Effect of low temperature fermentation on wine aroma

Get Complete Project Material File(s) Now! »

Chapter 4. Effect of low temperature fermentation on gene expression

Introduction

This chapter examines the influence of low temperature on yeast gene expression during the fermentation of SB, with the aim of building on previous knowledge of transcriptional changes during cold fermentation. Differentially regulated transcripts can indicate which proteins are increased or decreased and which biological pathways are influenced during the response to a particular stress.
Gene expression was analysed for a commercial wine strain, M2 (Enoferm, Lallemand), and four F₁ hybrids made between M2 and four genetically distinct strains: M2xY55, M2xS288c, M2xL-1528 and M2xBC187. Gene expression was measured during fermentation in SB juice at 25°C for M2, and at 12.5°C for M2 and all four F₁ hybrids, at two fermentation stages: when 2 % sugars had been consumed (cells proliferating exponentially), and when 70 % sugars had been consumed (cells had stopped proliferating).

Strain selection and fermentation for microarray sampling

The process of selecting the commercial wine yeast, M2, for microarray analysis and hybridisation is described in detail in Chapter 6, Section 6.2. The four strains which were hybridised with M2 were chosen from the set of 39 strains in Chapter 3, based on their ability to ferment in the cold, fertility of the F₁–derived spores, and their genetic diversity (for details on the screening and selection of the F₁ hybrids, see Chapter 6, Section 6.3). The chosen F₁ hybrids were: M2xY55, M2xS288c, M2xL-1528 and M2xBC187. These hybrids were also analysed for their aroma compound concentrations (see Chapter 5), and used as potential candidates for the introgression of cold tolerance genes into the M2 background (see Chapter 6).
Fermentation in 200 mL 07 SB juice, at 22 °Brix, was conducted at 12.5°C (cold) and 25°C (optimal) for the M2 Ura- parent strain and at 12.5°C (cold) for the four F₁ hybrid yeast strains. Uninoculated juice samples were included and fermentations were conducted in triplicate (resulting in 18 samples at 12.5°C and 6 samples at 25°C, 24 ferments in total). Fermentations were continuously shaken and their progress was monitored by measuring cumulative weight loss (g) at regular intervals (Figure 4-1). Although all strains eventually achieved more than 17 g weight loss, there were differences in their fermentation kinetics. At 12.5°C, M2xY55 and M2xS288c had a longer lag phase than the other strains, although M2xY55 subsequently finished fermentation well. M2 Ura- and M2xBC187 were the fastest to start fermentation. At the optimal temperature of 25°C, M2 Ura- initiated fermentation after only 24 h, compared to 100 h at 12.5°C.

Fermentation kinetics

The kinetic parameters V_max and A_max were estimated as described by Marullo et al. (2006) and are displayed in Figure 4-2. The V_max values (dCO₂/dt) for the 12.5°C fermentation were highest for the M2xBC187 F₁ hybrid (0.152 ± 0.005) and the M2 Ura- parent (0.146 ± 0.005). The other three hybrids had sequentially lower rates: M2xY55 (0.112 ± 0.004), M2xL-1528 (0.099 ± 0.003) and M2xS288c (0.087 ± 0.003). The A_max values (d²CO₂/dt²) were highest for M2xBC187 (0.023 ± 0.003), followed by M2xL-1528 (0.015 ± 0.001) and M2 Ura- (0.015 ± 0.001), M2xS288c (0.008 ± 0.002) and M2xY55 (0.005 ± 0.001). For the 25°C fermentation the V_max of M2 Ura- was 0.438 ± 0.006 and the A_max was much higher than at 12.5°C, at 0.023 ± 0.0002.

Extracting RNA from yeast cells for microarrays

From each fermentation, 1 x 10⁸ cells were harvested at two time points: early fermentation/late exponential growth (2 % cumulative weight loss, with a population of ~4-5 x 10⁷ cells mL^-1) and late fermentation/middle stationary growth (70 % cumulative weight loss, with a population of ~1-2 x 10⁸ cells mL^-1). These two stages will be referred to as exp and stat, respectively. Cells were rapidly snap frozen in liquid nitrogen to preserve their gene expression and RNA was extracted from the cell pellets, as described in Chapter 2, Section 2.5.7. In total, 36 samples were prepared for microarray analysis: five strains in triplicate at 12.5°C (exp and stat) and one in triplicate at 25°C (exp and stat).

Quantity and quality of extracted total RNA

The concentration and quality of the RNA extracted from the 36 samples was assessed using the NanoDrop® spectrophotometer ND-1000 and the Agilent 2100 BioAnalyzer®, respectively (see Chapter 2, Section 2.5.8). Samples were given chip ID codes (sample ID =RD0001 to RD0036, see Appendix A.1, Table A-1) and converted to cRNA by L. Williams (Centre for Genomics and Proteomics, SBS, University of Auckland). Table A-1 (Appendix A.1) also shows the RNA yields obtained for the 36 samples.
The BioAnalyzer® generated 36 electropherograms which were used to determine the quality (purity and integrity) of the extracted RNA, based on the appearance and ratios of the 18S and 28S rRNA peaks, and the appearance of any contaminating proteins. A representative electropherogram of one sample of extracted RNA is shown in Figure 4-3. All 36 samples displayed comparable results with clearly defined rRNA peaks and no protein contaminants.

Microarray pre-processing and analysis

Poly (A)+ RNA purification, cDNA synthesis, biotin-labelled cRNA synthesis and cRNA fragmentation were performed by L. Williams on the 36 RNA samples. Table A-1 (Appendix A.1) shows the cRNA concentrations and absorbance 260/280 values for each sample. All cRNA samples had total RNA yields below the recommended 20 µg in the 21 µL elution (Eukaryotic Arrays GeneChip® Expression Analysis and Technical Manual, Affymetrix, Santa Clara, California); however, after consultation with L. Williams it was decided that the amount and quality of the cRNA was adequate according to his experience. cRNA was hybridised to 5744 probesets on the Affymetrix GeneChip Yeast Genome 2.0 Arrays.
The additional 5021 probesets for the fission yeast S. pombe on the microarray were removed later. For each of the 11 probe pairs per probeset, there were both perfect match probes (PM) which specifically hybridise to their intended transcripts, and single base mismatch probes (MM) to measure non-specific hybridisation. Arrays were visualised, pre-processed and analysed using the statistical programme R with Bioconductor software (Version 2.2). The R code used to analyse the microarray data is presented in Appendix A.2. The 36 .CEL files containing raw array data (signal intensity at each probe) were imported and read into Bioconductor using the ReadAffy function and the quality of the 36 arrays was determined by visualising the log₂-transformed hybridisation signals for each chip, based on the methods by ALVORD et al. (2007). Most arrays showed good signal intensity and quality (see Appendix A.3, Figure A-1). Two arrays, RD0018 (MxY 12.5 stat 3) and RD0020 (MxS 12.5 exp 2), displayed a high signal:noise ratio, with an excess amount of cRNA hybridised to the chips, which would result in a decreased sensitivity to true differential expression. Two other chips, RD0008 (M2 12.5 exp 2) and RD0019 (MxS 12.5 exp 1) displayed low signal:noise ratio with low probe intensities, resulting in a lower than normal representation of rare transcripts, as shown in the boxplots of log₂-transformed signal intensities in Figure 4-4A.
RD0008 (M2 12.5 exp 2), RD0018 (MxY 12.5 stat 3), and RD0020 (MxS 12.5 exp 2) were excluded from gene expression analysis; however, RD0019 (MxS 12.5 exp 1) was retained in the experiment since at least two chips per sample were required for performing statistics on the gene expression data. Therefore, three conditions, M2 12.5 exp, MxY 12.5 stat and MxS 12.5 exp, were analysed in duplicate, not triplicate.
The conclusions from the assessment of chip quality, which were obtained by viewing the chip images, were confirmed by the boxplot representations, with the same four samples having high and low probe intensities. The remaining 33 samples were then background corrected, normalised (see boxplots in Figure 4-4B) and expression values calculated using robust multiarray averaging (RMA).

Overview of microarray results

This experiment had three aims: to look at differential expression between exponential and mid-late stationary phase; to elucidate yeast transcriptional changes during fermentation that were linked to cold fermentation; and to compare transcriptional changes between four genetically different F₁ hybrid yeast. The samples taken at 2 % (exp) and 70 % (stat) fermentation weight loss, represented different stages for the fermentations, as illustrated for M2 Ura- in Figure 4-5. The two fermentation stages were assumed to be comparable at the two temperatures, since the time points were taken at an equivalent stage of weight loss.

Abstract 
Acknowledgements 
List of Figures 
List of Tables 
Abbreviations
Chapter 1. Introduction 
1.1. General characteristics of S. cerevisiae
1.2. S. cerevisiae and alcoholic fermentation6
1.3. Low temperature fermentation
1.4. Aims and significance of the research
Chapter 2. Materials and methods
2.1. Laboratory reagents
2.2. Yeast strains
2.3. Yeast culture and handling
2.4. Yeast growth and fermentation experiments
2.5. Molecular biology techniques
2.6. Quantification of aroma compounds in wine
Chapter 3. Effect of temperature on the ability of genetically diverse S. cerevisiae strains to grow and ferment
3.1. Introduction
3.2. Initial yeast screening before testing for variation in growth and fermentation
3.3. Screening commercial wine yeast for cold fermentation parameters
3.4. Variation among strains for growth and fermentation across a range of temperatures
3.5. Discussion
3.6. Conclusion
Chapter 4. Effect of low temperature fermentation on gene expression 
4.1. Introduction
4.2. Strain selection and fermentation for microarray sampling
4.3. Quantity and quality of extracted total RNA
4.4. Microarray pre-processing and analysis .
4.5. Overview of microarray results
4.6. Comparison 1: Growth phase
4.7. Comparison 2: Temperature
4.8. Comparison 3: Strain
4.9. Discussion
4.10. Conclusion
Chapter 5. Effect of low temperature fermentation on wine aroma
5.1. Introduction
5.2. The effect of low temperature on the aroma compounds produced by M2 during microvinification
5.3. Strain differences between M2 and the four M2 F1 hybrids on aroma compounds produced
during low temperature fermentation
5.4. Genes differentially expressed during low temperature fermentation with roles in aroma
production
5.5. The effect of low temperature on the aroma compounds produced by M2 and three other S.
cerevisiae strains in different SB juices during microvinification
5.6. Discussion .
5.7. Conclusion
Chapter 6. Identification of genes important for low temperature fermentation 
6.1. Introduction
6.2. Selection of commercial wine yeast and four other S. cerevisiae strains for hybridisation,
backcrossing, microarrays and aroma compound analysis
6.3. Screening 12 F1 hybrids
6.4. Backcrossing of cold tolerance genes into M2
6.5. Identifying genes linked to cold fermentation using mapped progeny from a BY4716xRM11-1a
cross
6.6. Discussion
6.7. Conclusion
Chapter 7. Final discussion 
7.1. Introduction
7.2. Summary of key findings
7.3. Experimental components
7.4. Future research
7.5. Significance
7.6. Conclusion
References
Appendices
GET THE COMPLETE PROJECT
Effect of low temperature on Sauvignon blanc fermentation by Saccharomyces cerevisiae