Selection and validation of reference genes for quantitative real-time PCR studies during Saccharomyces cerevisiae alcoholic fermentation in presence of sulfite

Sulfur dioxide is extensively used during industrial fermentations and contributes to determine the harsh conditions of winemaking together with low pH, high sugar content and increasing ethanol concentration. Therefore the presence of sulfite has to be considered in yeast gene expression studies to properly understand yeast behavior in technological environments such as winemaking. A reliable expression pattern can be obtained only using an appropriate reference gene set that is constitutively expressed regardless of perturbations linked to the experimental conditions. In this work we tested 15 candidate reference genes suitable for analysis of gene expression during must fermentation in the presence of sulfite. New reference genes were selected from a genome-wide expression experiment, obtained by RNA sequencing of four Saccharomyces cerevisiae wine strains grown in enological conditions. Their performance was compared to that of the most common genes used in previous studies. The most popular software based on different statistical approaches (geNorm, NormFinder and BestKeeper) were chosen to evaluate expression stability of the candidate reference genes. Validation was obtained using other wine strains by comparing normalized gene expression data with transcriptome quantification both in the presence and absence of sulfite. Among 15 reference genes tested ALG9, FBA1, UBC6 and PFK1 appeared to be the most reliable while ENO1, PMA1, DED1 and FAS2 were the worst. The most popular reference gene ACT1, widely used for S. cerevisiae gene expression studies, showed a stability level markedly lower than those of our selected reference genes. Finally, as the expression of the new reference gene set remained constant over the entire fermentation process, irrespective of the perturbation due to sulfite addition, our results can be considered also when no sulfite is added to the must. © 2015 Elsevier B.V.