to old article view". Alternatively, view our Knowledge Base articles for additional help. Your feedback is important to us, so please let us know if you have comments or ideas for improvement.
Exogenous enzymes are an important tool of modern winemaking, and many industrial preparations are now commercially available. Enzyme preparations are applied for winemaking for various reasons, such as improving the extraction of grape skin components, releasing sensory active compounds from precursors and inhibition of spoilage microorganisms (Blättel et al., 2009; Ugliano, 2009). Enzymatic preparations containing β-glucanases are utilized extensively to facilitate the filtration of musts and wines coming from grapes affected by Botrytis cinerea, and to induce the release of mannoproteins and oligosaccharides from the cell walls of yeasts. As a prerequisite, these enzymes must function under harsh winemaking conditions, such as low pH and temperature, high concentrations of sugars and ethanol or the presence of polyphenols and sulphite (Zinnai et al., 2010).
The yeast Wickerhamomyces anomalus (formely Pichia anomala) has gained considerable importance for the wine industry. A strain FY-102, isolated from apple skin, was found to be antagonistic to the pathogenic grey mould Botrytis cinerea (Masih et al., 2000). Other strains have been found to inhibit the wine-spoiling yeast Brettanomyces bruxellensis (Passoth et al., 2006). W. anomalus is a constituent of the normal grape flora in early phases of alcoholic fermentation (Renouf et al., 2007). In contrast to the wine yeast Saccharomyces cerevisiae, aerobic metabolism of W. anomalus is not repressed in high sugar concentrations (Crabtree-negative). This means that oxygen availability, but not glucose concentration, is the main stimulus for the regulation of the central carbon metabolism (Fredlund et al., 2004). For this reason, W. anomalus may be a candidate to produce low-alcohol wines (Ertin and Campbell, 2001). The yeast can spoil wine by excessive production of acetic acid and ethyl acetate, but normally contributes positively to the wine aroma by the production of volatile compounds (Passoth et al., 2006). Another interesting feature is the yeast's ability to utilize and reduce the content of malic acid (Walker, 2011).
We recently described the novel W. anomalus strain A1 isolated from grape must, which hydrolysed a number of natural glycosides under oenological conditions (Sabel et al., 2014). This strain may thus be useful for winemaking by releasing aroma-active terpenes from glycosylated precursors. Here we report the purification and molecular/biochemical features of a stable multifunctional exo-β-glucanase (WaExg2) from strain AS1, which efficiently cleaves various glycosylated polyphenols and the polysaccharide laminarin. The characterized enzyme might have multiple uses in winemaking to increase levels of bioactive polyphenols, such as resveratrol, and to hydrolyse glucan slimes.
Materials and methods
Yeast strains and cultivation
W. anomalus AS1 (DSM 28943) has been deposited at the Leibniz Institute (DSMZ), German Collection of Microorganisms and Cell Cultures (Braunschweig, Germany). The commercial wine yeast Saccharomyces cerevisiae ‘Fermichamp’ No. 67J was purchased from Max F. Keller GmbH (Mannheim, Germany).
The yeasts were maintained on YEP medium (yeast extract 10 g/l, meat peptone 20 g/l, glucose 20 g/l). Solid medium was prepared with 2% w/v agar. The glucose in YEP medium was replaced by cellobiose (YEPC) for enzyme production by W. anomalus AS1 (Villena et al., 2006).
Quantitative determination of β-glucosidase activity was performed as described recently (Sabel et al., 2014). The release of yellow p-nitrophenol from β-pNPG was measured photometrically at 405 nm in a microplate reader (Titertek Multiscan Plus MKII); one unit corresponds to the amount of enzyme that catalyses the hydrolysis of 1 µmβ-pNPG/min (molar extinction coefficient ε of p-nitrophenol: 18 300 l/m/cm). Activity with α-pNPG, α-pNPR and β-pNPX was measured in an analogous manner.
Detection of yeast lytic activities
For the microscopic observation of enzymatic cell wall hydrolysis, protoplasts were generated following the protocols of Mateo and Di Stefano (1997) and Lomolino et al. (2001); 2 ml yeast culture (OD600 = 1.0) were centrifuged at 3000 × g for 5 min. The cell pellets were washed twice with deionized water and incubated for 15 min at 25°C in buffer A (sorbitol 182.2 g/l, dithiothreitol 4.6 g/l in 10 mm citrate phosphate buffer, pH 5.5). Afterwards, the suspensions were centrifuged at 1000 × g for 5 min and the cells resuspended in 2 ml buffer B (sorbitol 182.2 g/l, dithiothreitol 0.16 g/l, Na-EDTA 37.2 g/l in 10 mm citrate buffer, pH 5.5); 20 µl of this suspension were mixed with 180 µl purified glucanase WaExg2 (100 µg/ml) and incubated for 2 h at 37°C. Cell morphologies were observed by light microscopy (Zeiss Axiophot 2, Germany). Quantification of lytic activities was performed using a photometric test, as described previously (Blättel et al., 2009).
Determination of the protein content
Protein concentrations of the samples were determined with a bicinchoninic acid (BC)-related assay kit (Roti®-Quant Universal, Karlsruhe, Germany).
SDS–polyacrylamide gel electrophoresis (SDS–PAGE)
Electrophoretic conditions and glycoprotein staining have been described recently (Claus et al., 2014). Proteins were visualized using the Pierce Silver Stain Kit™ (Thermo Scientific, Germany). The PageRuler™ Plus prestained protein ladder (Thermo Scientific) was used as a molecular mass standard. The gels were incubated twice for 15 min in a 1% v/v Triton X solution with gentle shaking for activity staining. Subsequently, the gels were incubated in a 2 mm solution of MUF-Glc. Development of fluorescent bands was observed under UV light.
Analytic isoelectric focusing
Isoelectric focusing (IEF) and subsequent protein staining was carried out in Servalyt Precotes 3–10, following the recommendations of the manufacturer (Serva, Germany). IEF gels were stained with Serva Violet 17 (Serva) or by using a silver-staining kit for the visualization of proteins. Detection of β-glucosidase activity was performed with MUF-Glc.
Preparative isoelectric focusing
Purification of the extracellular exo-β-1,3-glucanase was achieved by preparative IEF, using the Rotofor™ Preparative IEF Cell (BioRad, Munich, Germany); 50 ml enzymatic active culture supernatant were dialysed for 4 days against deionized water and mixed with 1.5 ml 40% w/w ampholyte solution, pH 3.0–6.0 (Fluka, no. 11923). The electrode buffers were 0.5 m acetic acid (anode) and 0.25 m HEPES (cathode), respectively. The separation was performed under a preset power of 15 W until constant voltage was reached after 4–5 h at 10°C. The pHs of the 20 fractions obtained were measured, their protein content checked by SDS–PAGE and activity determined on MUF-Glc-containing agar plates. The two to four positive fractions were concentrated in 15 ml spin columns (10 kDa cut-off; Vivaspin, Germany), washed three times with deionized water and stored at –20°C. In some cases, a second purification step was performed using the MicroRotofor™ Cell (BioRad). Thereby, 3 ml of the concentrate were mixed with 60 µl 40% w/w ampholyte solution, pH 3.0–4.0 (Serva, no. 42922) and refocused in the small chamber at 10°C for 3–4 h at 1 W. The 10 fractions obtained were analysed as before.
Identification of the exo-β-1,3-glucanase
The purified proteins were blotted from SDS polyacrylamide gels to polyvinylidene fluoride (PVDF) membranes (Immobilon-P®, Millipore, Germany) by a semi-dry procedure according to the Millipore User Guide. The separated protein bands were excised, destained and then digested with trypsin. The separation and identification of peptide fragments by capillary liquid chromatography–mass spectrometry (LC–MS/MS) and database analysis have been described recently (Claus et al., 2014).
Thin-layer chromatography (TLC) of hydrolysis products
All substrates were tested at a concentration of 1 mg/ml in citrate–sodium phosphate buffer (0.1 m, pH 5.0) containing 10% w/v ethanol. After the addition of 25 µg/ml enzyme solution, the reaction assays were incubated for 48 h at 40°C. The enzymatic cleavage of glycosides and glycans was examined by TLC (Sabel et al., 2014).
Determination of glucose and fructose
Quantification of glucose and fructose released from the substrates was determined enzymatically using the d-glucose/d-fructose UV Test (no. 10139106035; R-Biopharm, Darmstadt, Germany).
Testing of wine-relevant parameters
The impact of different parameters on β-glucosidase activity was determined using the spectrophotometric standard test with 5 mm MUF-Glc as the substrate. Enzyme activity depending on the pH was tested by dissolving MUF-Glc in a universal buffer (Tris 6.5 g/l, maleic acid 2.32 g/l, citric acid 2.8 g/l, boric acid 1.26 g/l; the pH desired was adjusted with 0.1 m NaOH). Temperature dependence was tested at the range 10–80°C at pH 4.5. The effects of sugars, alcohol, metal cations and potassium metabisulphite were tested at the concentrations indicated at pH 3.5 and 40°C.
Amplification and sequencing of the WaEXG2 gene
In order to extract the DNA, 1.0 ml of a culture (about 5 × 107 cells) of W. anomalus AS1 grown overnight was harvested by centrifugation (10 000 × g) and washed in a solution of 0.9% NaCl. DNA isolation was performed using the DNeasy Blood and Tissue Kit (Qiagen, Hilden, Germany), according to the supplementary protocol for purification of total DNA from yeasts. Incubation with lyticase was carried out for 60 min and treatment with proteinase K for 3 h. The purified DNA was eluted from mini-spin columns with 100 µl PCR-grade water and stored at –20°C.
Polymerase chain reaction (PCR)
The WaEXG2 gene of W. anomalus AS1 was amplified using the primer pair exg2-F (5′-CACCTCTGTAGTCCATCTATG-3′) and exg2-R (5′-CAACCTTCTCGTATGAAACAG-3′), which were designed on the basis of two known EXG2 sequences (Accession Nos. AJ222862 and EF029071). Exg2-F anneals 204 bp upstream and exg2-R anneals 91 bp downstream of the EXG2 gene. Thus, an amplification product of 1579 bp was expected. PCR reaction mixtures (50 µl) contained 1 µl forward primer (10 µm), 1 µl reverse primer (10 µm), 1 µl dNTP mix (10 mm each), 2 µl MgCl2 (25 mm), 5 µl 10× reaction buffer S, 5 µl enhancer solution P and 1 µl Taq-DNA-polymerase (1 U/µl); 2 µl DNA from W. anomalus AS1 was used as a template. All PCR reagents were purchased from Peqlab (Erlangen, Germany); oligonucleotides were synthesized by Sigma-Aldrich (Steinheim, Germany).
The amplification was carried out in a Mastercycler gradient thermocycler (Eppendorf, Hamburg, Germany), using the following conditions: initial denaturation for 5 min at 95°C, followed by 35 cycles of 1 min at 94°C for denaturation, 1 min at 45°C for annealing and 1.5 min at 72°C for elongation; a final elongation step for 10 min at 72°C was performed.
The resulting PCR fragments were purified with the GeneJET PCR Purification Kit (Thermo Fisher Scientific, Waltham, USA). Sequencing was carried out by LGC Genomics (Berlin, Germany).
Homology modelling of WaExg2
A 3D structure of WaExg2 was modelled with the Swiss-Modell repository tool (http://swissmodel.expasy.org/repository). The crystal structure of an exo-β-1,3-glucanase from Candida albicans at a resolution of 1.9 Å (Cutfield et al., 1999) was used as the template.
Purification by preparative IEF
W. anomalus AS1 exhibited the highest extracellular β-glucosidase activity when grown as a shaken culture in complex YEPC medium at 30°C (Figure 1). A maximum was reached after 6 days of cultivation, which was followed by a rapid decline of activity. Yeast growth was fairly good at 20°C, but the β-glucosidase activity in the supernatants was low. Evaluation of this significant temperature-dependence of enzyme production warrants further investigation.
W. anomalus AS1 was grown for 6 days at 30°C for enzyme purification and the culture filtrates were dialysed and subjected to preparative IEF. The procedure was quite successful, as the activity was always limited to only four to five fractions (pI 3.8–4.2) in 10 separate runs. These fractions showed two main bands at about 35 and 50 kDa in SDS–PAGE without further treatment of the samples (Figure 2a). After concentration and ultradialysis (10 kDa spin columns) of the pooled factions, only the smaller 35 kDa band appeared (Figure 2a). We assume that the ampholytes used to build the pH gradient in the IEF may interact with the glycosidic enzyme and are responsible for the appearance of a second higher molecular form. After the removal of the ampholytes by ultradialysis, a single protein band was visible. When refocusing the concentrate in the Mini Rotofor chamber, two bands again appeared in the SDS gel (not shown). The ampholytes most probably cause an unfolding of the protein, because the same two bands were observed when heating the dialysed sample before SDS–PAGE, but a single 35 KDa band appeared without such treatment (not shown).
Only the 35 KDa band delivered glycosidase activity with MUG-Glc as a substrate after separation by SDS–PAGE and renaturation with Triton-X (Figure 2a). In addition, only one active form focused at ca. pI 4.0 in analytical IEF (Figure 2b). As sensitive silver staining was applied, the experiments underline the high purity of the enzyme preparation obtained by preparative electrofocusing. Periodic acid–Schiff (PAS) staining strongly suggested that the purified enzyme was glycosylated (Figure 3).
Identification and molecular features of WaExg2
The purified enzyme was separated electrophoretically and under native and denatured conditions for protein identification. The corresponding bands, with apparent molecular masses of 35 and 50 kDa, respectively, were subjected to in-gel digestion with trypsin. The resulting peptides were identified by LC–MS/MS and the spectra obtained compared with those of reference strains available in databases. The peptide pattern of the 50 kDa band perfectly matched the exo-β-1,3-glucanase (EC 18.104.22.168) of Pichia anomala strain K (syn. Hansenula anomala, Wickerhamomyces anomalus; NCBI nucleotide Accession No. AJ222862). The smaller protein band of 35 kDa yielded a similar result, confirming that the two molecular masses represent conformational variants of the same protein and not individual enzymes.
The identification was verified by PCR amplification and sequencing of the corresponding glucanase gene of W. anomalus AS1 (available under Accession No. HG514308 in the European Nucleotide Archive: http://www.ebi.ac.uk/ena/data/view/HG514308). Blast research (NCBI) with the amino acids sequence deduced revealed 100% identity with the exo-β-1,3-glucanases of Pichia anomala strain K (Accession No. O93983) and 99% identity with those of W. anomalus strains WRL-O76 (Accession No. AGC67022), BCU24 (Accession No. AFK33205), DSM 6766 (Accession No. AEZ66637), BS91 (Accession No. JQ734566) and YF07b (Accession No. ABK40520).
The gene WaEXG2 of W. anomalus AS1 codes for a protein of 427 amino acids (see supporting information, Fig. 1S). Amino acids 1–17 presumably present a signal peptide for enzyme secretion. The mature protein has a molecular mass of 47 456 Da with a calculated pI of 4.84. The somewhat higher mass of the protein in SDS–PAGE might be due to bound carbohydrates. A corresponding putative N-glycosylation site has been detected in the primary sequence (see supporting information, Fig. 1S). The family 5 glycoside hydrolase signature (Martin et al., 2007) was identified using bioinformatical tools. The five cysteine residues in the protein might build disulphide bonds (see supporting information, Fig. 2S) and confer a high compactness to the molecule. This structure explains the apparently smaller molecular mass (34.6 kDa) of the native enzyme in SDS–PAGE, whereas the chemically reduced and denaturated form is consistent with the theoretical mass. Two presumptive promoter regions with TATA-like elements could be identified in the 5′-flanking region of WaEXG2. These data are in good accordance with those published by Grevesse et al. (2003) for the EXG2 gene of P. anomala strain K. However, no significant similarities exist with a second exoglucanase gene, EXG1, detected in the genome of W. anomalus strains (Muccilli et al., 2013).
Enzymatic hydrolysis of selected glycosides and glycans was demonstrated by detection of cleavage products (glucose, aglycons) using TLC and enzymatic methods (Table 1; see also supporting information, Fig. 3S). The molecular approach mentioned above identified the purified enzyme of W. anomalus AS1 as an exoglucanase. This finding is in line with experimental data using the model substrate laminarin. This polysaccharide, which contains predominantly β-1,3-linked glucose units, is an excellent substrate for the purified enzyme. Only glucose monomers were detected by TLC (see supporting information, Fig. 3S), underlining that the enzyme acts as an exo-β-1,3-glucanase. Cellulose, a polymer consisting of β-1,4-linked glucose residues, was not hydrolysed. In addition, the enzyme efficiently cleaved small β-linked glycosides (arbutin, esculin, polydatin, salicin) and disaccharides (cellobiose, gentiobiose). Rutin was not attacked, presumably because of its low water solubility. Compounds with α-linkages (maltose, melibiose, saccharose, trehalose) were not hydrolysed, or only marginally.
Table 1. Hydrolytic cleavage of some natural and synthetic compounds by WaExg2
Amounts of glucose (mg/ml) detected after enzymatic hydrolysis of 1 mg/ml substrate are given in brackets. In the case of polydatin, the value has been calculated only for the glucose residue.
TLC (≤0.5), enzymatic (0.15)
β-1,3 + β-1,6 (15:1)
TLC (≥0.5), enzymatic (0.4)
TLC (≥0.5), enzymatic (0.28)
TLC (<0.5), enzymatic (0.04)
β-1,3-glucans with some structural relationship to laminarin are the predominant structure components of yeast cell walls (Salazar and Asenjo, 2007). The lytic capacity of WaExg2 against whole yeast cells was measured using a spectrophotometric test assay. Heat-treated cells of all six yeast species tested were lysed at an average degree of 30% (data not shown). By the microscopic observation of W. anomalus AS1 protoplasts generated by the action of WaExg2, we could confirm its activity as a cell wall hydrolase (Figure 4). The cell wall of S. cerevisiae ‘Fermi-Champ’ was obviously less sensitive against WaExg2, indicating some host-specificity of the enzyme.
Activity under wine-related conditions
The suitability of the newly described WaExg2 for practical purposes was tested under wine-related conditions (Figure 5, Table 2). WaExg2 was active under typical wine-related conditions, such as low pH (3.5–4.0), high sugar concentrations (up to 20% w/v), high ethanol concentrations (10–15% v/v), presence of sulphites (up to 2 mm) and various concentrations of cations (K, Ca, Mg, Mn, Cu, Zn). The enzyme was active under wine-making conditions (10–20°C), although only to 2–20% of the maximum activity at 60°C.
Table 2. Influence of metal cations and sulphite on the activity of the exo-β-1,3-glucanase of W. anomalus AS1
Determined at pH 3.5 with β-pNPG as the substrate.
An exoenzyme of high purity could be isolated from the supernatant of W. anomalus AS1 by a single preparative IEF step. The enzyme was identified as a member of the family 5 exo-β-1,3-glucanases (E.C.22.214.171.124) by LC–MS/MS of tryptic peptides and data analysis. Exo-β-1,3-glucanases catalyse the hydrolysis of β-glucan chains by sequentially cleaving glucose residues from the non-reducing end and releasing glucose as the sole hydrolysis product. Besides molecular data, the cleavage pattern of laminarin, releasing glucose as the sole hydrolysis product, assigns WaExg2 as exo-β-1,3-glucanase. Sometimes gentiobiose is also generated if the internal β-1,3-glucosidic linkages adjacent to the β-1,6-glucosyl residues in branched β-1,3, β-1,6 glucans, such as laminarin, are cleaved (Martin et al., 2007). We cannot exclude that small amounts of gentiobiose have been produced from laminarin which were below the detection limit of TLC and were not measured by the enzymatic assay. However, as gentiobiose is also a substrate for WaExg2 (Table 1), the determination of an intermediate disaccharide would be rather difficult. Similar to our results, only monosaccharides were produced from laminarin by a recombinant exo-β-1,3-glucanase from the yeast Cyberlindnera saturnus (formely Williopsis saturnus; Peng et al., 2011).
Besides the authentic exo-β-1,3-glucanase activity, WaExg2 exhibited β-glycosidase activity with a number of synthetic and natural glycosides. A similar broad substrate spectrum has already been reported with other yeast exo-β-1,3-glucanases, e.g. for Pichia pastoris (Xu et al., 2006). Suzuki et al. (2001) concluded that yeast exo-β-glucanases may be classified as a new type of β-glucanases or β-glucosidases. Accordingly, different physiological functions have been attributed to yeast exoglucanases, from involvement in cell wall morphogenesis and nutrient mobilization (Martin et al., 2007) to antagonistic toxins (Izgü et al., 2006, Friel et al., 2007). Exg2 enzymes are among 10 proteins representing the core secretome of different yeast species (Buerth et al., 2011).
The glucanase activity of WaExg2 is of considerable interest to winemaking. During vinification, bacteria and fungi can spoil wine quality by producing slimes consisting of β-1,3-glucans similar to laminarin. Must and wine filtration in these circumstances can be difficult or impossible. Therefore, there is a need for measures to degrade such exopolysaccharides. Blättel et al. (2011) described an novel bacterial β-1,3-endoglucanase with regard to its ability to hydrolyse β-1,3-glucan slimes in wine and cell walls of different yeast species. Similarly, in the present study, we demonstrate the capacity of WaExg2 to hydrolyse laminarin and yeast cells. The latter activity may also facilitate the release of beneficial cell wall components and cytoplasmic proteins from (sparkling) wine yeasts (Zinnai et al., 2010; Torresi et al., 2014).
The role of exo-β-1,3-glucanases from P. anomala strain K as antagonistic ‘toxins’ against fungi is a matter of debate. Results of Grevesse et al. (2003) ruled out any involvement, whereas Friel et al. (2007) demonstrated, by gene knock-out experiments, that PaExg2 is essential for the inhibition of the phytopathogenic fungus B. cinerea. However, it should be considered that enzyme systems for yeast cell lysis are usually a mixture of several different enzymes, including one or more β-1,3- and β-1,6-glucanases, proteases, mannanases or chitinases, which act synergetically to lyse the cell wall (Salazar and Asenjo, 2007). The possible contribution of WaExg2 as a biocontrol factor will be elucidated in future experiments.
Polyphenols in red wine, such as resveratrol, have gained increasing public and scientific interest, due to their presumed beneficial impact on human health (El Rayess, 2014). A major section of the polyphenols in nature are conjugated with sugars or organic acids, rendering them more hydrophilic and less bioavailable (Thilakarathna and Rupasinghe, 2013). The level of glycosylated forms of resveratrol, known as piceid or polydatin, have been found in red wines as much as 10-fold higher (El Rayess, 2014). As these modified forms are less bioactive, attempts have been made with β-glucosidases from different fungal sources to increase the trans-resveratrol content in wines by hydrolysing glycosylated precursors (El Rayess, 2014). WaExg2 released the polyphenolic aglycons from the model compounds arbutin, salicin, esculin and polydatin. In this respect, the glycosidase activites of W. anomalus AS1 might be useful for raising the level of polyphenols beneficial to health in wine.
As β-glucosidases are usually inhibited by glucose, the sugar tolerance of WaExg2 is one of its outstanding features. Similar beneficial properties were reported for β-glucosidases produced by a few other yeast strains (Swangkeaw et al., 2009; Baffi et al., 2011). In addition to sugar tolerance, WaExg2 proved to be stable under various wine-related conditions. Further studies are necessary to evaluate the influence of potential inhibitors (polyphenols, bentonite) on enzyme activity.
In conclusion, the multifunctional activities of WaExg2 have promising biotechnological potential. The species W. anomalus currently has Qualified Presumption of Safety (QPS) status from the European Safety Authority (ESFA), and this has benefits in terms of public perspectives on food biotechnology and the acceptability of novel microbiological products in food (Sundh and Melin, 2011).
The authors thank Dr Stefan Tenzer (Research Centre Immunology, Johannes Gutenberg University, Mainz, Germany) for mass spectroscopic protein identification. This study was financially supported by the Stiftung Rheinland-Pfalz für Innovation (Rhineland Palatinate, Germany; Project No. 961-38621/1051).