analyzed, 25 can be differentiated by theirPCR profile associated with δ sequences. Lavallée et al. (1994) also observed excellent discriminating power with this method while analyzing industrially produced commercial strains from Lallemand Inc. (Montreal, Canada). In addition, this method enables the identification of 25–50 strains per day; it is the quickest of the different strain identification techniques currently available. When used for native yeast strain identification in a given viticultural region, however, it seems to be less discriminating than karyotype analysis. PCR profiles of wild yeasts isolated in a given location often appear similar. They have several constant bands and only a small number of variable discriminating bands. Certain strains have the same PCR amplification profile while having different karyotypes. In a given location, the polymorphism witnessed by PCR associated with δ sequences is lower than that of the karyotypes. This method is therefore complementary to other methods for characterizing winemaking strains. PCR enables a rapid primary sorting of a native yeast population. Karyotype analysis refines this discrimination.
FIGURE 1.31 Electrophoresis in agarose gel (at 1.8%) of amplified fragments obtained from various commercial yeast strains. Band 1, F10; band 2, BO213; band 3, VL3c; band 4, UP30Y5; band 5, 522 D; band 6, EG8; band 7, L‐1597; band 8, WET 136; M, molecular weight marker; T, negative control.
Saccharomyces uvarum strains cannot be distinguished by this technique because their genome contains only a few Ty elements.
Lastly, because of its convenience and rapidity, PCR associated with δ sequences facilitates verification of the implantation of yeast starters used in winemaking. The analyses are conducted on the entire biomass derived from lees, placed beforehand in a liquid medium in a laboratory culture. The amplification profiles obtained are compared with inoculated yeast strain profiles. They are identical with a successful implantation, and different if the inoculation fails. Figure 1.32 gives examples of successful (yeasts B and C) and unsuccessful (yeasts A, D, and E) implantations. Contaminating strains have a different amplification profile than the yeast starter.
The detection threshold of a contaminating strain was studied in the laboratory by analyzing a mixture of two strains in variable proportions. In the example given in Figure 1.33, the contaminating strain is easily detected at 1%. In winery fermentations, however, several native strains can coexist with the inoculated strain in small proportions. When must undergoing fermentation or lees is analyzed by PCR, the yeast implantation rate is at least 90% when the amplification profiles of the lees and the yeast starter are identical.
1.9.6 PCR with Microsatellites
Microsatellites are tandem repeat units of short DNA sequences (1–10 nucleotides), i.e. in the same direction and dispersed throughout the eukaryote genome. The number of motif repetitions is extremely variable from one individual to another, making these sequences highly polymorphic in size. These regions are easily identified, thanks to knowledge of the full sequence of the S. cerevisiae genome. Approximately 275 sequences have been listed, mainly AT dinucleotides and AAT and AAC trinucleotides (Field and Wills, 1998; Hennequin et al., 2001; Perez et al., 2001). Furthermore, these sequences are allelic markers, transmitted to the offspring in a Mendelian fashion. Consequently, these are ideal genetic markers for identifying specific yeast strains, making it possible not only to distinguish between strains but also to arrange them in related groups. This technique has many applications in humans: paternity tests, forensic medicine, etc. In viticulture, this molecular identification method has already been applied to Vitis vinifera grape varieties (Bowers et al., 1999).
FIGURE 1.32 Electrophoresis in agarose gel (1.8%) of amplified fragments illustrating examples of yeast implantation tests (successful: yeasts B and C; unsuccessful: yeasts A, D, and E). Band 1, negative control; band 2, lees A; band 3, ADY A; band 4, lees B; band 5, ADY B; band 6, lees C; band 7, ADY C; band 8, lees D; band 9, ADY D; band 10, lees E; band 11, ADY E; M, molecular weight marker.
FIGURE 1.33 Determination of the detection threshold of a contaminating strain. T, negative control; band 1, strain A 70%, strain B 30%; band 2, strain A 80%, strain B 20%; band 3, strain A 90%, strain B 10%; band 4, strain A 99%, strain B 1%; M, molecular weight marker; band 5, strain A 99.9%, strain B 0.1%; band 6, strain A; band 7, strain B.
The technique consists in amplifying the region of the genome containing these microsatellites, and then analyzing the size of the amplified portion to a level of detail of one nucleotide by capillary electrophoresis. This size varies by a certain number of base pairs (approximately 8–40) from one strain to another, depending on the number of times the motif is repeated. A yeast strain may be heterozygous for a given locus, giving two different‐sized amplified DNA fragments. Using six microsatellites, Perez et al. (2001) were able to identify 44 different genotypes within a population of 51 strains of S. cerevisiae used in winemaking. Other authors (Gonzalez Techera et al., 2001; Hennequin et al., 2001, Klis et al., 2002) have shown that the strains of S. cerevisiae used in winemaking are weakly heterozygous for the loci studied. However, interstrain variability of the microsatellites is very high. The results are expressed in numerical values for the size of the microsatellite in base pairs or the number of repetitions of the motifs on each allele. These digital data are easy to interpret, unlike the karyotype images on agarose gel, which are not really comparable from one laboratory to another. Based on 41 microsatellites, Legras et al. (2005) selected six very discriminating and reproducible loci. They highlight the relationships between strains from different geographical zones or industrial environments.
Microsatellite analysis has also been used to identify the strains of S. uvarum (Masneuf‐Pomarède et al., 2007, 2016) and of S. kudriavzevii (Erny et al., 2012) used in winemaking. As the S. uvarum, S. kudriavzevii, and S. cerevisiae microsatellites have different amplification primers, this method provides an additional means of distinguishing between these species and their hybrids.
The development of new‐generation sequencing methods for yeast genomes has made sequences available for non‐Saccharomyces species. A typing method of yeast strains by microsatellite marker analysis is now offered for B. bruxellensis, T. delbrueckii, H. uvarum, and Starmerella bacillaris (Albertin et al., 2014a,b, 2016; Masneuf‐Pomarede et al., 2015, 2016). When applied to the study of a great number of yeast isolates, these methods help to better describe the genetic diversity and the population structure of winemaking yeasts. Factors influencing this structure as well as their life cycle and reproduction mode have also been described. From an applied point of view, this molecular typing method is a useful tool in winemaking yeast strain identification, ecological surveys, and quality control of industrial production batches.
1.9.7 Genome Sequencing
With the development of new‐generation sequencing methods, new approaches to yeast characterization have been suggested. The multilocus sequence
