Developing a novel selection method for alcoholic fermentation starters by exploring wine yeast microbiota from Greece

The selection of native yeast for alcoholic fermentation in wine focuses on ensuring the success of the process and promoting the quality of the final product. The purpose of this study was firstly to create a large collection of new yeast isolates and categorize them based on their oenological potential. Additionally, the geographical distribution of the most dominant species, Saccharomyces cerevisiae, was further explored. Towards this direction, fourteen spontaneously fermented wines from different regions of Greece were collected for yeast typing. The yeast isolates were subjected in molecular analyses and identification at species level. RAPD (Random Amplified Polymorphic DNA) genomic fingerprinting with the oligo-nucleotide primer M13 was used, combined with Matrix Assisted Laser Desorption Ionization-Time of Flight Mass Spectrometry (MALDI-TOF MS) technique. All yeast isolates were scrutinized for their sensitivity to killer toxin, production of non-desirable metabolites such as acetic acid and H2S, β-glucosidase production and resistance to the antimicrobial agent; SO2. In parallel, S. cerevisiae isolates were typed at strain level by interdelta - PCR genomic fingerprinting. S. cerevisiae strains were examined for their fermentative capacity in laboratory scale fermentation on pasteurized grape must. Glucose and fructose consumption was monitored daily and at the final point a free sorting task was conducted to categorize the samples according to their organoleptic profile. According to our results, among the 190 isolates, S. cerevisiae was the most dominant species while some less common non-Saccharomyces species such as Trigonopsis californica, Priceomyces carsonii, Zygosaccharomyces bailii, Brettanomyces bruxellensis and Pichia manshurica were identified in minor abundancies. According to phenotypic typing, most isolates were neutral to killer toxin test and exhibited low acetic acid production. Hierarchical Cluster Analysis revealed the presence of four yeast groups based on phenotypic fingerprinting. Strain level typing reported 20 different S. cerevisiae strains from which 65% indicated fermentative capacity and led to dry wines. Sensory evaluation results clearly discriminated the produced wines and consequently, the proposed yeast categorization was confirmed. A novel approach that employs biostatistical tools for a rapid screening and classification of indigenous wine yeasts with oenological potential, allowing a more efficient preliminary selection or rejection of isolates is proposed.