Development of microsatellite markers for Lachancea thermotolerans typing and population structure of wine-associated isolates

Microbiological Research 193 (2016) 1–10

Contents lists available at ScienceDirect

Microbiological Research journal homepage: www.elsevier.com/locate/micres

Development of microsatellite markers for Lachancea thermotolerans typing and population structure of wine-associated isolates Georgios Banilas a , Georgios Sgouros b,c , Aspasia Nisiotou b,∗ a b c

Department of Enology and Beverage Technology, Technological Educational Institute of Athens, Ag. Spyridonos Street, 12210, Greece Institute of Technology of Agricultural Products, ELGO DEMETER, 1 S. Venizelou Str., Lykovrysi, 14123, Greece Department of Molecular Biology and Genetics, Democritus University of Thrace Dragana, Alexandroupolis, 68100, Greece

a r t i c l e

i n f o

Article history: Received 30 April 2016 Received in revised form 9 August 2016 Accepted 13 August 2016 Available online 24 August 2016 Keywords: Yeast biodiversity Wine Lachancea thermotolerans Microsatellites SSRs Non-Saccharomyces

a b s t r a c t Lachancea (Kluyveromyces) thermotolerans is an important member of the grape/wine yeast community with great technological potential for the wine industry. Although several molecular marker techniques have been developed for typing different yeast species, no one has been designed so far for L. thermotolerans. Here we present a simple and efficient method based on a multilocus SSR analysis for molecular typing and genetic diversity assessment of L. thermotolerans isolates. Following whole genome screening, five polymorphic microsatellite markers were selected and tested on a panel of grape isolates from different vineyards of two geographically separated viticultural zones, Nemea and Peza, in Greece. The SSR method proved quite discriminatory as compared to tandem repeat-tRNA-PCR, a fingerprinting method for typing non-Saccharomyces yeasts. Genetic analysis based on SSR data revealed a clear structure between the populations of the two zones. Furthermore, significant differences were also detected in a number of phenotypic characters of enological interest. A positive correlation was observed between phenotypic and genotypic diversity. Taking together, present results support the microbial terroir concept in the case of L. thermotolerans in Greece, which is an important prerequisite for the exploitation of selected genotypes as fermentation starters with region-specific characters. © 2016 Elsevier GmbH. All rights reserved.

1. Introduction Lachancea thermotolerans (formerly Kluyveromyces thermotolerans) is an ascomycetous yeast species associated with fruits, Drosphila species and other plant-feeding insects (Ganter, 2006). It is considered a regular inhabitant of the grape/wine ecosystem, as it has been frequently encountered in grapes and fermenting musts in several viticultural regions worldwide (Jolly et al., 2003; Mills et al., 2002; Nisiotou et al., 2007; Torija et al., 2001). L. thermotolerans belongs to the group of the so-called non-Saccharomyces or ‘wild’ yeasts, originally derived from grapes, which due to their numerical supremacy in fresh must they commence the alcoholic fermentation. Although they are later replaced by the most alcohol tolerant S. cerevisiae, L. thermotolerans may exhibit further persistence and survive the elevated ethanol concentrations at the end of the fermentation course (Mills et al., 2002; Nisiotou et al., 2007). For long, the presence of non-Saccharomyces yeasts in winemaking has been viewed under skepticism, as they have been

∗ Corresponding author. E-mail address: [email protected] (A. Nisiotou). http://dx.doi.org/10.1016/j.micres.2016.08.010 0944-5013/© 2016 Elsevier GmbH. All rights reserved.

associated with stuck or sluggish fermentations and wines with unacceptable organoleptic characteristics (Ciani et al., 2010). However, their role has been recently revisited, since it has been shown that their activity may be also highly beneficial (Capozzi et al., 2015; Jolly et al., 2014). Therefore, different grape-related yeast species of various genera, like Candida, Metschnikowia, Pichia, Torulaspora, and Lachancea, have been evaluated as potential wine yeast starters in mixed cultures with S. cerevisiae (Benito et al., 2015; Comitini et al., 2011; Englezos et al., 2015). Selected strains are currently used in commercially produced mixed inocula for wine production. Among the non-Saccharomyces wine yeasts, L. thermotolerans has attracted much attention for the aroma compounds it may confer to wines, including significant amounts of 4-methyl4-sulfanylpentan-2-one (box-tree odor) and 3-sulfanylhexan-1-ol (grapefruit and passion fruit nuances) and for increased production of lactic acid and glycerol (Ciani and Ferraro, 1998; Comitini et al., 2011; Gobbi et al., 2013; Kapsopoulou et al., 2007; Soden et al., 2000; Zott et al., 2011). These properties have led to the development of a commercial L. thermotolerans dry yeast product, which according to the manufacturer produces lactic acid giving roundness and balanced acidity and confers increased flavor

2

G. Banilas et al. / Microbiological Research 193 (2016) 1–10

Fig. 1. Map of the vineyards in the viticultural zones of Nemea and Peza in Southern Greece. Genotypes (I–IX) detected in each vineyard are indicated in parentheses.

impact to wines (http://www.chr-hansen.com/food-cultures-andenzymes/wine/cards/collection-cards/speciality-yeast). Despite the importance of L. thermotolerans for the wine and food industry, our current knowledge on the ecology, distribution and population genetics of this species is limited. Recently, the genome diversity and evolution of L. thermotolerans was investigated by analyzing the mitochondrial (mt) genomes of 50 isolates from diverse geographical regions and ecological niches and a relatively low mt genetic diversity was found (Freel et al., 2014). Currently, much information at the genomic level is available, given that the genome of the type strain CBS 6340 has been completely sequenced (Talla et al., 2005). However, in order to facilitate strain discrimination and identification of wine-important non-Saccharomyces yeasts, there is an increasing need to develop methods for molecular marker-assisted genotyping. Such methods should ideally be highly discriminatory, affordable, simple and rapid, in order to track and monitor selected strains in complex microbial communities during the fermentation course. The need to develop such methods is further underlined by the increasing interest on ecological aspects of wine yeasts, mainly due to possible association of microbial mosaic to wine “terroir”. The recently emerging concept of ‘microbial terroir’, i.e. the association of native yeast populations with distinctive geographical wine phenotypes, may strengthen the originality and typicity of wines and deliver added value to the product. Although different molecular methods have been developed for typing wine yeasts, most of them have been applied to S. cerevisiae. To our knowledge, not any molecular marker method has been developed so far for L. thermotolerans typing. Microsatellites or simple sequence repeats (SSRs) are tandem repeats of a short DNA sequence motif, usually up to 6 bp long. They are characterized by a relatively high mutation rate, presenting high variability in the repeat number. Alleles are easy-to-score by means of PCR amplification and fragment-size analysis. SSRs provide several advantages over most other size-based molecular markers, among which are the high reproducibility within and among laboratories and the unambiguous scoring, while the analysis can be semi-automated. More importantly, unlike other fingerprinting methods such as RAPDs and AFLPs, SSRs are codominant markers, allowing extracting evolutionary relationships among individuals. They are thus widespread in the genome of most eukaryotes and

have proved particularly suitable markers for molecular fingerprinting, population genetics, and phylogenetic analysis in various taxa, including wine-related yeasts (Albertin et al., 2014; Goddard et al., 2010; Richards et al., 2009; Tofalo et al., 2013). With respect to the grape/wine ecosystem, SSRs have been mainly applied to S. cerevisiae (Legras et al., 2007; Richards et al., 2009; Schuller and Casal, 2007). Only recently they are beginning to expand to few non-Sacharomyces species, like Brettanomyces bruxellensis (Albertin et al., 2014), Candida zemplinina (Masneuf-Pomarede et al., 2015), and Hanseniaspora uvarum (Albertin et al., 2016). This is probably due to the lack of genome sequencing data available in most of the wild yeast species, a major limitation to a wider exploitation of SSRs. Here, taking advantage from the complete genome sequence of L. thermotolerans, a set of SSR markers was developed able to differentiate isolates of grape origin and to describe the genetic structure of populations from two geographically separated viticultural zones in Greece. A thorough analysis of enologically important characteristics of strains from different geographical regions was also conducted that showed important phenotypic to genotypic relatedness.

2. Materials and methods 2.1. Yeast isolation and species identification L. thermotolerans yeasts were isolated from fermented grape must originating from various vineyards of Nemea (Northern Peloponnese) or Peza (Crete) region, representing two major viticultural zones in Greece (Fig. 1, Table 1). In Nemea all vineyards have been cultivated with the red grapevine variety ‘Agiorgitiko’, whereas in Peza three varieties were sampled, the red ones ‘Mandilari’ and ‘Kotsifali’ and the white variety ‘Vilana’. After harvest, grapes were placed into sterile plastic bags and transferred at 4 ◦ C to the laboratory, where crushed with a stomacher and let to ferment spontaneously in sterile bottles. Yeasts were isolated at the end of the fermentation course after culture plating in Lysine agar medium (Oxoid, Unipath Ltd., Hampshire, UK). The type strain L. thermotolerans CBS6340 was also included in the analysis. All isolates were identified at the species level after PCR amplification using the universal primers ITS1 and ITS4 and sequencing analysis of the 5.8S-ITS

G. Banilas et al. / Microbiological Research 193 (2016) 1–10

3

Table 1 L. thermotolerans isolates analyzed in this study. Geographical origin

Grape cultivar

Vineyard

Fermentation sample

Code of isolate

Genotype group

Nemea

Agiorgitiko

Empatos

N-EM

Kalyvia

N-KA

Tripodi

N-TR

Psari Asprokampos

N-PS N-AS

Mandilari

Panorama

P-PN

Vilana

Patsideros

P-PTa

Kotsifali

Patsideros

P-PTb P-PTc

Vilana

Houdetsi-1

P-HOa

Kotsifali

Houdetsi-2

P-HO

Mandilari

Kalloni

P-KA

N-EM1 N-EM2 N-EM3 N-EM4 N-EM5 N-EM6 N-EM7 N-KA1 N-KA2 N-TR1 N-TR2 N-TR3 N-TR4 N-TR5 N-PS1 N-AS1 N-AS2 P-PN1 P-PN2 P-PN3 P-PN4 P-PN5 P-PN6 P-PN7 P-PT1 P-PT2 P-PT3 P-PT4 P-PT5 P-PT6 P-HO1 P-HO2 P-HO3 P-HO4 P-HO5 P-HO6 P-HO7 P-HO8 P-HO9 P-HO10 P-HO11 P-HO12 P-HO13 P-KA1 P-KA2 P-KA3 P-KA4

I I I I I II I II II III III III III III IV V V VI VI VI VI VI VI VI VI VI VI VI VII VII VIII VI IX IX IX IX IX IX IX IX IX IX IX VI VI VI VI

Peza

The complete genome sequence of Lachancea thermotolerans type strain CBS 6340 was downloaded from the Génolevures site (http://www.genolevures.org) and analyzed to identify potential microsatellite sequences using the Tandem Repeats Finder (TRF) software (https://tandem.bu.edu/trf/). Using the Primer3 software (http://primer3.sourceforge.net/) primer sets were designed in the flanking regions of microsatellite loci. The specificity of the primers was checked by BLAST search and verified by direct sequencing of the PCR products (Macrogen Inc., Seoul, Korea)

thermal cycler with an initial desaturation step at 94 ◦ C for 3 min followed by 35 cycles of 94 ◦ C for 30 s, 60 ◦ C for 30 s, 72 ◦ C for 40 s, and a final extension at 72 ◦ C for 10 min. PCR products of selected strains were initially run on 3% agarose gels for detection of polymorphic loci and selection of SSRs. Forty-seven L. thermotolerans isolates, along with the type strain, were then subjected to SSR fragment analysis using fluorescently labeled PCR products. The forward primers were labeled at the 5 -end with NED, VIC, or FAM fluorescent dye (Applied Biosystems, Foster City, CA). PCR products were multiplexed and run in an ABI 3730xl DNA Analyser (Macrogen Inc., Seoul, Korea) using the GeneScan 500 LIZ as size standard (Applied Biosystems). Data were visualized and allele fragment sizes were scored with the Peak ScannerTM ver. 1.0 (Applied Biosystems).

2.3. PCR amplification and fragment analysis

2.4. Fingerprinting by tandem repeat-tRNA

PCRs were carried out in a final volume of 50 ␮l containing 20 ng of template DNA, 0.2 ␮M of each primer, 0.2 mM of each deoxynucleoside triphosphate, and 1 U of Taq DNA polymerase (Kapa Biosystems) in the reaction buffer provided by the manufacturer of the enzyme. Amplification was performed in a Bio-Rad

For molecular fingerprinting of L. thermotolerans isolates, the tandem repeat-tRNA (TRtRNA) PCR method was applied using the TtRNASc (5 -GCTTCTATGGCCAAGTTG-3 )/ISSR-MB (CTCACAACAACAACAACA) primer pair as described by Barquet et al. (2012). Only unambiguous bands were scored and used for analysis.

rDNA region, following standard procedures described elsewhere (Nisiotou et al., 2007). 2.2. Microsatellite identification and PCR primer design

4

G. Banilas et al. / Microbiological Research 193 (2016) 1–10

Table 2 Microsatellite sequences initially evaluated, repetitive motifs, genomic sites and primers used for PCR amplification, according to the sequence of the type strain (CBS 6340). The SSRs chosen for the multilocus typing method are indicated in bold. SSR code

Motif (repeats)

Chromosome/ Position

Locus sequence

Primer Forward/ Primer Reverse (5 −3 )a

MA2

CA (11)

A/358081

ORF, XM 002551596

MB1

TA (11)

B/35805

Non-coding sequence

MB4

AGC (11)

B/564251

Non-coding sequence

MB5

CA (36)

B/588760

Non-coding sequence

MC7

CAGCAA (17)

C/554816

ORF, XM 002552455

MC8

TG (13)

C/621113

Intron, XM 002552491

MD3

CAA (17)

D/259537

ORF, XM 002552792

ME4

GA (20)

E/576050

Non-coding sequence

ME5

TA (11)

E/717328

Intron, XM 002553774

ME11

TG (19)

E/1381401

Intron, XM 002554109

MF10

CT (14)

F/1146792

Non-coding sequence

MF14

CTT (11)

F/1474150

ORF, XM 002554933

MG8

TCT (12)

G/983702

Non-coding sequence

MG10

GA (18)

G/1262701

Non-coding sequence

MH6

TGT (22)

H/372940

ORF, XM 002556014

MH7

TGT (12)

H/601386

ORF, XM 002556131

AATTTTACGAAGGGAGAGAGGG/ CTGCTGATGGTTTCTTCTGTGA CTTTGAACTTTCTCGCTTGCTT/ CTGATTACGGCAAATACACCAG GAGATCGGTGTGTATGACGCTA/ GATGGGTATTCTGGCTCACTTC CAGCAAGCCTAGAGAGGAGAGT/ TAGTGTGAGTGCGTGTATGTGC GTCCTGTTAGAAATGTCCCCAT/ GATTTTGATGGTGTTGTTGTGG GGTAGTAGAACTCCTCGACGCT/ ATAATAGGCTGGGGATTGGAC ACAAGAAAGCGAAGGAAAACAG/ CCCAGTAGAACGTGATTAAGCC TGGCCTCTTCTGTCTTTCCTAA/ CTCATCAACCAACACACTCCAT AGCGCAGAAAGTCCTGAAATAG/ GAAGTACGGTGAAGACTGGGAC CGGTTCTTAGCTTACCAACAGC/ ACTCGAACAGCCAGAGCTTAAC TGTAGTCGTGTTGGTTCATTGC/ GGAGCCTTGTGAGTGGATCTTA TTAGTACGCTTGTTTGTGCGTT/ TTAGTGTGCCTCAAGTCTCTGC CTGCCGAATTTTGTTAAGCTCT/ AGGATGATGCAAGTGAGGAAGT ACTCCACTGACGCTTCGTATTT/ GAAAACAAATCCCTTTCTCGTG CTTGCTGTTGTCGTAACCTCTG/ GAATCCCAATAATCTCACACCC ACATGGACTGGTAGTGCTGATG/ AAAGTTCAAGAGCGTCAAGACC

a

The Tms of primers ranged between 59.5 and 60.5 ◦ C.

2.5. Technological characterization of yeasts Ethanol resistance was evaluated by spot-inoculating yeast cells at 105 cfu/ml on agarised grape must plates containing 6, 8 or 10% (vol/vol) ethanol (Mauriello et al., 2009). Plates were incubated at 26 ◦ C for 72 h. For determining sulfur dioxide (SO2 ) resistance, cells were spot-inoculated at 105 cfu/ml on agarised grape must plates containing 50, 75, 100 or 150 mg/L SO2 (Mauriello et al., 2009) and incubated at 26 ◦ C for 48 h. An arbitrary scale with four levels (1–4) was used to evaluate growth with 1 corresponding to null, 2 to small, 3 to medium and 4 to high growth. H2 S production was determined on Bismuth Sulfite Glucose Glycine Yeast (BIGGY) agar. Inoculated plates (ca. 105 cells/ml) were incubated at 26 ◦ C for 48 h and the production of H2 S was revealed by pigmentation of colonies (Pando Bedrinana et al., 2010). The following scale was followed in order to evaluate H2 S production: white colour, 1 (no production); light yellow colour, 2 (low production); dark yellow colour, 3 (medium production); brown or black colour, 4 (high production). The decarboxylation of arginine, histidine, leucine, lysine, phenylalanine, tryptophan and tyrosine was estimated in YPD agar plates containing 0.006% bromocresol purple and 1% of the respective amino acid (Nikolaou et al., 2006). Yeasts were spot-inoculated at 105 cfu/ml and incubated at 25 ◦ C for 7 days. The formation of a purple halo around the yeast colony was evaluated as 1, no production; 2, low production; 3, medium production; or 4, high production. Flocculation phenotype was evaluated as previously described (Suárez Valles et al., 2008) with minor modifications as follows. Cells were incubated in 10 ml of pasteurized grape must (25 ◦ C, 72 h) and resuspended in 5 ml of Helm’s buffer (3 mM calcium chloride, 50 mM acetate–acetic buffer, pH 4.5). Flocculation capacity was spectrometrically estimated using the OD ratio

at 620 nm of the cell culture at time 0 and 10 min after stirring (OD10 × 100/OD0 ). 2.6. Microfermentations Fermentations were carried out at 25 ◦ C in 250 ml flasks containing 170 ml of pasteurized (100 ◦ C, 20 min) grape must with 215 g/L sugars (pH 3.6) in duplicate. The must was inoculated with L. thermotolerans strains at a final concentration of 106 cfu/ml. The flasks were closed with fermentation locks containing glycerol to allow only CO2 to escape from the system. The weight loss due to CO2 production was followed daily until the end of fermentation (constant weight for two consecutive measurements). Total acidity, volatile acidity and pH were estimated according to the Official European Union Method (EC, 2000). Sugars, ethanol and lactic acid were determined by enzymatic kits (R-Biopharm, Darmstadt, Germany). 2.7. Data analysis The discriminatory power (DP) of typing methods was calculated by Simpson’s index of diversity, based on the probability that two unrelated strains sampled from the test population will be placed into different typing groups, according to Hunter and Gaston (1988). Dendrograms were constructed by UPGMA clustering, based on Dice correlation coefficient, and compared with Mantel’s test available in the PAST software version 3.11 (Hammer et al., 2001). Genetic population structure and technological traits on agar plates were inferred by analysis of similarity (ANOSIM). Analysis of variance (ANOVA) was used to compare strain fermentation characteristics. The relationship between genetic distance and geography was examined by Mantel’s test. Permutational multivariate analysis of variance (PERMANOVA) was applied to test

G. Banilas et al. / Microbiological Research 193 (2016) 1–10

for phenotypic differences among groups of genotypes or between regions. ANOSIM and PERMANOVA tests were performed with 9999 permutations as routines included in the PAST software package. 3. Results 3.1. Selection of the SSR markers The genome of the L. thermotolerans type strain CBS 6340 (8 chromosomes, total size 10.4 Mb) was analyzed to identify tandem iterations of two or more nucleotides. Only microsatellites of at least ten repeats were retained for further analysis in order to increase the possibility for being polymorphic, given that the most important factor to affect mutation rate of a microsatellite locus seems to be its length (Ellegren, 2004). We also excluded compound microsatellites, i.e. loci containing two or more tandem repeats contiguously with short intervening non-repeat sequences, because they display complexity in defining different alleles and determining the nature of polymorphisms. As a result, 16 potentially perfect microsatellite loci (SSRs) were identified and retained for further analysis (Table 2). Primers were designed from the flanking regions of these SSRs and applied in PCR amplifications using the genomic DNA of the L. thermotolerans type strain as template. The PCR products were sequenced to confirm specific amplification and accuracy of the tandem repeat sequences. Length polymorphisms of the microsatellite loci were initially assessed through agarose gel electrophoresis in a panel of five L. thermotolerans strains, including the type strain and grape isolates from two geographically separated viticultural zones (Nemea and Peza). Only five SSR markers exhibited allele variation of fragment length polymorphism and were thus chosen for the multiplex SSR method and to study genetic diversity of vineyard isolates. These five SSR markers were located at 4 different chromosomes (A, E, D, and H). Three of them were di-nucleotides (MA2, ME4, ME11), while the other two SSRs (MD3 and MH6) consisted of trinucleotide motifs. Three SSRs were located in predicted putative coding sequences, MD3 within an open reading frame (ORF) coding for a hypothetical protein, MA2 within an ORF similar to S. cerevisiae Pho85 cyclin-dependent kinase gene, and MH6 was placed within a homolog of the S. cerevisiae Vps41 gene.

5

isolate could be characterized by a set of five alleles. Within the two natural populations analyzed from Nemea and Peza regions, a total of ten genotypes were recovered (genotypes I–IX, Table 1), while the type strain generated a distinct profile. The amplified alleles per SSR locus ranged from five (ME4) to eight (MH6). Fragment sizes of SSRs were well distributed in different size-ranges from 142 bp to 454 bp, thus allowing multiplexing of the amplicons by using different fluorescence dyes (Table 3). By combining the results obtained from the five SSRs, the numerical index of DP for the typing method was 0.818. Each geographical region possessed unique molecular patterns. A single genotype was retrieved from each fermentation vessel, except for samples N-EM and P-HOa from which two different genotypes per fermentation were recovered. The within region strain distribution differed between the two zones. In Nemea region, unique genotypes were encountered within each vineyard, except for genotype II that was encountered in both the nearby located vineyards of Empatos and Kalyvia. On the contrary, genotype VI was widely distributed throughout Peza region as it was isolated from 4 out of the 5 vineyards examined (Table 1, Fig. 1). The SSR data were used to estimate genetic distances and to cluster the isolates according to their genetic relatedness. A UPGMA dendrogram grouped the isolates in respect to their region of origin (Fig. 2). The population structure observed between Nemea and Peza isolates was further confirmed by ANOSIM, which showed highly significant differences between the two populations (R = 0.6508, p < 0.001). Most isolates from Peza were placed in a distinct cluster and were further subdivided into two typing subgroups. The Nemea isolates were also clustered together into 3 major typing subgroups. The geographically separated isolates N-AS1 and N-AS2 from Asprokampos vineyard in Nemea were grouped together and apart from the others isolates. Similarly, the single isolate from Psari vineyard (N-PS1) did not show any particular relationship to any other group, except from a relatively low similarity to the type strain. A positive relationship between genetic and geographical distance in Nemea vineyards was confirmed by Mantel’s test (R = 0.5465, p < 0.05). On the contrary, in Peza region not any correlation between genetic and geographic distances was observed (R = 0.023, p = 0.378). 3.3. Comparing SSRs with TRtRNA-PCR

3.2. SSR typing and population genetic analysis The multilocus SSR typing method was applied to 47 vineyard L. thermotolerans isolates from two geographically separated regions, Nemea in Peloponnese and Peza in the island of Crete. The SSRs were PCR-amplified with fluorescently labeled forward primers (Table 3) and run on a capillary electrophoresis system. All isolate/SSR combinations generated clear PCR fragments and no null alleles were detected. Replicate SSR analysis from independent DNA samples produced highly reproducible results, confirming the accuracy and reproducibility of the method. Interestingly, only single PCR fragments per SSR locus were obtained for all isolates examined, as would be expected for haploids or highly homozygous genotypes. Based on the combined data generated by the SSRs, each Table 3 Number of alleles and range of sizes for the five SSRs applied in typing natural L. thermotolerans populations. SSR code

Fluorescence dye

No. of alleles

Range of PCR fragment sizes (bp)

MA2 MD3 ME4 ME11 MH6

NED NED VIC 6-FAM 6-FAM

7 7 5 7 8

281–343 343–383 330–384 142–192 391–454

The strength of the novel SSR method for typing and detecting genetic diversity was compared to that of TRtRNA-PCR, a previously developed fingerprinting method for wine-associated non-Saccharomyces yeasts (Barquet et al., 2012). By using the TtRNASc/ISSR-MB primer pair in PCR amplifications, all L. thermotolerans isolates yielded clear banding patterns on agarose gels (Fig. 3). Analysis of the 47 individuals resulted into 18 total bands, of which 10 were polymorphic, giving a polymorphic value of 55.5%. Based on the polymorphic bands, six distinct molecular patters were obtained. The DP of the TRtRNA-PCR method was 0.592, which is a much lower value than the one recorder for SSRs (0.818). A UPGMA dendrogram based on TRtRNA-PCR showed clustering of isolates into two major groups according to the region of origin (data not shown). Although differences in clustering were detected between the UPGMA dendrograms generated by the two methods, Mantel test indicated a fairly good correlation (R = 0.6153, P < 0.001). 3.4. Phenotypic analysis Twenty-five isolates from different genotypic groups were randomly selected to represent the genetic diversity in each vineyard and region and examined in a number of phenotypic tests.

6

G. Banilas et al. / Microbiological Research 193 (2016) 1–10

Fig. 2. UPGMA dendrogram of the L. thermotolerans isolates based on the multilocus SSR analysis. The first letter of each code stands for the viticultural zone (N: Nemea or P: Peza), and the following two letters refer to the vineyard of origin, as indicated in Table 1. The type strain is CBS 6340.

Fig. 3. Representative agarose gel showing TRtRNA-PCR banding patterns obtained from different isolates from Nemea and Peza viticultural zones. M, 100 bp molecular marker.

Differences were observed between the two regional groups in ethanol or SO2 tolerance and H2 S or biogenic amine production, except for histidine decarboxylation for which all isolates showed negative reaction (Supplementary Fig. S1 in the online version at DOI: 10.1016/j.micres.2016.08.010). According to ANOSIM, SO2 tolerance was the technological trait with the highest discriminatory capacity between the two groups of isolates (R = 0.5427,

p < 0.0001), followed by tyrosine decarboxylation (R = 0.4596, p < 0.0001) and H2 S production (R = 0.4144, p < 0.0001). The other technological attributes showed lower between-group divergence (3.000 < R < 0.1309). In more detail, the vast majority of Peza isolates (93%) exhibited high growth at 100 ppm SO2 , whereas most Nemea isolates (80%) showed null to small growth. All Peza isolates produced medium to high amounts of H2 S, while the respective percentage in Nemea was 50%. Yeasts from both regions were highly resistant at 6% ethanol, except for 30% of Nemea isolates that showed medium growth. These Nemea isolates also showed low tolerance at 8% ethanol, whereas the other isolates showed medium to high resistance. The most frequently decarboxylated amino acid was tryptophan consumed rather weakly by all Peza and most (70%) of the Nemea yeasts, while 30% of Nemea strains showed medium to high decarboxylation capacity. Similar results were observed for arginine, with the majority of Peza isolates (73%) giving a weak positive reaction. The respective proportion for Nemea yeasts was 50%, while several strains (40%) gave moderately to highly positive reactions. With respect to the other amino acids, Peza isolates did not decarboxylate leucine, lysine nor phenylalanine, while only one isolate weakly reacted on tyrosine. On the other hand, 40% of Nemea isolates reacted moderately to highly on leucine, lysine and phenylalanine whereas 70% of strains were positive for tyrosine decarboxylation. Flocculation phenotype was more intense in Nemea than in Peza isolates (ANOVA, F = 15.17, p < 0.001). All, except from one, Peza yeasts showed low OD ratios, which are indicative of highly flocculent strains, ranging between 4.7443 and 23.9028 (average 10.9069 ± 5.0508). As opposed, Nemea isolates showed higher OD ratios than Peza, ranging from 16.3994 to 99.8531 (average 70.8013 ± 37.4016). Microfermentations revealed that Peza isolates, except for PPT5, had similar fermentation profiles to each other and formed a rather tight, well-separated group (Fig. 4). Nemea fermentation profiles showed higher within-group diversity than those of Peza. Significant differences (ANOVA, F = 122, p < 0.001) were observed in the total CO2 released, with the Nemea group producing higher quantities (11.3 ± 0.5 g/L) than Peza (9.8 ± 0.4 g/L). Accordingly, fermentation rate was significantly higher in Nemea than in Peza isolates (F = 57.1, p < 0.001). Significant differences were also observed in fermentation characteristics between the two geographical groups, except for volatile

G. Banilas et al. / Microbiological Research 193 (2016) 1–10

7

Fig. 4. Fermentation profiles of L. thermotolerans isolates from Peza (grey lines) and Nemea (black lines) in pasteurized grape must.

Table 4 Enological characteristics of L. thermotolerans strains in pasteurized grape must. Values within the same column without a common letter are significantly different according to Tukey’s HSD test (p < 0.05). Strain

Fermentation rate (g/day)a

Nemea isolates N-EM1 5.15 ± 0.24 a N-EM3 5.08 ± 0.01 a N-EM5 5.09 ± 0.10 a N-EM6 4.36 ± 0.01 bcd N-KA1 4.46 ± 0.19 bc 4.50 ± 0.14 b N-TR1 4.28 ± 0.03 bcde N-TR4 N-TR5 4.17 ± 0.05 bcdef 5.21 ± 0.13 a N-PS1 5.19 ± 0.11 a N-AS1 Peza isolates 3.86 ± 0.06 efgh P-PN1 3.75 ± 0.11 fgh P-PN2 3.72 ± 0.00 fgh P-PN3 3.74 ± 0.04 fgh P-PT1 4.04 ± 0.40 bcdefg P-PT4 5.19 ± 0.01 a P-PT5 3.85 ± 0.02 efgh P-HO1 3.57 ± 0.11 gh P-HO3 3.78 ± 0.12 fgh P-HO8 3.63 ± 0.01 gh P-HO9 3.53 ± 0.09 h P-HO10 3.65 ± 0.01 gh P-HO12 3.58 ± 0.02 gh P-HO13 3.91 ± 0.02 defgh P-KA1 3.96 ± 0.10 cdefgh P-KA2 a

Ethanol (g/L)

Residual sugars (g/L)

Glucose/Fructose

pH

Total acidity (g/L tartaric acid)

Volatile acidity (g/L acetic acid)

Lactic acid (g/L)

72 ± 1 ab 73 ± 0 ab 73 ± 0 ab 69 ± 2 bc 70 ± 0 b 73 ± 2 ab 75 ± 1 ab 75 ± 1 ab 71 ± 1 b 78 ± 2 a

60.2 ± 6.4 f 65.5 ± 0.2 ef 67.5 ± 1.8 def 73.1 ± 2.9 bcdef 77.1 ± 0.5 abcde 61.6 ± 9.6 f 65.5 ± 0.2 ef 67.4 ± 0.6 def 86.9 ± 2.1 a 76.8 ± 0.2 abcde

0.76 ± 0.03 f 0.78 ± 0.01 ef 0.79 ± 0.01 ef 0.81 ± 0.01 e 0.81 ± 0.00 e 0.87 ± 0.02 d 0.90 ± 0.00 bcd 0.90 ± 0.00 bcd 0.88 ± 0.00 cd 0.78 ± 0.02 ef

3.50 ± 0.00 c 3.51 ± 0.00 bc 3.52 ± 0.00 abc 3.56 ± 0.01 a 3.55 ± 0.01 ab 3.50 ± 0.00 c 3.52 ± 0.01 bc 3.51 ± 0.00 bc 3.44 ± 0.01 d 3.45 ± 0.01 d

5.2 ± 0.1 d 5.1 ± 0.1 d 5.6 ± 0.2 d 5.0 ± 0.1 d 5.0 ± 0.1 d 5.0 ± 0.0 d 5.6 ± 0.0 d 5.6 ± 0.0 d 5.8 ± 0.1 d 6.0 ± 0.1 d

0.35 ± 0.03 bcdf 0.35 ± 0.03 bcdf 0.35 ± 0.01 bcdf 0.34 ± 0.02 cdf 0.42 ± 0.04 abcd 0.38 ± 0.05 abcdf 0.46 ± 0.01 a 0.41 ± 0.00 abcd 0.31 ± 0.01 f 0.35 ± 0.04 bcdf

1.1 ± 0.2 e 1.1 ± 0.1 e 1.2 ± 0.1 e 1.0 ± 0.1 e 1.2 ± 0.3 e 1.7 ± 0.2 e 2.0 ± 0.1 e 2.1 ± 0.3 e 2.3 ± 0.1 e 2.6 ± 0.1 e

59 ± 1 d 61 ± 1 d 62 ± 1 d 62 ± 2 d 60 ± 6 d 72 ± 0 ab 61 ± 0 d 61 ± 1 d 63 ± 1 cd 63 ± 1 cd 62 ± 0 d 61 ± 1 d 61 ± 1 d 63 ± 1 cd 60 ± 1 d

80.6 ± 4.7 abc 87.5 ± 0.8 a 85.8 ± 1.1 ab 81.6 ± 2.9 abc 82.4 ± 5.0 abc 71.9 ± 2.5 cdef 81.8 ± 0.0 abc 78.6 ± 1.5 abcd 76.8 ± 2.1 abcde 79.4 ± 0.4 abcd 77.8 ± 3.5 abcde 76.8 ± 2.6 abcde 79.1 ± 2.3 abcd 82.0 ± 3.2 abc 84.7 ± 0.1 abc

0.92 ± 0.01 abc 0.94 ± 0.00 ab 0.92 ± 0.00 abc 0.92 ± 0.00 abc 0.90 ± 0.03 bcd 0.87 ± 0.01 d 0.92 ± 0.00 abc 0.94 ± 0.01 ab 0.93 ± 0.01 ab 0.94 ± 0.00 ab 0.93 ± 0.00 ab 0.96 ± 0.01 a 0.95 ± 0.01 a 0.93 ± 0.00 ab 0.94 ± 0.01 ab

3.20 ± 0.01 fgh 3.23 ± 0.01 ef 3.22 ± 0.00 ef 3.22 ± 0.01 ef 3.25 ± 0.04 e 3.52 ± 0.01 bc 3.18 ± 0.01 ghi 3.16 ± 0.00 hi 3.17 ± 0.00 ghi 3.16 ±0.00 hi 3.15 ± 0.00 i 3.16 ± 0.01 i 3.15 ± 0.01 i 3.21 ± 0.01 fg 3.2 ± 0.01 fg

13.7 ± 0.3 b 13.8 ± 0.4 b 13.9 ± 0.2 b 14.1 ± 0.1 b 12.3 ± 1.3 c 5.2 ± 0.1 d 15.6 ± 0.1 a 16.8 ± 0.1 a 15.6 ± 0.1 a 16.8 ± 0.0 a 16.8 ± 0.1 a 16.7 ± 0.4 a 16.6 ± 0.2 a 13.9 ± 0.3 b 14.2 ± 0.5 b

0.32 ± 0.01 df 0.35 ± 0.06 bcdf 0.41 ± 0.01 abcdf 0.39 ± 0.01 abcdf 0.35 ± 0.00 bcdf 0.31 ± 0.04 f 0.38 ± 0.02 abcdf 0.42 ± 0.01 abcd 0.41 ± 0.00 abcd 0.43 ± 0.01 abc 0.40 ± 0.03 abcdf 0.44 ± 0.01 ab 0.44 ± 0.01 abc 0.36 ± 0.01 abcdf 0.40 ± 0.01 abcdf

12.0 ± 0.0 cd 13.0 ± 0.2 bc 12.4 ± 0.4 cd 12.7 ± 0.2 c 10.7 ± 1.4 d 1.4 ± 0.1 e 15.4 ± 0.7 a 15.8 ± 0.2 a 14.8 ± 0.1 ab 16.6 ± 0.4 a 16.0 ± 0.2 a 15.0 ± 0.4 a 15.9 ± 0.6 a 12.1 ± 0.7 cd 12.5 ± 0.9 cd

The rate values refer to the first 3 days of fermentation.

acidity (Table 4). The highest difference was observed in the acidic character of the fermented product, followed by ethanol production and sugar consumption. Peza isolates conferred significantly higher

titratable acidity, lactic acid and lower pH (F = 191-196, p < 0.001) than Nemea. Nemea strains produced higher ethanol concentrations (F = 152, p < 0.001) and consumed sugars more effectively

8

G. Banilas et al. / Microbiological Research 193 (2016) 1–10

Fig. 5. Principal Component Analysis (PCA) of Nemea (open circles) and Peza (closed circles) isolates based on technological characteristics. Et: ethanol tolerance; SO2: SO2 resistance; H2S: H2 S production; Fl: flocculation phenotype; FR: fermentation rate; TA: titratable acidity; VA: volatile acidity; Arg, His, Leu, Lys, Phe, Try, Tyr: decarboxylation of arginine, histidine, leucine, lysine, phenylalanine, tryptophan and tyrosine, respectively. The principal components 1 and 2 account for 96.1% and 2.8%, respectively, of the total variation.

(F = 81, p < 0.001) than Peza yeasts. The higher glucose/fructose ratio of Peza isolates reveals a more fructophilic character for those strains compared to Nemea strains. Taken together, the overall physiological characteristics of isolates belonging to the same typing group were similar to each other but differed significantly from those of the other groups (PERMANOVA, F = 182.2, p < 0.001). Mantel’s test further revealed a positive correlation between phenotypic and genotypic diversity of the isolates (R = 0.6007, p < 0.001). With respect to the region of origin, clear phenotypic differences were also observed (F = 26.49, p < 0.001). According to PCA, Peza isolates formed a tight group, which was placed apart from Nemea isolates that formed two distinct clusters (Fig. 5). As an exception, a single isolate (P-PT5) was grouped within a Nemea cluster. It is worth noting, however, that isolate P-PT5 was genotypically more closely related to Nemea rather than to Peza isolates. 4. Discussion Although various molecular marker techniques are available for strain typing of wine-related yeasts, most of them have been developed for and applied to S. cerevisiae. However, the nonSaccharomyces yeasts may play important roles in the fermentation process and ultimately shape the wine characteristics and quality. L. thermotolerans is such a non-Saccharomyces yeast of great potential for the wine industry (Jolly et al., 2014). In order to better characterize different strains and to facilitate selection of elite genotypes, it is important to develop efficient typing methods for assessment of genetic diversity in natural L. thermotolerans populations. Here we present a novel multilocus SSR typing method for L. thermotolerans. By screening the whole genome sequence of L. thermotolerans numerous tandem repeats were identified consisting of combinations of two or more SSRs adjacent to each other with short in-between sequences. Out of a total 16 putative perfect microsatellites detected, five loci proved polymorphic in an initial screening on agarose gels and were thus selected for further application in natural populations from two distant regions in Greece. Most of the SSRs identified (10 out of 16) were located in noncoding sequences. This was not unexpected, since microsatellites are distributed randomly in the genome and thus most occur in

gene introns or other non-coding DNA (Tóth et al., 2000). However, numerous repeats may also exist in protein-coding genes or ORFs (Li et al., 2004). In this study, three out of the five selected SSRs were located in predicted coding sequences, i.e., a hypothetical protein, a putative Pho85 cyclin-dependent kinase and a putative Vps41, a vacuolar membrane protein subunit of the HOPS complex. It has been proposed that SSRs in transcribed sequences may confer to organism adaptation and evolution by generating extra variability (Levdansky et al., 2008). It is worth-noting, however, that in S. cerevisiae both Rho85 and Vps41 are non-essential genes. We then applied the multilocus SSR method to describe the genetic diversity of L. thermotolerans natural populations from two major viticultural regions in Greece. Interestingly, all isolates bore only one allele at each SSR locus. According to Freel et al. (2014) nearly all natural isolates of L. thermotolerans are haploids. It is therefore more likely that the present isolates represent haploid rather than homozygous genotypes. It is not clear yet how the ploidy state may influence yeast ecology and evolution, but data so far show that haploids adapt faster than diploids in all environments tested (Gerstein et al., 2011). This might be advantageous for the harsh and antagonistic environment of the grape/wine ecosystem, particularly during the alcoholic must fermentation. To our knowledge, not any molecular fingerprinting method has been applied so far for L. thermotolerans. Recently, Barquet et al. (2012) developed a PCR-based fingerprinting method, namely TRtRNA-PCR, to discriminate non-Saccharomyces yeasts at the subspecies level. Although this method has not been tested for typing L. thermotolerans, it proved to be highly discriminatory for Metschnikowia pulcherrima and other non-Saccharomyces species of grape and must origins. TRtRNA-PCR is based on the combination of two primers, one is complementary to a tandem repeat sequence and the other one (TtRNASc) is targeting a tRNA gene sequence randomly dispersed in the genome. We applied TRtRNA-PCR analysis in L. thermotolerans isolates and the results were compared with those obtained by the SSRs. The TRtRNA-PCR produced banding patterns with sufficient polymorphic bands to discriminate Peza from Nemea populations, but scoring was not easy and the banding patterns tend to vary by different DNA templates. The UPGMA dendrograms generated by the two methods were fairly

G. Banilas et al. / Microbiological Research 193 (2016) 1–10

comparable. However, the numerical index of DP of TRtRNA-PCR was much lower than that of the SSR method. Although some microbial species may exhibit chaotic population structure (Gerstein and Moore, 2011), it has been recently shown that populations of S. cerevisiae and C. zemplinina associated with grapevines exhibit regional delineations (Knight and Goddard, 2015; Masneuf-Pomarede et al., 2015). In the present study, a clear separation between the L. thermotolerans populations of Peza and Nemea was detected. These regions are geographically separated by the Aegean Sea and thus it is possible that the geographic isolation may have caused some degree of reproductive isolation in the two populations. The pattern of spatial distribution of isolates was shown to differ between the two zones. Unique genotypes were encountered within each vineyard in Nemea, with a fairly good correlation between genetic and geographical distances. In Peza, besides 3 unique genotypes in 3 respective vineyards, a cosmopolitan genotype was found to be a common constituent of the community. Present results point to clonal expansion of isolates in restricted regions as distinct vineyards, while well-adapted cosmopolitan strains may be also part of the ecosystem. Our findings are in agreement with those obtained for S. cerevisiae isolates in New Zealand (Goddard et al., 2010), a primarily clonal species (Ruderfer et al., 2006), suggesting that clonality does not necessarily preclude intraspecies genetic variation. It has been suggested that the genetic similarity among strains as described by molecular marker techniques does not necessarily coincide with physiological similarity (Pfliegler et al., 2014). In the present study, however, the physiological characteristics of isolates of the same typing group were similar to each other and differed significantly from those of the other groups. Mantel’s test further revealed a positive correlation between phenotypic and genotypic diversity of the isolates. Our results suggest that although physiological diversity may be higher than the ‘fingerprinting-diversity’, it seems that phenotypic differentiation is dictated by genotypic delineation. The spatial genetic differentiation of isolates observed here may support the microbial terroir concept, which states that different viticultural regions maintain different microbial communities, which in turn may contribute regional characters to wines (Gayevskiy and Goddard, 2012). This concept has been supported by experiments conducted using metagenomics at the early fermentation stage, which showed differences in the relative abundances of bacteria and fungi in different wine-growing regions in California (Bokulich et al., 2014). Recently, it was shown that genetically distinct wine-related S. cerevisiae sub-populations in New Zealand vineyards differentially affect wine phenotype (Knight et al., 2015). It is still unclear if the microbial terroir concept may also apply in non-Saccharomyces wine yeast species. If this is the case, it would be very important for the wine industry, since the so-called wild yeasts may dramatically contribute to the wine phenotype. Here we showed a defined biogeography for L. thermotolerans in Greece. Despite the intra-regional genetic variation of isolates, the genetic structure between Nemea and Peza populations coincides with distinct oenological phenotypes, a necessary prerequisite for the microbial terroir contribution to wines. Conflict of interest The authors declare that they have no conflict of interest. Acknowledgments This work was partially supported by the European Regional Development Fund (ERDF) and National Resources, under the

9

Operational Program Competitiveness and Entrepreneurship (EPAN II), Action “COOPERATION 2011” [grant number 11SYN 2 704] and by European Union’s Seventh Framework Programme (FP7/2007–2013) [grant number 315065]. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.

References Albertin, W., Panfili, A., Miot-Sertier, C., Goulielmakis, A., Delcamp, A., Salin, F., et al., 2014. Development of microsatellite markers for the rapid and reliable genotyping of Brettanomyces bruxellensis at strain level. Food Microbiol. 42, 188–195. Albertin, W., Setati, M.E., Miot-Sertier, C., Mostert, T.T., Colonna-Ceccaldi, B., Coulon, J., et al., 2016. Hanseniaspora uvarum from winemaking environments show spatial and temporal genetic clustering. Front. Microbiol. 6 (1569), http://dx.doi.org/10.3389/fmicb.2015.01569. Barquet, M., Martin, V., Medina, K., Perez, G., Carrau, F., Gaggero, C., 2012. Tandem repeat-tRNA (TRtRNA) PCR method for the molecular typing of non-Saccharomyces subspecies. Appl. Microbiol. Biotechnol. 93, 807–814. Benito, A., Calderón, F., Palomero, F., Benito, S., 2015. Combined use of selected Schizosaccharomyces pombe and Lachancea thermotolerans yeast strains as an alternative to the traditional malolactic fermentation in red wine production. Molecules 20, 9510–9523. Bokulich, N.A., Thorngate, J.H., Richardson, P.M., Mills, D.A., 2014. Microbial biogeography of wine grapes is conditioned by cultivar, vintage, and climate. Proc. Natl. Acad. Sci. U. S. A. 111, E139–E148. Capozzi, V., Garofalo, C., Chiriatti, M.A., Grieco, F., Spano, G., 2015. Microbial terroir and food innovation: the case of yeast biodiversity in wine. Microbiol. Res. 181, 75–83. Ciani, M., Ferraro, L., 1998. Combined use of immobilized Candida stellata cells and Saccharomyces cerevisiae to improve the quality of wines. J. Appl. Bacteriol. 85, 247–254. Ciani, M., Comitini, F., Mannazzu, I., Domizio, P., 2010. Controlled mixed culture fermentation: a new perspective on the use of non-Saccharomyces yeasts in winemaking. FEMS Yeast Res. 10, 123–133. Comitini, F., Gobbi, M., Domizio, P., Romani, C., Lencioni, L., Mannazzu, I., et al., 2011. Selected non-Saccharomyces wine yeasts in controlled multistarter fermentations with Saccharomyces cerevisiae. Food Microbiol. 28, 873–882. EC, 2000. Official Journal of the European Community. Community reference methods for the analysis of spirit drinks. Reg. EC No 2870/00. December 19, 2000. Brussels. Ellegren, H., 2004. Microsatellites: simple sequences with complex evolution. Nat. Rev. Genet. 5, 435–445. Englezos, V., Rantsiou, K., Torchio, F., Rolle, L., Gerbi, V., Cocolin, L., 2015. Exploitation of the non-Saccharomyces yeast Starmerella bacillaris (synonym Candida zemplinina) in wine fermentation: physiological and molecular characterizations. Int. J. Food Microbiol. 199, 33–40. Freel, K.C., Friedrich, A., Hou, J., Schacherer, J., 2014. Population genomic analysis reveals highly conserved mitochondrial genomes in the yeast species Lachancea thermotolerans. Genome Biol. Evol. 6, 2586–2594. Ganter, P.F., 2006. Yeast and invertebrate associations. In: Rosa, C.A., Gabor, P. (Eds.), Biodiversity and Ecophysiology of Yeasts. Springer, Berlin, pp. 303–370. Gayevskiy, V., Goddard, M.R., 2012. Geographic delineations of yeast communities and populations associated with vines and wines in New Zealand. ISME J. 6, 1281–1290. Gerstein, A.C., Moore, J.B., 2011. Small is the new big: assessing the population structure of microorganisms. Mol. Ecol. 20, 4385–4387, http://dx.doi.org/10. 1111/j.1365-294X.2011.05263.x. Gerstein, A.C., Cleathero, L.A., Mandegar, M.A., Otto, S.P., 2011. Haploids adapt faster than diploids across a range of environments. J. Evol. Biol. 24, 531–540. Gobbi, M., Comitini, F., Domizio, P., Romani, C., Lencioni, L., Mannazzu, I., et al., 2013. Lachancea thermotolerans and Saccharomyces cerevisiae in simultaneous and sequential co-fermentation: a strategy to enhance acidity and improve the overall quality of wine. Food Microbiol. 33, 271–281. Goddard, M.R., Anfang, N., Tang, R., Gardner, R.C., Jun, C., 2010. A distinct population of Saccharomyces cerevisiae in New Zealand: evidence for local dispersal by insects and human-aided global dispersal in oak barrels. Environ. Microbiol. 12, 63–73. Hammer, Ø., Harper, D.A.T., Ryan, P.D., 2001. PAST: paleontological statistics software package for education and data analysis. Palaeontol. Electronica 4 (1), 9. Hunter, P.R., Gaston, M.A., 1988. Numerical index of the discriminatory ability of typing systems: an application of Simpson’s index of diversity. J. Clin. Microbiol. 26, 2465–2466. Jolly, N.P., Augustyn, O.P.H., Pretorius, I.S., 2003. The occurrence of non-Saccharomyces yeast strains over three vintages in four vineyards and grape musts from four production regions of the Western Cape, South Africa. S. Afr. J. Enol. Vitic. 24, 35–42. Jolly, N.P., Varela, C., Pretorius, I.S., 2014. Not your ordinary yeast: non-Saccharomyces yeasts in wine production uncovered. FEMS Yeast Res. 14, 215–237.

10

G. Banilas et al. / Microbiological Research 193 (2016) 1–10

Kapsopoulou, K., Mourtzini, A., Anthoulas, M., Nerantzis, E., 2007. Biological acidification during grape must fermentation using mixed cultures of Kluyveromyces thermotolerans and Saccharomyces cerevisiae. World J. Microbiol. Biotechnol. 23, 735–739. Knight, S., Goddard, M.R., 2015. Quantifying separation and similarity in a Saccharomyces cerevisiae metapopulation. ISME J. 9, 361–370. Knight, S., Klaere, S., Fedrizzi, B., Goddard, M.R., 2015. Regional microbial signatures positively correlate with differential wine phenotypes: evidence for a microbial aspect to terroir. Sci. Rep. 5, 14233, http://dx.doi.org/10.1038/ srep14233. Legras, J.L., Merdinoglu, D., Cornuet, J.M., Karst, F., 2007. Bread, beer and wine: Saccharomyces cerevisiae diversity reflects human history. Mol. Ecol. 16, 2091–2102. Levdansky, E., Sharon, H., Osherov, N., 2008. Coding fungal tandem repeats as generators of fungal diversity. Fungal Biol. Rev. 22, 85–96. Li, Y.C., Korol, A.B., Fahima, T., Nevo, E., 2004. Microsatellites within genes: structure, function, and evolution. Mol. Biol. Evol. 2, 991–1007. Masneuf-Pomarede, I., Juquin, E., Miot-Sertier, C., Renault, P., Laizet, Y., Salin, F., et al., 2015. The yeast Starmerella bacillaris (synonym Candida zemplinina) shows high genetic diversity in winemaking environments. FEMS Yeast Res. 15, fov045, http://dx.doi.org/10.1093/femsyr/fov045. Mauriello, G., Capece, A., D’ Auria, M., Garde-Cerdan, T., Romano, P., 2009. SPME–GC method as a tool to differentiate VOC profiles in Saccharomyces cerevisiae wine yeasts. Food Microbiol. 26, 246–252. Mills, D.A., Johannsen, E.A., Cocolin, L., 2002. Yeast diversity and persistence in Botrytis-affected wine fermentations. Appl. Environ. Microbiol. 68, 4884–4893. Nikolaou, E., Soufleros, E.H., Oulompasi, E., Tzanetakis, N., 2006. Selection of indigenous Saccharomyces cerevisiae strains according to their oenological characteristics and vinification results. Food Microbiol. 23, 205–211. Nisiotou, A.A., Spiropoulos, A.E., Nychas, G.J.E., 2007. Yeast community structures and dynamics in healthy and Botrytis-affected grape must fermentations. Appl. Environ. Microbiol. 73, 6705–6713. Pando Bedrinana, R., Querol Simon, A., Suarez Valles, B., 2010. Genetic and phenotypic diversity of autochthonous cider yeasts in a cellar from Asturias. Food Microbiol. 27, 503–508.

Pfliegler, W.P., Horváth, E., Kállai, Z., Sipiczki, M., 2014. Diversity of Candida zemplinina isolates inferred from RAPD, micro/minisatellite and physiological analysis. Microbiol. Res. 169, 402–410, http://dx.doi.org/10.1016/j.micres. 2013.09.006. Richards, K., Goddard, M., Gardner, R., 2009. A database of microsatellite genotypes for Saccharomyces cerevisiae. Antonie Van Leeuwenhoek 96, 355–359. Ruderfer, D.M., Pratt, S.C., Seidel, H.S., Kruglyak, L., 2006. Population genomic analysis of outcrossing and recombination in yeast. Nat. Genet. 38, 1077–1081, http://dx.doi.org/10.1038/ng1859. Schuller, D., Casal, M., 2007. The genetic structure of fermentative vineyardassociated Saccharomyces cerevisiae populations revealed by microsatellite analysis. Antonie Van Leeuwenhoek 91, 137–150. Soden, A., Francis, I.L., Oakey, H., Henschke, P.A., 2000. Effects of co-fermentation with Candida stellata and Saccharomyces cerevisiae on the aroma and composition of Chardonnay wine. Aust. J. Grape Wine Res. 6, 21–30. ˜ Suárez Valles, B., Pando Bedrinana, R., Lastra Queipo, A., Mangas Alonso, J.J., 2008. Screening of cider yeasts for sparkling cider production (Champenoise method). Food Microbiol. 25, 690–697, http://dx.doi.org/10.1016/j.fm.2008.03. 004. Tóth, G., Gáspari, Z., Jurka, J., 2000. Microsatellites in different eukaryotic genomes: survey and analysis. Genome Res. 10, 967–981. Talla, E., Anthouard, V., Bouchier, C., Frangeul, L., Dujon, B., 2005. The complete mitochondrial genome of the yeast Kluyveromyces thermotolerans. FEBS Lett. 579, 30–40. Tofalo, R., Perpetuini, G., Schirone, M., Fasoli, G., Aguzzi, I., Corsetti, A., et al., 2013. Biogeographical characterization of Saccharomyces cerevisiae wine yeast by molecular methods. Front. Microbiol. 4, 166, http://dx.doi.org/10.3389/fmicb. 2013.00166.eCollection2013. Torija, M.J., Rozes, N., Poblet, M., Guillamon, J.M., Mas, A., 2001. Yeast population dynamics in spontaneous fermentations: comparison between different wine-producing areas over a period of three years. Antonie Van Leeuwenhoek 79, 345–352. Zott, K., Thibon, C., Bely, M., Lonvaud-Funel, A., Dubourdieu, D., Masneuf-Pomarede, I., 2011. The grape must non-Saccharomyces microbial community: impact on volatile thiol release. Int. J. Food Microbiol. 151, 210–215.