Comparative analysis of stress response patterns of Brettanomyces bruxellensis and Pichia guilliermondii from Patagonian grapes and wines.

SAEZ JULIETA; LOPES, CHRISTIAN; CONSTANZO, MARCOS; SANGORRÍN, MARCELA

ROSARIO

Congreso; SAMIGE; 2008

SAMIGE

Brettanomyces bruxellensis and Pichia guilliermondii yeast species are associated with wine spoilage. B. bruxellensis contaminations are related to finished wines, while P. guilliermondii ones occur during the first stages of fermentation. Consequently, these species are put under different environmental stress conditions that influence appreciably their growth and, consequently, their spoilage potential capacities. Some of the stress factors affecting differentially these species are: high sugar concentration and low temperatures at the initial stages, low pH and free SO2 concentration along the complete process, and high ethanol levels and temperatures during the middle and end stages. Biological interactions as the killer-sensitive ones are also an important factor affecting yeast viability during fermentation process. In this work, the response patterns of both B. bruxellensis and P. guilliermondii spoilage yeasts to physical-chemical and biological stress conditions occurring during wine fermentation was analyzed. Yeast isolates bearing different molecular patterns were tested in their tolerance to: pH, temperature, and sugar, free SO2 and ethanol concentrations, by colony growth of different dilutions on agar plates conditioned according the stress factor analyzed. Killer sensitivity was evaluated against 20 indigenous yeasts belonging to three species usually detected in wine fermentations by means of the growth inhibition method in YPD-MB agar plates. A matrix value of 1 or 0 was assigned to denote the presence or absence of colony growth respectively, as well as for the sensitivity or resistance to each particular killer strain respectively. Relationships among the strains were evaluated by Principal Coordinates Analysis (PCoA). PCoA (50% of total variability) obtained for P. guilliermondii using stress response patterns grouped the isolates into two clusters. Cluster I isolates exhibited low ethanol and high sugar tolerances, and they became from grapes or unfermented grape juice. Cluster II joined isolates showing high ethanol tolerance and average tolerance to high sugar concentrations; these isolates were mainly obtained from fermenting musts. In general, the strains showing high ethanol tolerance also evidenced a high resistance to killer yeast according to PCoA of killer response patterns (60% of total variability). PCoA (51% of total variability) using B. bruxellensis stress response patterns showed three clusters significantly related to the isolates origin. Isolates from fermentation vat 56 (grouped in cluster I), exhibited lower ethanol tolerance than the isolates from vat 46 (cluster II). Cluster III showed only one isolate from vat 56 characterized by its high ethanol tolerance. The same clustering pattern according to vat origin was also evidenced when killer sensitivity patterns were analyzed by PCoA (57% of the total variability). According to our results, the different wines stored in the vats could have selected different B. bruxellensis strains. EP.56.35 could represent the more dangerous isolate because of its high tolerance to ethanol and resistance against indigenous killer yeasts. We observed both a species-specific and an isolate-specific stress response among studied spoilage yeasts. Stress response patterns could be interesting in the identification of those potentially more dangerous isolates for the wine industry.