Differences between flocculating yeast and regular industrial yeast in transcription and metabolite profiling during ethanol fermentation.

Objectives: To improve ethanolic fermentation performance of self-flocculating yeast, difference between a flocculating yeast strain and a regular industrial yeast strain was analyzed by transcriptional and metabolic approaches.
Results: The number of down-regulated (industrial yeast YIC10 vs. flocculating yeast GIM2.71) and up-regulated genes were 4503 and 228, respectively. It is the economic regulation for YIC10 that non-essential genes were down-regulated, and cells put more "energy" into growth and ethanol production. Hexose transport and phosphorylation were not the limiting-steps in ethanol fermentation for GIM2.71 compared to YIC10, whereas the reaction of 1,3-disphosphoglycerate to 3-phosphoglycerate, the decarboxylation of pyruvate to acetaldehyde and its subsequent reduction to ethanol were the most limiting steps. GIM2.71 had stronger stress response than non-flocculating yeast and much more carbohydrate was distributed to other bypass, such as glycerol, acetate and trehalose synthesis. Conclusions: Differences between flocculating yeast and regular industrial yeast in transcription and metabolite profiling will provide clues for improving the fermentation performance of GIM2.71.

Introduction

Bioethanol production by Saccharomyces cerevisiae is currently, by volume, the single largest fermentative process in industrial biotechnology. The major portion of total expenditure in today’s bioethanol industry is allotted to feedstock costs (Galbe et al., 2007). A global research effort is under way to expand the substrate range of S. cerevisiae to include nonfood feedstocks, such as Jerusalem artichoke. Jerusalem artichoke (Helianthus tuberosus L.) can grow well in non-fertile land and is resistant to frost, drought, salt-alkaline and plant diseases (Yu et al., 2011). It is superior to the other inulin-accumulating crops in terms of its output of biomass production, inulin content, and tolerance of a relatively wide range of environmental conditions. The tuber yield of Jerusalem artichokes can be up to 90 t/ha resulting in 5–14 t carbohydrates/ha (Stephe et al., 2006). Besides its economic value, it also has a function of soil remediation, such as salt adsorption. To date, Jerusalem artichoke has predominantly been cultivated in North America, Northern Europe, Korea, Australia, New Zealand and China (Li et al., 2013). The principle storage carbohydrate of Jerusalem artichoke is inulin, which consists of linear chains of β-2, 1-linked d-fructofuranose molecules terminated by a glucose residue. It preserves carbohydrate in a 9:1 average ratio of fructose to glucose. Improving of fermentation performance with Jerusalem artichoke would have significant impacts on profits in large scale ethanol production.

Flocculating yeast separated from fermentation broth by self-flocculating at the end of fermentation and was re-used in consecutive fermentation, and therefore high density cell was obtained without increasing operating costs. High density cells exponentially shortened the fermentation time and increased cells resistance to ethanol stress (Li et al., 2009a). This work provides the first demonstration of the differences in transcriptic and metabolic profiles between flocculating yeast and regular industrial yeast. The result will provide clues to improve fermentative performance of flocculating yeast.

Materials and methods

Strain and cell culture

Industrial S. cerevisiae YIC10 is presented by Bincheng alcohol company (Shandong Province, China), self-flocculating S. cerevisiae GIM2.71 is obtained from Guangdong Microbiology Culture Center. Yeasts were grown overnight before inoculated in fresh medium (1% yeast extract, 2% peptone, 0.4% glucose, 3.6% fructose, ratio of fructose/glucose is 9 in order to stimulate hydrolysates of Jerusalem artichoke) to an initial OD600 of 0.1. Samples for microarray analysis were collected at exponential growth phase (7 h) and total RNA was then isolated. Samples for monitoring cell growth and fermentation were taken at 0, 2, 4, 5, 6, 7, 8, 10, 12, 14, 16, 18, 20, 21 and 23 h.

RNA extraction

After the sample was taken, it was immediately centrifuged at 4000 rpm for 3 min at 4 °C, the cells were then stored in liquid nitrogen until total RNA was extracted. Total RNA was extracted using Yeast RNAiso Kit (TaKaRa, Japan) after partially thawing the samples on ice, and RNA was purified using NucleoSpin Extract II kits (Machery-Nagel, Germany) according to the manufacturers’ instructions. Then total RNA was assessed by formaldehyde agarose gel (1.2%, w/v) electrophoresis and was quantitated spectrophotometrically (A260 nm/A280 nm ⩾ 1.80).

DNA microarray assays

An aliquot of 2 μg of total RNA was used to synthesize double-stranded cDNA, and cDNA was used to produce biotin-tagged cRNA by MessageAmpTM II aRNA Amplification Kit (Ambion, USA). The resulting biotin-tagged cRNA were fragmented to strands of 35–200 bases in length according to the protocols from Affymetrix. The fragmented cRNA was hybridized to Affymetrix GeneChip Yeast Genome 2.0 Arrays. Hybridization was performed at 45 °C using Affymetrix GeneChip Hybridization Oven 640 for 16 h. The GeneChip arrays were washed and then stained by Affymetrix Fluidics Station 450 followed by scanning with Affymetrix GeneChip Scanner 3000.

Microarray data processing

Hybridization data were analyzed using Affymetrix GeneChip Command Console Software. An invariant set normalization procedure was performed to normalize different arrays using DNA-chip analyzer 2010 (http://www.dchip.org, Harvard University). A multiclass method for analysis of microarray software (Significant Analysis of Microarray method, developed by Stanford University) was used to identify significant differences. Genes with false discovery rate <0.05 and a fold-change >2 were identified as differentially expressed genes. Differentially expressed genes were clustered hierarchically using Gene Cluster 3.0 (Stanford University). Gene ontology (GO) analysis of differentially expressed genes was done with DAVID (http://david.abcc.ncifcrf.gov/list.jsp).

Real-Time quantitative PCR

Based on microarray results, seven genes (HXT1-7) were selected for quantitative transcription analysis. The primers used in RT-qPCR analyses are listed in Table 1. Real-Time quantitative PCR (RT-qPCR) was performed according to the method described by Ye et al. (2009). ACT1 was used as an internal reference for normalizing gene expression (Liu et al., 2007).

Table 1

Genes and primers used in RT-qPCR.

Gene	ID	Primer Sequence 5′ → 3′	Amplicon (bp)
ACT1	F	AACATCGTTATGTCCGGTGGT	144
	R	ACCACCAATCCAGACGGAGTA
HXT1	F	GTTGCTTTCGGTGGTTTCAT	101
	R	TCGTGGTGCTTCATACCAAA
HXT2	F	ATTCGCTACTAGCCGCGTT	140
	R	TTGGCTTTGCTGGGAGTTCA
HXT3	F	GGCCGACCAAGTACTTACCA	85
	R	ACCGAAGGCAACCATAACAC
HXT4	F	TACCGTTTTCACTGCTGTCG	145
	R	GGAAGCAGCACCCCATAATA
HXT5	F	TCTGAAGTGTCGCCTAAGCA	139
	R	ATGGTACCCTCCATTGGACA
HXT6/7	F	GGGCTGTTTGGTCTTCATGT	94
	R	TTCTTCCCACATGGTGTTGA

Metabolites preparation and analysis

Intracellular and extracellular metabolites including glucose, fructose, ethanol, glycerol, acetate and trehalose were prepared by methods reported by our previous study (Li et al., 2009b). Samples were analyzed by a high-performance liquid chromatography (HPLC, Waters, USA) system with an Aminex HPX-87H column (Bio-Rad), 2414 refractive index detector and 515 HPLC pump. Column was kept at 50 °C and 5 mM H2SO4 was used as eluent at a flow rate of 0.5 ml/min.

Results and discussion

Fermentation behavior

YIC10 was superior to GIM2.71 in cell growth rate, sugar consumption and ethanol production performance (Fig. 1). YIC10 and GIM2.71 reached their highest ethanol yield at 12 h (16.2 g/L) and 21 h (16.0 g/L), respectively. Both strains showed indeed a similar behavior in terms of ethanol yield.

Figure 1

Fermentative performance of YIC10 and GIM2.71. Cell growth (a), fructose (b), glucose (c) and ethanol concentration (d) were determined.

Overview and GO analysis of microarray data

Microarray analysis showed that the number of down-regulated (YIC10 vs. GIM2.71) and up-regulated genes were 4503 and 228, respectively. It is the economic regulation for YIC10 that non-essential genes were down-regulated, and cells put more “energy” into growth and ethanol production. GO analysis was carried out with the up-regulated genes and the significant GO terms obtained were sorted according to their corresponding GO categories (Table 2). According to that analysis, most of genes focused on monosaccharide, hexose and glucose metabolic process, generation of precursor metabolites and energy and ion transport (Table 2), which indicated that these pathways may have some contributions for fermentative performance.

Table 2

The GO analysis of up-regulated genes (Top 10).

Term	Count	%	P-value
Metabolism
DNA metabolic process	26	13.3	2.90 × 10−2
Monosaccharide metabolic process	15	7.7	4.40 × 10−4
Hexose metabolic process	13	6.6	1.70 × 10−3
Glucose metabolic process	11	5.6	3.60 × 10−3
Cellular carbohydrate catabolic process	10	5.1	1.80 × 10−3
Hexose catabolic process	8	4.1	4.20 × 10−3
Monosaccharide catabolic process	8	4.1	6.20 × 10−3
Alcohol catabolic process	8	4.1	9.60 × 10−3
Glucose catabolic process	7	3.6	9.60 × 10−3
Oxidoreduction coenzyme metabolic process	7	3.6	2.80 × 10−2
Energy
Generation of precursor metabolites and energy	16	8.2	6.20 × 10−2
Energy reserve metabolic process	5	2.6	6.80 × 10−2
Transport
Ion transport	17	8.7	4.20 × 10−3
Cation transport	15	7.7	5.10 × 10−4
Metal ion transport	9	4.6	1.40 × 10−2
Di-, tri-valent inorganic cation transport	7	3.6	7.90 × 10−3
Transition metal ion transport	7	3.6	1.90 × 10−2
Carboxylic acid transport	7	3.6	4.40 × 10−2
Siderophore transport	5	2.6	2.00 × 10−4
Anion transport	5	2.6	3.20 × 10−2
Iron assimilation by chelation and transport	4	2	9.00 × 10−4
Siderophore-iron transport	4	2	9.00 × 10−4
Protein
Protein modification by small protein conjugation or removal	11	5.6	2.70 × 10−2
Protein modification by small protein conjugation	10	5.1	1.80 × 10−2

Hexose transport

Gene expression analysis using RT-qPCR method was well corresponded with microarray means (Fig. 2a). Transport is suggested as the rate-limiting step of glycolysis in metabolic control analysis and transport exerts a high degree of control on glycolytic flux (Oehlen et al., 1994). The results showed that the detected transporter genes were all down-regulated in YIC10 vs. GIM2.71 comparisons, except HXT5 (Fig. 2a). It was consistent with the report that HXT5 was regulated by the growth rate of cells, where the growth rate of YIC10 was significantly higher than GIM2.71. However, different from HXT5, HXT1-4 and HXT6/7 were regulated by extracellular glucose (Diderich et al., 2001). Investigations using single transport mutants also showed that Hxt1-4, 6 and 7 are the major hexose transporters in yeast transporting glucose and fructose (Reifenberger et al., 1997, Reifenberger et al., 1995). Furthermore, analysis of intracellular glucose and fructose showed that both sugars levels were always higher in GIM2.71 than in YIC10 (Fig. 2b and c), which was consistent with the higher expression of major genes involved in hexose transporter. It concluded that hexose transport was not the limiting-step in sugar consumption and ethanol production for GIM2.71, compared to YIC10.

Figure 2

Expression of genes encoding hexose transport (A) and intracellular fructose (B) and glucose (C) levels.

Central carbon metabolism

Once sugars have been imported into cells, they are phosphorylated by one of three sugar kinases, Hxk1, Hxk2 and Glk1. Glucose and fructose are both phosphorylated by hexokinases Hxk1 and Hxk2 but with different efficiencies, and the glucokinase Glk1 phosphorylates glucose but not fructose (Rodriguez et al., 2001). The three genes were all down-regulated in YIC10 to GIM2.71 comparisons, which indicated that hexose phosphorylation was not the limiting steps in sugar consumption and ethanol production for GIM2.71.

Most genes in central carbon metabolism were down-regulated, only 3-phosphoglycerate kinase encoding genes PGK1, pyruvate decarboxylase encoding genes PDC6, alcohol dehydrogenase encoding genes ADH5 were up-regulated (Fig. 3). During S. cerevisiae growth on fermentable carbon sources, six PDC genes were identified out of which three structural genes (PDC1, PDC5 and PDC6) were encoded for active Pdc enzymes, independently (Milanovic et al., 2012). Pdc6p is the predominant isoenzyme form that catalyzes an irreversible reaction in which pyruvate is decarboxylated to acetaldehyde. Additionally, there are four genes (ADH1, ADH3, ADH4 and ADH5) that encode alcohol dehydrogenases involved in ethanol synthesis. ADH5 gene product is the major enzyme that is responsible for converting acetaldehyde to ethanol. It suggested that the most limiting steps of ethanol fermentation were the reaction of 1,3-disphosphoglycerate to 3-phosphoglycerate, the decarboxylation of pyruvate to acetaldehyde and its subsequent reduction to ethanol.

Figure 3

Expressions of genes involved in central carbon metabolism. “+” and “–” before regulation fold represented up-regulated and down-regulated genes, respectively.

Expression of genes involved in glycerol and its intracellular level

Glycerol was the major by-product in ethanol fermentation. The first step in glycerol synthesis is the most important as glycerol-3-phosphate dehydrogenase (encoded by GPD1 and GPD2) activity controls the amount of glycerol produced. In this experiment, GPD1 and GPD2 were down-regulated significantly, and other genes involved in glycerol both synthesis (RHR2 and HOR2) and degradation (GUT1 and GUT2) were all down-regulated (Fig. 4a). Intracellular metabolic analysis showed that glycerol was at relatively low levels both for YIC10 and GIM2.71 at the onset of fermentation, whereas it was accumulated 83-fold compared to its initial level in GIM2.71 when ethanol was exponentially synthesized and carbon resource was exhausted (Fig. 4b). And this response was significantly stronger than YIC10.

Figure 4

Expression of genes involved in glycerol (a), acetate (c) and trehalose (e), and intracellular levels of glycerol (b), acetate (d) and trehalose (f).

Expression of genes involved in acetate and its intracellular level

Among genes encoding acetate synthesis, only ALD4 was up-regulated and the other three genes (ALD2, ALD5 and ALD6) were down-regulated (Fig. 4c). It was reported that the deletion of ALD4 had no effect on the amount of acetate formed (Remize et al., 2000). Intracellular metabolic analysis showed that acetate in YIC10 was always at a relatively low level, whereas acetate in GIM2.71 was accumulated quickly at late-logarithmic phase (Fig. 4d).

Expression of genes involved in trehalose and its intracellular level

Genes both were encoding trehalose synthesis (TPS1 and TPS2) and hydrolysis (ATH1 and NTH1) were all down-regulated (Fig. 4e). The intracellular trehalose in YIC10 was always at a relatively low level throughout the fermentation, whereas trehalose in GIM2.71 was accumulated rapidly at 10 h and 16 h (Fig. 4f).

Glycerol, acetate and trehalose were significantly accumulated in response to environmental stress in GIM2.71. Glycerol formation is the results of redox balance and stress response (Nevoigt and Stahl, 1997) and the observed differences suggest that the two strains could have a different stress response. This hypothesis is also supported by the formation of acetate, another significant redox-driven product, and the accumulation of trehalose, other potential stress protectants like glycerol.

Achieving high fermentative performance is a major challenge, particularly when it comes to modifications of the central carbon metabolism which is inherently coupled to energy and redox issues. Glycerol is the major by-product accounting for up to 5% of the carbon in S. cerevisiae ethanolic fermentation. Decreasing glycerol formation may redirect part of the carbon toward ethanol production (Nissen et al., 2000). Pagliardini et al. (2010) reported that fine-tuning the glycerol synthesis pathway allowed the strains to keep their initial ethanol tolerance.

Conflict of interest

The authors declare that they have no conflict of interest.