H. Yamagishi, Brewing Research & Development Laboratory, Asahi Breweries, Ltd, 1–1-21, Midori, Moriya-machi, Kitasoma-gun, Ibaraki, 302–0106, Japan (e-mail: email@example.com).
In order to differentiate brewing from non-brewing yeasts, a specific polymerase chain reaction (PCR) which targeted the open reading frame of FLO1 was employed. Non-brewing yeasts include ‘non-brewing Saccharomyces yeasts’ and ‘non-Saccharomyces yeasts’. The molecular sizes of the PCR products differed between brewing and non-brewing Saccharomyces yeasts. No FLO1 PCR products were obtained from non-Saccharomyces yeasts. Specific PCR, using oligonucleotide primers that targeted the region between the 5S and 26S rRNA genes, could be used to differentiate brewing yeasts from some non-brewing yeasts. These PCR products were digested with restriction enzymes, Scr FI and Msp I. Different restriction profiles were obtained from brewing and non-brewing yeasts which could not be differentiated using specific PCR of rDNA. These results suggest that it is possible to identify brewing from non-brewing yeasts using specific PCR of FLO1 and rDNA, and detection of restriction fragment polymorphism of rDNA.
If you can't find a tool you're looking for, please click the link at the top of the page to "Go to old article view". Alternatively, view our Knowledge Base articles for additional help. Your feedback is important to us, so please let us know if you have comments or ideas for improvement.
In the brewing industry, microbiological purification of brewing yeast strains from non-brewing ‘wild’ yeast strains is necessary to maintain consistency in fermentation performance and in overall product quality. It was reported that contamination with non-brewing yeasts such as Hansenula, Pichia, Candida, Brettanomyces, and other species of the genus Saccharomyces could cause beer spoilage and result in haze and off-flavours. Traditional methods for detection, identification and characterization of yeast based on biochemical, morphological and physiological criteria tend to be time-consuming (several days to a week) and can produce inconclusive or incomplete results. In particular, it is difficult to distinguish brewing yeasts phenotypically from yeasts of the genus Saccharomyces, such as S. cerevisiae, S. bayanus and S. diastaticus, because brewing yeasts themselves belong to this genus. Therefore, contamination with non-brewing yeasts of the genus Saccharomyces is a serious concern. For this reason, there is great interest in developing a method for differentiating brewing from non-brewing yeasts of the genus Saccharomyces.
The polymerase chain reaction (PCR) provides a means for rapidly identifying yeast species; several studies have used restriction fragment length polymorphism (RFLP) in rRNA genes (rDNA) ( Vilgay & Hester 1990). Molina et al. (1993) reported RFLP in the rDNA of closely related Saccharomyces species. Random amplified polymorphic DNA-polymerase chain reaction (RAPD-PCR) and specific PCR using δ sequence primers have been used to characterize lager yeasts, ale yeasts and non-brewing yeasts ( Coakley et al. 1996 ; Laidlaw et al. 1996 ; Tompkins et al. 1996 ). From these reports, it is clear that brewing and non-brewing yeasts cannot be distinguished absolutely using RAPD-PCR and specific PCR using δ sequence primers. This study was carried out to determine whether rDNA PCR products exhibited amplification length polymorphisms, and whether variations in fingerprints of digested rDNA PCR products could be used to distinguish brewing yeasts from non-brewing ‘wild’ yeasts. It was reported that the FLO1 gene contains four families of repeated sequences ( Teunissen et al. 1993 ; Watari et al. 1994 ). Furthermore, there have been reports that Saccharomyces yeasts have genes homologous to FLO1 ( Bidard et al. 1994 ; Bussey et al. 1995 ). Based on these reports, it is speculated that there are polymorphisms of the FLO1 gene among yeast strains. The study also examined whether polymorphisms of the FLO1 gene and its homologues among yeasts species and strains could be used to distinguish brewing yeasts from non-brewing yeasts.
Materials and methods
Yeast species, strains and medium
Yeast species and strains used in this study are described in Table 1. For yeast cell growth, YPD medium (1% yeast extract, 2% peptone, 2% glucose) was used.
Table 1. Yeast species and strains used in this study
Yeast species and strains
Genotype and source
Saccharomyces cerevisiae ABXL-1D
MAT a FLO1 from Yeast Genetic Stock Center
S. cerevisiae ABXR-11B
MAT a FLO5 from Yeast Genetic Stock Center
S. cerevisiae YIY261
MAT a FLO8 from Dr I. Yamashita
S. cerevisiae X-2180-1A
MAT a from Yeast Genetic Stock Center
S. cerevisiae KA31
Dipolid, laboratory yeast from Dr A. Toe
Brewing yeast TF1
Top-fermenting yeast from our own stock
Brewing yeast BF1
Bottom-fermenting yeast from our own stock
Brewing yeast BF2
Bottom-fermenting yeast from our own stock
Carlsberg bottom-fermenting yeast No. 1, S. carlsbergensis type strain
S. pastorianus type strain
S. bayanus type strain
S. paradoxus type strain
S. logos type strain
S. willianus type strain
1103 from our stock strain
B. anomalus type strain
C. utilis IAM0626
DNA isolation, Polymerase Chain Reaction, Restriction Digestion and RFLP analysis, Southern analysis
Genomic DNA of yeasts was prepared according to the method of Hoffman & Winston (1987). A final concentration of 1–50 ng μl−1 target DNA was used in each reaction mixture. All PCR reactions were performed under the manufacturers specifications (Perkin-Elmer Cetus) using a Gene Amp PCR System 2400 (Perkin-Elmer Cetus, USA). For the amplification of the FLO1 gene, primers FL1 (5′ CCAAAAATGACAATGCCTCATCGCTAT 3′) and FLR2 (5′ CCATTGCTAGGATAGAATGGGGTAATAATTGGACG 3′) were used. The region amplified by these primers spanned base position −6 to 3344 on the FLO1 gene. For the amplification of rDNA, primers 5S2 (5′ CACCGTTTCCCGTCCGATC 3′) and EST2 (5′ CTGAACGCCTCTAAGTC 3′) were used. These primers, designed for amplification of rDNA, were previously described by Molina et al. (1993) . The region amplified by these PCR primers spanned base position 3110 on the 26S rRNA to base position 27 on the 5S rRNA. PCR products were digested with restriction enzymes Scr FI and Msp I. PCR products were electrophoresed on 1% agarose gel in 1 × TAE buffer. Restriction fragments of PCR products were subjected to electrophoresis on 3% NuSieve 3 : 1 agarose gel (FMC Bioproducts, USA) in 0·5 × TAE buffer. Gels were stained in ethidium bromide and photographed. Southern hybridization was done with the ECL direct labelling system according to the supplier’s instruction (Amersham Life Science Ltd, UK).
The flocculation assay used was a modification of Helm’s method ( Helm et al. 1953 ).
The results of PCR which targeted the FLO1 gene varied among Saccharomyces yeasts.
Studies were carried out to determine whether the polymorphisms of the FLO1 gene could be detected using PCR ( Fig. 1). Specific PCR was performed with primers which target the open reading frame (ORF) of FLO1. From S. cerevisiae ABXL-1D (MAT a FLO1), a 3·2 kbp band was amplified ( Fig. 1, lane 6). This result agreed with the sequence reports of FLO1 ( Teunissen et al. 1993 ; Watari et al. 1994 ). From S. cerevisiae YIY261 (MAT a FLO8 STA1), a 3·2 kbp band was also amplified ( Fig. 1, lane 8), but from the two bottom-fermenting yeasts ( Fig. 1, lanes 2 and 3), one top-fermenting yeast ( Fig. 1, lane 4), laboratory yeast KA31 (MAT α/MAT a his3/his3 leu2/leu2 trp1/trp1 ura3/ura3) ( Fig. 1, lane 5) and ABXR-11B (MAT a FLO5) ( Fig. 1, lane 7), amplified bands showed varied patterns and the molecular sizes were lower than that of S. cerevisiae ABXL-1D (FLO1). These observations suggest polymorphisms of the FLO1 gene among Saccharomyces yeasts. Flocculation grades of these strains are described in Table 2. The molecular sizes of FLO1 PCR products did not correlate with flocculation grades. Further, specific PCR was performed in the Saccharomyces yeasts ( Fig. 2). FLO1 PCR products were amplified from S. cerevisiae X-2180–1 A, S. carlsbergensis type strain, S. pastorianus IFO 2003 and S. pastorianus type strain ( Fig. 2, lanes 2–5), but FLO1 PCR products were not amplified from S. bayanus type strain and S. paradoxus type strain ( Fig. 2, lanes 6 and 7). The results from Southern hybridization showed that all these species and strains have the FLO1 gene or its homologues (data not shown). Furthermore, FLO1 PCR was performed on brewing and non-brewing (Saccharomyces, Hansenula, Brettanomyces and Candida) yeasts to ascertain whether brewing yeasts could be distinguished from non-brewing yeasts using polymorphisms of FLO1 and its homologue genes. In this study, non-brewing yeast species described by Back (1994) were chosen. In ‘The Yeasts, a Taxonomic Study, 4th edn’, many Saccharomyces yeast species were re-classified to S. cerevisiae ( Vaughan-Martini & Martini 1998), but in this study, the names of the yeast species are those used in ‘Farbatlas und Getrankebiologie’ which was described by Back (1994). From each of the two bottom-fermenting yeasts, a 1·8 kbp band was amplified ( Fig. 3, lanes 2 and 3). From the top-fermenting yeast, the molecular size of the amplified band was 1·2 kbp ( Fig. 3, lane 4). Among three strains of S. bayanus, the molecular sizes of the amplified bands varied ( Fig. 3, lanes 6–8); two bands were amplified from one strain ( Fig. 3, lane 6), one band was amplified from another strain ( Fig. 3, lane 7), and no bands were amplified from S. bayanus type strain ( Fig. 3, lane 8). More work is needed to define further the variations of FLO1 PCR products among S. bayanus strains. Among three strains of S. logos and three strains of S. willianus, amplified bands showed varied patterns ( Fig. 3, lanes 9–14) but were obtained from all tested strains. From non-Saccharomyces yeasts such as Hansenula, Brettanomyces and Candida, no bands were amplified (data not shown). These results demonstrate that it is possible to distinguish Saccharomyces yeasts which are classified to S. cerevisiae and its sibling species from non-Saccharomyces yeasts, such as Hansenula, using PCR which targets FLO1. In this study, several bands were amplified from brewing and non-brewing yeasts of the genus Saccharomyces. Previously, Watari et al. (1994) reported that only one band was hybridized from laboratory yeasts using a FLO1-specific probe. However, Bidard et al. (1994) reported that the FLO5 gene was highly homologous to the FLO1 gene. In this study, it was observed that bands with different molecular sizes were amplified among certain brewing and non-brewing yeasts of the genus Saccharomyces. This variation may be due to the polymorphisms of the FLO1 gene and its homologue. However, same-sized bands were obtained from different strains of the genus Saccharomyces. For example, the molecular sizes of bands amplified from two bottom-fermenting yeasts, S. pastorianus, one strain of S. bayanus and one strain of S. logos, were the same ( Fig. 3, lanes 2, 3, 5, 7 and 9). Therefore, it is difficult to differentiate brewing yeasts from non-brewing yeasts of the genus Saccharomyces definitively. In this study, there were polymorphisms of FLO1 and its homologue genes in the genus Saccharomyces. Using these polymorphisms, it is impossible to delineate yeast species exactly, but the technique is useful for strain-typing among S. cerevisiae and its sibling species.
Table 2. Flocculation grade of laboratory and brewing yeasts
Saccharomyces cerevisiae KA31
S. cerevisiae ABXL-1D (FLO1)
S. cerevisiae ABXR-11B (FLO5)
S. cerevisiae YTY261 (FLO8)
Bottom-fermenting yeast BF1
Bottom-fermenting yeast BF2
Top-fermenting yeast TF1
Brewing yeasts could be differentiated from non-brewing yeasts using restriction fragment length polymorphisms of the rRNA gene.
PCR amplification of the 3′ external transcribed and intergenic spacers of the ribosomal DNA repeat unit was performed for various yeast species including Saccharomyces, Hansenula, Brettanomyces and Candida. The results are shown in Figs 4 and 5. The molecular sizes of the PCR products were the same among the Saccharomyces yeasts except for TF1 ( Fig. 4, lanes 2–14). The two strains of H. anomala ( Fig. 5, lanes 3 and 5) exhibited slightly larger bands than brewing yeast BF1 ( Fig. 5, lane 2), but the other two strains of H. anomala ( Fig. 5, lanes 4 and 6) had a band equal to BF 1 ( Fig. 5, lane 2). All four strains of Brettanomyces ( Fig. 5, lanes 7–10) had a larger band than brewing yeast BF1 ( Fig. 5, lane 2). All four strains of Candida utilis ( Fig. 5, lanes 11–14) had bands almost equal to the brewing BF1 strain ( Fig. 5, lane 2). From the other three strains of Candida yeasts ( Fig. 5, lanes 15–17), no band, or various sized bands were observed. Restriction profiles resulting from digestion of the PCR products with the enzymes Scr FI and Msp I are shown in Figs 6, 7 and 8. The two bottom-fermenting yeasts exhibited equal restriction profiles of Scr FI and Msp I ( Fig. 6, lanes 2 and 3, and Fig. 7, lanes 2 and 3). The top-fermenting yeast had a slightly different restriction profile from the two bottom-fermenting yeasts ( Fig. 6, lane 4, and Fig. 7, lane 4). From the three strains of S. bayanus various restriction profiles were obtained on digestion with Scr FI and Msp I ( Fig. 6, lanes 6–8, Fig. 7, lanes 6–8). All three strains of S. logos had similar restriction profiles ( Fig. 6, lanes 9–11, Fig. 7, lanes 9–11) but differed from the brewing yeasts ( Fig. 6, lanes 2–4, Fig. 7, lanes 2–4). All three strains of S. willianus had the same restriction profiles ( Fig. 6, lanes 12–14, Fig. 7, lanes 12–14) but differed from the brewing yeasts ( Fig. 6, lanes 2–4, Fig. 7, lanes 2–4). The four strains of H. anomala exhibited similar restriction profiles ( Fig. 8, lanes 5–8) but differed from brewing yeasts ( Fig. 8, lanes 2–4). From the four strains of Candida utilis various restriction profiles were obtained ( Fig. 8, lanes 9–12), and these strains also differed from the brewing yeasts ( Fig. 8, lanes 2–4). RFLP of rDNA were also examined in four strains of S. ellipsoideus and five strains of S. diastaticus (data not shown). All restriction profiles of these strains were different from those of brewing yeasts. Thus, it was possible to identify non-brewing from brewing yeasts using a combination of specific PCR of FLO1 and rDNA, and detection of restriction fragment polymorphisms of rDNA.
Several genes are reportedly involved in yeast flocculation ( Teunissen et al. 1993 ; Bidard et al. 1994 ; Watari et al. 1994 ; Bussey et al. 1995 ). The FLO1 gene was the first such gene cloned ( Teunissen et al. 1993 ; Watari et al. 1994 ), and its deduced amino acid sequence was found to encode a lectin-like protein. The FLO5 gene was cloned next and was reported to be highly homologous to FLO1 ( Bidard et al. 1994 ). Furthermore, nucleotide sequencing of S. cerevisiae chromosome I ( Bussey et al. 1995 ) revealed several FLO1 homologous genes. On the other hand, brewing yeasts are reported to have ‘mixed’ genomes; they are not only polyploid but also contain different genomic sets ( Vaughan-Martini & Martini 1987). For these reasons, it is speculated that brewing yeasts have polymorphisms of the FLO1 gene different from those of laboratory yeast strains. In fact, Kobayashi et al. (1995) reported that bottom-fermenting yeasts yielded four hybridized bands for the FLO1 gene by Hind III digestion. A remarkable feature of the FLO1 gene is that it contains four families of repeated sequences composed of 18, two, three and three repeats. It may be that these repeated sequences cause the polymorphisms of the gene. From sequence reports of the FLO1 gene ( Teunissen et al. 1993 ; Watari et al. 1994 ), it was speculated that only the 3·2 kbp band would be amplified. However, the 1·8 kbp band was amplified from two bottom-fermenting yeasts ( Fig. 3, lanes 2 and 3). It is not clear whether FLO1 polymorphisms affect yeast flocculation, an important characteristic of brewing yeasts. Kobayashi et al. (1995) cloned a new flocculation gene, Lg-FLO1, from a bottom-fermenting yeast, and this gene caused brewing strain-type flocculation which is inhibited not only by mannose but also by maltose and glucose. In this study, we examined whether brewing yeasts could be differentiated from non-brewing yeasts using polymorphisms of the FLO1 gene. Among Saccharomyces yeasts, bands of different molecular sizes were obtained from same species, and bands with the same molecular size were obtained from different species. For example, the amplified FLO1 PCR products from five yeast strains, BF1 ( Fig. 3, lane 2), BF2 ( Fig. 3, lane 3), S. pastorianus CBS 1538 ( Fig. 3, lane 5), S. bayanus IFO 0676 ( Fig. 3, lane 7) and S. logos IFO 1225 ( Fig. 3, lane 9), were of the same size. These results demonstrate that absolute differentiation of brewing from non-brewing yeasts is difficult using only FLO1 PCR. It has been reported that a remarkable feature of the putative Flo1 protein is that it contains many repeated sequences: 18 repeats of 45 amino acid residues, two repeats of 20 residues, three repeats of 51 residues, and three repeats of nine residues ( Teunissen et al. 1993 ; Watari et al. 1994 ). These repeated sequences could have caused polymorphisms of FLO1. The FLO1 PCR products were not amplified from S. bayanus type strain ( Fig. 2, lane 6, Fig. 3, lane 8), but they were amplified from two strains of S. bayanus. Further work is needed to determine why these differences occurred.
In this study, we made certain that Saccharomyces yeasts had a FLO1 gene homologue by Southern analysis, and no FLO1 PCR DNA fragment was amplified from non-Saccharomyces yeasts. Gene cloning of non-Saccharomyces yeasts revealed the DNA encoding ORFs to have relatively high sequence identities ( Sudbery 1994). For example, the ODCase (orotidine-5′-phosphate decarboxylase) of S. cerevisiae and H. anomala were 73% identical ( Ogata et al. 1992 ). However, with this degree of homology, it is unlikely that a PCR fragment could be amplified. Therefore, the fact that the FLO1 PCR DNA fragment was not amplified from non-Saccharomyces yeasts should be expected. Further research is needed to determine whether there are FLO1 analogue genes in non-Saccharomyces yeasts. Hansenula anomala has been used for the treatment of waste-water ( Yoshizawa 1978). As flocculation is important for the treatment of waste-water ( Saito et al. 1990 ), it is of particular interest to determine whether there are FLO1 analogue genes in H. anomala or not.
rDNA (rRNA gene) is conserved more than those of the DNA sequences encoding ORFs of usual proteins ( Mao et al. 1982 ). For example, 5S rRNA of S. cerevisiae and Klyveromyces lactis had 100% sequence identity ( Miyazaki 1977), whereas ODCase had only 72% ( Mizukami & Hishinuma 1988). With such a high degree of homology in rDNA, it would be expected that the PCR fragment of rDNA could be amplified. The rDNA PCR products of Saccharomyces yeasts were of the same size, whereas those of Brettanomyces and other yeasts varied in size. However, the molecular sizes of rDNA PCR products of Saccharomyces yeasts, two strains of H. anomala and C. utilis were similar. Consequently, non-brewing yeasts could not be completely differentiated from brewing yeasts based only on specific PCR of rDNA. We therefore examined RFLP of rDNA and observed that the restriction profiles of rDNA differed between brewing and non-brewing yeasts. Using this technique, we could discriminate brewing from non-brewing yeasts.
The authors thank Dr Ichiro Yamashita for donating S. cerevisiae YIY261. They also thank Mr Yutaka Miyamoto and Dr Seizou Yabuuchi for valuable discussion and input.