Construction of Novel Saccharomyces cerevisiae Strains for Bioethanol Active Dry Yeast (ADY) Production

The application of active dry yeast (ADY) in bioethanol production simplifies operation processes and reduces the risk of bacterial contamination. In the present study, we constructed a novel ADY strain with improved stress tolerance and ethanol fermentation performances under stressful conditions. The industrial Saccharomyces cerevisiae strain ZTW1 showed excellent properties and thus subjected to a modified whole-genome shuffling (WGS) process to improve its ethanol titer, proliferation capability, and multiple stress tolerance for ADY production. The best-performing mutant, Z3-86, was obtained after three rounds of WGS, producing 4.4% more ethanol and retaining 2.15-fold higher viability than ZTW1 after drying. Proteomics and physiological analyses indicated that the altered expression patterns of genes involved in protein metabolism, plasma membrane composition, trehalose metabolism, and oxidative responses contribute to the trait improvement of Z3-86. This work not only successfully developed a novel S. cerevisiae mutant for application in commercial bioethanol production, but also enriched the current understanding of how WGS improves the complex traits of microbes.


Introduction
Bioethanol, a long considered suitable alternative to fossil fuel, has shown a sharp increase in production since 2006 [1,2]. Saccharomyces cerevisiae strains are the main microbes used for ethanol production because of their capability to efficiently convert hexose to ethanol and carbon dioxide [3][4][5]. Traditionally, yeast cells are cultured and propagated until the cell number desired for ethanol fermentation is achieved. This process begins with slant cultivation followed by gradual expansion of the cultivation scale in liquid medium to increase the volume. As an alternative to this hassle, many factories use commercial active dry yeast (ADY) to initiate ethanol fermentation. The application of ADY in fuel ethanol production cuts down initial lag time for fermentation and reduces yeast culture plant numbers and the risk of bacterial contamination [6].
To improve the economic and social benefits of bioethanol, novel technologies, such as very-high-gravity (VHG) and hightemperature ethanol fermentation, have been proposed [7,8]. While these approaches can reduce energy, condensate water consumption and increase fermentation efficiency, they also impose many stresses (e.g. osmotic pressure, high temperature, and high concentration of ethanol) on yeast cells. During the air-drying and rehydration of ADY, cells are exposed to fluctuating osmotic pressures and other stressors [9][10][11][12]. Thus, only S. cerevisiae strains that are able to tolerate multiple stresses can be made into commercial ADY for ethanol fermentation under stressful conditions.
Since the publication of the S. cerevisiae genome in 1996, many functional genomics studies have been carried out to investigate how this model unicellular organism responds and adapts to environmental stresses [13][14][15][16]. Tolerance to even a single stressor is usually controlled by gene networks that are difficult to modify through genetic engineering. To overcome this difficulty, global and comprehensive gene modification techniques have been demonstrated for improving the stress tolerance and fermentation properties of strains. Wholegenome shuffling (WGS) is such an approach in which gene recombination is achieved at different positions throughout the whole genome by recursive protoplast fusion or crossing [17]. In this process, genetic diversity is generated and resultant mutants with the desired phenotypes can be selected using specific screening methods [18,19]. Pinel et al. reported that genome shuffling through recursive population mating improved the tolerance of S. cerevisiae to spent sulfite liquor [20]. Shi et al. demonstrated that the thermotolerance of the species can be significantly enhanced using WGS [8]. Our previous studies have also suggested that WGS can improve the ethanol yield of S. cerevisiae under VHG conditions [21,22]. These reports suggest that WGS is a powerful tool for selection of the strain of interest, probably improved.
In the present study, an excellent industrial S. cerevisiae strain, ZTW1, was selected and subjected to a modified WGS process to construct a novel yeast strain that is applicable in commercial bioethanol ADY production. After three rounds of genome shuffling, we obtained mutants with improved tolerance to multiple stresses and ethanol titer compared with ZTW1. A comparative study of the mutant Z3-86 and its parent strain ZTW1 was also performed to elucidate the mechanisms underlying the improved traits of the mutant.

Strains and media
S. cerevisiae strains NY1300 (CICC 1300), NY1308 (CICC 1308), R12, A1-3 (Angel active dry yeast), ZK2 (CCTCC AY2012002), and ZTW1 (CCTCC M2013061) were used in this study. Yeast extract peptone dextrose (YPD) medium (10 g/L yeast extracts, 20 g/L peptone, and 20 g/L glucose, pH 5.5) was used for yeast cultivation. The fermentation medium was made from cassava flour using the double enzyme hydrolysis method. Briefly, a certain amount of cassava flour and water (ratio, 1:2) was mixed and α-amylase (8 U/g cassava) was added to the cassava flour mix. The cassava mash was kept in an autoclave at 95 °C for 4 h for full liquefaction. Then glucoamylase (200 U/g cassava flour) was used to saccharify the mash at 60 °C for 2 h. α-amylase (20,000 U/mL) and glucoamylase (100,000 U/mL) were donated by Novozymes,China. Urea (NH 2 CONH 2 ; Sinopharm, Shanghai, China) of 0.5% concentration was added to the cassava mash as a nitrogen source for the yeast.

Drying process
Yeast cells were cultured in 500 mL YPD at 30 °C at 200 rpm for 24 h with the initial OD (A600) of 0.1. Cells were centrifugally separated and four-fifth of the supernatant was removed. Sorbitan monoestearate and NaOH was added to the yeast cream as described by Garre et al. [10]. A Büchner funnel combined a vacuum pump was used to filtrate the residual supernatant of yeast cream to obtain a yeast cake. Then, the cake was extruded through a sieve (1mm diameter) to obtain yeast strands. The strands were dried in a thermostat drier at 42 °C to loss intracellular water. The dry weight of yeast cells were determined by drying to constant weight at 100 °C for 6 h.

Modified genome shuffling process
Genome shuffling was conducted by treatment with methyl benzimidazole-2-yl-carbamate (MBC) and conduction of three rounds of sporulation and hybridization. MBC is a antimitotic drug and can cause random chromosomal aberration in S. cerevisiae cells [23] and thus used to produce mutants with genomic structural variations from ZTW1 in this study. Parent strain ZTW1 was grown in 25 mL YPD medium [initial OD (A600)=0.05] with 50 μg/mL MBC (Adamas-beta, Shanghai, China) to an OD (A600) of 2 and then collected by centrifugation. Cells (~3×10 8 ) were then transferred into 25 mL sporulation medium (1% NaAc, 0.2% yeast extract, 0.1% glucose, and 0.2% KCl, at pH 6.0) for culture at 28 °C for 5 d, after which the spores (~5×10 8 ) were purified and crossed randomly in 50 mL YPD medium for 36 h. The hybrids (~3×10 9 ) were then centrifuged collected and dried until 92% of their water had been eliminated as described above. The dried cells were cultured in liquid YPD at 30 °C for 2 h for rehydration. The rehydrated cells were spread onto YPD plates containing multiple stressors (8% ethanol (v/v), 0.7 M NaCl, and cultured at 36°C) and the first 120 colonies to appear were randomly selected for ethanol fermentation test. The ten hybrids with the highest ethanol titers were selected for further genome shuffling. Procedures for the two succeeding rounds of sporulation and hybridization were similar to those adopted for the first round.

Fermentation and metabolites
Yeast cells were pre-cultured in 50 mL YPD for 24 h at 30 °C and harvested. Cells were transferred to a cassava mash at a concentration of ~2×10 7 cells/mL. Anaerobic fermentation was performed in 500 mL Erlenmeyer flasks under regular (cassava mash containing 224.5 g/L glucose at 33 °C for 60 h) or stressful (concentration of glucose was increased to 280 g/L and the fermentation temperature was 35 °C) conditions. The concentrations of ethanol, glucose, glycerol, and acetic acid were measured by a high-performance liquid chromatography (HPLC) instrument using an Aminex HPX-87H column (Bio-Rad Laboratories, Hercules, CA, USA) at 60 °C with 5 mmol H 2 SO 4 as the mobile phase at a flow rate of 0.6 ml min −1 . Peaks were detected by a refractive index detector (410 differential refractometer, Waters) and were identified and quantified by comparison to retention times of authentic standards.

Determination of phenotypic stability of the selected mutant
The selected mutant was first grown in 25 mL YPD medium at 30 °C at 200 rpm with the initial cell density of 1.5×10 6 /mL for 18 h. Cells were transferred to a fresh 25 mL YPD (initial cell concentration was 1.5×10 6 /mL) and cultured at 30 °C at 200 rpm for 12. Such passage was repeated for sixty times. The cells from the twentieth, fortieth, and sixtieth generation were diluted and spread onto YPD plates. Five isolates appeared on the plates from each generation were randomly picked for ethanol fermentation and stress tolerance test.

Isobaric tags for relative and absolute quantitation (iTRAQ) and pathway enrichment analysis
Yeast cells were grown in 25 mL YPD medium for 18 h with an initial OD (A600) of 0.05 to the early stationary phase at which the genes related stress response are greatly induced. The protein content of the yeast cells was extracted by the YeastBuster TM Protein Extraction Reagent (Novagen, Darmstadt, Germany) and quantified by a BCA protein assay kit (Pierce, Rockford, IL, USA). Protein samples were reduced, alkylated, digested with trypsin, and labeled with iTRAQ regents according to the standard iTRAQ protocol (Applied Biosystems, Foster City, CA, USA). Mascot 2.3.02 (Matrix Science, London, United Kingdom) software was used to identity proteins by searching the yeast protein database in Saccharomyces Genome Database (SGD).

Physiological and biochemical determination
Yeast cells were grown in 25 mL YPD medium under the same condition for iTRAQ experiment. Cells were collected by centrifugation for physiological and biochemical factors extraction. The composition of fatty acids in the strains was analyzed by gas chromatography using a GC FOCUS instrument equipped with a DSQ II MS detector (Thermo, USA) on a DB-5 MS capillary column (J&W Scientific Inc., Folson, CA, USA) as previously described [21]. Ergosterol content was measured using an HPLC system with a reverse-phase column [24]. Prior to HPLC determination, the alkaline hydrolysis of cells (~3×10 8 ) in an 3 mL alcohol alkali solution at 95 °C was performed for 3 h followed by extraction in n-heptane. Trehalose was extracted from chilled cells (~1×10 8 ) with 3 mL cold 0.5 M trichloroacetic acid (TCA), and the extract was measured for carbohydrate content using the anthrone method [25]. Glutathione (GSH) of yeast cells (~3×10 8 ) was extracted from cells with 2 mL 50% ethanol aqueous solution for 2 h, and the extract was measured using a commercial assay kit (Nanjing Jiancheng Biological Products, Nanjing, China). The contents of ergosterol, trehalose, and glutathione are expressed as mg per g dry weight. Catalase (CAT) and superoxide dismutase (SOD) activities were measured as previously described [25].

Malondialdehyde (MDA) and membrane integrity measurement
Yeast cells (~1×10 8 ) were collected and resuspended in 1 mL of 10% TCA. A vortex mixer was used to break up the cells, and the supernatant was used to detect MDA (lipid peroxidation product) through the thiobarbituric acid-reactive species (TBARS) method [10). Cell membrane integrity was observed using a confocal laser scanning microscope (CLSM) after staining with 20 μg/mL fluorescein diacetate (FDA) and 10 μg/mL propidium iodide (PI) for 0.5 h.

Real-time quantitative PCR (RT-qPCR)
Total RNA from yeast cells samples was extracted using the hot phenol method and then reverse-transcribed into cDNA as previously described [26]. Primer design (Table S1) and RT-qPCR experiments were performed using an ABI Prism 7500 instrument (Applied Biosystems, Foster City, CA, USA). Each sample was tested in triplicate in a 96-well plate (Axygen, Union City, CA, USA) with a final reaction volume of 20 µl using the SYBR ® PrimeScript ® RT-PCR kit (Takara, Dalian, Shandong, China). The expression activity or the relative copy number of genes was quantified and normalized using the method of Vandesompele et al with ALG9, TAF10, UBC6 and TFC1 as the reference gene [27].

Comparison of ethanol fermentation performances and stress tolerances of widely used industrial bioethanol S. cerevisiae strains
Higher ethanol titer and lower byproduct (such as glycerol and acetic acid) formations are preferred traits for bioethanol S. cerevisiae strains. The strains (A1-3, NY1300, NY1308, R12, ZK2, and ZTW1) selected in this study are all industrial S. cerevisiae strains used for bioethanol fermentation. Strain ZTW1, which was first isolated from corn mash in our laboratory and used for industrial-scale bioethanol fermentation since 2006, produced more ethanol than the other strains (Table 1). Another advantageous trait of ZTW1 is its lower production of glycerol (0.0337g/g glucose and 0.0354g/g glucose under regular and stressful condition, respectively; Table 1), a main byproduct of ethanol fermentation. The NY1308 and A1 strains showed better tolerance to dehydration and rehydration than other strains, maintaining cell viabilities of 58.4% and 54.2%, respectively (Table 1). These results demonstrate that while the strains are all used in bioethanol production, they show significantly diverse performances in ethanol fermentation and stress tolerance under different conditions.

WGS to improve the ethanol yields and stress tolerance related to ADY production of ZTW1
Although ZTW1 shows higher ethanol titer than commercial ADY strains and other industrial S. cerevisaie strains, it was initially consider unqualified for application as an ADY because of its low tolerance to dehydration (Table 1). We sought to address this issue by applying a modified WGS technique to the strain. Unlike in previously reported WGS methods, we first used MBC to treat parent cells and produce genomic variations, after which three rounds of sporulation and hybridization were performed to shuffle the variations in the mutant genomes. Table 2 shows that the ethanol yield and tolerance to desiccation of ZTW1 gradually increased during WGS. The mutant Z3-86, which was obtained from the third round of WGS, produced the highest ethanol among the mutants subjected to stressful conditions ( Table 2). The survival ratio of this mutant was 2.15-fold higher than that of ZTW1 after elimination of 92% of its water after drying and slightly higher than that of A1 ( Figure 1A; t test, P < 0.05).
Stress tolerance is crucial for ADY because yeast cells would encounter multiple stressors during ADY production and ethanol fermentation process. Consistent with its improved final ethanol concentration and higher viability after dehydration and rehydration (Table 2), Z3-86 was more resistant to ethanol, osmotic pressure, heat, and H 2 O 2, compared with ZTW1. Figure 1b shows that Z3-86 exhibits viability values higher than those of ZTW1 by 78.1%, 38.5%, 77.8%, and 81.1% after treatment with ethanol, osmotic pressure, heat, and H 2 O 2 , respectively (t test, P < 0.01). Z3-86 also surpassed the viability of A1 with comparative increases in survival rate of 26.7%, 12.5%, 17.1%, and 13.5% after exposure to ethanol, osmotic pressure, heat, and H 2 O 2 , respectively (Figure 1b; t test, P < 0.05). These improvements indicate that Z3-86 is more suitable for ADY production than other industrial strains and demonstrates that, in fact, our breeding strategy improves the multiple stress tolerance of yeast. After subculture of Z3-86 for sixty generations as described in the section of Materials and Methods, no significant differences were observed in ethanol titer and tolerances to drying of the progenies (Table S2), indicating Z3-86 can keep a phenotypic stability during serial propagation process. At present, this mutant is being applied in the pilot-scale production of bioethanol ADY in cooperation with TianGuan Technology Co., Ltd., China. Optimization of the ADY production process and the performance of Z3-86 in industrial ethanol fermentation will be discussed in our future reports.

Proteome analysis to disclose the differently expressed genes between Z3-86 and ZTW1
To investigate how the breeding strategy confers desirable traits to Z3-86, iTRAQ analysis was used to compare the proteomes of this mutant with those of ZTW1. A total of 479 genes showed different expression levels between these two strains (Table S3, P < 0.05). Using RT-qPCR, we determined the expression levels of 20 differently expressed genes (DEGs) detected in the iTRAQ experiment. Most of these genes showed consistent expression patterns at the mRNA and protein levels but with changing ratios (Table S1).   (Table 3; hypergeometric test, P < 0.05). A total of 180 downregulated genes were enriched in pathways for oxidative phosphorylation, endocytosis, RNA transport, and vitamin B6 Figure 1. Viabilities of strains A1, ZTW1, and Z3-86 after drying and application of certain stressors. Ratios of viable cells of the three strains after dehydration described in the Material and methods section (a). The dired cells were then diluted serially and the appropriate dilutions were plated onto solid YPD medium to count the colony-forming. The sample without drying was used as the control. Comparison of the tolerance of A1, ZTW1, and Z3-86 to ethanol treatment [20% (v/v) for 2 h], osmotic pressure (4 M NaCl for 2 h), heat (55 °C for 5 min), and H 2 O 2 (20 mM for 1 h) (b). Yeast cells were precultured in YPD at 30 °C for 18 h with the initial OD(A600) of 0.05. The cells (~3×10 7 ) were collected and transferred into 2 mL YPD (for heat tolerance test) or YPD with indicated stressors. The treated cells were diluted and plated onto solid YPD medium for viability calculation. "a" indicates significant difference between Z3-86 and ZTW1 at P < 0.01; "b" indicates significant difference between Z3-86 and A1 at the P < 0.05 level, using t test. The bars indicated the standard deviation here and other figures.  (Table 3, hypergeometric test, P < 0.05). These results suggest that the modified WGS process developed in this study can alter the global transcription pattern of yeast cells, thereby offering a genetic basis for changing the traits of the mutant strain. In the next section, we discuss the possible mechanisms responsible for the different phenotypes of ZTW1 and Z3-86, mainly focusing on DEGs or physiological factors involved in the protein metabolism, membrane composition, trehalose metabolism, and oxidative response of yeast.

Physiological mechanisms underlying the improved traits of Z3-86 compared with ZTW1
Protein metabolism and growth capability. A fairly large number of upregulated genes in Z3-86 that are located in the ribosome and proteasome pathways, which function in protein synthesis and degradation, respectively. Many genes involved in the biosynthesis of secondary metabolites, such as amino acids and nicotinate, were also upregulated in this mutant strain (Table 3). These findings show that the physiological state of Z3-86 tends to promote cell growth and reproduction. In agreement with this speculation, Z3-86 showed higher proliferative capability than ZTW1 (Figure 2). At the stationary phase (40 h), Z3-86 formed 15.4% more cells than ZTW1 and 7.1% less cells than A1 (Figure 2; t test, P < 0.05), although no significant differences were observed in the dry weights of their biomass (data not shown). These improvements in the growth capability and stress tolerance of Z3-86 are beneficial to ADY production because viable cell count/g yeast solid is an important technical index for finished ADY [7].

Plasma membrane integrity and membrane composition
Previous studies report that the plasma membrane is the main target for injury when yeast cells are exposed to environmental stresses [28]. During dehydration and rehydration, water moving across the membrane can largely alter the cell volume and surface-to-volume ratio of yeast cells. As a response, endocytosis and lipid phase transition occur, making the membrane more unstable. Membrane permeabilization is thought to be a critical event leading to cell death during ADY production [29]. To assess the membrane integrity and viability of ZTW1 and Z3-86 after drying, FDA-PI dual fluorescent staining method was used. If the membrane integrity was lost, PI can enter the cell, and make the cell be colored with red by binding the nucleic acid. Viable cells can convert non-fluorescent FDA into a green fluorescent compound, and make themselves be colored green. Figure 3a shows that most of the cells of ZTW1 and Z3-86 were viable and colored with green before drying. Nevertheless, more than half of the cells of these two strains lost their membrane integrity and were stained with PI after drying (Figure 3b). The proportion of ZTW1 and Z3-86 cells stained with PI was 87% and 75%, respectively, indicating the higher ability of the mutant strain to maintain its membrane integrity (Figure 3b).
Membrane composition was previously reported to be closely related to stress-induced membrane permeabilization and cell death [29,30]. Because lipids are more sensitive to osmotic pressure and more compressible than proteins, osmotic pressure induces fatty acyl chains to pack more tightly, reducing membrane fluidity and endangering its integrity [31]. Fatty acid compositions of ZTW1 and Z3-86 were detected under normal and osmotic conditions (1 M NaCl was used as hypertonic dehydrator). Without osmotic pressure, no distinct differences in the four main fatty acid ratios were observed between the two strains (Table 4). By contrast, application of osmotic pressure increased the contents of unsaturated fatty acids (palmitoleic acid, oleic acid, and linoleic acid) in both strains, resulting in 17.6% and 24.6% increases in the unsaturation indices of ZTW1 and Z3-86, respectively (Table 4; t test, P < 0.05). Z3-86 showed a higher proportion of C 16:1 and a lower proportion of C 18:1 compared with ZTW1 (Table 4; t test, P < 0.05). Higher palmitoleic acid (C 16:1 ) contents may be more conducive to the membrane fluidity of Z3-86 under hyperosmolality stress because this fatty acid has a shorter carbon chain than oleic acid.
Ergosterol has been demonstrated to play a role in membrane stabilization by suppressing the phase transition of phospholipid bilayers [32]. iTRAQ data revealed that two genes involved in ergosterol synthesis, HMG1 (HMG-CoA reductase annotated in the SGD database; catalyzes the conversion of HMG-CoA to mevalonate, which is a rate-limiting step in sterol biosynthesis) and ERG9 (encoding farnesyl-diphosphate farnesyl transferase), were upregulated in Z3-86. Compared with ZTW1, the mutant strain accumulated 14.6% more ergosterol (Figure 4a). Additional ergosterol in the culture medium enhanced ergosterol contents in ZTW1 and Z3-86 by 18.4% and 16.1%, respectively (Figure 3a), and the survival ratio of these two strains was accordingly enhanced by 8.5% and 6.4%, respectively (Figure 4b). These results indicate that an altered membrane composition may partially explain the changed traits of the mutant Z3-86.

Intracellular trehalose content.
Trehalose is present in a wide variety of organisms and can protect proteins and cellular membrane from inactivation and denaturation caused by many stressful conditions. In addition to TSL1 and NTH1 genes, which were detected in the iTRAQ experiments, other four  genes (PGM1, TPS1, TPS2, and TPS3) involved in trehalose metabolism also showed higher expression levels in Z3-86 than in ZTW1, as determined by RT-qPCR (Figure 5a). Osmotic pressure greatly induced trehalose synthesis in both ZTW1 and Z3-86, but the mutant generally accumulated more trehalose than ZTW1 at different time points (Figure 5b). At the 8 h time point, the trehalose contents in ZTW1 and Z3-86 respectively reached 10.8% and 13.5% of their dry weights (Figure 5b). After being cultured in YPD medium with exogenous trehalose, an increased intracellular trehalose content was observed in both the strains, although Z3-86 synthesized more trehalose than ZTW1 (Figure 5c). The survival ratio of ZTW1 and Z3-86 was correspondly increased by 8.3% and 6.8%, respectively, after air-drying (Figure 5d). These results were consistent with the findings of Gadd et al. that trehalose-enriched yeast cells have stronger tolerance to dehydration and desiccation than cells without enrichment [33].
The protective role of trehalose may be embodied by its substitution of water as well as its binding to the polar head groups of phospholipids and proteins to maintain the biological functions of these molecules under stressful conditions [34]. Thus, increased accumulation of intracellular trehalose may contribute to the higher cell viability of Z3-86 after dehydration and rehydration. Oxidative damage and anti-oxidative factors. Reactive oxidative oxygen (ROS) caused by the drying process is another main stress for ADY strains [10,35,36]. ROS is injurious to biomolecules such as DNA and protein and can induce lipid peroxidation in yeast cells [37,38]. Under regular conditions, low levels of MDA resulting from lipid peroxidation formed in both ZTW1 and Z3-86. Dehydration led to a significant increase in MDA levels in both strains, although Z3-86 showed 28.0% lower MDA concentrations than ZTW1 (Figure 6a). The mutant also exhibited superior growth compared with Z1 when grown on YPD plates with H 2 O 2 ( Figure 6b). These results corroborate the conclusion that Z3-86 is more tolerant to oxidative stress than its parental strain.
The radical and peroxide scavenging enzymes (such as SOD and CAT), the thioredoxin system, and the glutathione/ glutaredoxin system help maintain the reduced environment of yeast cell and defend the cell against oxidative stress caused by multiple stressors [39]. As shown in Table 5, compared with ZTW1, Z3-86 has slightly lower SOD and similar CAT activities; these findings suggest that the mutant strain may not show greater antioxidant potential in terms of antioxidant enzymes. iTRAQ data showed that TRX2 gene, which is involved in the thioredoxin system, and GSH2 and GLR1 genes, which are involved in the glutathione/glutaredoxin system, were significantly upregulated in Z3-86 (Table S3). An additional seven genes upregulated in Z3-86 were found to be involved in the pentose phosphate pathway, which functions in NADPH formation and serves as an electron donor for oxidized glutaredoxins and GSH (Table 2 and Table S3). Consistent with these findings, Z3-86 showed 38.2% and 27.5% more GSH than ZTW1 under both normal and osmotic conditions  of sufficient water, only molecular antioxidants such as GSH and trehalose can alleviate oxidative stress when water is insufficient [40]. These molecular antioxidants immobilize the cytoplasm to a glassy state, thereby preventing chemical reactions, molecular diffusion, and conformational changes in biomolecules and protecting cell membranes under water deficiency [41]. These observations suggest that improved antioxidative abilities related to trehalose, thioredoxin, and glutathione production are important mechanisms in the improved multiple stress tolerance of Z3-86.

Conclusion
In the present study, a modified WGS approach was developed to improve the traits of S. cerevisiae ZTW1 related to ADY production. The mutant strain Z3-86 was determined to be qualified as a commercial ADY because of its improved multiple stress tolerance. Comparative analysis of Z3-86 and ZTW1 indicated that the increased protein metabolism capability, membrane integrity, trehalose content, and antioxidative activities of Z3-86 are important mechanisms underlying its altered phenotypes. This work reports the construction of a novel bioethanol ADY for the first time and enriches the current understanding of how WGS improves the complex traits of microbes.
Construction of Novel Bioethanol ADY Stains PLOS ONE | www.plosone.org   Supporting Information Table S1. Primers used to determine the expressions of genes through RT-qPCR. A total of 20 genes were selected to determined their expression levels at mRNA levels using RT-qPCR. These primers were designed using the software Primer Premier 5.0. Most of these genes had consistent expression patterns at the mRNA and protein levels.
(DOC) Table S2. Phenotypic comparison of the progenies of Z3-86 after serial passage. A total of 15 progenies of Z3-86 were selected for ethanol fermentation and survival rate test after drying. There were no significant differences in ethanol titer and tolerances to drying of the progenies.