Identification and characterization of Saccharomyces cerevisiae and Saccharomyces paradoxus strains isolated from Croatian vineyards

Letters in Applied Microbiology 2002, 35, 305 310 Identification and characterization of Saccharomyces cerevisiae and Saccharomyces paradoxus strains isolated from Croatian vineyards S. Redžepović 1, S. Orlić 1, S. Sikora 1, A. Majdak 2 and I.S. Pretorius 3 1 Department of Microbiology, 2 Department of Viticulture & Enology, Faculty of Agriculture, University of Zagreb, Zagreb, Croatia and 3 Institute for Wine Biotechnology, Stellenbosch University, Stellenbosch, South Africa 350 01: received 30 November 2001, revised 3 April 2002 and accepted 17 June 2002 S. REDŽEPOVIĆ,S.ORLIĆ, S. S I K O R A, A. M A J D A K A N D I. S. P R E T O R I U S. 2002. Aims: The identification, differentiation and characterization of indigenous Saccharomyces sensu stricto strains isolated from Croatian vineyards and the evaluation of their oenological potential. Methods and Results: A total of 47 Saccharomyces sensu stricto strains were isolated from Chardonnay grapes and identified by physiological and molecular genetic methods. By using the standard physiological and biochemical tests, six isolates were identified as Saccharomyces cerevisiae and 41 as Saccharomyces paradoxus. However, PCR RFLP analyses of the internal transcribed spacer (ITS1) region of the 18S ribosomal DNA identified 12 of the isolates as S. cerevisiae and 35 as S. paradoxus. Fermentation trials in a grape juice medium showed that these isolates ferment vigorously at 18 C and display tolerance to high levels of ethanol. None of these isolates appeared to produce either hydrogen sulphide or killer toxins. Conclusion: Saccharomyces paradoxus, possessing potentially important oenological characteristics, occurs in much higher numbers than S. cerevisiae in the indigenous population of Saccharomyces sensu stricto strains in Croatian vineyards. Significance and Impact of the Study: This study forms an essential step towards the preservation and exploitation of the hidden oenological potential of the untapped wealth of yeast biodiversity in the Croatian grape-growing regions. The results obtained demonstrate the value of using molecular genetic methods, such as PCR RFLP analyses, in conjunction with the traditional taxonomic methods based on phenotypic characteristics in such ecotaxonomic surveys. The results also shed some light on the ecology and oenological potential of S. paradoxus, which is considered to be the natural parent species of the domesticated species of the Saccharomyces sensu stricto group. INTRODUCTION Various yeast species found on the grape and on winery surfaces participate in spontaneous fermentations. Yeasts of the genera Kloeckera, Hanseniaspora and Candida predominate in the early stages of the fermentation, followed by several species of Metschnikowia and Pichia in the middle stages, when the ethanol rises to 3 4% (Pretorius et al. 1999). The latter stages of spontaneous wine fermentations Correspondence to: S. Orlić, Department of Microbiology, Faculty of Agriculture, University of Zagreb, Svetosimunska 25, 10000 Zagreb, Croatia (e-mail: sandi.orlic@agr.hr). invariably are dominated by the alcohol-tolerant strains of the Saccharomyces sensu stricto group of yeasts. This taxon consists of four yeast species, namely Saccharomyces bayanus, Saccharomyces cerevisiae, Saccharomyces paradoxus and Saccharomyces pastorianus. Since S. cerevisiae is found almost exclusively in man-made fermentation environments (Vaughan-Martini and Martini 1987) and is universally preferred for initiating wine fermentations, it became known as the wine yeast (Pretorius 2000). However, some wine yeast starter culture strains also belong to the species of S. bayanus. In contrast to the strong link of the latter two members of the Saccharomyces sensu stricto group with winemaking, strains of S. pastorianus are usually associated with the ª 2002 The Society for Applied Microbiology

306 S. REDŽEPOVIĆ ET AL. production of lager beer (Vaughan-Martini and Martini 1987; Turakainen et al. 1993; Rodrigues de Sousa et al. 1995), whereas strains of S. paradoxus generally are not perceived as being associated with the production of either wine or beer. However, this general perception of S. paradoxus strains should not prevent ecological and taxonomic surveys of S. paradoxus populations associated with vineyards and an evaluation of the oenological potential of such strains. Previous ecological, taxonomic and genetic studies have led to the hypothesis that S. paradoxus is the natural parent species of the domesticated species of the Saccharomyces sensu stricto group, namely S. cerevisiae (Vaughan-Martini 1989; Vaughan-Martini and Martini 1998). This hypothesis is supported by the results of DNA reassociation studies, which showed that approximately 50% of the nucleotide sequences of the S. paradoxus and S. cerevisiae genomes are homologous, while the relatedness to S. bayanus and S. pastorianus is less than 20% (Vaughan-Martini 1989). The aim of this study was to isolate and characterize the sibling species of S. paradoxus and S. cerevisiae from the grape-growing region in the north-western part of Croatia. By investigating the ecology and oenological potential of indigenous strains of these Saccharomyces sensu stricto yeasts, this study opens up the possibility that, in the future, some winemakers might prefer to use individual, or a mixture of, indigenous yeasts as starter culture strains to reflect the yeast biodiversity of the Croatian wine-producing region. Such an approach could promote wine s most enthralling and fascinating characteristic, namely its diversity of style. MATERIALS AND METHODS Yeast isolation Vineyards were sampled near the Jazbina research centre, which is located in the north-western region of Croatia. The isolation of Saccharomyces yeasts from Chardonnay grapes was done essentially as described by Martini et al. (1980) and Van der Westhuizen et al. (2000). Yeast identification Four type strains of the Saccharomyces sensu stricto group, namely S. bayanus DBVPG 6171 (CBS 380), S. cerevisiae DBVPG 6173 (CBS 1171), S. paradoxus DBVPG 6411 (CBS 432) and S. pastorianus DBVPG 6047 (CBS 1538) were used as reference strains during the identification of the isolates. The identification of the isolates was carried out according to conventional yeast identification methods based on the morphology, sporulation and fermentation characteristics, as well as the assimilation of a range of carbon sources (Kregervan Rij 1984). For the identification of the various species within the Saccharomyces sensu stricto group, the isolates were tested for their ability to utilize mannitol and maltose as carbon sources, and to grow on a vitamin-free medium and on YEPG agar plates (1% yeast extract, 2% glucose, 1% peptone, 1Æ5% agar) at 37 C (Vaughan-Martini and Martini 1998). Isolates that grew at 37 C were considered as S. bayanus. Saccharomyces paradoxus was identified on the basis of growth on mannitol and maltose, while the isolates that did not grow on these sugars were considered to be S. cerevisiae. The identification of the isolates was then verified by molecular genetic methods. These entailed the amplification of the internal transcribed spacer (ITS1) region of the 18S ribosomal DNA (rdna) (White et al. 1990) of the isolates by the polymerase chain reaction (PCR) technique, and the analysis of the restriction fragment length polymorphism (RFLP) of the digested PCR-generated products. DNA amplification by the PCR technique was carried out according to Smole-Možina et al. (1997). The NS1 ITS2 primer pair with the following sequences was used to amplify the ITS1 region of the 18S rdna: NS1 5 -GTAGTCA- TATGCTTGTCTC-3 and ITS2 5 -GCTGCGTTCTT- CATCGATGC-3 (White et al. 1990). The PCR products were digested with two restriction endonucleases, namely HaeIII and MspI, and separated by electrophoresis. Yeast characterization Strains 1080 (nonkiller strain) and 1079 (a K 1 killer strain) were used as negative and positive indicator strains to determine the killer activity of the isolated strains. YEPG agar plates, buffered at ph 4Æ5, were used to detect zones of growth inhibition caused by the K 1 toxin secreted by the killer yeast. The nonkiller strain 1080 was used as a sensitive lawn on the indicator agar plates, while the killer strain 1079 was used as a positive control. The plates were incubated for 48 h at 25 C and then examined to note killer activity. The production of H 2 S was determined by the yeast colour formation on BIGGY agar (Difco, Detroit, MI, USA) plates. The degree of browning associated with yeast growth on the BIGGY agar plates was scored using the following scale: 1 ¼ white, 2 ¼ cream, 3 ¼ light brown, 4 ¼ brown, 5 ¼ dark-brown, 6 ¼ black. Fermentation trials were conducted in 300-ml Erlenmeyer flasks each containing 100 ml of Chardonnay grape juice (with a sugar content of 93 Oe or 22Æ21 Brix, a total acidity of 7Æ9 gl )1 and ph 3Æ2). Pasteurized must (95 ml) was inoculated with 5 ml of actively fermenting Saccharomyces sensu stricto strains, covered with sterile Vaseline and incubated at 18 C. Completion of fermentation was evaluated by weighing the flasks in relation to CO 2 loss. Fermentation was considered to be completed when the

SACCHAROMYCES STRAINS ISOLATED FROM VINEYARDS 307 weight remained constant. Standard methods for wine analysis were used to determine the ph, ethanol content and levels of volatile acidity. A commercial wine yeast starter culture strain (CS2) from Lallemand was used as a standard fermenting strain during the microvinification trials. RESULTS A total of 47 Saccharomyces sensu stricto strains were isolated from Chardonnay grapes and identified by the conventional taxonomic methods, as well as by molecular genetic analyses. By using the standard physiological and biochemical tests, six isolates were identified as S. cerevisiae and 41 isolates were identified as S. paradoxus. However, PCR RFLP analyses of the ITS1 region of the 18S rdna identified 12 of these isolates as S. cerevisiae and 35 as S. paradoxus. During the process of identification of the yeast isolates, we have come to realize that the standard WL nutrient medium is unsuitable for the differentiation of the four species belonging to the Saccharomyces sensu stricto group. Moreover, our results also confirmed the observation of Rodrigues de Sousa et al. (1995) that the fermentation of maltose, assimilation of D-mannitol and growth at 37 C are variable reactions and that these tests therefore are also inappropriate for the delineation of Saccharomyces species. In an attempt to overcome the ambiguity of the results obtained by the conventional taxonomic methods, the smallsubunit rrna-coding DNA with sequences in the neighbouring ITS region was amplified in all indigenous isolates and type strains. The size of the amplified fragment was about 2 kbp (Smole-Možina et al. 1997). These 2 kbp PCR-generated rdna fragments were digested with HaeIII or MspI, and separated by gel electrophoresis. Figure 1 depicts the PCR RFLP analysis of the18s rdna of the Saccharomyces sensu stricto type strains, whereas Fig. 2 shows examples of the digested PCR products derived from some of the indigenous S. cerevisiae and S. paradoxus strains isolated from the Chardonnay grapes. It was evident that, while the S. paradoxus strains exhibited similar sugar fermentation patterns to those of S. cerevisiae, they displayed (a) 1 2 3 4 5 (b) 5 4 3 2 1 1600 bp 1320 bp 700 bp 500 bp 360 bp 300 bp 200 bp Fig. 1 PCR RFLP analyses of the ITS1 region of the 18S ribosomal DNA of the Saccharomyces sensu stricto type strains. The PCR products were digested with HaeIII (a) and MspI (b), and loaded onto each of the recast 6% poly (NAT) gels as follows. Lane 1: molecular size marker 1 kbp DNA ladder (Life Technologies, Basel, Switzerland); lane 2: S. cerevisiae type strain; lane 3: S. bayanus type strain; lane 4: S. paradoxus type strain; lane 5: S. pastorianus type strain

308 S. REDŽEPOVIĆ ET AL. (a) 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 (b) Fig. 2 PCR RFLP analyses of the ITS1 region of the 18S ribosomal DNA of the indigenous Saccharomyces sensu stricto strains isolated from Chardonnay grapes gathered in vineyards located in the north-western region of Croatia. The PCR products were digested with HaeIII (a) and MspI (b), and loaded onto each of the recast 6% poly (NAT) gels as follows. Lane 1: RO 54; lane 2: RO 66; lane 3: RO 83; lane 4: RO 102; lane 5: RO 134; lane 6: RO 82; lane 7: RO 3; lane 8: RO 108; lane 9: RO 78; lane 10: RO 9; lane 11: RO 45; lane 12: RO 106; lane 13: RO 17; lane 14: RO 6; lane 15: RO 46; lane 16: RO 67; lane 17: RO 86. The RO numbers represent the strain designations of the isolater distinctive RFLP patterns. The discrepancy between the results from the physiological tests and the RFLPs confirmed that the combined use of conventional taxonomic methods and molecular genetic techniques is essential for the accurate identification of yeasts (Pretorius et al. 1999). Of particular importance in this study is the fact that S. paradoxus, which is normally associated with oak species (Quercus robur or Quercus mongolica) in Europe, Far East Asia and North America (Naumov 1985, 1996; Naumov et al. 1992, 1998), appeared to occur in much higher numbers than S. cerevisiae in the indigenous population of Saccharomyces sensu stricto strains in Croatian vineyards. Since the most distinctive feature of S. paradoxus is that it can sporulate readily on rich media, such as malt agar (Phaff et al. 1956), it was surprising to find that about 75% of the yeast isolates from the Croatian Chardonnay grapes sporulated without difficulty on rich YPD medium. While Török et al. (1996) and Mortimer (2000) reported similar results, they did not positively identify their sporulating isolates as S. paradoxus; instead, they only presumed that their isolates belonged to the species of S. paradoxus. In contrast to the latter studies, we were able to positively identify the S. paradoxus isolates from the Jazbina vineyards by using the conventional identification methods in conjunction with the PCR RFLP analyses of the ITS1 region of the 18S rdna. Our observation that S. paradoxus occurs in higher numbers in vineyards than S. cerevisiae provides indirect support for the premise that S. paradoxus is the natural parent species of the domesticated species of the Saccharomyces sensu stricto group (Vaughan-Martini 1989; Vaughan- Martini and Martini 1998), and that S. cerevisiae, which is widely considered as the wine yeast, is found almost exclusively in man-made habitats, such as wineries (Vaughan- Martini and Martini 1987). It is known that some strains that belong to S. cerevisiae and S. paradoxus (Vaughan-Martini and Martini 1998) produce a homogeneous fermentation profile (Rainieri et al. 1999). We have also found that the S. cerevisiae and S. paradoxus isolates generally fermented vigorously, that they were resistant to relatively high levels of ethanol, and that they gave higher ethanol yields (ranging from 10Æ2 to 11Æ95% v v) from the grape sugars (mainly glucose and fructose) than S. pastorianus and S. bayanus control strains. Surprisingly, the least vigorous and least ethanol-resistant strain was the type strain of S. cerevisiae, which is in agreement with the findings of Rainieri et al. (1999). In addition to the primary role played by any wine yeast starter culture strain to catalyse the efficient and complete conversion of grape sugars into alcohol, it is also important that they do not produce off-flavours. Owing to their high volatility, reactivity and potency at very low threshold levels, sulphur-containing compounds, such as H 2 S, have a profound effect on the flavour of wine (Pretorius 2000). From a winemaking point of view, it was therefore encouraging to

SACCHAROMYCES STRAINS ISOLATED FROM VINEYARDS 309 find that all of the isolates produced low concentrations of H 2 S. Furthermore, none of the isolates produced more than 0Æ8 gl )1 volatile acidity, thereby indicating that they would not negatively affect the sensorial quality of wine. With respect to the possible production of killer toxins that may interfere with the growth of Saccharomyces yeasts during the process of winemaking, we have found that none of the isolates produced any zymocidal peptide that could inhibit the growth of the indicator strain, 1080. However, it must be pointed out that, in most strains of S. cerevisiae, the killer toxin is encoded by linear double-stranded RNA genomes encapsulated in viral-like particles (Bevan and Makover 1963), while the killer characteristic of S. paradoxus is presumably encoded by chromosomal genes (Naumov 1985). DISCUSSION The taxon Saccharomyces sensu stricto appears to consist of a group of species or a species complex of closely related yeasts, rather than four distinct species separated by welldefined boundaries. From a practical point of view, it is therefore even more problematic to differentiate wild Saccharomyces sensu stricto strains than is the case with the respective type strains of the four species belonging to this complex. This study has demonstrated the value of the combined use of conventional taxonomic methods (morphological observations, sugar fermentation and assimilation patterns) and molecular genetic techniques (RFLP analyses of the 18S rdna) to identify wild strains of the Saccharomyces sensu stricto complex. We have shown that when it is cumbersome to differentiate wild strains of S. cerevisiae and S. paradoxus on the basis of conventional phenotypic characteristics, such as their ability to grow on D-mannitol as sole carbon source and to ferment maltose, PCR RFLP analyses of the ITS1 region of their 18S rdna help to sharpen the boundary between these two sibling species. Although similar results were reported previously with type and laboratory strains of S. cerevisiae and S. paradoxus, this is the first practical assessment of the value of PCR RFLP analyses of the ITS1 region of the 18S rdna of a large collection of wild strains of these two closely related species. This study is in line with the viewpoint that taxonomic surveys should not result in meaningless catalogues or inventories of yeast names, devoid of any real biological relevance (Pretorius et al. 1999). Instead, our results have shed more light on the ecology and enological potential of the Saccharomyces sensu stricto complex of species, thereby contributing to the fundamental understanding of yeast communities in the niche habitats of Croatian vineyards. Thanks to the important observation concerning the apparent prevalence of S. paradoxus in vineyards, this study has now been expanded into a larger taxonomic survey within the ecological framework of the grape-growing region of Croatia. Since there is a growing need amongst leading winemakers for yeast starter culture strains that are better adapted to the different wine-producing regions of the world with their respective grape varietals, viticultural practices, winemaking techniques and styles of wine, this study provides an important step towards the preservation and exploitation of the hidden oenological potential of the untapped wealth of yeast biodiversity in the Croatian grapegrowing regions. ACKNOWLEDGEMENTS The authors thank A. Vaughan-Martini (Universitá Degli Studi di Perugia, Perugia, Italy) for the type strains of Saccharomyces sensu stricto and G.A. Farris (Universitá Degli Studi di Sassari, Sassari, Sardinia, Italy) for the killer indicator strains. REFERENCES Bevan, E.A. and Makover, M. (1963) The physiological basis of killer character in yeast. In: Genetics Today Xth International Congress for Genetics ed. Geets, S.J. pp. 53 58. Oxford: Pergamon Press. Kreger-van Rij, N.J.W. (1984). The Yeast A Taxonomic Study, 3rd edn. Amsterdam: Elsevier Science. Martini, A., Federici, F. and Rosini, G. (1980) A new approach to the study of yeast ecology of natural substrates. Canadian Journal of Microbiology 26, 856 859. Mortimer, R. (2000) Evolution and variation of the yeast (Saccharomyces) genome. Genome Research 10, 403 409. Naumov, G.I. (1985) Comparative genetics of yeast. XXIII: Unusual inhertiance of toxin production in Saccharomyces paradoxus Batschinskaya. Genetika 21, 1794 1978. Naumov, G.I. (1996) Genetic identification of biological species in the Saccharomyces sensu stricto complex. Journal of Industrial Microbiology 17, 295 302. Naumov, G.I., Naumova, E.S. and Korhola, M. (1992) Genetic identification of natural Saccharomyces sensu stricto yeast from Finland, Holland and Slovakia. Antonie van Leewenhoek 61, 237 243. Naumov, G.I., Naumova, E.S. and Sniegowski, P.D. (1998) Saccharomyces paradoxus and Saccharomyces cerevisiae are associated with exudates of North American oaks. Canadian Journal of Microbiology 44, 1045 1050. Phaff, H.J., Miller, M.W. and Shifirne, M. (1956) The taxonomy of yeast isolated from Drosophila in the Yosemite region of California. Antonie van Leeuwenhoek 30, 132 140. Pretorius, I.S. (2000) Tailoring wine yeast for the new millennium: novel approaches to the ancient art of winemaking. Yeast 16, 675 729. Pretorius, I.S., Van der Westhuizen, T.J. and Augustyn, O.P.H. (1999) Yeast biodiversity in vineyards and wineries and its importance to the South African industry a review. South African Journal of Enology and Viticulture 20 (2), 61 74.

310 S. REDŽEPOVIĆ ET AL. Rainieri, S., Zambonelli, C., Hallsworth, J.E., Pulvirenti, A. and Giudici, P. (1999) Saccharomyces uvarum, a distinct group within Saccharomyces sensu stricto. FEMS Microbiology Letters 177 (1), 177 185. Rodrigues de Sousa, H., Madeira-Lopes, A. and Spencer-Martins, I. (1995) The significance of active fructose transport and maximum temperature for growth in the taxonomy of Saccharomyces sensu stricto. Systematic and Applied Microbiology 18, 44 51. Smole-Možina, S., Dlauchy, D., Deak, T. and Raspor, P. (1997) Identification of Saccharomyces sensu stricto and Torulaspora yeast by PCR ribotyping. Letters in Applied Microbiology 24, 311 315. Török, T., Mortimer, R.K., Romano, P., Suzzi, G. and Polsinelli, M. (1996) Quest for wine yeast an old story revisited. Journal of Industrial Microbiology 17, 303 313. Turakainen, H., Aho, S. and Korhola, M. (1993) MEL gene polymorphism in the genus Saccharomyces. Applied and Environmental Microbiology 59, 2622 2630. Van der Westhuizen, T.J., Augustyn, O.P.H. and Pretorius, I.S. (2000) Geographical distribution of indigenous Saccharomyces cerevisiae strains isolated from vineyards in the coastal regions of the Western Cape in South Africa. South African Journal of Enology and Viticulture 1, 21 24. Vaughan-Martini, A. (1989) Saccharomyces paradoxus, a newly separated species of the Saccharomyces sensu stricto complex based upon ndna ndna homologies. Systematic and Applied Microbiology 12, 119 122. Vaughan-Martini, A. and Martini, A. (1987) Three newly delimited species of Saccharomyces sensu stricto. Antonie van Leeuwenhoeck 53, 77 84. Vaughan-Martini, A. and Martini, A. (1998) Saccharomyces Meyen ex Reess. In The Yeast, A Taxonomic Study, 4th edn. ed. Kurtzman, C.P. and Fell, J.W. pp. 358 372. Amsterdam: Elsevier Science. White, T.J., Bruns, T., Lee, S. and Taylor, J. (1990) Amplification and direct sequencing of fungal ribosomal RNA genes for phylogenetics. In PCR Protocols: A Guide to Methods and Applications ed. Innis, M.A., Gelfand, D.H., Sninsky, J.J. and White, T.J. pp. 315 322. London: Academic Press.