Transcriptional response of ethanol-stressed vs. non-stressed culture

The yeast Saccharomyces cerevisiae is well known for its high ethanol production performances. An original fermentation process that allows the yeast S. cerevisiae to produce in less than 45 h more than 150 g/l ethanol (i.e. 18.9°GL) was set up in our laboratory [1]. Under this condition, the yeast cells induce a dynamic process to adapt to increased ethanol concentration by a mechanism that is likely different to the stress response triggered by sudden ethanol addition to exponentially growing cells [2]. Kinetic analysis of the growth curve identified two main phases: a growth phase that ended up at 90 g/l ethanol and then an uncoupling phase during which non-growing cells kept producing ethanol. This latter phase is also characterized with an increased loss of viability. In order to investigate on a genome scale the expression changes occurring during this process, gene expression was quantified using DNA chips technology at six different time-points during fed-batch fermentation. [1] Alfenore et al, Appl. Microbiol. Biotechnol. 60 : 67-72, 2002. [2] Alexandre H. et al., FEBS Lett. 498(1) : 98-103, 2001. We measure the changes in the gene expression of ethanol stressed culture at five different time-points during fed-batch fermentation compared to a common reference consisting of exponentially growing yeast cells ( sample number 1 : growing cells ; low ethanol concentration of 17 g/l ; specific growth rate of 0.3 h-1) . The sets corresponded to sample number 2 : growing cells/ethanol concentration of 60 g/l ; sample number 3 : before growth arrest/ethanol concentration of 90 g/l ; sample number 4 : growth arrest/ethanol concentration of 95 g/l ; sample number 5 : 1 hour after growth arrest/ethanol concentration of 100 g/l and sample number 6 : uncoupling phase/ethanol concentration of 125 g/l. For each sample, total RNAs from one yeast culture (no biological replicate) were extracted three times (technical replicates -extract). For labelling, we labelled the common reference with dCTP-Cy5 and labelled the sample with dCTP-Cy3 and hybridized cDNA on three independent microarrays, given rise to six data value per gene (each gene is duplicate in the slide).