Application of the Different Electrophoresis Techniques to the Detection and Identification of Lactic Acid Bacteria in Wines

were separated with 8% polyacrylamide gels containing a 32 to 50% urea-formamide gradient. Electrophoresis was performed for 10 min at 20 V, and 16 h at 60 V at a constant temperature of 60 °C. The DGGE gels were stained in ethidium bromide after the electrophoresis and then were visualized with UV transiluminattion (GelDoc, Bio-Rad). Blocks of polyacrylamide gels contained selected and in 20 μ L sterile and at ºC to make One microliter used PCR Oenococcus oeni comparison of the patterns from and genotypes from four genotypes. of these four genotypes one to three years, in some cases with a high frequency of appearance, significance of study in order these genotypes a future selection process.


Introduction
The microorganisms play an essential role in winemaking since a mixed culture of numerous microorganisms including fungal, yeast, and bacteria species are involved in this process and are the responsible for the final quality of the wine (Bisson et al., 1993). Therefore, in order to control the fermentation processes knowing and understanding the complex microbiota involved in them is necessary.
Yeast are able to convert sugar from grapes into ethanol and many other changes that lead to wine. Lactic acid bacteria (LAB) that are often present on the surface of the grapes and can represent significant populations in musts (Lonvaud-Funel, 1999) play dual roles in wine fermentations: as wine spoilage agents and as the main effectors of malolactic fermentation (MLF). Numerous studies have been conducted on the LAB that occur on grapes, grape musts and wines and it is generally agreed that a succession of species happens during the different stages of winemaking and conservation of wines (Ribéreau-Gayon et al., 2006). Most bacterial species present in wine fermentations have been identified by traditional microbiological techniques involving cultivation. However, as it was observed with microbial ecology studies of other environments, cultivation-dependent methods often exhibit biases resulting in an incomplete representation of the true present bacterial diversity (Amann et al., 1995;Hugenholtz et al., 1998). Applications of cultureindependent molecular techniques to monitor the microbial successions of various food and beverage fermentations have revealed microbial constituents and microbial interactions not witnessed by previous plating analyses (Giraffa & Neviani, 2001).
Denaturing gradient gel electrophoresis (DGGE) and temperature gradient gel electrophoresis to separate bacterial 16S ribosomal DNA (rDNA) amplicons are common culture independent methods employed to characterize microbial communities from specific environmental niches (Lopez et al., 2003;Muyzer & Smalla, 1998). These approaches are attractive since they enable to detect individual species as well as to get overall profiling of community structure changes with time.
Otherwise, ecology, interactions and development of the different bacterial strains during alcoholic fermentation (AF) and MLF are still a field of active research. Efficient and precise methods of strain identification and discrimination have been developed during the last years, either to prepare well-defined starters of biotechnological interest in winemaking, or to quickly assess the presence of certain strains in a wine, or to gain insight of such a complex ecosystem as wine.
For all these reasons, the aims of this work were: (a) applying the DGGE, PFGE and RAPD-PCR techniques to the analysis of the LAB species diversity and the intraspecific diversity of Oenococcus oeni in a winery of La Rioja region during three consecutive vintages, (b) getting a better knowledge of bacterial ecology throughout both AF and MLF in wine elaborated with Tempranillo, the classic red grape variety of Spain and native of Appellation of Origin Rioja, (c) evaluating the occurrence of genotypes from commercial Oenococcus oeni strains between the autochthonous Oenococcus oeni strains isolated from non inoculated fermenting wines, and (d) contributing to the maintenance of the LAB biodiversity in Rioja red wines.

Wine production and wine samples
Traditional red wine fermentations from c.v. Tempranillo local grapes of 2006, 2007 and 2008 vintages at one winery of the Spanish northern region of Rioja were studied. Winemaking practices were the typical of this wine-producing area: AFs were conducted in the presence of grape skins, seeds and stalks, after the addition of sulphur dioxide and until the residual reducing sugar content was under 2 g/L. At this final point of AF, wines were drawn off into tanks and were allowed to undergo spontaneous MLF with the endogenous microbiota. The sampled winery had never used commercial starters for MLF. One fermentation tank was sampled in each vintage. Wine samples were collected aseptically for chemical and microbiological analysis at different times: must (stage 1), tumultuous AF (density around 1,025; stage 2), at final AF (< 2 g/L glucose + fructose; stage 3), initial MLF (consumption of 10% of the initial malic acid; stage 4), tumultuous MLF (consumption of 60% of the initial malic acid; stage 5) and at final MLF (L-malic acid concentration < 0.5 g/L; stage 6).

Commercial Oenococcus oeni starter samples
Sixteen commercial starter cultures employed to induce MLF derived from six different companies were analyzed. These commercial cultures were selected between the most frequently used in Spain.

Chemical analysis of the musts and wines
Alcohol degree, pH, total acidity, volatile acidity, reducing sugars, free and total sulphur dioxide and L-malic and L-lactic acid content were measured according to the European Community Official Methods (European Community, 1990).
Every commercial lyophilized starter culture was hydrated in saline solution (0.9% NaCl) and then 100 μL aliquot from the appropriate dilution was plated at the surface of modified MRS agar without pymaricine. Because of their low viability in laboratory conditions (Maicas et al., 1999a;Maicas et al., 1999b) glycine (40 mM) and ethanol (10% v/v) were added to this medium. The plates were incubated for at least 10 days at 30 ºC under anaerobic atmosphere (Gas Pak System, Oxoid Ltd.) and five colonies were isolated from each one.

Species identification
Species identification was carried out by previously recommended methods, which included bacteria morphology, Gram staining, and catalase (Holt et al., 1994). Oenococcus oeni, Lactobacillus plantarum and Lactobacillus brevis species were confirmed by the speciesspecific PCR method (Beneduce et al., 2004;Zapparoli et al., 1998). In case of the identification of other species, PCR amplification of partial 16S rRNA genes was performed with WLAB1 and WLAB2 as previously described (Lopez et al., 2007). PCR products were sequenced by Macrogen Inc. (Seoul, South Korea) and sequences were used for comparison to the data in GenBank using the Basic Local Alignment Search Tool (BLAST) (Altschul et al., 1990).

Oenococcus oeni typification by PFGE
PFGE was carried out according to the method described by Birren et al. (1993) with some modifications (Lopez et al., 2007) for agarose block preparation. Because of the difficulty of typing commercial strains first of all these cells underwent to fifteen minutes of ultrasounds, moreover a higher quantity of lysozime (100 μL/block) was added and incubated for 2 h. Macrorestriction analysis was performed with two endonucleases: SfiI, following the method reported by López et al. (2007), and ApaI by the method reported by Larisika (2008) with the following modifications for optimal separation of fragments: 1.2% (w/v) agarose gels were submitted to 24 h with a pulse ramping between 0.5 and 20 s at 14 ºC and 6 V/cm in a CHEF DRII apparatus (Bio-Rad).

Oenococcus oeni typification by RAPD-PCR
RAPD-PCR was carried out following the procedure described by Ruiz et al. (2010b) with some modifications: MgCl 100 mM, dNTP 50 mM and primer M13 100 mM. RAPD-PCR reaction was developed in a total volume of 50 μL and it was carried out with a Perkin Elmer, GeneAmp PCR System 2400 thermocycler. 20 μL of amplified products were resolved by electrophoresis in a 1.4% agarose gel in 0.5x TBE (45 mM Tris base, 89 mM, boric acid, 2.5 mM EDTA pH 8) for 3 h at 70 V.

Numerical analysis of PFGE and RAPD-PCR images
The conversion, normalization and further processing of images were carried out by InfoQuest TM FP software version 5.10 (Bio-Rad, USA). Comparison of the obtained PFGE patterns was performed with Pearson's product-moment correlation coefficient and the Unweighted Pair Group Method using Arithmetic averages (UPGMA). Comparison of the pulse types from the PFGE and RAPD was made by composite data set comparison with average molecular analysis by Unweighted Pair Group Method using Arithmetic averages (UPGMA) .

Direct DNA extraction from wines samples
A volume of 10 mL of each must or wine sample was centrifuged (30 min, 10000xg, 4 ºC). The supernatant was discarded and 1.2 mL of saline solution (NaCl 0.9%) and 2.4 mL of zirconium hydroxide (7 g/L) were added to the pellet to facilitate pelleting of the bacteria in wine (Lucore et al., 2000). After horizontal shaking during 10 min at room temperature, the suspension was again centrifuged (10 min, 500xg, 7 ºC) and finally DNA was purified from the cell pellet by using a PowerSoil ® DNA isolation kit (MO BIO Laboratories, Inc., Carlsbad, CA USA) as per the manufacturer's instructions.

PCR conditions
PCR was performed using an Applied Biosystem, GeneAmp ® PCR System 2700 thermocycler at a final volume of 50 μL. To amplify the region V4 to V5 of 16S rDNA gene, primers WLAB1 and WLAB2 GC were used as López et al. described (2003). Moreover, primers rpoB1, rpoB1o, and rpoB2 GC were employed to amplify the region of the rpoB gene as it was described by Renouf et al. (2006) with the next modifications: 0.5 μM of each primer, 1 mM dNTP mix and 0.5 μL of PfuUltra II Fusion HS DNA Polymerase (Stratagene).

PCR-DGGE analysis
The separation of the respective PCR products was performed with the D-CODE TM universal mutation detection system (Bio-Rad, Hercules, Calif.). PCR products obtained from WLAB1-WLAB2 GC primers were run on 8% (wt/V) polyacrilamide gels in a running buffer containing 2 M Tris base, 1 M Glacial acetic acid and 50 mM EDTA pH 8 (TAE), and a denaturing gradient from 35 to 55% of urea and formamide. The electrophoresis was performed at 20 V for 10 min, and 80 V for 18 h at a constant temperature of 60 °C. PCR products generated with the rpoB1, rpoB1o, and rpoB2 GC primers were separated with 8% polyacrylamide gels containing a 32 to 50% urea-formamide gradient. Electrophoresis was performed for 10 min at 20 V, and 16 h at 60 V at a constant temperature of 60 °C. The DGGE gels were stained in ethidium bromide after the electrophoresis and then were visualized with UV transiluminattion (GelDoc, Bio-Rad). Blocks of polyacrylamide gels which contained selected DGGE bands were excised and later incubated overnight in 20 μL of sterile and pure water at 4 ºC to make DNA bands diffuse to the liquid. One microliter of this solution was used to reamplify the PCR product.

DNA sequencing and phylogenetic analysis
PCR products were sequenced by Macrogen Inc. (Seoul, South Korea). The quality and characteristics of the obtained sequences were analyzed with the software InfoQuest TM FP 5.10, only those ones considered as appropriate were used for comparison to the GenBank database with the Basic Local Alignment Search Tool (BLAST) (Altschul et al., 1990). After this preliminary study, our sequences and their homologous ones (obtained from the Nucleotide Database: http://www.ncbi.nlm.nih.gov/nuccore) were assembled and submitted to phylogenetic and evolutionary analysis with MEGA version 4.0.2 (Tamura et al., 2007). The Neighbor-Joining analysis (Saitou & Nei, 1987) allowed to get information about the relations between the gotten sequences and the reliability of the identifications provided by the Nucleotide Database. The bootstrap test was based on 1000 replicates (Felsenstein, 1985).
The evolutionary distances were computed using the Maximum Composite Likelihood method (Tamura et al., 2007) that allowed to calculate the equivalent units to the base substitutions per site.

Oenological parameters of wine samples and fermentation development
Results for analytical composition of wines during three vintages are displayed in Table 1. Data were within the usual range of Tempranillo wines from this Spanish region (González-Arenzana et al., 2012b). After completion AF, alcohol content ranged between 13.0% and 14.0%, pH was between 3.32 and 3.64 and free SO2 level was between 4.24 and 18.1 mg/L. During MLF a decrease in total acidity and a subsequent increase in pH were observed. In addition, an increase in volatile acidity was noted as it was expected. The wine from 2006 vintage showed less restrictive parameters for microbial growth, so it presented higher pH and the lowest values of alcohol content and SO2.  Community, 1990;Lopez et al., 2008). The development of the MLF was related to the viable population of LAB and there was a relation between bacterial population and decrease in L-malic acid (data not shown). Important differences in MLF duration were observed between vintages and MLF completion lasted for 21, 239 and 136 days in 2006, 2007 and 2008 vintages, respectively. Different temperatures at each vintage (wine temperature below 12 ºC after AF in 2007) and the lack of temperature control in the winery were the determinant factors in these differences, but factors such as pH, composition of the wine and the interaction with other microorganisms implicated in the fermentation could also influence, as it has been reported by other authors (du Plessis et al., 2004;Lonvaud-Funel, 1999;Reguant et al., 2005a;Reguant et al., 2005b). Figure 1 shows the number of isolates of the viable LAB species identified at each stage and year of vinification. A total of 251 LAB isolates were recovered and identified as belonging to eight different species. The greatest diversity of LAB species was detected during the AF.

Culture dependent microbiological analysis
Oenococcus oeni was present in all studied stages of the fermentation process except in 2007. It was isolated in must and tumultuous AF in 2006 and 2008 vintages, and it was the only species isolated at MLF in the three years, being therefore the predominant species, followed by Lactobacillus plantarum, Leuconostoc mesenteroides and Lactobacillus mali. The other non-Oenococcus oeni species appeared at stages 1-2 in variable rates in the three vintages. A similar distribution of species has been also reported by other authors (Fugelsang & Edwards, 2007;Ruiz et al., 2010) and they also concluded that Oenococcus oeni was the main responsible species for MLF. The diversity of species found at each year was different, being the number of species isolated in 2007 almost double that in 2006 and 2008, and missing Oenococcus oeni until the end of the 2007 AF, a fact that could also have influence in the MLF duration, as it was indicated above.

Culture independent microbiological analysis and comparison with culture dependent method
PCR-DGGE analysis of the sampled wine fermentations in the three studied vintages using primers WLAB1/2 (16S rDNA-based primer sets) and primers rpoB1/1o/2 (β subunit of the  The sequences obtained from the DNA excised DGGE-bands of each sample and their homologous ones from Nucleotide Database (Altschul et al., 1990) constituted a tree for each studied gene (Figure 3). Figure 3a shows a tree based on 16S rDNA gene composed by four ramifications or branches belonging to the genus Oenococcus, Lactobacillus, Weissella and Leuconostoc; and Figure 3b shows a tree based on rpoB gene composed by three branches belonging to Oenococcus, Leuconostoc and Lactobacillus.  (Felsenstein, 1985). Numbers over the branches are bootstrap values (1000 repetitions). Escherichia coli and Listeria inocua were used as outgroups. Scale bar represents (calculated) distance. Reference strains are closely related to the sequenced bands and accession number of each gene is indicated. Band number isolated from fermenting wines are indicated by the sampled stage of vinification (1, must; 2, tumultuous AF; 3, final AF; 4, initial MLF; 5, tumultuous MLF; 6, final MLF) the letter of the position in the gel and the isolation year.
The species identification at each stage of vinification in the three studied years with culture independent techniques (16S rDNA/PCR-DGGE and rpoB/PCR-DGGE) and culture dependent method (plating on modified MRS) are recovered in Table 2. A total of fourteen different LAB species were identified in the three studied vintages by traditional and culture independent methods. PCR-DGGE analysis allowed to identify nine species in comparison to the eight ones detected by culture in plate of the sampled wine. Thus, Fructobacillus ficulneus, Fructobacillus tropaeoli, Lactobacillus buchneri, Leuconostoc pseudomesenteroides and Weissella were not detected in the employed culture medium, while Lactobacillus coryniformis, Lactobacillus brevis, Lactobacillus mali, Lactobacillus uvarum and Pediococcus parvulus were not detected by PCR-DGGE. Therefore, results obtained by both methods were complementary and demonstrated the importance of using a combined analytical approach to explore microbial communities as other authors have concluded in different ecological niches (Iacumin et al., 2009). In relation with DGGE analysis, PCRamplified bands from 16S rDNA gene gave better results than the rpoB bands amplification, with eight and three identified species, respectively. Nevertheless, the results obtained with the two genes were again complementary so Lactobacillus buchneri was only detected by rpoB/PCR-DGGE.
Results about diversity of LAB species found at each year and stage of vinification were very similar to those described above for culture dependent method. The greatest diversity was detected again during AF, opposite to MLF were only two species were present, Oenococcus oeni and Leuconostoc mesenteroides. The presence of Leuconostoc mesenteroides in Fructobacillus ficulneus

PFGE analysis of the strains of this study
Identification of the Oenococcus oeni strains of this study was successfully achieved by PFGE of DNA digested with SfiI and each strain presented a characteristic PFGE pattern. Digestions with ApaI enzyme were not more discriminating than SfiI restriction (data not shown). Cluster analysis and visual inspection of the PFGE patterns from the 187 Oenococcus oeni isolates recovered from wine fermentations in the three years gave a total number of 37 distinct genotypes (Figure 4).   Comparing coincident genotypes for the three vintages, it was observed that between the genotypes isolated in 2006 vintage two were found in 2007 (genotypes 18 and 20) and four in 2008 (genotypes 3, 13, 17 and 18). Moreover, two genotypes isolated in 2007 (genotypes 18 and 25) were also detected in 2008 vintage. Only one genotype (18) was identified in the three studied years. The frequency of participation of each genotype varied from year to year, thus dominant genotypes one year were minority or not present at other one which suggested the adaptation of Oenococcus oeni strains to the winery conditions every year. Similar results were reported by other authors in studies about bacteria and yeast populations (Gutierrez et al., 1999;Reguant et al., 2005a;Ruiz et al., 2010;Santamaría, 2009).
Interestingly, no genotype was isolated in all fermentation stages so fourteen genotypes appeared only at AF (stages 1-3), six were present at all MLF stages (4, 5 and 6) and three of them were also detected at the end of AF. Most fermentation stages showed mixed Oenococcus oeni strains populations, which confirmed that several Oenococcus oeni strains occurred in a single spontaneous MLF (Lopez et al., 2007;Renouf et al., 2009;Ruiz et al., 2010). The number of different genotypes identified at each stage ranged from 0 to 5, and from 3 to 9, at stages 1-3 and 4-6, respectively. Genotypes 9 and 18 in 2006 vintage, 20 and 26 in 2007, and 13 in 2008 were the predominant ones during MLF. Three out of these genotypes (13, 18 and 20) could be considered as interesting Oenococcus oeni strains for the selection of new malolactic starter cultures as individual or mixed strains, because in addition to be dominant in most of the MLF stages at each vintage, they were isolated at more than one year in quality wines. Figure 5. Frequency of appearance (%) of Oenococcus oeni genotypes at each stage (1, must; 2, tumultuous AF; 3, final AF; 4, initial MLF; 5, tumultuous MLF; 6, final MLF) and vintage. Genotypes without color only appeared once, genotypes with grey color appeared only in one year at more than one stage and genotypes with color appeared in more than one vintage.

Comparison of the PFGE and RAPD-PCR profiles from the wine fermentation strains and commercial strains
The thirty-seven genotypes of the indigenous Oenococcus oeni strains from wine fermentations and fourteen PFGE and RAPD-PCR patterns obtained of strains from several commercial cultures were submitted to comparison by bioinformatics and visual analysis (data not shown). Figure 6 shows that genotypes I, II, III and IV from commercial starter cultures resulted indistinguishable from four indigenous genotypes (23, 13, 21 and 18, respectively), despite commercial malolactic cultures had never been employed in the sampled cellar.   Therefore, despite two of these strains had been previously considered as interesting for the selection of new malolactic starter cultures, the possible identical strain identification with already marketed strains suggested reject these two indigenous Oenococcus oeni isolates from a future selection process regardless of their oenological properties as malolactic starter cultures.

Conclusion
This study has been a contribution to a better description of the LAB ecology along the process of Tempranillo wines winemaking.
The study about the microbial diversity of viable LAB populations showed that the species diversity was higher at the AF stage where eight different species were identified. Oenococcus oeni was detected during AF in variable proportions and it became the majority species during spontaneous MLF.
This work allowed to increase the endogenous strain collection of LAB isolated from fermenting wines of the Appellation of Origin Rioja what meant a contribution to the preservation of biodiversity and wine peculiarity of this region and a starting point for future research.
The analysis of the total LAB populations by culture independent techniques (PCR-DGGE) showed that the species diversity detected along the winemaking process was higher than the one found by the study of viable LAB, identifying up to nine different LAB species. The LAB species variability was also higher at the previous stages to the MLF. Once spontaneous MLF started this variability was greatly reduced, with Oenococcus oeni and Leuconostoc mesenteroides as the only detected species.
The results obtained with culture dependent and independent techniques were complementary so in studies conducted in microbial ecology they both should be used to achieve a broader view of the studied ecosystem.
PFGE has shown to be a suitable method for strain differentiation, for monitoring individual strains and determining which strains actually survive and carry out MLF. The results of Oenococcus oeni typification indicated the high diversity of indigenous Oenococcus oeni strains responsible for MLF of the wines of this study and the complexity of the ecology involved in a fermentating wine. The frequency of participation of each genotype varied from year to year, thus dominant genotypes one year were minority or not present at other one, which suggested the adaptation of Oenococcus oeni strains to the winery conditions every year.
Several genotypes could be considered as interesting Oenococcus oeni strains for the selection of new malolactic starter cultures as individual or mixed strains because, in addition to be isolated at more than one year in quality wines, they were dominant in most of the MLF stages at each vintage. The comparison of the patterns from commercial cultures and the genotypes from indigenous Oenococcus oeni strains showed four indistinguishable genotypes. The presence of these four genotypes for one to three years, and in some cases with a high frequency of appearance, demonstrated the significance of this study in order to exclude these genotypes from a future selection process.