Yeast Community Structures and Dynamics in Healthy and Botrytis-Affected Grape Must Fermentations

APPLIED AND ENVIRONMENTAL MICROBIOLOGY, Nov. 2007, p. 6705 6713 Vol. 73, No. 21 0099-2240/07/$08.00 0 doi:10.1128/aem.01279-07 Copyright 2007, American Society for Microbiology. All Rights Reserved. Yeast Community Structures and Dynamics in Healthy and Botrytis-Affected Grape Must Fermentations Aspasia A. Nisiotou, 1 Apostolos E. Spiropoulos, 1,2 and George-John E. Nychas 1 * Laboratory of Microbiology and Biotechnology of Foods, Department of Food Science and Technology, Agricultural University of Athens, Iera Odos 75, 118 55 Athens, Greece, 1 and Arkas SA, Artemisio, Ancient Mantinia, Tripoli Arcadia, Greece 2 Received 9 June 2007/Accepted 19 August 2007 Indigenous yeast population dynamics during the fermentation of healthy and Botrytis-affected grape juice samples from two regions in Greece, Attica and Arcadia, were surveyed. Species diversity was evaluated by using restriction fragment length polymorphism and sequence analyses of the 5.8S internal transcribed spacer and the D1/D2 ribosomal DNA (rdna) regions of cultivable yeasts. Community-level profiles were also obtained by direct analysis of fermenting samples through denaturing gradient gel electrophoresis of 26S rdna amplicons. Both approaches revealed structural divergences in yeast communities between samples of different sanitary states or geographical origins. In all cases, Botrytis infection severely perturbed the bioprocess of fermentation by dramatically altering species heterogeneity and succession during the time course. At the beginning and middle of fermentations, Botrytis-affected samples possessed higher levels of biodiversity than their healthy counterparts, being enriched with fermentative and/or spoilage species, such as Zygosaccharomyces bailii and Issatchenkia spp. or Kluyveromyces dobzhanskii and Kazachstania sp. populations that have not been reported before for wine fermentations. Importantly, Botrytis-affected samples exposed discrete final species dominance. Selection was not species specific, and two different populations, i.e., Saccharomyces cerevisiae in samples from Arcadia and Z. bailii in samples from Attica, could be recovered at the end of Botrytis-affected fermentations. The governing of wine fermentations by Z. bailii is reported for the first time and could elucidate the origins and role of this particular spoilage microbe for the wine industry. This is the first survey to compare healthy and Botrytis-affected spontaneous fermentations by using both culture-based and -independent molecular methods in an attempt to further illuminate the complex yeast ecology of grape must fermentations. Grape must fermentations are governed by dynamic yeast species assemblages, whose physiognomy is influenced by several chemical constraints and biological attributes. Among these factors, the chemical composition of the must, particular enological practices, and primary species richness/heterogeneity in the grape juice have been shown to determine yeast succession during the fermentation course and the final species dominance (12, 13). Accumulating data reveal additional factors that may designate the boundaries of yeast inhabitants, such as the grape variety and the geographical location of the vineyard (6, 12, 31, 33, 38). Biological invasions also influence resident yeast biota and may stimulate forces that shape community diversity, yet their impact on must fermentation has been underestimated in earlier literature (12, 20). Botrytis is an important grape pathogen that dramatically changes the physicochemical state of berries and consequently may alter the structure of the yeast community (12, 27). It has also been postulated that Botrytis infection may affect species succession during alcoholic fermentation (12, 20), though descriptive inspections of the yeast ecology of Botrytis-affected fermentations are quite scarce. Earlier studies suggested that in Botrytis-affected wine fermentations, the array of yeast succession usually starts with an * Corresponding author. Mailing address: Laboratory of Microbiology and Biotechnology of Foods, Department of Food Science and Technology, Agricultural University of Athens, Iera Odos 75, 118 55 Athens, Greece. Phone and fax: 30 210 5294693. E-mail: gjn@aua.gr. Published ahead of print on 31 August 2007. increased incidence of weakly fermentative yeasts, such as Hanseniaspora and Candida spp., which are later obscured by Saccharomyces species (9, 14). Similar results were also obtained by Mills et al. (25), who applied different molecular techniques to explore yeast diversity in commercial Botrytisaffected fermentations. However, in most of these studies, industrial enological practices were implemented, such as the clarification of juice prior to fermentation or the addition of sulfur, which could severely perturb the succession of indigenous yeast populations. Moreover, direct comparative studies between healthy and Botrytis-affected fermenting samples have not been conducted yet, and thus it is difficult to evaluate potential influences of Botrytis on the process. In the present study, the indigenous yeast population dynamics in Botrytis-affected spontaneous (noninoculated) fermentations were analyzed. In order to evaluate the impact of Botrytis infection on yeast ecology of fermenting musts, parallel fermentations with noninfected (healthy) counterparts from the same vineyard were also conducted and compared. Grapes of four grapevine varieties from vineyards in the Attica and Arcadia regions were surveyed. Different culture media were applied to reveal the biodiversity in yeast populations, and PCR-restriction fragment length polymorphism (PCR-RFLP) analysis was combined with sequence analyses of the 5.8S internal transcribed spacer (ITS) region and the D1/D2 domain of ribosomal DNA (rdna) to identify a total of 1,463 isolates. PCR-denaturing gradient gel electrophoresis (PCR-DGGE) was also employed on fermenting samples, without previous cultivation of yeasts on plates. 6705

6706 NISIOTOU ET AL. APPL. ENVIRON. MICROBIOL. Sample code TABLE 1. Physicochemical characteristics and origins of grape juice samples a Variety, status of berries Location of vineyard Initial sugar content (g/liter) Initial ph value Final ethanol content (% vol/vol ) Mh Mavroliatis, healthy Attica 180 3.6 11.5 Mb Mavroliatis, Botrytis infected Attica 288 4.0 11.0 Sh Sefka, healthy Attica 182 3.8 11.3 Sb Sefka, Botrytis infected Attica 163 3.6 10.8 Lh Limnio, healthy Attica 186 3.6 11.0 Lb Limnio, Botrytis infected Attica 270 3.8 11.0 F1h Moschofilero, healthy Arcadia 202 3.3 10.2 F1b Moschofilero, Botrytis infected Arcadia 187 3.0 11.4 F2h Moschofilero, healthy Arcadia 198 3.3 10.3 F2b Moschofilero, Botrytis infected Arcadia 185 2.9 11.2 a Reported values represent averages for duplicate analyses. MATERIALS AND METHODS Grape must fermentations. Healthy (h) or profoundly Botrytis-infected (b) bunches belonged to the Mavroliatis (M), Sefka (S), Limnio (L), and Moschofilero (F1 and F2) grapevine cultivars. M, S, and L samples were collected from the experimental vineyard of the Agricultural University of Athens, Attica, Greece (latitude, 37 58 ; longitude, 23 32 ; height, 30 m above sea level), and F1 and F2 samples were collected from two neighboring commercial vineyards in Arcadia, Peloponnesus, Greece (37 31, 22 23 ; 655 m above sea level) (Table 1). Grapes were immediately transferred to the laboratory in sterile bags, where they were crushed aseptically, and approximately 1.5 liters of grape juice was transferred to 2-liter flasks. Fermentations were carried out in duplicate at 20 C until sugars were depleted ( 0.4%). D-Glucose/D-fructose and ethanol contents were determined by using appropriate enzymatic kits (Boehringer Mannheim/ R-Biopharm, Darmstadt, Germany). ELISA for Botrytis detection. A plate-trapped antigen enzyme-linked immunosorbent assay (ELISA) was applied to confirm the presence of Botrytis in infected grape samples essentially as described previously (8). Briefly, juice samples were diluted 1/10 in phosphate-buffered saline (PBS; 0.8% NaCl, 0.02% KCl, 0.02% KH 2 PO 4, 0.115% Na 2 HPO 4, ph 7.2) and centrifuged at 12,000 g for 3 min, and the supernatants were transferred to microtiter wells. Three replicate wells were filled for each grape sample. PBS and the washings of a Botrytis cinerea culture were also used to coat wells, as negative and positive controls, respectively. After 30 min of incubation, wells were washed with PBST (PBS with 0.05% Tween 20) and then incubated for 1 h with the Botrytis-specific monoclonal antibody BC-12.CA4 (23). Afterward, wells were incubated for another hour with horseradish phosphatase-conjugated anti-mouse immunoglobulin G antibody diluted 1:1,000 in PBST. After each single step, wells were washed three times with PBST. Finally, wells were incubated with the substrate tetramethylbenzidine until a color reaction was visible. Microbial sampling and enumeration. Samples from the fermenting juices were taken at different time intervals and successively diluted from 10 1 to 10 6 in Ringer s solution. For the enumeration and isolation of total yeasts and non-saccharomyces, Saccharomyces, and Dekkera/Brettanomyces species, 100 l of each dilution was plated in triplicate on Wallerstein laboratory nutrient agar (Oxoid Ltd.), lysine medium agar (Oxoid Ltd.), ethanol sulfite agar (17), and Dekkera/Brettanomyces differential medium (32), respectively. To prevent growth of bacteria, chloramphenicol (Sigma) was added to the media at 100 mg liter 1. Twenty to thirty colonies per agar plate derived from a single fermentation were randomly selected at three different stages, i.e., the beginning (stage BF); the middle (stage MF), when about 50% of sugars were consumed; and the end of the course (stage EF), when sugars were depleted. Yeasts were stored at 80 C until further analysis. DNA extraction. For genomic DNA extraction, yeast cells were grown overnight in YPD broth (1% yeast extract, 2% bacteriological peptone, 2% glucose, ph 6.2) at 30 C in a rotary shaker. Cells were collected by centrifugation at 8,000 g for 1 min, resuspended in 300 l of breaking buffer (2% Triton X-100, 1% sodium dodecyl sulfate, 100 mm NaCl, 10 mm Tris, ph 8, 1 mm EDTA, ph 8), and transferred to 2-ml tubes containing 0.3 g of 0.5-mm-diameter glass beads (Sigma). Cell suspensions were subjected to vortex mixing for 2 min after the addition of 300 l phenol-chloroform-isoamyl alcohol (25:24:1 [vol/vol/vol]). Three hundred microliters of TE buffer (10 mm Tris, 1 mm EDTA, ph 7.6) was added, and the bead-cell mixture was centrifuged at 12,000 g for 10 min. The supernatant was transferred to a fresh 1.5-ml tube, and DNAs were precipitated by the addition of 2.5 volumes ethanol, followed by centrifugation at 12,000 g for 10 min. The DNA pellet was washed with 70% ethanol and resuspended in water. For DGGE analysis of fermentation samples, DNA extraction was performed as described previously (26). PCR amplification. The 5.8S ITS rdna region of yeast isolates was amplified using the primers ITS1 (5 -TCCGTAGGTGAACCTGCGG-3 ) and ITS4 (5 -T CCTCCGCTTATTGATATGC-3 ) (39), and the D1/D2 domain of the 26S rdna gene was amplified using the primer pair NL1 (5 -GCATATCAATAAG CGGAGGAAAAG-3 ) and NL4 (5 -GGTCCGTGTTTCAAGACGG-3 ) (19). PCRs were performed in a total volume of 50 l containing 10 ng of template DNA, 20 pmol of each primer, a 100 M concentration of each deoxynucleoside triphosphate, and 1 U of DyNAzyme EXT DNA polymerase (Finnzymes, Oy, Finland) in the incubation buffer provided by the manufacturer of the enzyme. PCR products for DGGE analysis of fermentation samples were generated using the primers NL1, with a GC clamp (5 -CGCCCGCCGCGCGCGGCGGGCGG GGCGGGGGCATATCAATAAGCGGAGGAAAAG-3 ), and LS2 (5 -ATTC CCAAACAACTCGACTC-3 ) (4). Reactions were performed in a total volume of 50 l containing 10 ng of template DNA, 30 pmol of each primer, a 200 M concentration of each deoxynucleoside triphosphate, 2 mm of MgCl 2, and 1.5 U of DyNAzyme EXT DNA polymerase (Finnzymes, Oy, Finland) in the incubation buffer provided by the manufacturer of the enzyme. Amplification was achieved in a PTC-200 Peltier thermal cycler (MJ Research, Waltham, MA) programmed as follows: 94 C for 3 min and 35 cycles of 94 C for 30 s, 52 C for 30 s, and 74 C for 2 min, followed by 74 C for 10 min. PCR products were separated by gel electrophoresis on a 1.0% (wt/vol) agarose gel, detected by ethidium bromide staining, and photographed under UV light with a chargecoupled device camera (Sony, Japan) or a GelDoc system (Bio-Rad). Sizes of fragments were determined using a standard molecular size marker (100-bp ladder; New England Biolabs). RFLP analysis. For restriction reactions of the 5.8S ITS region, approximately 500 ng of PCR product was incubated for1hat37 C with 10 U of HinfI, HaeIII, HhaI, DraI (Takara, Japan), or DdeI (New England Biolabs) restriction endonuclease. Restriction fragments were separated by gel electrophoresis on a 3% (wt/vol) agarose gel, detected by ethidium bromide staining, and photographed. Sizes of fragments were estimated using a standard molecular size marker (100-bp ladder; New England Biolabs). DGGE analysis. Sequence-dependent separation of yeast PCR amplicons was performed with a DCode universal mutation detection system (Bio-Rad, Hercules, CA) and 8% polyacrylamide gels (bisacrylamide, 37.5:1) with a denaturing gradient of 30 to 60% urea and formamide as described previously (25). DGGE bands generated by direct PCR on fermenting samples were identified by comigration with reference patterns of representative isolates and by direct sequencing after gel extraction. DNA bands were excised using sterile blades, soaked in 40 l of water, and incubated overnight at 4 C. Eluted DNAs were PCR amplified with the primers NL1 and LS2. Sequence analysis. PCR products from the 5.8S ITS region and the D1/D2 domain of representative isolates per distinct restriction pattern and the partial D1/D2 domains in gel-extracted DGGE bands were purified using a QIAquick PCR purification kit (QIAGEN, Germany) according to the manufacturer s instructions. By using forward (NL1 or ITS1) and reverse (NL4 or ITS4) primers, both DNA strands were directly sequenced by Macrogen, using an ABI 3730 XL automatic DNA sequencer. BLAST searches of sequences were performed at the National Center for Biotechnology Information (NCBI) GenBank data library.

VOL. 73, 2007 YEASTS IN HEALTHY AND BOTRYTIS-AFFECTED MUST SAMPLES 6707 FIG. 1. Representative restriction patterns of the 5.8S ITS region of yeast isolates obtained with HaeIII (lanes 1 to 7) or DdeI (lanes 8 to 17). Lanes: M, 100-bp molecular marker; 1, C. lusitaniae; 2 and 3, C. zemplinina; 4, Kazachstania sp.; 5, L. thermotolerans; 6, K. dobzhanskii; 7, P. fermentans; 8to11,H. opuntiae; and 12 to 17, H. uvarum. Clustal X (1.83) software (http://www-igbmc.u-strasbg.fr/bioinfo) was used to perform sequence alignments among sequences of the isolates and homologous sequences available in GenBank. Nucleotide sequence accession numbers. Nucleotide sequences have been deposited in the NCBI GenBank data library under accession numbers EF620859 to EF620862. RESULTS ELISA was applied to verify Botrytis infection of grape samples. Strong signals, equivalent to the intensity of the positive control, were observed for Botrytis-infected samples (Mb, Sb, Lb, F1b, and F2b), while a negligible or no signal was detected for healthy samples (Mh, Sh, Lh, F1h, and F2h) (data not shown). Sugar contents and ph values were recorded for the grape musts, as shown in Table 1. Identification of isolates based on RFLP and sequence analyses of the 5.8S ITS region. A total of 1,463 yeasts isolated from three distinct stages of the fermentation course, i.e., the beginning (stage BF), the middle (stage MF), and the end (stage EF), were analyzed by PCR-RFLP analysis of the 5.8S Profile PCR product size (bp) ITS rdna region (Fig. 1). Using the restriction endonucleases HinfI, HaeIII, HhaI, DdeI, and DraI, 19 different banding profiles were generated (profiles I to XIX) (Table 2). Profile comparisons between isolates and published strains (10, 27) assigned groups of isolates to the species Hanseniaspora uvarum, Hanseniaspora guilliermondii, Hanseniaspora opuntiae, Issatchenkia terricola, Issatchenkia occidentalis, Issatchenkia orientalis, Clavispora lusitaniae, Metschnikowia pulcherrima, Pichia fermentans, Pichia anomala, Pichia guilliermondii, Pichia angophorae, Saccharomyces cerevisiae, Zygosaccharomyces bailii, Kluyveromyces dobzhanskii, Candida diversa, Candida glabrata, Candida zemplinina, and Lachancea thermotolerans. Sequencing of the 5.8S ITS region confirmed the presence and positions of experimental restriction sites. Sequence alignments and phylogenetic analyses (data not shown) verified the previous identifications, except for the species C. diversa (profile IX) and P. angophorae (profile XV), for which there were no sequence data available in GenBank. Surprisingly, the 5.8S ITS sequences of isolates corresponding to profile XV did not exhibit any significant homology to those of other published strains, including different Pichia species. The highest similarity value scored was as low as 77% and referred to Kluyveromyces hubeiensis strain AS 2.1536. The highest similarity value scored for group IX was also low (75%) and referred to Saturnispora ahearnii strain NRRL Y-7555. Due to these low similarities, the identification of these groups of isolates was based on sequencing of the D1/D2 domain, as described below. Identification of isolates based on sequence analysis of the D1/D2 domain. The D1/D2 domains of the 26S rdna genes of representative isolates from groups I to XIX were amplified and sequenced. Previous identifications based on the 5.8S ITS region were confirmed. The D1/D2 domain of group IX clearly assigned the isolates to C. diversa, as RFLP analysis suggested. Although profile XV (Table 2) was previously assigned to P. angophorae, sequence alignments and further phylogenetic analysis (data not shown) clearly placed this group of isolates within the genus Kazachstania, being most closely related to the recently described species Kazachstania zonata and Kazachstania gamospora (16). However, the level of sequence divergence detected was relatively high (at least five noncontig- TABLE 2. Sizes of 5.8S ITS rdna amplicons and restriction fragments of yeast isolates Restriction fragment size(s) (bp) HhaI HaeIII HinfI DdeI DraI Species I 750 320, 310, 115 750 340, 190, 170, 60 290, 180, 90, 85, 75, 50 H. uvarum II 750 320, 310, 115 750 340, 190, 170, 60 360, 180, 85, 70, 50 420, 150, 130, 30 H. guilliermondii III 750 320, 310, 120 750 340, 190, 170, 60 360, 180, 85, 70, 50 420, 300, 30 H. opuntiae IV 460 120, 95, 95, 70, 60 290, 130 240, 110, 100 I. terricola V 500 200, 190, 70, 50 400, 90 220, 140, 140 I. orientalis VI 480 240, 100, 80, 70 320, 100, 55 260, 110, 110 I. occidentalis VII 480 210, 105, 105, 60 480 240, 240 310, 120 C. zemplinina VIII 880 390, 170, 160, 135 670, 220 360, 270 C. glabrata IX 450 160, 100, 70, 70 400 210, 190 C. diversa X 450 170, 100, 100, 80 340, 80, 30 250, 200 P. fermentans XI 620 560, 60 600 310, 300 P. anomala XII 590 295, 255 390, 100, 80 315, 270 P. guilliermondii XIII 770 310, 270, 95, 90 700, 70 330, 220, 160, 60 Z. bailii XIV 400 205, 100, 95 280, 100 200, 190 M. pulcherrima XV 750 320, 320, 100 750 400, 340 Kazachstania sp. XVI 880 385, 365 320, 230, 180, 150 365, 155 S. cerevisiae XVII 680 300, 285, 90 300, 190, 100, 90 355, 325 L. thermotolerans XVIII 730 285, 185, 160, 80 625, 75 245, 180, 115, 105, 75 K. dobzhanskii XIX 380 210, 90, 80 380 190, 190 C. lusitaniae

6708 NISIOTOU ET AL. APPL. ENVIRON. MICROBIOL. uous substitutions over 554 nucleotides) to assign isolates to either of the above species (18). Therefore, this group of isolates is referred to hereafter as Kazachstania sp. Yeast population dynamics and species heterogeneity. Figure 2 presents the sequential development of Saccharomyces and non-saccharomyces yeast populations during the fermentation courses. Significant differences in kinetic patterns were observed between healthy and Botrytis-affected samples over a geographical region. S. cerevisiae developed in healthy but not Botrytis-affected musts from the Attica region. In samples from Arcadia, S. cerevisiae governed the Botrytis-affected fermentations, whereas non-saccharomyces yeasts fermented the healthy samples. No Dekkera/Brettanomyces yeasts were recovered from Dekkera/Brettanomyces differential medium. Structural alterations of yeast communities between healthy and Botrytis-affected samples at different stages of fermentation were revealed by using RFLP and sequence analyses (Fig. 3). Species compositions also differed between samples of the same sanitary status originating from distinct vineyards. In samples from the Attica region, H. uvarum was dominant at the beginning of the course in all healthy fermentations, followed by H. opuntiae, P. guilliermondii, and I. terricola. H. uvarum also prevailed in Botrytis-affected fermentations, though it was less abundant than in healthy samples, as other species were growing along on occasion, including H. opuntiae, H. guilliermondii, M. pulcherrima, Z. bailii, L. thermotolerans, I. terricola, I. orientalis, C. lusitaniae, and C. zemplinina. Species heterogeneity changed as the fermentation proceeded. At stage MF, H. uvarum populations declined in all samples, and several other species emerged. Particularly in healthy samples, S. cerevisiae, C. glabrata, P. anomala, and C. zemplinina developed. In Botrytis-affected samples, most of the species encountered at stage BF also persisted in the middle of the course. Kazachstania sp. and C. zemplinina emerged in all samples, and Z. bailii counted for a considerable proportion of the total yeast population in samples Mb and Sb. At stage EF, S. cerevisiae completed the fermentation in healthy samples. Several non-saccharomyces yeasts were also recovered, albeit as smaller populations, including C. zemplinina, L. thermotolerans, and P. guilliermondii. Z. bailii was the dominant species in Botrytis-affected samples. In most cases, C. zemplinina was also encountered. Kazachstania sp. and I. orientalis persisted throughout the fermentation course in samples Sb and Mb, respectively. In samples originating from Arcadia, H. uvarum was dominant at the beginning of healthy fermentations (Fig. 3). Botrytis-affected samples also harbored H. uvarum, but the dominant population was C. zemplinina. The yeast community was enriched with I. terricola and three more species, i.e., C. diversa, K. dobzhanskii, and P. fermentans, that were not detected in samples from the Attica region. At the middle stage, C. zemplinina represented the dominant population in both healthy and Botrytis-affected fermentations. Like the case at stage BF, Botrytis-affected samples exhibited more species richness than their healthy counterparts. In particular, K. dobzhanskii and I. occidentalis evolved, while I. terricola and P. fermentans persisted. At the end of the course, C. zemplinina was the only species recovered from healthy samples, whereas S. cerevisiae completed the fermentation in Botrytis-affected samples. DGGE analysis. Most previously isolated species could be discriminated by their DGGE profiles, with the exception of H. uvarum, H. opuntiae, and H. guilliermondii, which comigrated in the gels (Fig. 4). M. pulcherrima generated a faint band, while the amplicon of C. lusitaniae appeared as a smear rather than a clear band. Direct PCR-DGGE of samples from different stages of fermentation revealed diversification in yeast communities between healthy and Botrytis-affected samples (Fig. 5). All species of relatively high abundance, as estimated by plating analysis, were also detected by PCR-DGGE in different stages of fermentations. Some quantitatively minor species isolated from plates at the BF and MF stages, such as P. guilliermondii, P. anomala, L. thermotolerans, and C. glabrata, were not detected in the gels, whereas I. terricola, I. occidentalis, and Kazachstania sp. could be recovered. In some cases, species that were not previously isolated from plates were observed. For instance, C. zemplinina was detected in gels for samples Mb and Sb at the beginning of fermentations. Similarly, bands corresponding to Z. bailii and I. terricola appeared in samples Sb and Lb, respectively. Aureobasidium pullulans was identified only by DGGE analysis of both healthy samples from Arcadia. At stage MF, I. terricola appeared in sample F1b, I. occidentalis appeared in sample F2b, and C. zemplinina was recovered in healthy and Botrytis-affected samples from Attica and Arcadia, respectively. DISCUSSION An integrated approach combining both culture-based and -independent techniques was applied to compare yeast communities in healthy and Botrytis-affected grape must fermentations. A relatively high level of biodiversity, including 11 genera and 19 species, was uncovered by using different culture media and 5.8S ITS-RFLP analysis of 1,463 isolates. This methodology was efficient in identifying isolates at the species level in that only one group of isolates, belonging to a Kazachstania sp., was misidentified and was classified by sequencing of the D1/D2 domain. Although in beverage fermentations the spectrum of resident species may be described accurately by the inclusion of selective/differential media and variations in growth conditions (1), it has also been proposed that plate culturing techniques reveal a portion rather than the whole amount of yeast diversity in a natural wine ecosystem (4, 5, 24). PCR-DGGE may serve as a culture-independent approach and was employed recently to assess the structure of yeast communities in grape/wine samples (25, 29). In this study, direct PCR-DGGE provided informative fingerprints for Botrytis-affected and healthy samples and described a similar picture of community structures to that uncovered by the culturebased approach. All species that served as controls could be discriminated by DGGE analysis of 26S rdna fragments, except for H. uvarum, H. opuntiae, and H. guilliermondii, which were clearly identified by 5.8S ITS-RFLP analysis as described previously (27). It is known that these species are very closely related to each other and therefore are difficult to discriminate (2). Particularly in the D1/D2 domain, they have only a few nucleotide differences downstream of the LS2 primer, at the 3 end, and hence their resolution was not feasible by use of the NL1/LS2 primer pair. Some quantitatively minor populations of the cultivable community, for instance, certain non-saccharomyces yeasts at the EF stage in healthy samples from Attica,

FIG. 2. Growth profiles of S. cerevisiae (Œ) and non-saccharomyces (E) populations during the alcoholic fermentation of healthy (Mh, Sh, Lh, F1h, and F2h) and Botrytis-affected (Mb, Sb, Lb, F1b, and F2b) grape musts. Reported values represent the averages of duplicate analyses. 6709

FIG. 3. Yeast species heterogeneity in WL nutrient agar of healthy (Mh, Sh, Lh, F1h, and F2h) and Botrytis-affected (Mb, Sb, Lb, F1b, and F2b) grape musts at the beginning (stage BF), middle (MF), and end (EF) of the fermentation courses. M, S, and L samples originated from Attica, while F1 and F2 samples originated from Arcadia. The number of isolates examined per case (n) is indicated above each column. 6710

VOL. 73, 2007 YEASTS IN HEALTHY AND BOTRYTIS-AFFECTED MUST SAMPLES 6711 FIG. 4. DGGE analysis of 26S rdna products of yeast species isolated during fermentation. Lanes: 1, I. occidentalis; 2,P. fermentans; 3, C. diversa; 4, C. glabrata; 5, I. orientalis; 6, I. terricola; 7, M. pulcherrima; 8, Z. bailii; 9, C. lusitaniae; 10, Kazachstania sp.; 11, L. thermotolerans; 12, C. zemplinina; 13, P. guilliermondii; 14, K. dobzhanskii; 15, P. anomala; 16, S. cerevisiae; 17, H. uvarum; 18, H. opuntiae; and 19, H. guilliermondii. did not appear on DGGE gels. This is probably because the method may not feasibly detect species present at populations of 10 3 to 10 4 CFU/g grapes or CFU/ml must or when their abundance is 10 2 -fold lower than those of other species (25, 29). Direct PCR-based methods may also encounter difficulties in uncovering the total diversity within mixed populations, mostly due to species-specific recalcitrance to cell lysis or various levels of hybridization efficiency and primer specificity (40). M. pulcherrima, for example, has been found to exhibit poor PCR efficiency with the NL1/LS2 primer set (25). This was observed in the present study for M. pulcherrima and also for C. lusitaniae and could serve as a possible explanation for not detecting these species in DGGE gels, though they were members of the cultivable yeast community. Nevertheless, the results sustain the merit of direct PCR-DGGE as a fast method to describe yeast ecology and, most importantly, to detect species whose growth is not favored under certain culture conditions. For instance, the presence of C. zemplinina at stage BF in Botrytis-affected samples from Attica could be detected only by PCR-DGGE. In addition, the method was capable of detecting A. pullulans, a species that did not belong to the present cultivable populations, possibly because it was outgrown on plates by faster-growing species, such as H. uvarum and C. zemplinina. Alternatively, the DGGE band could represent amplifiable DNAs from nonviable cells persisting in the must. The present results propose the implementation of a polyphasic approach, rather than the application of a single methodology, for better resolving the spectrum of yeast populations during a spontaneous alcoholic fermentation. Yeast heterogeneity in Botrytis-affected fermentations differed from that in healthy counterparts by an increased number of resident species, particularly at the early and middle stages of fermentation. It has previously been assumed that a more complex yeast community structure is encouraged in damaged grapes (20). More recently, it was observed that Botrytis infection stimulates a higher level of diversity of yeasts, and the community is likely enriched with fermentative and/or spoilage species (27). Accordingly, a broader reservoir of FIG. 5. DGGE profiles of healthy (Mh, Sh, Lh, F1h, and F2h) and Botrytis-affected (Mb, Sb, Lb, F1b, and F2b) samples at the beginning (stage BF), middle (MF), and end (EF) of fermentations. Abbreviations: H.spp, Hanseniaspora spp.; C.z, C. zemplinina; A.p, A. pullulans; Z.b, Z. bailii; I.t, I. terricola; K.sp, Kazachstania sp.; S.c, S. cerevisiae; I.o, I. occidentalis; I.or, I. orientalis. yeasts in the Botrytis-infected grapes could explain the initial species richness observed in Botrytis-affected fermentations compared to that in healthy ones. Numerous studies have shown the dominance of S. cerevisiae during spontaneous wine fermentations (7, 30, 35, 37). However, it should be considered that most of these experiments were conducted under industrial conditions in wine cellars or plants, where the environment was occupied by S. cerevisiae strains selected over the years (22). In addition, enological practices such as grape must sulfiting and/or clarification disturb spontaneity, providing strong selection towards S. cerevisiae strains. In the absence of these conditions, the initial species richness and heterogeneity have been suggested to dictate population dynamics, allowing other robust fermenting

6712 NISIOTOU ET AL. APPL. ENVIRON. MICROBIOL. species to evolve (3, 12, 28). This is exemplified by different ethanol-tolerant Candida species that equal or overwhelm S. cerevisiae in dominating the end of fermentation (3, 12). Consistent with this, in the present study two non-saccharomyces species, C. zemplinina and Z. bailii, prevailed in healthy and Botrytis-affected fermentations from the Arcadia and Attica regions, respectively. Damaged grape berries may be very rich depositories of S. cerevisiae, in that one in four is S. cerevisiae positive, while at the same time only 1 in 1,000 sound berries carries the wine yeast (26). This could explain the evolvement of S. cerevisiae in the Botrytis-affected samples from Arcadia. In the case of the Attica region, however, Z. bailii dominated Botrytis-affected fermentations despite the fact that S. cerevisiae prevailed in healthy samples. This should not be unexpected, since communities of damaged or Botrytis-infected grapes are generally more complex than those of sound berries and may also harbor elevated populations of Zygosaccharomyces species (12, 27). Z. bailii is a highly ethanol-tolerant yeast, and its increased incidence at the beginning of fermentations of Botrytis-affected samples from Attica, but not in healthy samples, probably points towards its final prevalence (3, 28). The potential dominance of Z. bailii in must fermentations is very important for wine technology, since it constitutes a real threat for product quality and preservation (15, 20). Despite such an enological significance, the development and succession of Z. bailii during alcoholic fermentation have not been described before. This is probably because it is rarely found in fermenting grape juice, for as yet unknown reasons (12). Fleet (12) also proposed that it grows slower than other wine yeasts and consequently is often outcompeted or inhibited by factors produced by other species. This is the first report to show that Z. bailii may dominate grape must fermentations and that Botrytis-affected juices could serve as potential vehicles for its introduction into the winery. The predominance of H. uvarum at early stages of must fermentations has been well documented (7, 37). Our results are in general agreement with this statement and further propose the evolvement of C. zemplinina at early and mid stages, particularly in Botrytis-affected fermentations, which is also in accordance with previous reports (25, 34). It is noteworthy that besides H. uvarum, two more Hanseniaspora species, i.e., H. opuntiae and H. guilliermondii, could be present at relatively high counts in different samples. H. opuntiae has only recently been described (2), and its role in alcoholic fermentation has not yet been elucidated. In this study, it was observed that although H. uvarum gradually declines during the time course, H. opuntiae persists, accounting for equal or even higher proportions of cells at the middle stage. Further research is needed to illuminate a possible implication of H. opuntiae in the occasional longer persistence of Hanseniaspora species in must fermentations, as previously described (7, 11, 36). Accumulating data point towards a further persistence of non-saccharomyces yeasts in wine fermentations than was previously suggested (7, 11, 35). In accordance with these findings, several non-saccharomyces yeasts could survive the elevated ethanol concentrations at the end of the course, including I. orientalis, L. thermotolerans, P. guilliermondii, Kazachstania sp., C. zemplinina, and Z. bailii. The genus Kazachstania within the Saccharomycetaceae family was first described in 1971 on the basis of a single species, namely Kazachstania viticola, isolated from fermenting grapes in Kazakhstan (42). Since then, different Kazachstania species have been isolated from bamboo litter, mushrooms, and decaying corn silage (16, 21, 41), but never from grapes or fermenting musts. Here we report the occurrence of Kazachstania sp. populations in three different grape must fermentations. The present isolates are phylogenetically distant from K. viticola and rather unrelated to any other Kazachstania species described so far. In accordance with the previous correlation of Kazachstania species with deteriorating plant material, Kazachstania sp. populations were encountered solely in Botrytis-affected samples. It is possible that these populations lurk in infected grapes, contributing to the decomposition of plant biomass. Although these organisms were not detected at the early stages of the course, changes in environmental conditions, such as ethanol accumulation, at the mid and final stages may provide selection towards Kazachstania sp. that can survive at ethanol concentrations of up to 11% (vol/vol). In conclusion, the implementation of culture-dependent and -independent approaches to describe yeast communities in fermenting musts from Greece revealed a relatively broad spectrum of resident yeast species. New members of the winerelated ecosystem are presented, and their weights in the complex microbial consortium of fermentations are worth further evaluation. Botrytis-affected samples showed increased species richness and altered heterogeneity at the early stages of fermentation. These structural differences influenced the succession of species during the course and determined, at least partially, the final species dominance. Assessment of biological attributes of the system, such as possible interactive associations between yeasts and bacteria, would help us to understand this dynamic consortium in more depth. ACKNOWLEDGMENTS We gratefully acknowledge R. Burns (SAPS, Homerton College, Cambridge, United Kingdom) for the generous gifts of Botrytis-specific monoclonal and phosphatase-conjugated antibodies. We also thank L. Cocolin and K. Rantsiou for helpful advice on the DGGE technique. A.A.N. was awarded a scholarship from the State Scholarship Foundation (IKY). REFERENCES 1. Boundy-Mills, K. 2006. Methods for investigating yeast biodiversity, p. 67 100. In R. Carlos and P. Gábor (ed.), Biodiversity and ecophysiology of yeasts. Springer, Heidelberg, Germany. 2. Cadez, N., G. A. Poot, P. Raspor, and M. T. Smith. 2003. Hanseniaspora meyeri sp. nov., Hanseniaspora clermontiae sp. nov., Hanseniaspora lachancei sp. nov. and Hanseniaspora opuntiae sp. nov., novel apiculate yeast species. Int. J. Syst. Evol. Microbiol. 53:1671 1680. 3. Clemente-Jimenez, J. M., L. Mingorance-Cazorla, S. Martíınez-Rodríguez, F. J. Las Heras-Vázquez, and F. Rodríguez-Vico. 2004. Molecular characterization and oenological properties of wine yeasts isolated during spontaneous fermentation of six varieties of grape must. Food Microbiol. 21:149 155. 4. Cocolin, L., L. F. Bisson, and D. A. Mills. 2000. Direct profiling of the yeast dynamics in wine fermentations. FEMS Microbiol. Lett. 189:81 87. 5. Cocolin, L., M. Manzano, S. Rebecca, and G. Gomi. 2002. Monitoring of yeast population changes during a continuous wine fermentation by molecular methods. Am. J. Enol. Vitic. 53:24 27. 6. Combina, M., A. Elía, L. Mercado, C. Catania, A. Ganga, and C. Martinez. 2005. Dynamics of indigenous yeast populations during spontaneous fermentations of wines from Mendoza, Argentina. Int. J. Food Microbiol. 99:237 243. 7. Combina, M., L. Mercado, P. Borgo, A. Elia, V. Jofré, A. Ganga, C. Martinez, and C. Catanis. 2005. Yeasts associated to Malbec grape berries from Mendoza, Argentina. J. Appl. Microbiol. 98:1055 1061.

VOL. 73, 2007 YEASTS IN HEALTHY AND BOTRYTIS-AFFECTED MUST SAMPLES 6713 8. Dewey, F. M., S. E. Ebeler, D. O. Adams, A. C. Noble, and U. M. Meyer. 2000. Quantification of Botrytis in grape juice determined by a monoclonal antibody-based immunoassay. Am. J. Enol. Vitic. 51:276 282. 9. Donèche, B. J. 1993. Botrytized wines, p. 327 351. In G. H. Fleet (ed.), Wine microbiology and biotechnology. Harwood Academic Publishers, Chur, Switzerland. 10. Esteve-Zarzoso, B., C. Belloch, F. Uruburu, and A. Querol. 1999. Identification of yeasts by RFLP analysis of the 5.8S rrna gene and the two ribosomal internal transcribed spacers. Int. J. Syst. Bacteriol. 49:329 337. 11. Fleet, G. H. 1990. Growth of yeast during wine fermentation. J. Wine Res. 1:211 223. 12. Fleet, G. H. 2003. Yeast interactions and wine flavour. Int. J. Food Microbiol. 86:11 22. 13. Fleet, G. H., and G. M. Heard. 1993. Yeasts growth during fermentation, p. 27 54. In G. H. Fleet (ed.), Wine microbiology and biotechnology. Harwood Academic Publishers, Chur, Switzerland. 14. Fleet, G. H., S. Lafon-Lafourcade, and P. Ribereau-Gayon. 1984. Evolution of yeasts and lactic acid bacteria during fermentation and storage of Bordeaux wines. Appl. Environ. Microbiol. 48:1034 1038. 15. Fugelsang, K. C. 1997. Wine microbiology, p. 68 85. Chapman & Hall, New York, NY. 16. Imanishi, Y., K. Ueda-Nishimura, and K. Mikata. 2007. Two new species of Kazachstania that form ascospores connected by a belt-like intersporal body: Kazachstania zonata and Kazachstania gamospora. FEMS Yeast Res. 7:330 338. 17. Kish, S., R. Sharf, and P. Margalith. 1983. A note on a selective medium for wine yeasts. J. Appl. Microbiol. 55:177 179. 18. Kurtzman, C. P., and J. W. Fell. 2006. Yeast systematics and phylogeny implications of molecular identification methods for studies in ecology, p. 11 30. In R. Carlos and P. Gábor (ed.), Biodiversity and ecophysiology of yeasts. Springer, Heidelberg, Germany. 19. Kurtzman, C. P., and C. J. Robnett. 1998. Identification and phylogeny of ascomycetous yeasts from analysis of nuclear large subunit (26S) ribosomal DNA partial sequences. Antonie Leeuwenhoek 73:331 371. 20. Loureiro, V., and M. Malfeito-Ferreira. 2003. Spoilage yeasts in the wine industry. Int. J. Food Microbiol. 86:23 50. 21. Lu, H. Z., Y. Cai, Z. W. Wu, J. H. Jia, and F. Y. Bai. 2004. Kazachstania aerobia sp. nov., an ascomycetous yeast species from aerobically deteriorating corn silage. Int. J. Syst. Evol. Microbiol. 54:2431 2435. 22. Martini, A. 1993. The origin and domestication of the wine yeast Saccharomyces cerevisiae. J. Wine Res. 4:165 176. 23. Meyer, U. M., and F. M. Dewey. 2000. Efficacy of different immunogens for raising monoclonal antibodies to B. cinerea. Mycol. Res. 104:979 987. 24. Millet, V., and A. Lonvaud-Funel. 2000. The viable but non-culturable state of wine microorganisms during storage. Lett. Appl. Microbiol. 30:136 141. 25. Mills, D. A., A. J. Eric, and L. Cocolin. 2002. Yeast diversity and persistence in Botrytis-affected wine fermentations. Appl. Environ. Microbiol. 68:4884 4893. 26. Mortimer, R., and M. Polsinelli. 1999. On the origins of the wine yeast. Res. Microbiol. 150:199 204. 27. Nisiotou, A. A., and G. J. Nychas. 2007. Yeast populations residing on healthy or Botrytis-infected grapes from a vineyard in Attica, Greece. Appl. Environ. Microbiol. 73:2765 2768. 28. Povhe Jemec, K., N. Cadez, T. Zagorc, V. Bubic, A. Zupec, and P. Raspor. 2001. Yeast population dynamics in five spontaneous fermentations of Malvasia must. Food Microbiol. 18:247 259. 29. Prakitchaiwattana, C. J., G. H. Fleet, and G. M. Heard. 2004. Application and evaluation of denaturing gradient gel electrophoresis to analyse the yeast ecology of wine grapes. FEMS Yeast Res. 4:865 877. 30. Pramateftaki, P. V., P. Lanaridis, and M. A. Typas. 2000. Molecular identification of wine yeasts at species or strain level: a case study from two vine-growing areas of Greece. J. Appl. Microbiol. 89:236 248. 31. Raspor, P., D. M. Milek, J. Polanc, S. S. Možina, and N. Čadež. 2006. Yeasts isolated from three varieties of grapes cultivated in different locations of the Dolenjska vine-growing region, Slovenia. Int. J. Food Microbiol. 109:97 102. 32. Rodrigues, N., G. Gonçalves, M. Malfeito-Ferreira, and V. Loureiro. 2001. Development and use of a differential medium to detect yeasts of the genera Dekkera/Brettanomyces. Int. J. Food Microbiol. 90:588 599. 33. Sabate, J., J. Cano, B. Esteve-Zarzoso, and J. M. Guillamón. 2002. Isolation and identification of yeasts associated with vineyard and winery by RFLP analysis of ribosomal genes and mitochondrial DNA. Microbiol. Res. 157: 1 8. 34. Sipiczki, M. 2003. Candida zemplinina sp. nov., an osmotolerant and psychrotolerant yeast that ferments sweet botrytized wines. Int. J. Syst. Evol. Microbiol. 53:2079 2083. 35. Torija, M. J., N. Rozés, M. Poblet, J. M. Guillamón, and A. Mas. 2001. Yeast population dynamics in spontaneous fermentation: comparison between two different wine-producing areas of a period of three years. Antonie Leeuwenhoek 79:345 352. 36. Török, T., R. K. Mortimer, P. Romano, G. Suzzi, and M. Polsinelli. 1996. Quest for wine yeasts an old story revisited. J. Ind. Microbiol. 17:303 313. 37. van Keulen, H., D. G. Lindmark, E. K. Zeman, and W. Gerlosky. 2003. Yeasts present during spontaneous fermentation of Lake Erie Chardonnay, Pinot Gris and Riesling. Antonie Leeuwenhoek 83:149 154. 38. Vezinhet, F., J. N. Hallet, M. Valade, and A. Poulard. 1992. Ecological survey of wine yeast strains by molecular methods of identification. Am. J. Enol. Vitic. 43:83 86. 39. White, T. J., T. Bruns, S. Lee, and J. Taylor. 1990. Amplification and direct sequencing of fungal ribosomal RNA genes for phylogenetics, p. 315 322. In M. A. Innis, D. H. Gelfand, J. J. Sninsky, and T. J. White (ed.), PCR protocols, a guide to methods and applications. Academic Press, San Diego, CA. 40. Wintzingerode, F., U. Gobel, and E. Stackebrandt. 1997. Determination of microbial diversity in environmental samples: pitfalls of PCR-based rrna analysis. FEMS Microbiol. Rev. 21:213 229. 41. Wu, Z. W., and F. Y. Bai. 2005. Kazachstania aquatica sp. nov. and Kazachstania solicola sp. nov., novel ascomycetous yeast species. Int. J. Syst. Evol. Microbiol. 55:2219 2224. 42. Zubkova, R. D. 1971. Genus novum Saccharomycetacearum e Kazachstania. Bot. Mater. Gerb. Inst. Bot. Akad. Nauk. Kazakh. SSR 7:53 56.