To clone and characterize two related intracellular esterases from Oenococcus oeni and Lactobacillus hilgardii under wine-like conditions.
To clone and characterize two related intracellular esterases from Oenococcus oeni and Lactobacillus hilgardii under wine-like conditions.
The published genome sequences for O. oeni and Lact. hilgardii were used to identify, clone and purify putative esterase genes from these species designated EstCOo8 and EstC34, respectively. Both esterases are members of family V of lipolytic enzymes. However, EstC34 contains an SGSLG nucleophilic elbow structural motif instead of the usual GGSLG motif which is conserved in other lactic acid bacteria. Both esterases exhibited greatest specificity for C2–C4 pNP-linked substrates and retained activity under wine-like conditions. EstCOo8 had an optimum temperature, pH, and ethanol concentration of 40°C, 5·5 and 6% (v/v), respectively. Whereas EstC34 had an optimum temperature, pH and ethanol concentration of 50°C, 5·0 and 10% (v/v), respectively.
Both esterases were stable and retained activity under conditions that would be encountered in wine. They have the potential to reduce short-chain ethyl esters such as ethyl acetate.
This study provides information that might help improve the performance of LAB during malolactic fermentation in wine in the future, either by strain selection, optimization or direct enzyme addition.
Lactobacillus spp., Pediococcus spp. and Oenococcus oeni are commonly isolated from wine, the latter being typically used in winemaking to carry out malolactic fermentation (MLF) and deacidify wine (Davis et al. 1985; Carr et al. 2002). Lactobacillus hilgardii has been identified as the most predominant Lactobacillus spp. in wine. Rodas et al. (2005) isolated 178 Lactobacillus spp. strains from wine, 71 of these being Lact. hilgardii. Although many Lactobacillus spp. can carry out MLF in wine, only a few can survive the alcoholic fermentation (Dicks and Endo 2009). Lact. hilgardii is resistant to high acid and alcohol contents (Dicks and Endo 2009); however, Lact. hilgardii can produce undesirable metabolites such as biogenic amines, and this species is generally regarded as a wine spoilage bacterium (Fernandez et al. 2010). However, this does not preclude its use as a source of enzymes which could be potentially utilized under the harsh conditions found in wine and perhaps other industrial settings. In fact, a recent review by du Toit et al. (2011) illustrates the potential of using Lactobacillus spp. such as Lact. hilgardii to carry out MLF and alter wine aroma compounds and invites further investigations into their use as MLF starter cultures. Due to the ability of Lact. hilgardii to survive in the harsh wine environment, wine isolate Lact. hilgardii CSCC-5489 was chosen as a potential source of novel esterases.
Oenococcus oeni is the species most commonly used to carry out MLF in wine. It is acidophilic and indigenous to wine and similar environments. While there are now three strains of O. oeni with published genome sequences (CP000411·1, AAUV00000000, ACSE00000000·1), there is still limited information on the function of esterase genes and their potential contribution to food and beverage aroma. Previous studies have highlighted the potential of wine-associated lactic acid bacteria (LAB) as a source of novel purified esterase enzymes for use as additives in winemaking (Matthews et al. 2004; Sumby et al. 2009). Analogous characterization of other flavour-related genes and enzymes may not only have practical implications for processes using LAB but are also of fundamental interest.
Aroma compounds and the quantities in which they are present can play an important role in determining the quality of fermented food products such as wine (Sumby et al. 2010). A large number of aroma compounds are formed during the primary fermentation by yeast; after this, LAB can contribute by increasing and decreasing the ester concentration during MLF (Avedovech et al. 1992; Bartowsky and Henschke 1995; Maicas et al. 1999; Delaquis et al. 2000; Pozo-Bayón et al. 2005; Ugliano and Moio 2005). As a group, esters are a quantitatively significant constituent of beverages such as wine (total of 100 mg l−1) (Sumby et al. 2010). The fact that most esters are present in wine at concentrations around their threshold value implies that minor concentration changes might have a dramatic effect on wine flavour. For this reason, an understanding of the hydrolysis and synthesis of esters in winemaking and how these may be manipulated is essential.
Ester hydrolysis and synthesis can be catalysed by esterases (Inoue et al. 1997; Bornscheuer 2002; Liu et al. 2004; Torres et al. 2008). These enzymes commonly contain a catalytic triad composed of Ser, His, and Asp/Glu residues and a nucleophilic elbow structural motif (GXSXG, in which X is a variable residue), which contains the active-site serine residue (Arpigny and Jaeger 1999; Sumby et al. 2010). They also contain an oxyanion hole, of which two residues donate their backbone amide protons to stabilize the substrate in the transition state (Pleiss et al. 2000). While extensive research has been carried out on the yeast enzymes responsible for ester formation by wine strains of Saccharomyces cerevisiae, esterase activity for wine-related LAB is not well documented. Most characterization of esterases in LAB has focused on dairy isolates (Fenster et al. 2000; Fernandez et al. 2000; Fenster et al. 2003a,b; Choi et al. 2004). Parallel work in a wine context is limited despite general acceptance of the importance of esters in wine. Our characterization of esterase EstB28 from O. oeni wine strain Ooeni28 was the first report of the characterization of such enzymes at the biochemical or genetic level in this organism (Sumby et al. 2009). Until recently, most evidence that wine LAB possess esterase activity came from wine volatile profiling studies which investigated the changes in concentration of individual esters during MLF (Maicas et al. 1999; Delaquis et al. 2000; Ugliano and Moio 2005). Such changes in ester concentration were strain specific and had the potential to greatly affect the final aroma of wine.
Our survey of the esterase activities of whole LAB cells found variations within species and even greater variation between the genera (Matthews et al. 2006), with O. oeni showing greatest activity towards the p-nitrophenyl (pNP)-linked substrates tested. Lact. hilgardii strain Lac34 showed the highest activity of the lactobacilli strains tested. More recently, we determined the esterase activities of whole O. oeni, lactobacilli and pediococci cells under conditions with some relevance to wine (Matthews et al. 2007). At least partial resistance to the harsh conditions used was observed, thereby demonstrating a necessary requirement of any enzyme intended for application in analogous environments. To more completely characterize esterases of LAB, the enzymes and their structural genes must be fully investigated.
As an extension of our previous investigations into the esterases of wine-related LAB, two putative esterase genes, designated estCOo8 and estC34, were identified based on sequenced genomes in O. oeni (AAUV00000000) and Lact. hilgardii (ACGP00000000). The estC gene was also sequenced in twenty strains of O. oeni to determine whether the presence of a stop codon in this gene could be correlated with a decrease in whole-cell activity. This study represents an effort to dissect the complex array of ester synthesis and hydrolysis activities in whole-cells by cloning, heterologous expression, partial purification and biochemical characterization of two related esterase proteins from O. oeni and Lact. hilgardii. With a view to applying such an esterase under conditions found in wine and perhaps other industrial settings, enzyme function under the harsh physicochemical conditions frequently encountered in wine was examined.
All lactic acid bacteria strains were isolated from commercially available freeze-dried LAB starter cultures for use in winemaking. The strains used in this study were 20 O. oeni strains; Ooeni1, Ooeni2, Ooeni3, Ooeni4, Ooeni6, Ooeni8, Ooeni9, Ooeni10, Ooeni11, Ooeni12, Ooeni13, Ooeni14, Ooeni17, Ooeni20, Ooeni21, Ooeni22, Ooeni23, Ooeni28, Ooeni30, Ooeni31 and one Lact. hilgardii strain Lac34 (Matthews et al. 2006). Oenococcus oeni strain Ooeni8 isolated from CHR Hansen Viniflora and Lact. hilgardii CSCC-5489 (Lac34) purchased from the Australian Starter Culture Research Centre were used as the source of estC genes for cloning experiments. In preparation for experiments, LAB were cultured from the stocks in de Man-Rogosa-Sharp (MRS) medium (Amyl Media) supplemented with 20% (v/v) apple juice. Escherichia coli DH5α was used as a host strain for cloning and maintenance of plasmids. E. coli BL21(λDE3) was used as a host for expression of the estC genes under the control of the T7 promoter. E. coli transformants were grown at 37°C in either Luria-Bertani broth (LB) or Terrific broth (for expression only), which contained 100 μg ml−1 of ampicillin when required. All bacteria were stored at −80°C in 30% (v/v) glycerol.
Isolation of DNA from LAB was carried out using the UltraClean microbial DNA isolation kit (MoBio). DNA was extracted from 20 strains of O. oeni for analysis of estC sequence. Primers targeting the open reading frame (ORF) of the putative O. oeni and Lact. hilgardii esterase genes were designed (Table 1) based on the published ATTC BAA_1163 O. oeni genome sequence (AAUV00000000) and previously published LAB esterase sequences. PCR, plasmid preparation, agarose gel electrophoresis and DNA manipulations were performed as described previously (Sumby et al. 2009). T4 DNA ligase and other DNA-modifying enzymes were purchased from New England Biolabs Inc. and were used as recommended. Bacterial transformation was performed using chemically competent E. coli cells and heat shock at 42°C for 2 min.
|Gene||Primer sequence (5'–3')a||Restriction site|
The estC gene sequences of O. oeni strains and of both strands of the plasmid constructs were confirmed by nucleotide sequencing, with labelling and capillary separation on the AB 3730xl performed by the Australian Genome Research Facility (Brisbane). Either sequencing primers were identical to those used for cloning (Table 1) without the multiple cloning site, or in the case of plasmid construct sequencing, T7 promoter and terminator primers were used. Nucleotide sequencing results were analysed using Chromas Lite ver. 2·01 (Technelysium Pty. Ltd, Australia), BioEdit (Hall 1999) and NCBI Blast (www.ncbi.nlm.nih.gov/BLAST).
Plasmids pET14.b//estCOo8 and pET14.b//estC34 were constructed by inserting the amplified estC esterase genes into the multiple cloning site of pET14.b (Novagen; Merck Millipore, Kilsyth, Vic., Australia). The ligation mixture was used to transform E. coli DH5α cells. A positive clone was selected on LBA plates (LB broth with 2% bacteriological agar) containing ampicillin, and the presence of the insert was initially confirmed by whole-cell PCR using the T7 promoter and terminator primers (Novagen). Positive clones were then sequenced, and the sequenced plasmid was then transformed into E. coli BL21(λDE3). Target proteins were produced using this clone and contained a polyhistidine tag and a thrombin cleavage site attached to the N-terminal region for purification purposes.
Escherichia coli BL21(λDE3) harbouring the pET14.b plasmid containing the genes of interest was used as the source of the recombinant enzyme. Induction of cells and purification of recombinant esterases was performed as described previously (Sumby et al. 2009), with the exception that the recombinant enzymes were eluted at 0·5 ml min−1 using elution buffer (50 m mol l−1 sodium phosphate, 300 mmol l−1 sodium chloride, 150 mmol l−1 imidazole, pH 7·0). Chromatography data were recorded using LP Data View software (Bio-Rad Laboratories Pty., Ltd, Gladesville, NSW, Australia). The molecular mass of the purified recombinant esterase was determined by sodium dodecyl sulphate (SDS)-polyacrylamide gel electrophoresis with a 5% (w/v) stacking gel and a 12% (w/v) separating gel, which was made and run according to the manufacturer's protocols (Bio-Rad).
Esterase assays were conducted using pNP-butanoate substrate, with the exception of the substrate specificity trials, in which eight different pNP-linked ester substrates were used, as described below. The influence of pH (3·0–8·0), temperature (10–60°C) and ethanol (2–22%, v/v) on esterase activity was determined following the previously published method (Sumby et al. 2009). To determine the specificity of EstCOo8 and EstC34 for different ester substrates, eight pNP-linked esters with different carbon chain lengths were used: pNP-acetate, -butanoate, -octanoate and -decanoate (Sigma; Sigma-Aldrich Pty. Ltd, Sydney, NSW, Australia); pNP-hexanoate (TCI); and pNP-dodecanoate, -tetradecanoate and -octadecanoate (Fluka; Sigma-Aldrich Pty. Ltd, Sydney, NSW, Australia). Assays were carried out as described previously (Sumby et al. 2009). All assays and controls were performed in triplicate, and activities were quantified by comparison with p-nitrophenol standards.
Analysis of variance of the data was conducted using Microsoft Excel and GraphPad Prism, ver. 5·01 (GraphPad Software, Inc., La Jolla, CA, USA, www.graphpad.com).
Based on the published sequences of LAB lipolytic enzymes, ORFs encoding a putative esterase gene in O. oeni and a putative esterase gene in Lact. hilgardii were identified and successfully amplified from strains Ooeni8 and Lac34. These esterases were originally identified by sequence homology using the published nucleotide sequence for O. oeni ATCC BAA-1163 and Lact. hilgardii ATCC 8290. Nucleotide sequencing of the esterase from O. oeni Ooeni8 revealed a 771-bp ORF which was designated estCOo8 (Fig. 1). Nucleotide sequencing of the esterase from Lact. hilgardii Lac34 revealed a 780-bp ORF which was designated estC34 (Fig. 2). The translated ORFs reveal sequence which could encode a protein of 29·0 kDa or 29·1 kDa, respectively. The translated proteins show 61% sequence identity with each other (76% positive).
The deduced amino acid sequences of EstCOo8 and EstC34 were analysed for key lipolytic enzyme motifs and nucleophilic elbow pentapeptides were identified (Figs 1 and 2). EstCOo8 belongs to family V of bacterial lipolytic proteins (Arpigny and Jaeger 1999) with a putative nucleophilic elbow GGSLG. The deduced amino acid sequence for EstC34 differs from the standard GXSXG motif at this site and is instead SGSLG. Based on sequence homology EstC34 also belongs to family V of bacterial lipolytic proteins (Arpigny and Jaeger 1999). The putative catalytic Asp/Glu and his residues of EstCOo8 and EstC34 were identified (Figs 1 and 2), based on sequence alignments with related sequences from a variety of bacterial strains based on an NCBI blast search. The sequence patterns conserved around the active-site residues of family V enzymes aligned with EstCOo8 and EstC34 are presented in Fig. 3.
The estCOo8 sequence was aligned with the O. oeni strain ATCC BAA-1163 genome sequence using NCBI Blast. The results indicate differences at 22 nucleotides in Ooeni8 compared with ATCC BAA-1163, resulting in 13 changes when translated to the amino acid sequence (Fig. 1). The deduced amino acid sequence of the putative esterase gene was compared with other known and predicted esterase sequences available from GenBank (NCBI database). The published esterase to which EstCOo8 showed highest homology was EstC from Lact. casei LILA (58% identity) (Fenster et al. 2003a). The EstC34 sequence was aligned with the Lact. hilgardii ATCC 8290 genome sequence using NCBI Blast. The results indicate no differences in Lac34 nucleotide sequence compared with Lact. hilgardii ATCC 8290 (Fig. 2). The deduced amino acid sequence of the putative esterase gene was compared with other known and predicted esterase sequences available from GenBank (NCBI database). The published esterase to which EstC34 showed highest homology was EstC from Lact. casei LILA (99% identity) (Fenster et al. 2003a). There are only two differences in the deduced amino acid sequence of EstC34 compared with EstC from Lact. casei LILA.
The estC gene in O. oeni strain PSU-1 contains a termination codon at position 83 of the translated amino acid sequence. To analyse the occurrence of this throughout O. oeni strains the estC gene was sequenced in 20 strains from a variety of sources, and three strains with published sequences were included in the analysis. Of these, only five of the sequenced strains and one published strain do not have a stop codon in the middle of the sequence (Table 2). The full estC sequence of Ooeni8 and the amino acid translation is shown in Fig. 1, along with the position of the stop codon in 20 of the sequenced strains. The acquired sequence data were compared with previously published results for these strains (Matthews et al. 2006). There is no correlation between having a termination codon in the middle of estC and whole-cell ester hydrolysis activity. This suggests that the estC gene is not essential for the observed ester hydrolysis activity of the whole cell.
|No Termination codon||Termination codon at position 83 (before active site)||Termination codon at position 102 (after active site)|
|Ooeni8, Ooeni13, Ooeni14, Ooeni20, Ooeni23, ATTC BAA_1163||Ooeni9, Ooeni12, Ooeni30, Ooeni28, DSMZ_20252, DSMZ_20255, PSU_1, AWRI B429_1604||Ooeni1, Ooeni2, Ooeni3, Ooeni4, Ooeni6A, Ooeni10, Ooeni11, Ooeni17, Ooeni21, Ooeni22, Ooeni31|
The putative esterase genes were cloned into pET14.b, and once positive clones had been sequenced, one was chosen for further experiments. The E. coli BL21(λDE3) cells containing the plasmid of interest were induced and cell extracts obtained for analysis by SDS-polyacrylamide gel electrophoresis. The induced BL21(λDE3) cells containing pET14.b//estCOo8 or pET14.b//estC34 overexpressed a protein of approximately 29 kDa (Fig. 4). Crude cell extracts were assayed for activity (data not shown), and BL21(λDE3) containing pET14.b//estCOo8 or pET14.b//estC34 had 25-fold greater activity than BL21(λDE3) containing pET14.b. EstCOo8 and EstC34 were purified from the whole-cell lysate to electrophoretic homogeneity using affinity chromatography (Fig. 4). The purified enzymes could be stored, without significant loss of activity, for several weeks at –80°C in elution buffer supplemented with glycerol (10%, v/v) (data not shown).
Having established that the product of the cloned estCOo8 and estC34 genes possessed esterase activity, further characterization of these products commenced. Substrate specificity was determined using pNP-linked esters of various acyl chain lengths (C2 to C18) at 37°C. Both esterases had greatest activity on substrates with a chain length of up to C10, although EstC34 activity was more restricted to pNP-acetate (C2) and pNP-butanoate (C4) (Figs 5a and 6a). Minimal activity was detected with the chain lengths of C12 to C18. The influence of pH, in the range of 3·0–8·0, on esterase activity was tested at 37°C (Figs 5b, 6b). Although both enzymes were active at all of the pH values tested, the highest activity was observed at pH 5·5 for EstCOo8 (Fig. 5b), and pH 5·0 for EstC34 (Fig. 6b). The esterases were also active at temperatures of between 10 and 60°C, with the highest activity being observed at 40°C for EstCOo8 (Fig. 5c) and 50°C for EstC34 (Fig. 6c). The influence of ethanol on esterase activity was tested at 37°C, and highest activity was observed at an ethanol concentration of 6% (Fig. 5d) with EstCOo8 and 10% with EstC34 (Fig. 6d).
This research focused on the cloning and biochemical characterization of two esterase enzymes designated EstCOo8 and EstC34. Both esterases are members of family V of lipolytic enzymes. This group of enzymes originates from a wide variety of bacteria, including mesophilic bacteria and both psychrophilic and thermophilic organisms (Arpigny and Jaeger 1999). The putative catalytic triad of EstCOo8 and EstC34 composed of Ser 95, Asp 209 and His 236 was identified based on sequence alignments with other family V esterases and lipases. EstCOo8 and EstC34 showed no potential signal sequences (Nakai and Horton 1999), suggesting that these enzyme are located in the cytoplasm.
The esterases of O. oeni Ooeni8 (EstCOo8) and Lact. hilgardii Lac34 (EstC34) were overexpressed in E. coli and purified using affinity chromatography. The purified recombinant enzymes were then characterized under physiochemical conditions relevant to winemaking. The potential role of the EstC enzymes, when used as an additive in wine, in aroma development would be partly dependent on the sensitivity of the enzyme to environmental conditions encountered therein. Low pH values, in the range 3·0–3·6, are sought for both grape juice and wine and are monitored throughout the winemaking process. At these values, both EstC esterases were stable, with EstCOo8 retaining 12% and EstC34 retaining 20% of the relative activity, respectively (Figs 5b and 6b). The EstC esterases were also active at temperatures of between 10 and 60°C, and EstCOo8 retained 57% of its activity at 20°C and 36% at 10°C relative to that observed under optimal conditions (Fig. 5c), while EstC34 retained 49% of its activity at 20°C and 30% at 10°C relative to that observed under optimal conditions (Fig. 6c). This suggests that both enzymes have the potential to affect the ester profile of wine even at these low temperatures as may be encountered during wine storage or the MLF. Unlike EstB28 (Sumby et al. 2009), EstC activity did not continually increase with increasing ethanol concentration. EstCOo8 had highest activity at an ethanol concentration of 6%, and EstC34 had highest activity at an ethanol concentration of 10%. Nevertheless at 12% ethanol, a typical level found in wines, activity was still at 106% for EstCOo8 and 140% for EstC34 relative to 0% ethanol (Figs 5d and 6d).
The substrate specificity of both EstCOo8 and EstC34 revealed greater selectivity for pNP esters of short-chain fatty acids. EstC34 demonstrated a marked drop-off in relative activity towards pNP esters of medium-chain fatty acids compared with EstCOo8. The specificity of these esterase for pNP-acetate and pNP-butanoate is similar to that of EstC from Lact. casei LILA (Fenster et al. 2003a), which was also most active against short-chain fatty acids. Examples of such esters in wine include ethyl acetate and ethyl butanoate. The removal of these from wine is not necessarily desirable, as they can contribute to the fruity aroma of wine at low concentrations (Sumby et al. 2010). However, the removal of excess ethyl acetate, which at high concentrations (≥100 mg ml−1) contributes a solvent-like aroma to wine, would be an advantage. The role that EstCOo8 and EstC34 plays in ester synthesis will also need to be investigated to determine the usefulness of adding whole cells of selected strains or else a preparation of the purified enzyme to wine with a view to selectively modifying the ester profile.
To determine whether EstC from O. oeni could be important in explaining the differences in esterase activity that is observed in strains of this bacterium sequence data were compared with previous whole-cell activity experiments (Matthews et al. 2006). There was no correlation between strains containing a stop codon in the middle of estC and whole-cell activity, implying that the presence of an EstC homologue is not necessary for high whole-cell activity. Whether estC has an effect on strain-specific differences in ester hydrolysis and synthesis in wine is yet to be determined. EstC34 has an altered nucleophilic elbow motif, with the first conserved Gly residue replaced with a Ser (SGSLG). It is unclear how much of an effect this has on enzyme activity, but it is noteworthy that the relative activity of EstC34 was 2·2-fold lower than EstCOo8.
As EstC34 had the lower relative activity compared with EstCOo8, a Blast search using protein sequences from previously published characterized esterases was preformed to check for the presence of other esterase genes in Lact. hilgardii. The results showed no homologues in Lact. hilgardii of; EstA from L. lactis subsp. lactis (AF157484), EstA from Lact. helveticus CNRZ32 (AF136284) or EstI from Lact. casei CL96 (AY251019). The search however determined up 3 putative esterases showing homology to EstB from Lact. casei LILA (AF494421). The three sequences showing homology to EstB, EEI23149 (29·3kDa), EEI24546 (32kDa) and EEI25207 (26·8kDa), have 30, 33 and 20% identity to EstB from Lb. casei, respectively. All three have a putative nucleophilic elbow motif (GXSXG) and a GGGX N-terminal oxyanion hole residue, which suggests that they are putative esterase genes. The presence of these three EstB homologues could explain why Lac34 had the highest activity of all lactobacilli tested (Matthews et al. 2006). However, Lac34 activity was still lower than O. oeni, a fact that could be due to the lack of an EstA homologue (two in O. oeni) (Sumby et al. 2012).
To conclude, EstC34 is the first esterase from the wine-associated species Lact. hilgardii to be characterized under conditions relevant to winemaking. Based on the findings reported for the artificial substrates used in this study, it appears that EstCOo8 and EstC34 will retain at least partial activity under wine-like conditions of pH, temperature and ethanol. Both the O. oeni Ooeni8 and Lact. hilgardii Lac34 esterases are potential candidates for altering the ester profile of wine and could play an important role in ester profile modifications during the malolactic fermentation. Both of the two EstC esterases have similar potential for altering the ester profile of wine and will need to be examined further along with other characterized esterase enzymes to confirm this. Also a study comparing the ester profile in wine after MLF with O. oeni strain Ooeni8 and with strains producing a truncated EstC could explain some of the strain-specific differences in ester profile after MLF. As most esterases are also capable of ester synthesis, investigation of the role that this enzyme might play in wine would be of interest.
The research in this article was undertaken as part of project UA 05/01 supported by Australia's grape growers and winemakers through their investment body, the Grape and Wine Research Development Corporation, with matching funds from the Australian Government. K.S. is supported by a Grape and Wine Research Development Corporation scholarship. The University of Adelaide is a member of the Wine Innovation Cluster in Adelaide (wineinnovationcluster.com).