Characterization and evaluation of indigenous Saccharomyces cerevisiae strains isolated from South African vineyards

[摘要] ENGLISH SUMMARY: The bioconversion of grape juice into wine by simply allowing the yeasts, associated withgrape berries and winery equipment, to ferment the sugars to ethanol, carbon dioxide andother minor, but important metabolites, is an ancient process. The art and science ofwinemaking has been extensively studied since the time when Louis Pasteur demonstrated,for the first time, the relationship between yeast and alcoholic fermentation. It is nowrecognized that the fermentation of grape must and production of premium quality wines is acomplex ecological and biochemical process involving the sequential development ofmicrobial species, as affected by a particular environment. This complex heterogeneousmicrobiological process includes the interaction of many microbial species, represented byfungi, yeasts, lactic acid bacteria and acetic acid bacteria, as well as the mycoviruses andbacteriophages affecting these grape-associated microorganisms. However, of all thesedifferent microbes and viruses, yeast represent the heart of the harmonious biochemicalinteraction with the musts derived from the various varieties of Vitis species which, in turn,are largely products of their respective genetic make-ups and the terroir. These yeasts aresignificant in winemaking because they not only conduct the alcoholic fermentation, but canalso spoil wine during conservation in the cellar and after packaging, and they affect winequality through the production of fermentation metabolites and through autolysis. A soundunderstanding of yeast systematics, biogeography and ecology is therefore essential toendeavours to preserve and exploit the hidden oenological potential of the untapped wealthof yeast biodiversity in our wine-producing regions. One of the main thrusts of this kind ofeco-taxonomic survey is to determine the actual contribution of the indigenous strains of theso-called wine yeast (Saccharomyces cerevisiae) and wild yeasts (non-Saccharomycesspecies) to the sensory properties of wines and to eventually develop new starter culturestrains for guided fermentations, including mixed starter cultures tailored to reflect thecharacteristics of a given wine region.Against this background, a comprehensive, long-term biogeographical survey and straindevelopment programme was launched. This dissertation represents the first phase of thislong-overdue research programme aimed to systematically catalogue yeasts in differentclimatic zones of the 350-year-old wine-producing regions of the Western Cape and todevelop new yeast starter cultures that would further increase the quality of South Africanwine. The specific aims of this dissertation included (i) the evaluation of yeast fingerprintingtechniques for their suitability to accurately and rapidly differentiate amongst S. cerevisiaestrains; (ii) the isolation and characterization of S. cerevisiae strains from the coastal regionsof the Western Cape; (iii) to determine the natural population dynamics of S. cerevisiaestrains in selected vineyards over a four-year period; (iv) to make a preliminary determinationof the possible effect that these indigenous S. cerevisiae isolates may have on wine flavour,and (v) to breed new starter culture strains with improved characteristics.Eighteen strains of S. cerevisiae used for commercial production of wine in South Africa werecharacterized by means of long-chain fatty acid analysis, randomly amplified polymorphicDNA (RAPD-PCR) and electrophoretic karyotyping (CHEF-DNA analysis). Variations in DNAprofiles of the strains were apparent in the number, position and intensity of bands. It wasfound that electrophoretic karyotyping, as a single technique, seemed to be the most usefulmethod to be used for routine fingerprinting. However, it was proposed that the combineduse of these three techniques would provide the most reliable means of differentiatingamongst wine yeast strains.Two of these fingerprinting techniques, CHEF-DNA and RAPD-PCR analysis, were used todetermine the geographic distribution of indigenous S. cerevisiae strains isolated from localvineyards. Grapes were aseptically harvested from 13 sites in five areas in the coastalregions of the Western Cape during 1995. These sites were Groot Constantia andBuitenverwachting in the Constantia area; Jordan, Lievland, Mont Fleur and Nietvoorbij in the Stellenbosch area; Vergelegen in Somerset West; De Rust, Oak Valley, White Hall andWildekrans in the Elgin/Bot River area; and Bouchard Finlayson and Hamilton Russel in theHermanus area. After fermentation, 30 yeast colonies per sample were isolated andexamined for the presence of S. cerevisiae. Five sampling sites yielded no S. cerevisiaestrains. Electrophoretic karyotyping revealed the presence of 46 unique karyotypes in eightof the remaining sites. No dominant strain was identified and each site had its own uniquecollection of strains. The number of strains per site varied from two to 15. Only in four casesdid one strain appear at two sites, while only one instance of a strain occurring at three siteswas recorded. All sites contained killer and sensitive strains, however, killer strains did notalways dominate. Commercial strains were recovered from three sites. Although commercialyeasts dominated the microflora at two sites, it appears that fears of commercial yeastsultimately dominating the natural microflora seem to be exaggerated.As an extension of the 1995 survey samples were taken from the same locations at GrootConstantia, Buitenverwachting, Jordan, Lievland, Mont Fleur, Vergelegen, BouchardFinlayson and Hamilton Russel during 1996 to 1998. This was done in an effort to assesshow the natural population dynamics of S. cerevisiae are affected over the long term byabiotic factors. Thirty colonies per site were isolated and the S. cerevisiae strains werecharacterized by electrophoretic karyotyping. The identity of strains appearing at more thanone site in the same, or different years, was confirmed by RAPD-PCR analysis. Strainnumbers per site varied over the four-year study period. Weather conditions resulting insevere fungal infestations and heavy applications of chemical sprays during 1996 and early1997 dramatically reduced the numbers of S. cerevisiae strains recovered during 1997. Areturn to normal weather patterns during mid 1997 resulted in a gradual recovery of theindigenous population as noted during the 1998 harvest. Indications are that some of thestrains isolated are widespread in the study area and may represent yeasts typical of thearea. Again, commercial wine yeast strains were recovered in only a few instances and thelikelihood that commercial yeasts will eventually replace the natural yeast microflora invineyards therefore seems remote.As a preliminary study to determine the possible effect of these indigenous S. cerevisiaestrains on wine flavour, 33 of the indigenous yeasts were allowed to ferment Chenin blancwine in laboratory fermentations. The juice was analyzed. The ability to form esters, fattyacids and higher alcohols was compared to that of two local commercial yeasts. None of theindigenous strains were found to be suitable for fermenting white must at 15°C. Their abilityto ferment red musts at much higher temperatures still needs to be assessed. Furthermore,differences noted indicate that some of these strains show potential to be included in ourextensive yeast-breeding programme as this would broaden the genetic pool.In parallel with the search to isolate and identify indigenous S. cerevisiae strains with goodoenological potential, an extensive selection and breeding programme with cultures from ourstrain collection was undertaken. The aim of this programme was to generate new strainsthat are better suited to New World winemaking styles and conditions prevailing in SouthAfrica. As a result, 145 hybrids, differentiated by elecrophoretic karyotyping and long-chainfatty acid analysis, were produced. Fifty-eight of the hybrids were able to ferment juice todryness at 15°C in less than 21 days during microvinification trials. Five of the strains werereleased for commercial use after extensive industrial-scale evaluation. Based on thesuccess of these interstrain hybridizations, the breeding programme will now be expanded toinclude some of the indigenous S. cerevisiae strains.In conclusion, it is only when we have a much better understanding of yeast biodiversity,biogeography, ecology and the interaction within yeast communities that we will be able tooptimally harness the genetic pool in our strain development programme, aimed to benefitboth the wine producer and the consumer.

[发布机构] Stellenbosch University