Metabolic and genomic characterisation of stress-tolerant industrial Saccharomyces cerevisiae strains from TALENs-assisted multiplex editing

Metabolic and genomic characterisation of stress-tolerant industrial Saccharomyces cerevisiae... Abstract TALENs-assisted multiplex editing (TAME) toolbox was previously established and used to successfully enhance ethanol stress tolerance of Saccharomyces cerevisiae laboratory strain. Here, the TAME toolbox was harnessed to improve and elucidate stress tolerances of S. cerevisiae industrial strain. One osmotolerant strain and one thermotolerant strain were selected from the mutant library generated by TAME at corresponding stress conditions, and exhibited 1.2-fold to 1.3-fold increases of fermentation capacities, respectively. Genome resequencing uncovered genomic alterations in the selected stress-tolerant strains, suggesting that cell wall and membrane-related proteins might be major factors behind improved tolerances of yeast to different stresses. Furthermore, amplified mitochondrial DNA might also have an important impact on increased stress tolerance. Unexpectedly, none of predesigned target potential TALENs modification sites showed any genomic variants in sequenced genomes of the selected strains, implicating that the improved stress tolerances might be due to indirect impacts of genome editing via TALENs rather than introducing genomic variants at potential target sites. Our findings not only confirmed TAME could be a useful tool to accelerate the breeding of industrial strain with multiple stress tolerance, but also supported the previous understandings of the complicated mechanisms of multiple stress tolerance in yeast. TALENs, multiplex genome editing, Saccharomyces cerevisiae, industrial strain, stress tolerance, genome resequencing INTRODUCTION As important industrial strain, Saccharomyces cerevisiae has been broadly used in ethanol production due to its general robustness and high productivity (Kavscek et al.2015). To reduce production costs, industrial fermentations prefer to be carried out at high temperature and high concentration of glucose, thus producing high concentration of ethanol (Abdel-Banat et al.2010). However, the industrial fermentation conditions would cause thermal and hyperosmotic stresses and ethanol toxicity to yeast cells, which lead to that massive biological macromolecules are damaged and arrested cell growth or even cell death, eventually reducing ethanol production efficiency. Therefore, it is necessary to improve stress tolerance of S. cerevisiae (Deparis et al.2017). Stress tolerance is a complicated phenotype, which is regulated by dozens of, or even hundreds of, genes simultaneously. For instance, in order to adapt to elevated temperature, S. cerevisiae changes the expression level of thousand genes to regulate cAMP-dependent protein kinase A signalling pathway, cell cycle, lipid metabolism and so on (Jarolim et al.2013; Satomura et al.2016). Previous studies indicate that it is difficult to improve yeast stress tolerance by regulating one or several functional genes (de Nadal, Ammerer and Posas 2011). To overcome this obstacle, many traditional technologies are developed, such as mutagenesis, genome shuffling, adaptive evolution, etc. Sequential mutagenesis has been used to improve multiple stress tolerance in yeast strains (Kumari and Pramanik 2012). To improve yeast stress tolerance, genome shuffling requires at least two strains, and mostly relies on a mutagenised pool of a single strain or a natural pool of strains with different desirable phenotypes (Shi, Wang and Wang 2009; Snoek et al.2015). Adaptive laboratory evolution, which could accumulate desirable genomic and physiological changes in cell populations during long-term selection under specified growth conditions (Dragosits and Mattanovich 2013), is exploited to obtain thermotolerant or ethanol stress-tolerant strains (Caspeta et al.2014; Voordeckers et al.2015). Overall, these traditional approaches are feasible for the breeding of stress-tolerant strains, but laborious and relatively inefficient. Targeted genome editing is an alternative approach for improving desirable traits of yeast, especially TALENs and CRISPR/Cas9 systems which are considered as effective platforms because of their powerful and multiplex genome editing capacities (Alexander 2018). For targeted genome editing, endonuclease FokI or Cas9 generate a double-strand break (DSB) at targeted DNA, which can be repaired by endogenous DNA repair pathways, thus introducing variable-length insertion/deletion mutations or specific sequence alterations (Miller et al.2011; Tsai et al.2014). Recently, we established a TALENs-assisted multiplex editing (TAME) toolbox to target the conserved TATA box and GC box at 66 potential modification sites of S. cerevisiae genome and probably influence 98 genes (Zhang et al.2015), and ethanol stress tolerance of S. cerevisiae laboratory strain was successfully enhanced by TAME. Here, to improve osmotolerance and thermotolerance of S. cerevisiae industrial strain, TAME toolbox was applied to generate a mutant library from an industrial strain ScY01 in several days. After screening at hyperomostic or thermal stress conditions, one osmotolerant strain and one thermotolerant strain were obtained, respectively, and stably exhibited 1.2-fold to 1.3-fold increases of fermentation capacities at corresponding stress conditions. Genome resequencing analysis revealed genomic changes in coding regions affecting the functions of encoded proteins or in intergenic regions probably influencing transcriptional gene expression, which might be genetic factors that resulted in improved stress tolerance of strains. Our results indicated that the TAME toolbox was a useful approach for improving complicated and desirable traits of S. cerevisiae industrial strain, such as stress tolerances, and metabolic and genome characterisation of stress-tolerant strains further confirmed the previously elucidated mechanisms underlying different stress tolerances of yeast. MATERIALS AND METHODS TALENs plasmid construction In our previous study (Zhang et al.2015), plasmids pYES2/CT-GC and p313-GAL-TA (Table 1) with the URA3 and HIS3 autotrophic marker genes, respectively, have been constructed to express the TALENs pair recognizing the GGGCGG and TATAAA sequence, respectively, which allow to induce multiplex editing in the genome of S. cerevisiae laboratory strains. To achieve similar multiplex genome editing in S. cerevisiae industrial strain, two new plasmids pYES2/CT-GGGCGG-GAL1-ZeoR and pRS313-TATAAA-GAL1-KanMX4 were constructed by inserting expression cassettes of the ZeoR and KanMX drug-resistant marker genes into the plasmids pYES2/CT-GC and p313-GAL-TA, respectively, using the ClonExpress® II One Step Cloning Kit (Vazyme Biotech Co., Ltd, China). Specifically, the linear vector of pYES2/CT-GC was PCR amplified using the primer pair pYES2-F/pYES2-R, the ZeoR expression cassette was PCR amplified using the primer pair Zeocin-F/Zeocin-R from the plasmid pIS438 (Sadowski, Lourenco and Parent 2008) and these two PCR fragments were fused together by following the instruction of the kit. Similarly, the linear vector of p313-GAL-TA was PCR amplified using the primer pair pRS313-F/pRS313-R, and the KanMX expression cassette was PCR amplified using the primer pair KanMX4-F/KanMX4-R from the plasmid pUG6 (Guldener et al.1996). The primers used in this study were listed in Table 2. All recombinant plasmids were sequenced by BGI (Shenzhen, China) to verify the insertion of the ZeoR and KanMX expression cassettes. Table 1. Plasmids and S. cerevisiae strains used in this study. Plasmids or strains Description Reference or source Plasmids pYES2/CT-GC 2μ plasmid, autotrophic URA3 gene Zhang et al. (2015) p313-GAL-TA CEN plasmid, autotrophic HIS3 gene Zhang et al. (2015) pYES2/CT-GGGCGG-GAL1-ZeoR 2μ plasmid, ZeoR used for genome editing This study pRS313-TATAAA-GAL1-KanMX4 CEN plasmid, KanMX4, used for genome editing This study Strains ScY01 Evolved thermotolerant strain, diploid Shui et al. (2015) ScY001T Osmotolerant strain, diploid This study ScY033T Thermotolerant strain, diploid This study Plasmids or strains Description Reference or source Plasmids pYES2/CT-GC 2μ plasmid, autotrophic URA3 gene Zhang et al. (2015) p313-GAL-TA CEN plasmid, autotrophic HIS3 gene Zhang et al. (2015) pYES2/CT-GGGCGG-GAL1-ZeoR 2μ plasmid, ZeoR used for genome editing This study pRS313-TATAAA-GAL1-KanMX4 CEN plasmid, KanMX4, used for genome editing This study Strains ScY01 Evolved thermotolerant strain, diploid Shui et al. (2015) ScY001T Osmotolerant strain, diploid This study ScY033T Thermotolerant strain, diploid This study View Large Table 1. Plasmids and S. cerevisiae strains used in this study. Plasmids or strains Description Reference or source Plasmids pYES2/CT-GC 2μ plasmid, autotrophic URA3 gene Zhang et al. (2015) p313-GAL-TA CEN plasmid, autotrophic HIS3 gene Zhang et al. (2015) pYES2/CT-GGGCGG-GAL1-ZeoR 2μ plasmid, ZeoR used for genome editing This study pRS313-TATAAA-GAL1-KanMX4 CEN plasmid, KanMX4, used for genome editing This study Strains ScY01 Evolved thermotolerant strain, diploid Shui et al. (2015) ScY001T Osmotolerant strain, diploid This study ScY033T Thermotolerant strain, diploid This study Plasmids or strains Description Reference or source Plasmids pYES2/CT-GC 2μ plasmid, autotrophic URA3 gene Zhang et al. (2015) p313-GAL-TA CEN plasmid, autotrophic HIS3 gene Zhang et al. (2015) pYES2/CT-GGGCGG-GAL1-ZeoR 2μ plasmid, ZeoR used for genome editing This study pRS313-TATAAA-GAL1-KanMX4 CEN plasmid, KanMX4, used for genome editing This study Strains ScY01 Evolved thermotolerant strain, diploid Shui et al. (2015) ScY001T Osmotolerant strain, diploid This study ScY033T Thermotolerant strain, diploid This study View Large Table 2. Primers used in this study. Name Sequence (5΄ → 3΄) Application pYES2-F ATCCACATGTGTTTTTAGTA Amplifying the linear vector of pYES2/CT-GC without URA3 expression cassette pYES2-R ATCTTGACTGATTTTTCCAT Zeocin-F AAAATCAGTCAAGATCATATGCCCACACACCATAG Amplifying the ZeoR expression cassette Zeocin-R AAAACACATGTGGATGATATCAGCTTGCAAATTAA pRS313-F CCGGGCACGGATTAGAAGCC Amplifying the linear vector of p313-GAL-TA without HIS3 expression cassette pRS313-R GGGATCCACTAGTTCTAGAG KanMX4-F GAACTAGTGGATCCCCCCGGGCGTACGCTGCAGGT Amplifying the KanMX expression cassette KanMX4-R CTAATCCGTGCCCGGCGGCCGATCGATGAATTCGA Name Sequence (5΄ → 3΄) Application pYES2-F ATCCACATGTGTTTTTAGTA Amplifying the linear vector of pYES2/CT-GC without URA3 expression cassette pYES2-R ATCTTGACTGATTTTTCCAT Zeocin-F AAAATCAGTCAAGATCATATGCCCACACACCATAG Amplifying the ZeoR expression cassette Zeocin-R AAAACACATGTGGATGATATCAGCTTGCAAATTAA pRS313-F CCGGGCACGGATTAGAAGCC Amplifying the linear vector of p313-GAL-TA without HIS3 expression cassette pRS313-R GGGATCCACTAGTTCTAGAG KanMX4-F GAACTAGTGGATCCCCCCGGGCGTACGCTGCAGGT Amplifying the KanMX expression cassette KanMX4-R CTAATCCGTGCCCGGCGGCCGATCGATGAATTCGA View Large Table 2. Primers used in this study. Name Sequence (5΄ → 3΄) Application pYES2-F ATCCACATGTGTTTTTAGTA Amplifying the linear vector of pYES2/CT-GC without URA3 expression cassette pYES2-R ATCTTGACTGATTTTTCCAT Zeocin-F AAAATCAGTCAAGATCATATGCCCACACACCATAG Amplifying the ZeoR expression cassette Zeocin-R AAAACACATGTGGATGATATCAGCTTGCAAATTAA pRS313-F CCGGGCACGGATTAGAAGCC Amplifying the linear vector of p313-GAL-TA without HIS3 expression cassette pRS313-R GGGATCCACTAGTTCTAGAG KanMX4-F GAACTAGTGGATCCCCCCGGGCGTACGCTGCAGGT Amplifying the KanMX expression cassette KanMX4-R CTAATCCGTGCCCGGCGGCCGATCGATGAATTCGA Name Sequence (5΄ → 3΄) Application pYES2-F ATCCACATGTGTTTTTAGTA Amplifying the linear vector of pYES2/CT-GC without URA3 expression cassette pYES2-R ATCTTGACTGATTTTTCCAT Zeocin-F AAAATCAGTCAAGATCATATGCCCACACACCATAG Amplifying the ZeoR expression cassette Zeocin-R AAAACACATGTGGATGATATCAGCTTGCAAATTAA pRS313-F CCGGGCACGGATTAGAAGCC Amplifying the linear vector of p313-GAL-TA without HIS3 expression cassette pRS313-R GGGATCCACTAGTTCTAGAG KanMX4-F GAACTAGTGGATCCCCCCGGGCGTACGCTGCAGGT Amplifying the KanMX expression cassette KanMX4-R CTAATCCGTGCCCGGCGGCCGATCGATGAATTCGA View Large Generation and screening of mutant strains by TALENs-mediated multiplex genome editing The procedure to generate and screen stress-tolerant mutants of industrial S. cerevisiae strain with TAME was described in Fig. 1A. To facilitate improving stress tolerances of S. cerevisiae industrial strain by TALENs-mediated multiplex genome editing, the plasmids pYES2/CT-GGGCGG-GAL1-ZeoR and pRS313-TATAAA-GAL1-KanMX4 expressing the TALENs that were designed to recognise the GGGCGG and TATAAA sequence, respectively (Zhang et al.2015) were used to induce multiplex editing in the yeast genome. These two TALENs-expressing plasmids were first transformed into the previously evolved thermotolerant industrial stain ScY01 in our lab (Shui et al.2015), which is able to grow and ferment well at 40°C. The positive transformants were selected on YPD plates (per litre, 10 g yeast extract, 20 g peptone, 20 g glucose and 20 g agar) containing 70 μg mL−1 zeocin and 400 μg mL−1 geneticin G418. To induce TALENs, the transformants with TALENs-expressing plasmids were inoculated into 20 mL YP medium (per litre, 10 g yeast extract, 20 g peptone) containing 10 g L−1 galactose, 70 μg mL−1 zeocin and 200 μg mL−1 geneticin G418 in 50-mL flasks with an initial OD600 of 0.2, cultured at 200 rpm at 30°C for 24 h, and subcultured for serial transfers every 24 h for 7 days, thus allowing accumulation of TALENs-mediated multiplex genomic mutations. The resulting mutant library was further subcultured in liquid YPD medium for serial transfers every 12 h for 10 days to lose the TALENs-expressing plasmids, which could eliminate disturbing effects caused by the bindings of TALENs on the genome. To screen osmotolerant mutants, one aliquot of the mutant library was plated on YP plates with 400 g L−1 glucose producing hyperosmotic stress and grown at 40°C for 3 days until colonies appeared. Total 218 colonies were separately inoculated into 200 μL liquid YP medium containing 400 g L−1 glucose on 96-well plates and grown at 40°C for 48 h. The colony showing the best cell growth in contrast to the parent strain ScY01 and any other colonies from our experiment at hyperosmotic condition was named ScY001T (Table 1). To screen thermotolerant mutants, the other aliquot of the mutant library was plated on YP plates containing 200 g L−1 glucose and grown at 42°C for 4 days. Total 671 colonies were separately inoculated into 200 μL liquid YP medium containing 200 g L−1 glucose on 96-well plates and grown at elevated temperature of 42°C for 24 h. The colony showing the best cell growth in contrast to the parent strain ScY01 and any other colonies from our experiment at thermal stress condition was named ScY033T (Table 1). These two more stress-tolerant industrial S. cerevisiae strains were assessed for subsequent physiological characterisation and functional genomic elucidation. Figure 1. View largeDownload slide Screening of industrial S. cerevisiae by TALENs-assisted multiplex editing (TAME) for improving stress tolerance. (A) Diagram of TAME for improvement of industrial yeast stress tolerance. (B) Screening of the mutant strains for improved osmotolerance at 40°C using YP media with 400 g L−1 glucose. The cell growth of the parent strain ScY01 reached an OD600 of 1.13. (C) Screening of the mutant strains for improved thermotolerance at 42°C using YP media with 200 g L−1 glucose. The cell growth of the parent strain ScY01 reached an OD600 of 0.75. The relative cell growths in (B) and (C) were calculated in contrast to the parent strain ScY01 at the certain stress condition. The red lines in (B) and (C) represent the same cell growth to the parent strain ScY01 at the same test conditions. Figure 1. View largeDownload slide Screening of industrial S. cerevisiae by TALENs-assisted multiplex editing (TAME) for improving stress tolerance. (A) Diagram of TAME for improvement of industrial yeast stress tolerance. (B) Screening of the mutant strains for improved osmotolerance at 40°C using YP media with 400 g L−1 glucose. The cell growth of the parent strain ScY01 reached an OD600 of 1.13. (C) Screening of the mutant strains for improved thermotolerance at 42°C using YP media with 200 g L−1 glucose. The cell growth of the parent strain ScY01 reached an OD600 of 0.75. The relative cell growths in (B) and (C) were calculated in contrast to the parent strain ScY01 at the certain stress condition. The red lines in (B) and (C) represent the same cell growth to the parent strain ScY01 at the same test conditions. Characterisation of fermentation capacity To characterise fermentation capacities of stress-tolerant industrial S. cerevisiae strains, yeast cells on YPD plates were grown in 100-mL flasks containing 50 mL YP media with 200 g L−1 glucose at 30°C overnight (∼15 h). Cells were harvested by centrifugation and then inoculated into fermentation media. The initial OD600 used for fermentations was 0.5. YP medium containing 400 g L−1 glucose was used for testing fermentation capacity at hyperosmotic conditions, and fermentations were performed at 40°C. Fermentations at thermal stress conditions were carried out at 42°C using YP medium containing 200 g L−1 glucose. Fermentations at moderate stress conditions were conducted at thermal stress condition of 40°C using 200 g L−1 glucose, or 37°C using 400 g L−1 glucose, or at normal temperature of 30°C using 400 g L−1 glucose. YP medium containing 200 g L−1 glucose and supplemented with 2%–5% (v/v) or 3%-8% (v/v) ethanol as specified in the text was used for testing fermentation capacity at ethanol stress conditions, and fermentations were conducted at 30°C and 40°C, respectively. Cell growth was monitored by measuring OD600. Concentrations of glucose and ethanol were measured by high-performance liquid chromatography on an Agilent 1260 system (Agilent, USA) equipped with a refractive index detector and a Phenomenex RFQ fast acid column (100 mm × 7.8 mm ID) (Phenomenex Inc., Torrance, CA, USA). The column was eluted with 0.01 N H2SO4 at a flow rate of 0.6 mL min−1 at 55°C. Spot assay for heat-shock survival Heat-shock assays and spot assays of cell survival were performed as described previously with some modifications (Gibney et al.2013; Jarolim et al.2013). Yeast colonies were grown in 10 mL tubes containing 3 mL YP medium with 20 g L−1 glucose at 30°C with shaking at 200 rpm overnight (∼15 h). Cells were harvested by centrifugation, and inoculated into 25 mL YP medium with 200 g L−1 glucose in 50 mL flasks, to achieve an initial OD600 value of 0.2. The cell cultures were grown to early log phase at 30°C with shaking at 220 rpm for 5 h. Two aliquots containing an appreciate amount of cells were harvested and resuspended in 1 mL supernatants to obtain cell suspensions with an OD600 of 5.0. One aliquot of cell suspension was placed on ice as a pre-heat shock control. The other aliquots of cell suspension were transformed to 10 mL tube, and incubated at 50°C for 30 min with shaking at 200 rpm, and immediately chilled on ice for 5 min. Both pre-heat shock and heat-shock aliquots of cell suspensions with an OD600 of 5.0 were diluted to OD600 of 1.0, 0.3, 0.1, 0.03, 0.01, and 5 μL samples at each dilution spotted on YPD plates (YP medium with 20 g L−1 glucose) which were incubated at 30°C for 48 h. Genome resequencing and data analysis Genomic DNA isolation and the sequencing libraries of the osmotolerant strain ScY001T, the thermotolerant strain ScY033T and the parent strain ScY01 were constructed and sequenced on Illumina HiSeq 4000 using 150-bp paired-end sequencing by GeneDenovo Co. (Guangzhou, Guangdong, China). A mean of 21.2 million 150-bp clean reads was generated for each library. The S. cerevisiae S288c genome as a reference was downloaded from RefSeq at the NCBI (sequence assembly version R64, RefSeq assembly accession: GCF_000146045.2) including 16 chromosomes and the mitochondrial genome. The Genome Analysis Toolkit (GATK v3.5) Best Practices pipeline was used to detect single nucleotide polymorphisms (SNPs) and insertion/deletion (InDels) (McKenna et al.2010; DePristo et al.2011). The cleaned reads were mapped to the reference genome using the mapping tools BWA-mem (version 0.7.13) (Li and Durbin 2009), providing an average sequencing depth of ×185 and 98% sequencing coverage for each library. And variants were then called using GATK HaplotypeCaller. For genome analysis, reads with mapping quality ≥30 were included. Initially called SNPs were filtered with a minimum read depth of 10 and a quality score threshold of 20 (Clevenger et al.2015). Variants annotation was performed using the package ANNOVAR (Wang, Li and Hakonarson 2010). The affected genes of detected genic non-synonymous and intergenic variants (SNPs and InDels) in comparing ScY001T with ScY01 as well as ScY033T with ScY01 are available in supplementary material (Supplementary files 1–4), and were further analysed for enrichment in Gene Ontology (GO) terms using DAVID Bioinformatics Resources 6.8 with a P-value cut-off of less than 0.05 and a Benjamini false discovery rate cut-off of less than 0.5 (Huang da, Sherman and Lempicki 2009) (Table S1, Supporting Information). Copy number variants (CNV) corresponding to potential large-scale chromosomal duplications and loss were examined by using CNV-seq package in R (Xie and Tammi 2009). The results were plotted as log-ratio plots with calculated P-values for genomic and mitochondrial genome, respectively. The genome sequencing raw data were deposited in the NCBI Sequence Read Archive under the accession number SRP127218. Determination of cell wall structure and cell membrane integration Cells were grown in 100-mL flasks containing 50 mL YP media with 200 g L−1 glucose at 30°C overnight (∼15 h), and then inoculated into media with an initial OD600 of 0.5 after centrifugation. For treatments at hyperosmotic stress condition, cells were cultured at 400 g L−1 glucose and 40°C for 48 h. For treatments at thermal stress condition, cells were cultured at 200 g L−1 glucose and 42°C for 36 h. As controls, cells were correspondingly cultured at normal condition using 200 g L−1 glucose at 30°C for 48 or 36 h. Cells were then harvested by centrifugation and washed twice with 0.1 M phosphate buffered solution (pH 7.4). Cell pellets were resuspended in phosphate buffered solution. One aliquot of cell suspension was used for monitoring cell wall structure by electron microscope. The other aliquots were used for monitoring membrane integration by flow cytometry. For scanning electron microscopy (SEM) analysis, samples were fixed and treated as previously described (Niu et al.2017). Inspection and photomicrographs were performed with a scanning electron microscope (SU8010, Hitachi, Ltd., Japan) operating at a voltage of 1.0 kV. For transmission electron microscopy (TEM) analysis, samples were fixed and processed as previously reported (Guan et al.2012). After contrasting with uranyl acetate and lead citrate, the samples were examined with transmission electron microscope (HT7700, Hitachi, Ltd., Japan). Determination of membrane integration by flow cytometry analysis was conducted according to the previously reported method (Khan et al.2011). Briefly, cells were stained using propidium iodide (PI), which penetrates cells with severe membrane lesions only, and make the cell be coloured with red by binding the nucleic acid. After staining for 30 min, monoparametric detection of PI fluorescence was performed using a MoFlo XDP flow cytometer (Beckman Coulter, Brea, CA, USA) with a wavelength of 488/620 nm. RESULTS AND DISCUSSION Screening of stress-tolerant industrial S. cerevisiae by TAME The TALENs recognizing the critical and conserved GC box (GGGCGG) and TATA box (TATAAA) was previously designed to target 66 potential TALENs-mediated modification sites in the genome of S. cerevisiae and probably influence 98 genes, thus allowing to induce genomic modification at multiple sites (Zhang et al.2015). By applying this pair of TALENs, a TAME toolbox was established and used to successfully improve ethanol tolerance of S. cerevisiae laboratory strain BY4741 in a short amount of time (Zhang et al.2015). Here, this TAME toolbox was harnessed to improve tolerances of S. cerevisiae industrial strain ScY01 to hyperosmotic and high-temperature stresses. After TAME treatment for seven rounds, a mutant library of the industrial strain ScY01 was generated and subjected to screen stress-tolerant mutants at hyperosmotic and high-temperature stress conditions, respectively (Fig. 1). On the one hand, among total 218 colonies screened at hyperosmotic stress condition, 12.4% of cells showed more than 1.2-fold increase of cell growth in contrast to the parent strain ScY01, whereas 9.6% showed less than 0.8-fold decrease of cell growth than ScY01 (Fig. 1B). The colony with the best cell growth showed a 1.3-fold increase in contrast to the initial strain ScY01, and named ScY001T for further evaluation of osmotolerance and other stress characterisation as well as genomic analysis. On the other hand, among total 671 colonies screened at high-temperature condition, 6.6% of cells showed more than 1.2-fold increase of cell growth than the parent strain ScY01, whereas 33.4% showed less than 0.8-fold decrease of cell growth than ScY01 (Fig. 1C). The colony with the best cell growth showed a 1.4-fold increase in contrast to the parent strain ScY01, and named ScY033T for further evaluation of thermotolerance and other stress characterisation as well as genomic analysis. Physiological characterisation of stress-tolerant industrial S. cerevisiae To evaluate whether fermentation capacities of the selected stress-tolerant mutants were also improved at stress conditions, the selected osmotolerant strain ScY001T and thermotolerant strain ScY033T were compared with the parent strain ScY01 at hyperosmotic and high-temperature conditions, respectively. At hyperosmotic condition caused by high concentration of 400 g L−1 glucose, ScY001T consumed 144.3 ± 1.0 g L−1 glucose in 84 h, which was 1.3 times as much as ScY01 (Fig. 2A). Meanwhile, ScY001T produced 61.9 ± 0.3 g L−1 ethanol in 84 h, which was 1.2 times more than that of ScY01. On the other hand, ScY033T consumed 108.9 ± 2.6 g L−1 glucose in 84 h at 42°C, which was 1.2 times as much as ScY01. Meanwhile, ScY033T produced 46.3 ± 0.2 g L−1 ethanol in 84 h, which was 1.3 times more than that of ScY01 (Fig. 2B). These results indicated that the selected stress-tolerant strains could also obtain enhanced fermentation capacities at stress conditions. However, glucose consumption of both ScY001T and ScY01 got slow after fermenting for 72 h at hyperosmotic condition, leading to more than half of the residual glucose left in the media. Both ScY033T and ScY01 consumed only a little more than half of the glucose at higher temperature (42°C). These results might be due to the combination inhibitory effects of high-temperature and hyperosmotic stresses, since fermentations at hyperosmotic condition were performed at 40°C instead of normal temperature of 30°C and fermentations at thermal stress condition were conducted using a relatively high concentration of glucose of 200 g L−1 glucose and a higher temperature of 42°C. To verify this assumption, we compared the fermentation capacities of the stress-tolerant and parent strains at moderate stress conditions (Fig. S1, Supporting Information). The parent strain ScY01 has been previously evolved to grow and ferment well at 40°C (Shui et al.2015). Thus, when fermentations were conducted using 200 g L−1 glucose at 40°C, the stress-tolerant strains ScY001T and ScY033T showed similar fermentation capacities to the parent strain ScY01 and completely consumed glucose within 48 h (Fig. S1A, Supporting Information). When fermentations at hyperosmotic condition were performed at a moderate temperature of 37°C and a normal temperature of 30°C (Fig. S1B and C, Supporting Information), much more glucose was consumed in contrast to 42°C (Figs 2A and 3B). Furthermore, the osmotolerant strain ScY001T obviously consumed glucose more quickly than ScY01 and ScY033T at 37°C (Fig. S1B, Supporting Information). By contrast, at 30°C, ScY001T consumed glucose slightly faster at the beginning of fermentation but eventually slower than ScY01 and ScY033T (Fig. S1C, Supporting Information). Therefore, these results confirmed the combination inhibitory effects of high-temperature and hyperosmotic stresses. Figure 2. View largeDownload slide Physiological characterisation of industrial S. cerevisiae mutants with improved stress tolerance. (A) Characterisation of osmotolerant industrial S. cerevisiae mutants. Fermentations of the osmotolerant strain ScY001T and parent strain ScY01 were performed at 40°C using YP media with 400 g L−1 glucose. (B) Characterisation of thermotolerant industrial S. cerevisiae mutants. Fermentations of the thermotolerant strain ScY033T and parent strain ScY01 were performed at 42°C using YP media with 200 g L−1 glucose. Initial OD600 of 0.5 was used for all the fermentations. Data represent the mean and standard error of duplicate cultures at each condition. Figure 2. View largeDownload slide Physiological characterisation of industrial S. cerevisiae mutants with improved stress tolerance. (A) Characterisation of osmotolerant industrial S. cerevisiae mutants. Fermentations of the osmotolerant strain ScY001T and parent strain ScY01 were performed at 40°C using YP media with 400 g L−1 glucose. (B) Characterisation of thermotolerant industrial S. cerevisiae mutants. Fermentations of the thermotolerant strain ScY033T and parent strain ScY01 were performed at 42°C using YP media with 200 g L−1 glucose. Initial OD600 of 0.5 was used for all the fermentations. Data represent the mean and standard error of duplicate cultures at each condition. Figure 3. View largeDownload slide Physiological characterisation of the osmotolerant strain ScY001T and the thermotolerant strain ScY033T at other stress conditions. (A) Fermentation profiles of the osmotolerant strain ScY001T and parent strain ScY01 at thermal stress condition of 42°C using YP media with 200 g L−1 glucose. (B) Fermentation profiles of the thermotolerant strain ScY033T and parent strain ScY01 at hyperosmotic condition at 40°C using YP media with 400 g L−1 glucose. Initial OD600 of 0.5 was used for all the fermentations. Data represent the mean and standard error of duplicate cultures at each condition. Figure 3. View largeDownload slide Physiological characterisation of the osmotolerant strain ScY001T and the thermotolerant strain ScY033T at other stress conditions. (A) Fermentation profiles of the osmotolerant strain ScY001T and parent strain ScY01 at thermal stress condition of 42°C using YP media with 200 g L−1 glucose. (B) Fermentation profiles of the thermotolerant strain ScY033T and parent strain ScY01 at hyperosmotic condition at 40°C using YP media with 400 g L−1 glucose. Initial OD600 of 0.5 was used for all the fermentations. Data represent the mean and standard error of duplicate cultures at each condition. During the industrial production process, S. cerevisiae often suffers from multiple stresses. Thus, strains resistant to multiple stresses would be desirable for their industrial applications. To test the physiological response of the selected stress-tolerant strains to other stress conditions, fermentation capacity of the selected osmotolerant strain ScY001T was evaluated at thermal stress condition, while fermentation capacity of the selected thermotolerant strain ScY033T was detected at hyperosmotic condition. Furthermore, both ScY001T and ScY033T were evaluated for fermentation capacities at other stress conditions. At thermal stress condition, osmotolerant strain ScY001T consumed 144.3 ± 1.0 g L−1 glucose and produced 39.5 ± 0.2 g L−1 ethanol in 60 h at 42°C, which were 1.1 times as much as that of ScY01 (Fig. 3A). At hyperosmotic stress condition, thermotolerant strain ScY033T consumed 117.3 ± 4.1 g L−1 glucose and produced 43.6 ± 2.6 g L−1 ethanol in 84 h, which were 1.1 times as much as that of ScY01 (Fig. 3B). These results suggested that the selected osmotolerant strain might acquire slightly increased thermotolerance and the selected thermotolerant strain might also acquire slightly increased osmotolerance. To evaluate the ethanol tolerance of ScY001T and ScY033T, we performed fermentations using media supplemented with 0% (v/v), 3%(v/v), 6% (v/v) or 8% (v/v) ethanol concentration at 30°C (Fig. S2, Supporting Information) and with 0% (v/v), 2% (v/v), 3% (v/v) or 5% (v/v) ethanol concentration at 40°C (Fig. S3, Supporting Information). At normal condition using 200 g L−1 glucose at 30°C, glucose was depleted within 18 h, and ScY001T and ScY033T exhibited similar fermentation capacities to ScY01 (Fig. S2A, Supporting Information). With the increasing concentration of ethanol supplemented in media, fermentation capacities of these strains were gradually hampered at either 30°C or 40°C (Figs S2 and S3, Supporting Information), and completely inhibited at the condition of 5% (v/v) ethanol at 40°C. Overall, both ScY001T and ScY033T did not exhibit greatly different fermentation capacities from parent ScY01 at ethanol stress condition (Figs S2 and S3, Supporting Information), indicating that the screening process of both osmotolerant and thermotolerant mutants might have no significant impact on enrichment of ethanol-tolerant mutants. Furthermore, the physiological responses of the selected stress-tolerant strains to short-term heat shock were monitored by measuring cell viability through spot assay before and after heat shock treatment. As for cells without heat shock treatment, both the selected stress-tolerant strains ScY001T and ScY033T showed similar cell viabilities to the parent strain ScY01 (Fig. 4A). Inherently, both ScY001T and ScY033T acquired increased long-term thermotolerance at prolonged stress conditions (Figs 2 and 3), but they both showed dramatically decreased cell viabilities after heat shock treatment in contrast to ScY01 (Fig. 4B). This result implicated that there might be a trade-off mechanism for acquiring long-term stress tolerance and maintaining short-term stress tolerance, such as heat shock response. The heat shock response is appropriately considered to be evolutionarily selected to prevent damage caused by an anticipated stress rather than to promote recovery from an existing insult (Verghese et al.2012). In contrast to potential lethal damages caused by short-term severe stresses, long-term moderate stresses lead to sustained but nonlethal impacts on physiological activities of cells, which might provide cells a chance to be adapted to stress conditions. It has been reported that adaptively evolved thermotolerant S. cerevisiae strain show trade-off at ancestral temperatures and preadaptation to other stresses (Caspeta and Nielsen 2015), and even expressed heat stress response at normal temperature of 30°C (Caspeta, Chen and Nielsen 2016). Furthermore, S. cerevisiae has distinct regulatory mechanism of thermotolerance at long-term thermal stress from the heat shock response at short-term thermal stress (Shui et al.2015). Figure 4. View largeDownload slide The physiological response of the osmotolerant strain ScY001T and the thermotolerant strain ScY033T to heat shock treatment. (A) Spot assay of cells before heat shock treatment. (B) Spot assay of cells after heat shock treatment. Strains ScY001T, ScY033T and ScY01 were grown in biological duplicates to early-log phase in YP medium containing 20 g L−1 glucose. Cells were harvested and resuspended to 5.0 OD600. One aliquot of the 5.0 OD600 cell suspension without heat shock treatment and the other aliquot after treated at 50°C for 30 min were serially diluted to the OD600 indicated and spotted on YPD plates after. The plates were incubated for 2 days at 30°C and imaged. Figure 4. View largeDownload slide The physiological response of the osmotolerant strain ScY001T and the thermotolerant strain ScY033T to heat shock treatment. (A) Spot assay of cells before heat shock treatment. (B) Spot assay of cells after heat shock treatment. Strains ScY001T, ScY033T and ScY01 were grown in biological duplicates to early-log phase in YP medium containing 20 g L−1 glucose. Cells were harvested and resuspended to 5.0 OD600. One aliquot of the 5.0 OD600 cell suspension without heat shock treatment and the other aliquot after treated at 50°C for 30 min were serially diluted to the OD600 indicated and spotted on YPD plates after. The plates were incubated for 2 days at 30°C and imaged. Taken together, aiming to further improve fermentation capacities at multiple stress conditions, more efforts of strain breeding remained to be undertaken. TAME toolbox would be a promising approach to accelerate the breeding of multiple stress-tolerant strains, especially by performing several rounds of TAME treatment and switching the screening conditions between different stresses. Functional genomic elucidation of stress-tolerant industrial S. cerevisiae To uncover the genomic alterations behind the increased stress tolerances of the selected strains, both mutant strains ScY001T and ScY033T as well as the parent strain ScY01 were subjected to whole-genome resequencing. By comparing with the reference genome of strain S228c, the common and different SNPs and InDels were detected in ScY001T versus ScY01 as well as ScY033T versus ScY01 (Table 3). Among the different SNPs and InDels between ScY001T and ScY01, total 200 genes with non-synonymous variants were annotated to result in mutant or non-functional proteins (Supplementary file 1); meanwhile, total 709 genes with intergenic variants were predicted to be modulated in terms of gene expressions in the osmotolerant mutant ScY001T (Table 3, Supplementary file 2). Among the different SNPs and InDels between ScY033T and ScY01, total 205 genes with non-synonymous variants were annotated to result in mutant or non-functional proteins (Supplementary file 3); meanwhile, total 707 genes with intergenic variants were predicted to be altered in terms of gene expressions in the thermotolerant mutant ScY033T (Table 3, Supplementary file 4). Remarkably, a majority of these variants were classified as heterozygous in either or both of stress-tolerant strains and the parent strain, where more than one allelic variant was genotyped by genome sequencing (Supplementary file 1–4). This might be due to the diploid nature of these strains (Shui et al.2015). Similarly, a large portion of differential variants were also observed to be heterozygous in another industrial diploid strain of S. cerevisiae strain Ethanol Red and its evolved strain ISO12 (Wallace-Salinas et al.2015). However, the difference was that the percentage of heterozygous variants slightly increased in the previously reported evolved strain ISO12 (Wallace-Salinas et al.2015), but obviously decreased in the stress-tolerant strains ScY001T and ScY033T in this study (Supplementary file 1–4). It seemed that more alleles became homozygous during the TAME treatment, which might be beneficial to eliminate the deleterious effects of heterozygous mutations causing haploinsufficiency in diploid strains (Deutschbauer et al.2005). For instance, Ste50, which is an adaptor protein for osmosensory signalling pathway (Jansen et al.2001), was detected to have a heterozygous mutation of D109G in the parent strain ScY01 and become homozygous wild-type Ste50 in the osmotolerant strain ScY001T (Supplementary file 1). It has been reported that mutations in the SAM domain of Ste50 at position 30 to 104 cause signalling defects in the pathways for mating, filamentous growth and osmotolerance (Jansen et al.2001). Here, the mutation D109G is localised near the SAM domain, implicating its adverse effect on osmotolerance, which was eliminated because of the homozygous wild-type Ste50 in ScY001T. Furthermore, compared with the stress-tolerant mutants derived from industrial diploid strains, the evolved mutants derived from laboratory haploid strains were not reported to have heterozygous mutations (Caspeta et al.2014; Satomura et al.2016). Thus, the identified point mutations affecting the functions of ERG3 and CDC25 could be easily verified for their positive effects on stress tolerance by reconstructing them in the parent strains (Caspeta et al.2014; Satomura et al.2016). By contrast, it would be difficult to verify the effects of the identified genomic mutations on stress tolerance in industrial diploid strains due to their homo-/heterozygosity, but worthwhile to make further efforts in the future. Table 3. Results of the variant calling and analysis. Variants (SNPs & InDels) Count ScY001T vs. ScY01 ScY033T vs. ScY01 General statistics Total number (vs. S288c) 69487 & 7010 69529 & 6991 Background variants 67782 & 6496 67811 & 6493 Mutant variants 1705 & 514 1718 & 498 Location of mutant variants Intergenic variants 812 & 410 838 & 387 Coding region variants 861 & 94 846 & 102 Unknown 32 & 10 34 & 9 Effect of coding region variants Synonymous effects 480 & 0 466 & 0 Non-synonymous effects 381 & 94 380 & 102 Genes with non-synonymous effects 177 & 48 a 178 & 49 b Genes with intergenic variants 380 & 485 c 401 & 474 d Variants (SNPs & InDels) Count ScY001T vs. ScY01 ScY033T vs. ScY01 General statistics Total number (vs. S288c) 69487 & 7010 69529 & 6991 Background variants 67782 & 6496 67811 & 6493 Mutant variants 1705 & 514 1718 & 498 Location of mutant variants Intergenic variants 812 & 410 838 & 387 Coding region variants 861 & 94 846 & 102 Unknown 32 & 10 34 & 9 Effect of coding region variants Synonymous effects 480 & 0 466 & 0 Non-synonymous effects 381 & 94 380 & 102 Genes with non-synonymous effects 177 & 48 a 178 & 49 b Genes with intergenic variants 380 & 485 c 401 & 474 d a aThe total number of genes with non-synonymous SNPs or/and InDels in ScY001T vs. ScY01 is 200. b bThe total number of genes with non-synonymous SNPs or/and InDels in ScY033T vs. ScY01 is 205. c cThe total number of genes with intergenic SNPs or/and InDels in ScY001T vs. ScY01 is 709. d dThe total number of genes with intergenic SNPs or/and InDels in ScY033T vs. ScY01 is 707. View Large Table 3. Results of the variant calling and analysis. Variants (SNPs & InDels) Count ScY001T vs. ScY01 ScY033T vs. ScY01 General statistics Total number (vs. S288c) 69487 & 7010 69529 & 6991 Background variants 67782 & 6496 67811 & 6493 Mutant variants 1705 & 514 1718 & 498 Location of mutant variants Intergenic variants 812 & 410 838 & 387 Coding region variants 861 & 94 846 & 102 Unknown 32 & 10 34 & 9 Effect of coding region variants Synonymous effects 480 & 0 466 & 0 Non-synonymous effects 381 & 94 380 & 102 Genes with non-synonymous effects 177 & 48 a 178 & 49 b Genes with intergenic variants 380 & 485 c 401 & 474 d Variants (SNPs & InDels) Count ScY001T vs. ScY01 ScY033T vs. ScY01 General statistics Total number (vs. S288c) 69487 & 7010 69529 & 6991 Background variants 67782 & 6496 67811 & 6493 Mutant variants 1705 & 514 1718 & 498 Location of mutant variants Intergenic variants 812 & 410 838 & 387 Coding region variants 861 & 94 846 & 102 Unknown 32 & 10 34 & 9 Effect of coding region variants Synonymous effects 480 & 0 466 & 0 Non-synonymous effects 381 & 94 380 & 102 Genes with non-synonymous effects 177 & 48 a 178 & 49 b Genes with intergenic variants 380 & 485 c 401 & 474 d a aThe total number of genes with non-synonymous SNPs or/and InDels in ScY001T vs. ScY01 is 200. b bThe total number of genes with non-synonymous SNPs or/and InDels in ScY033T vs. ScY01 is 205. c cThe total number of genes with intergenic SNPs or/and InDels in ScY001T vs. ScY01 is 709. d dThe total number of genes with intergenic SNPs or/and InDels in ScY033T vs. ScY01 is 707. View Large Based on GO Cellular Component enrichment analysis (Huang da, Sherman and Lempicki 2009), the affected genes of non-synonymous variants in both ScY001T and ScY033T showed similar enrichments, and were mostly enriched for integral component of membrane and cell wall-related proteins (Fig. 5A, Table S1, Supporting Information). Besides the enrichment for cell wall and membrane-related components, some affected genes were enriched for vacuole. GO Molecular Function enrichment analysis indicated that the affected genes of non-synonymous variants in ScY001T were specifically enriched for structural constituent of cell wall, while those in ScY033T were specifically enriched for receptor activity proteins (Fig. 5A, Table S1, Supporting Information). Furthermore, according to GO Biological Process enrichment analysis, only the affected genes of non-synonymous variants in ScY001T were found to be enriched in biological processes including fungal-type cell wall organisation, transmembrane transport and flocculation-related processes proteins (Fig. 5A; Table S1, Supporting Information). Membrane-associated stress proteins include not only chaperones but also other proteins (Horvath et al.2008). Wallace-Salinas et al. (2015) reported that non-synonymous variants were significantly enriched in GO terms related to cell periphery, membranes and cell wall during the adaptive evolution of an industrial S. cerevisiae strain to combined heat and hydrolysate stress. Our observations were consisted with the previous findings, suggesting that cell wall and membrane-related proteins might be major genomic targets behind improved tolerances of industrial S. cerevisiae strain to different stresses. Figure 5. View largeDownload slide Enriched Gene Ontology (GO) terms of genes influenced by genic nonsynonymous (A) and intergenic (B) variants in mutant strains ScY001T (blue line) and ScY033T (red line). The affected genes of detected genic non-synonymous and intergenic variants (SNPs and InDels) in comparing ScY001T with ScY01 as well as ScY03T with ScY01 were analysed for enrichment in GO terms using DAVID Bioinformatics Resources 6.8 (Huang da et al.2009). The percentage associated with each GO terms including GO Biological Process (BP), GO Cellular Component (CC), GO Molecular Function (MF) was calculated as the percentage of genes involved in the corresponding GO terms among the pool of genes that were influenced by genic non-synonymous or intergenic variants. Figure 5. View largeDownload slide Enriched Gene Ontology (GO) terms of genes influenced by genic nonsynonymous (A) and intergenic (B) variants in mutant strains ScY001T (blue line) and ScY033T (red line). The affected genes of detected genic non-synonymous and intergenic variants (SNPs and InDels) in comparing ScY001T with ScY01 as well as ScY03T with ScY01 were analysed for enrichment in GO terms using DAVID Bioinformatics Resources 6.8 (Huang da et al.2009). The percentage associated with each GO terms including GO Biological Process (BP), GO Cellular Component (CC), GO Molecular Function (MF) was calculated as the percentage of genes involved in the corresponding GO terms among the pool of genes that were influenced by genic non-synonymous or intergenic variants. Furthermore, variations in non-coding regions have been investigated to have an important impact on phenotypic diversity given their influence on gene expression level (Connelly et al.2013). Based on GO term enrichment analysis, the affected genes of intergenic variants in ScY033T and ScY001T were also found to have enrichments for cell wall-related cellular components, molecular functions and biological processes (Fig. 5B, Table S1, Supporting Inormation), suggesting that gene expression of cell wall-related proteins might be reshaped in the stress-tolerant strains due to these intergenic variants. Remarkably, both the affected genes of intergenic variants in ScY001T and ScY033T were enriched for the biological process of response to stress, and those in ScY001T had a specific enrichment for cellular components of cytoplasmic stress granule. These results implicated that altering gene expression of stress response-related genes might be another genomic targets behind improved tolerances of industrial S. cerevisiae strain to different stresses. Satomura et al. (2016) also reported that evolved thermotolerant strains showed highly upregulated gene expression involved in response to stress and heat, due to a CDC25 point mutation that led to the downregulation of the cAMP-dependent protein kinase A signalling pathway. However, the effects of intergenic variants were too complicated to be verified. DNA CNV—amplification or deletion of DNA segments—is also an important source of genetic variation, changing the original number of DNA copies that could lead to phenotypic variations. Thus, CNVs of the selected stress-tolerant strains were assessed for not only nuclear genomic DNA but also mitochondrial genomic DNA (mtDNA) (Figs 6 and 7). Amplification-type CNVs seemed to be prevalent on mtDNA of ScY001T (Fig. 6A). Distribution histograms of amplification-type (blue in Fig. 6B) and deletion-type (red in Fig. 6B) CNVs clearly displayed that amplification-type and deletion-type CNVs were equally distributed across nuclear chromosomes while the majority of CNVs were amplification-type on mtDNA (Fig. 6B). Similarly, amplification-type CNVs of ScY033T were also observed on mtDNA instead of nuclear chromosomes (Fig. 7). These observations implicated that the stress-tolerant mutants ScY001T and ScY033T might have more mtDNA copies or more mitochondria than the parent strain ScY01. Mitochondria are the sites for producing ATP through respiration, and also seen to regulate nuclear gene expression and cellular functions (Whelan and Zuckerbraun 2013). Previous study reported that ATP is an important factor for yeast cells to maintain normal physiological levels at stress conditions (Postmus et al.2011). The mechanism underlying mtDNA amplifications in the stress-tolerant strains remained to be discovered. Figure 6. View largeDownload slide Copy number variation (CNV) between ScY001T and ScY01 in nuclear and mitochondrial genomes. (A) A log2 ratio plot for CNV in mitochondrial genome generated by CNV-seq. The red and blue coloured plots represent amplification-type and deletion-type CNVs, respectively. The red and blue colour gradients represent the P-value calculated on each of the ratios where an increase in brightness shows a decrease in P-value. (B) Log2 ratio distribution for CNVs on nuclear genome (left) and mitochondrial genome (right). Figure 6. View largeDownload slide Copy number variation (CNV) between ScY001T and ScY01 in nuclear and mitochondrial genomes. (A) A log2 ratio plot for CNV in mitochondrial genome generated by CNV-seq. The red and blue coloured plots represent amplification-type and deletion-type CNVs, respectively. The red and blue colour gradients represent the P-value calculated on each of the ratios where an increase in brightness shows a decrease in P-value. (B) Log2 ratio distribution for CNVs on nuclear genome (left) and mitochondrial genome (right). Figure 7. View largeDownload slide Copy number variation (CNV) between ScY033T and ScY01 in nuclear and mitochondrial genomes. (A) A log2 ratio plot for CNV in mitochondrial genome generated by CNV-seq. The red and blue coloured plots represent amplification-type and deletion-type CNVs, respectively. The red and blue colour gradients represent the P-value calculated on each of the ratios where an increase in brightness shows a decrease in P-value. (B) Log2 ratio distribution for CNVs on nuclear genome (left) and mitochondrial genome (right). Figure 7. View largeDownload slide Copy number variation (CNV) between ScY033T and ScY01 in nuclear and mitochondrial genomes. (A) A log2 ratio plot for CNV in mitochondrial genome generated by CNV-seq. The red and blue coloured plots represent amplification-type and deletion-type CNVs, respectively. The red and blue colour gradients represent the P-value calculated on each of the ratios where an increase in brightness shows a decrease in P-value. (B) Log2 ratio distribution for CNVs on nuclear genome (left) and mitochondrial genome (right). In our previous study (Zhang et al.2015), ethanol tolerance of a laboratory strain BY4741 was rapidly improved by coupling TAME-accelerated genome evolution with the screening method of using ethanol as selective stress. However, when applying the same strategy to the industrial strain ScY01, we did not obtain any mutations with enhanced ethanol tolerance (data not shown). These differential results implicated that the consequence of TALENs-mediated genome editing in the industrial strain might be different from that in the laboratory strain. Furthermore, genome sequencing of the ethanol-tolerant laboratory strain showed that 7 sites (10.6%) of a total of 66 potential TALENs modification sites contained the TALENs-induced InDels (Zhang et al.2015). Unexpectedly, in this study, none of potential predesigned TALENs modification sites were found to have any genomic variants in sequenced genomes of ScY001T and ScY033T. Thus, the improved stress tolerances of TAME-treated cells might be due to non-target genomic modifications (Zhang et al.2015) and indirect impacts of genome editing via TALENs rather than introducing genomic variants at potential predesigned target sites. Compared with S. cerevisiae laboratory strains, the genome editing via CRISPR/Cas9 in an industrial strain is relatively difficult to be established because of low transformation and genome editing efficiencies (Stovicek, Borodina and Forster 2015). Similar difficulties might exist in genome editing via TALENs in an industrial strain. Furthermore, it has been reported that oligonucleotide-directed gene editing activates damage response pathway and replication fork stress in mammalian cells (Bonner et al.2012). Although there were no genomic variants introduced into industrial strains by genome editing via TALENs, DSBs generated by TALENs might induce DNA replication stress thereby triggering stress responses of yeast cells. Alteration of cell wall structure and cell membrane integration in stress-tolerant strains When cells are exposed to various environmental stresses, cell wall and membrane act as the first barrier against external stresses. Some recent demonstrated that cell wall remodelling and alteration of membrane composition and structure seem to be the primary mechanisms required for protection against cell damage in stress-tolerant strains (Caspeta et al.2014; Wallace-Salinas et al.2015; Kitichantaropas et al.2016). Furthermore, genomic variants of stress-tolerant strains in this study converged on genes related to cell wall, cell periphery and cell membrane (Fig. 5, Supplementary file 5). Thus, we determined and compared cell wall structure and cell membrane integration of stress-tolerant strains ScY001T and ScY033T with the parent strain ScY01 at hyperosmotic and thermal stress conditions using electron microscope and flow cytometry. Based on the inspection by SEM, the stress-tolerant strains, especially ScY033T, had more rough surfaces than ScY01 at hyperosmotic stress condition (Fig. 8A), and TEM analysis further recognised fimbriate cell surfaces of ScY001T and ScY033T in comparison with ScY01 (Fig. S4, Supporting Information). On the other hand, at thermal stress condition, ScY001T and ScY033T showed slightly more surfaces than ScY01 according to SEM analysis (Fig. 8A), and their relatively short hairy surfaces were observed by using TEM analysis (Fig. S4, Supporting Information). These results suggested that cell walls of the stress-tolerant strains ScY001T and ScY033T might be remodeled in comparison with the parent strain ScY01, leading to a more robust wall to protect cells against hyperosmotic or thermal stresses. Furthermore, genomic variants related to cell wall might be responsible for cell wall remodelling of the stress-tolerant strains. For instance, compared with ScY01, both ScY001T and ScY033T had non-synonymous variants related to cellular surface properties such as adhesion (AGA1, FIG2), and biofilm and flocculation (FLO1, FLO5, FLO9) (Supplementary file 5), which were also observed to be mutant in previously reported thermotolerant strain (Wallace-Salinas et al.2015). Besides, SED1 encoding major stress-induced structural GPI-cell wall glycoprotein also showed non-synonymous mutations in ScY001T and ScY033T (Supplementary file 5). Interestingly, 16 and 14 of 24 PAU genes encoding structural constituent of cell wall, which might be differentially induced and possess specific roles for the adaptation of S. cerevisiae to certain environmental stresses (Luo and van Vuuren 2009), were found to have intergenic mutations in ScY001T and ScY033T, respectively (Supplementary file 5). This observation further confirmed that cell wall remodelling of ScY001T and ScY033T might be due to these genomic variants. In addition, the same stress-tolerant strain showed more rough cell surface at hyperosmotic stress condition than at thermal stress condition, indicating more activated cell wall remodelling at hyperosmotic stress condition. Figure 8. View largeDownload slide Scanning electronic microscope images (A) and percentage of PI-stained cells (B) of the stress-tolerant (ScY001T and ScY033T) and parent (ScY01) strains at different conditions. Hyperosmotic stress condition was performed using 400 g L−1 glucose at 40°C, and cells were cultured for 48 h. Thermal stress condition was conducted at 42°C using 200 g L−1 glucose, and cells were cultured for 36 h. As controls, cells were correspondingly cultured at normal condition using 200 g L−1 glucose at 30°C for 48 or 36 h. The percentage of PI-stained cells was analysed and calculated by flow cytometer. Data represent the mean and standard error of duplicate cultures at each condition. Figure 8. View largeDownload slide Scanning electronic microscope images (A) and percentage of PI-stained cells (B) of the stress-tolerant (ScY001T and ScY033T) and parent (ScY01) strains at different conditions. Hyperosmotic stress condition was performed using 400 g L−1 glucose at 40°C, and cells were cultured for 48 h. Thermal stress condition was conducted at 42°C using 200 g L−1 glucose, and cells were cultured for 36 h. As controls, cells were correspondingly cultured at normal condition using 200 g L−1 glucose at 30°C for 48 or 36 h. The percentage of PI-stained cells was analysed and calculated by flow cytometer. Data represent the mean and standard error of duplicate cultures at each condition. Besides cell wall remodelling, cell membrane integration might be also modified due to genomic variants in ScY001T and ScY033T (Supplementary file 5). Flow cytometric monitoring of PI uptake has been well established to inspect cell membrane integration (Davey and Hexley 2011). Under control conditions without stress, percentages of PI-stained cells were lower than 5% for both the stress-tolerant strains ScY001T and ScY033T and the parent strain ScY01 (Fig. 8B). By contrast, percentage of PI-stained cells of ScY01 increased to 16.6 ± 0.5% at hyperosmotic stress condition and 19.6 ± 0.2% at thermal stress condition, respectively. Compared with ScY01, ScY001T showed slightly lower percentage of PI-stained cells at hyperosmotic stress condition and apparently lower percentage (10.9 ± 0.1%) of PI-stained cells at thermal stress condition. On the other hand, ScY033T showed significantly lower percentage (6.0 ± 0.1%) of PI-stained cells at hyperosmotic stress condition but similar percentage of PI-stained cells at thermal stress condition in contrast to ScY01. These results suggested that the stress-tolerant strains might develop more robust cell membranes to different extent. Meanwhile, many genes (Supplementary file 5), which encode membrane-associated proteins, membrane transporters, integral component of membrane, etc., were found to have non-synonymous variants in the stress-tolerant strains ScY001T and ScY033T, which might be beneficial to cell membrane integration at stress conditions. Overall, the stress-tolerant strains ScY001T and ScY033T seemed to remodel cell wall and alter cell membrane integration to different extent by developing related genomic variants, thus protecting cells from adverse stresses. Further efforts, however, are required to clarify the precise molecular mechanisms underlying stress tolerance of these strains. In the future, further improvement of transformation and genome editing efficiencies of S. cerevisiae industrial strains would be beneficial to the application of the TAME toolbox in the breeding of multiple stress-tolerant strains. TAME-introduced genomic mutations were generated and accumulated when the TALENs-induced DSBs were being fixed through the NHEJ (nonhomologous end joining) pathway of DSB repair. In yeast, two major competing pathways including homologous recombination (HR) and NHEJ are involved in DSB repair (Aylon and Kupiec 2004), but the NHEJ efficiency is pretty low (Li et al.2011). It was recently reviewed that inhibiting critical NHEJ proteins, such as Ligase IV, a serine/threonine protein kinase DNA-PK responsible for initiating the NHEJ pathway and the heterodimeric Ku complex for binding DSB ends, could enhance HR-mediated genome editing (Pawelczak et al.2018). On the contrary, inhibition of HR activity would increase the NHEJ efficiency, thereby providing a promising approach to improve the TALENs-mediated genome editing efficiency and increase the rate of TALENs-introduced genomic mutations in industrial S. cerevisiae strains. In addition, population genomics studies reported that phenotypic variation of S. cerevisiae isolates correlates with genomic variation (Liti et al.2009). Therefore, comparative genome analysis between S. cerevisiae industrial and lab strains would identify the genomic variants that might determine the differential consequence of TAME treatment in these strains, thus providing clues to further improve the TAME efficiency in industrial strains by modulating these genomic variants. SUPPLEMENTARY DATA Supplementary data are available at FEMSYR online. Acknowledgements We thank Dr Guoqiang Zhang for constructing the TALENs-expressing plasmids with drug-resistant marker genes. We thank Lixian Wang and Huanhuan Zhai (Technical Support Center, Tianjin Institute of Industrial Biotechnology, Chinese Academy of Sciences) for technical assistance in flow cytometry and electron microscope, respectively. FUNDING This work was supported by the National Science Foundation of China (31470214 and 31700077), the National Science Foundation of Tianjin (16JCYBJC43100) and the Science and Technology Support Program of Tianjin, China (15PTCYSY00020), and funding from the Science and Technology Foundation for Selected Overseas Chinese Scholar of Tianjin to Yuping Lin. Conflict of interest. None declared. REFERENCES Abdel-Banat BM , Hoshida H , Ano A et al. High-temperature fermentation: how can processes for ethanol production at high temperatures become superior to the traditional process using mesophilic yeast? Appl Microbiol Biot 2010 ; 85 : 861 – 7 . Google Scholar CrossRef Search ADS Alexander WG . A history of genome editing in Saccharomyces cerevisiae . Yeast 2018 ; 35 : 1 – 6 . Google Scholar CrossRef Search ADS Aylon Y , Kupiec M . DSB repair: the yeast paradigm . DNA Repair (Amst) 2004 ; 3 : 797 – 815 . Google Scholar CrossRef Search ADS PubMed Bonner M , Strouse B , Applegate M et al. DNA damage response pathway and replication fork stress during oligonucleotide directed gene editing . Mol Ther- Nucl Acids 2012 ; 1 : e18 . Google Scholar CrossRef Search ADS Caspeta L , Chen Y , Ghiaci P et al. Altered sterol composition renders yeast thermotolerant . Science 2014 ; 346 : 75 – 8 . Google Scholar CrossRef Search ADS PubMed Caspeta L , Nielsen J . Thermotolerant yeast strains adapted by laboratory evolution show trade-off at ancestral temperatures and preadaptation to other stresses . mBio 2015 ; 6 : e00431 . Google Scholar CrossRef Search ADS PubMed Caspeta L , Chen Y , Nielsen J . Thermotolerant yeasts selected by adaptive evolution express heat stress response at 30°C . Sci Rep 2016 ; 6 : 27003 . Google Scholar CrossRef Search ADS PubMed Clevenger J , Chavarro C , Pearl SA et al. Single nucleotide polymorphism identification in polyploids: a review, example, and recommendations . Mol Plant 2015 ; 8 : 831 – 46 . Google Scholar CrossRef Search ADS PubMed Connelly CF , Skelly DA , Dunham MJ et al. Population genomics and transcriptional consequences of regulatory motif variation in globally diverse Saccharomyces cerevisiae strains . Mol Biol Evol 2013 ; 30 : 1605 – 13 . Google Scholar CrossRef Search ADS PubMed Davey HM , Hexley P . Red but not dead? Membranes of stressed Saccharomyces cerevisiae are permeable to propidium iodide . Environ Microbiol 2011 ; 13 : 163 – 71 . Google Scholar CrossRef Search ADS PubMed de Nadal E , Ammerer G , Posas F . Controlling gene expression in response to stress . Nat Rev Genet 2011 ; 12 : 833 – 45 . Google Scholar CrossRef Search ADS PubMed Deparis Q , Claes A , Foulquie-Moreno MR et al. Engineering tolerance to industrially relevant stress factors in yeast cell factories . FEMS Yeast Res 2017 ; 17 : DOI: 10.1093/femsyr/fox036 . DePristo MA , Banks E , Poplin R et al. A framework for variation discovery and genotyping using next-generation DNA sequencing data . Nat Genet 2011 ; 43 : 491 – 8 . Google Scholar CrossRef Search ADS PubMed Deutschbauer AM , Jaramillo DF , Proctor M et al. Mechanisms of haploinsufficiency revealed by genome-wide profiling in yeast . Genetics 2005 ; 169 : 1915 – 25 . Google Scholar CrossRef Search ADS PubMed Dragosits M , Mattanovich D . Adaptive laboratory evolution—principles and applications for biotechnology . Microb Cell Fact 2013 ; 12 : 64 . Google Scholar CrossRef Search ADS PubMed Gibney PA , Lu C , Caudy AA et al. Yeast metabolic and signaling genes are required for heat-shock survival and have little overlap with the heat-induced genes . Proc Natl Acad Sci USA 2013 ; 110 : E4393 – 402 . Google Scholar CrossRef Search ADS PubMed Guan B , Lei J , Su S et al. Absence of Yps7p, a putative glycosylphosphatidylinositol-linked aspartyl protease in Pichia pastoris, results in aberrant cell wall composition and increased osmotic stress resistance . FEMS Yeast Res 2012 ; 12 : 969 – 79 . Google Scholar CrossRef Search ADS PubMed Guldener U , Heck S , Fiedler T et al. A new efficient gene disruption cassette for repeated use in budding yeast . Nucleic Acids Res 1996 ; 24 : 2519 – 24 . Google Scholar CrossRef Search ADS PubMed Horvath I , Multhoff G , Sonnleitner A et al. Membrane-associated stress proteins: more than simply chaperones . BBA-Biomembranes 2008 ; 1778 : 1653 – 64 . Google Scholar CrossRef Search ADS PubMed Huang da W , Sherman BT , Lempicki RA . Systematic and integrative analysis of large gene lists using DAVID bioinformatics resources . Nat Protoc 2009 ; 4 : 44 – 57 . Google Scholar CrossRef Search ADS PubMed Jarolim S , Ayer A , Pillay B et al. Saccharomyces cerevisiae genes involved in survival of heat shock . G3 2013 ; 3 : 2321 – 33 . Google Scholar CrossRef Search ADS PubMed Jansen G , Buhring F , Hollenberg CP et al. Mutations in the SAM domain of STE50 differentially influence the MAPK-mediated pathways for mating, filamentous growth and osmotolerance in Saccharomyces cerevisiae . Mol Genet Genomics 2001 ; 265 : 102 – 17 . Google Scholar CrossRef Search ADS PubMed Kavscek M , Strazar M , Curk T et al. Yeast as a cell factory: current state and perspectives . Microb Cell Fact 2015 ; 14 : 94 . Google Scholar CrossRef Search ADS PubMed Khan A , Ahmad A , Akhtar F et al. Induction of oxidative stress as a possible mechanism of the antifungal action of three phenylpropanoids . FEMS Yeast Res 2011 ; 11 : 114 – 22 . Google Scholar CrossRef Search ADS PubMed Kitichantaropas Y , Boonchird C , Sugiyama M et al. Cellular mechanisms contributing to multiple stress tolerance in Saccharomyces cerevisiae strains with potential use in high-temperature ethanol fermentation . AMB Expr 2016 ; 6 : 107 . Google Scholar CrossRef Search ADS Kumari R , Pramanik K . Improvement of multiple stress tolerance in yeast strain by sequential mutagenesis for enhanced bioethanol production . J Biosci Bioeng 2012 ; 114 : 622 – 9 . Google Scholar CrossRef Search ADS PubMed Li H , Durbin R . Fast and accurate short read alignment with Burrows-Wheeler transform . Bioinformatics 2009 ; 25 : 1754 – 60 . Google Scholar CrossRef Search ADS PubMed Li T , Huang S , Zhao X et al. Modularly assembled designer TAL effector nucleases for targeted gene knockout and gene replacement in eukaryotes . Nucleic Acids Res 2011 ; 39 : 6315 – 25 . Google Scholar CrossRef Search ADS PubMed Liti G , Carter DM , Moses AM et al. Population genomics of domestic and wild yeasts . Nature 2009 ; 458 : 337 – 41 . Google Scholar CrossRef Search ADS PubMed Luo Z , van Vuuren HJ . Functional analyses of PAU genes in Saccharomyces cerevisiae . Microbiology 2009 ; 155 : 4036 – 49 . Google Scholar CrossRef Search ADS PubMed McKenna A , Hanna M , Banks E et al. The genome analysis toolkit: a MapReduce framework for analyzing next-generation DNA sequencing data . Genome Res 2010 ; 20 : 1297 – 303 . Google Scholar CrossRef Search ADS PubMed Miller JC , Tan S , Qiao G et al. A TALE nuclease architecture for efficient genome editing . Nat Biotechnol 2011 ; 29 : 143 – 8 . Google Scholar CrossRef Search ADS PubMed Niu L , Nomura K , Iwahashi H et al. Petit -High Pressure Carbon Dioxide stress increases synthesis of S -Adenosylmethionine and phosphatidylcholine in yeast Saccharomyces cerevisiae . Biophys Chem 2017 ; 231 : 79 – 86 . Google Scholar CrossRef Search ADS PubMed Pawelczak KS , Gavande NS , VanderVere-Carozza PS et al. Modulating DNA repair pathways to improve precision genome engineering . ACS Chem Biol 2018 ; 13 : 389 – 96 . Google Scholar CrossRef Search ADS PubMed Postmus J , Tuzun I , Bekker M et al. Dynamic regulation of mitochondrial respiratory chain efficiency in Saccharomyces cerevisiae . Microbiology 2011 ; 157 : 3500 – 11 . Google Scholar CrossRef Search ADS PubMed Sadowski I , Lourenco P , Parent J . Dominant marker vectors for selecting yeast mating products . Yeast 2008 ; 25 : 595 – 9 . Google Scholar CrossRef Search ADS PubMed Satomura A , Miura N , Kuroda K et al. Reconstruction of thermotolerant yeast by one-point mutation identified through whole-genome analyses of adaptively-evolved strains . Sci Rep 2016 ; 6 : 23157 . Google Scholar CrossRef Search ADS PubMed Shi DJ , Wang CL , Wang KM . Genome shuffling to improve thermotolerance, ethanol tolerance and ethanol productivity of Saccharomyces cerevisiae . J Ind Microb Biotech 2009 ; 36 : 139 – 47 . Google Scholar CrossRef Search ADS Shui W , Xiong Y , Xiao W et al. Understanding the mechanism of thermotolerance distinct from heat shock response through proteomic analysis of industrial strains of Saccharomyces cerevisiae . Mol Cell Proteomics 2015 ; 14 : 1885 – 97 . Google Scholar CrossRef Search ADS PubMed Snoek T , Picca Nicolino M , Van den Bremt S et al. Large-scale robot-assisted genome shuffling yields industrial Saccharomyces cerevisiae yeasts with increased ethanol tolerance . Biotechnol Biofuels 2015 ; 8 : 32 . Google Scholar CrossRef Search ADS PubMed Stovicek V , Borodina I , Forster J . CRISPR–Cas system enables fast and simple genome editing of industrial Saccharomyces cerevisiae strains . Metab Eng Commun 2015 ; 2 : 13 – 22 . Google Scholar CrossRef Search ADS Tsai SQ , Wyvekens N , Khayter C et al. Dimeric CRISPR RNA-guided FokI nucleases for highly specific genome editing . Nat Biotechnol 2014 ; 32 : 569 – 76 . Google Scholar CrossRef Search ADS PubMed Verghese J , Abrams J , Wang Y et al. Biology of the heat shock response and protein chaperones: budding yeast (Saccharomyces cerevisiae) as a model system . Microbiol Mol Biol R 2012 ; 76 : 115 – 58 . Google Scholar CrossRef Search ADS Voordeckers K , Kominek J , Das A et al. Adaptation to high ethanol reveals complex evolutionary pathways . PLoS Genet 2015 ; 11 : e1005635 . Google Scholar CrossRef Search ADS PubMed Wallace-Salinas V , Brink DP , Ahren D et al. Cell periphery-related proteins as major genomic targets behind the adaptive evolution of an industrial Saccharomyces cerevisiae strain to combined heat and hydrolysate stress . BMC Genomics 2015 ; 16 : 514 . Google Scholar CrossRef Search ADS PubMed Wang K , Li M , Hakonarson H . ANNOVAR: functional annotation of genetic variants from high-throughput sequencing data . Nucleic Acids Res 2010 ; 38 : e164- . Google Scholar CrossRef Search ADS PubMed Whelan SP , Zuckerbraun BS . Mitochondrial signaling: forwards, backwards, and in between . Oxid Med Cell Longev 2013 ; 2013 : 1 – 10 . Google Scholar CrossRef Search ADS Xie C , Tammi MT . CNV-seq, a new method to detect copy number variation using high-throughput sequencing . BMC Bioinformatics 2009 ; 10 : 80 . Google Scholar CrossRef Search ADS PubMed Zhang G , Lin Y , Qi X et al. TALENs-assisted multiplex editing for accelerated genome evolution to improve yeast phenotypes . ACS Synth Biol 2015 ; 4 : 1101 – 11 . Google Scholar CrossRef Search ADS PubMed © FEMS 2018. This article is published and distributed under the terms of the Oxford University Press, Standard Journals Publication Model (https://academic.oup.com/journals/pages/about_us/legal/notices) http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png FEMS Yeast Research Oxford University Press

Metabolic and genomic characterisation of stress-tolerant industrial Saccharomyces cerevisiae strains from TALENs-assisted multiplex editing

Loading next page...
 
/lp/ou_press/metabolic-and-genomic-characterisation-of-stress-tolerant-industrial-Hs08IkjK3s
Publisher
Oxford University Press
Copyright
© FEMS 2018.
ISSN
1567-1356
eISSN
1567-1364
D.O.I.
10.1093/femsyr/foy045
Publisher site
See Article on Publisher Site

Abstract

Abstract TALENs-assisted multiplex editing (TAME) toolbox was previously established and used to successfully enhance ethanol stress tolerance of Saccharomyces cerevisiae laboratory strain. Here, the TAME toolbox was harnessed to improve and elucidate stress tolerances of S. cerevisiae industrial strain. One osmotolerant strain and one thermotolerant strain were selected from the mutant library generated by TAME at corresponding stress conditions, and exhibited 1.2-fold to 1.3-fold increases of fermentation capacities, respectively. Genome resequencing uncovered genomic alterations in the selected stress-tolerant strains, suggesting that cell wall and membrane-related proteins might be major factors behind improved tolerances of yeast to different stresses. Furthermore, amplified mitochondrial DNA might also have an important impact on increased stress tolerance. Unexpectedly, none of predesigned target potential TALENs modification sites showed any genomic variants in sequenced genomes of the selected strains, implicating that the improved stress tolerances might be due to indirect impacts of genome editing via TALENs rather than introducing genomic variants at potential target sites. Our findings not only confirmed TAME could be a useful tool to accelerate the breeding of industrial strain with multiple stress tolerance, but also supported the previous understandings of the complicated mechanisms of multiple stress tolerance in yeast. TALENs, multiplex genome editing, Saccharomyces cerevisiae, industrial strain, stress tolerance, genome resequencing INTRODUCTION As important industrial strain, Saccharomyces cerevisiae has been broadly used in ethanol production due to its general robustness and high productivity (Kavscek et al.2015). To reduce production costs, industrial fermentations prefer to be carried out at high temperature and high concentration of glucose, thus producing high concentration of ethanol (Abdel-Banat et al.2010). However, the industrial fermentation conditions would cause thermal and hyperosmotic stresses and ethanol toxicity to yeast cells, which lead to that massive biological macromolecules are damaged and arrested cell growth or even cell death, eventually reducing ethanol production efficiency. Therefore, it is necessary to improve stress tolerance of S. cerevisiae (Deparis et al.2017). Stress tolerance is a complicated phenotype, which is regulated by dozens of, or even hundreds of, genes simultaneously. For instance, in order to adapt to elevated temperature, S. cerevisiae changes the expression level of thousand genes to regulate cAMP-dependent protein kinase A signalling pathway, cell cycle, lipid metabolism and so on (Jarolim et al.2013; Satomura et al.2016). Previous studies indicate that it is difficult to improve yeast stress tolerance by regulating one or several functional genes (de Nadal, Ammerer and Posas 2011). To overcome this obstacle, many traditional technologies are developed, such as mutagenesis, genome shuffling, adaptive evolution, etc. Sequential mutagenesis has been used to improve multiple stress tolerance in yeast strains (Kumari and Pramanik 2012). To improve yeast stress tolerance, genome shuffling requires at least two strains, and mostly relies on a mutagenised pool of a single strain or a natural pool of strains with different desirable phenotypes (Shi, Wang and Wang 2009; Snoek et al.2015). Adaptive laboratory evolution, which could accumulate desirable genomic and physiological changes in cell populations during long-term selection under specified growth conditions (Dragosits and Mattanovich 2013), is exploited to obtain thermotolerant or ethanol stress-tolerant strains (Caspeta et al.2014; Voordeckers et al.2015). Overall, these traditional approaches are feasible for the breeding of stress-tolerant strains, but laborious and relatively inefficient. Targeted genome editing is an alternative approach for improving desirable traits of yeast, especially TALENs and CRISPR/Cas9 systems which are considered as effective platforms because of their powerful and multiplex genome editing capacities (Alexander 2018). For targeted genome editing, endonuclease FokI or Cas9 generate a double-strand break (DSB) at targeted DNA, which can be repaired by endogenous DNA repair pathways, thus introducing variable-length insertion/deletion mutations or specific sequence alterations (Miller et al.2011; Tsai et al.2014). Recently, we established a TALENs-assisted multiplex editing (TAME) toolbox to target the conserved TATA box and GC box at 66 potential modification sites of S. cerevisiae genome and probably influence 98 genes (Zhang et al.2015), and ethanol stress tolerance of S. cerevisiae laboratory strain was successfully enhanced by TAME. Here, to improve osmotolerance and thermotolerance of S. cerevisiae industrial strain, TAME toolbox was applied to generate a mutant library from an industrial strain ScY01 in several days. After screening at hyperomostic or thermal stress conditions, one osmotolerant strain and one thermotolerant strain were obtained, respectively, and stably exhibited 1.2-fold to 1.3-fold increases of fermentation capacities at corresponding stress conditions. Genome resequencing analysis revealed genomic changes in coding regions affecting the functions of encoded proteins or in intergenic regions probably influencing transcriptional gene expression, which might be genetic factors that resulted in improved stress tolerance of strains. Our results indicated that the TAME toolbox was a useful approach for improving complicated and desirable traits of S. cerevisiae industrial strain, such as stress tolerances, and metabolic and genome characterisation of stress-tolerant strains further confirmed the previously elucidated mechanisms underlying different stress tolerances of yeast. MATERIALS AND METHODS TALENs plasmid construction In our previous study (Zhang et al.2015), plasmids pYES2/CT-GC and p313-GAL-TA (Table 1) with the URA3 and HIS3 autotrophic marker genes, respectively, have been constructed to express the TALENs pair recognizing the GGGCGG and TATAAA sequence, respectively, which allow to induce multiplex editing in the genome of S. cerevisiae laboratory strains. To achieve similar multiplex genome editing in S. cerevisiae industrial strain, two new plasmids pYES2/CT-GGGCGG-GAL1-ZeoR and pRS313-TATAAA-GAL1-KanMX4 were constructed by inserting expression cassettes of the ZeoR and KanMX drug-resistant marker genes into the plasmids pYES2/CT-GC and p313-GAL-TA, respectively, using the ClonExpress® II One Step Cloning Kit (Vazyme Biotech Co., Ltd, China). Specifically, the linear vector of pYES2/CT-GC was PCR amplified using the primer pair pYES2-F/pYES2-R, the ZeoR expression cassette was PCR amplified using the primer pair Zeocin-F/Zeocin-R from the plasmid pIS438 (Sadowski, Lourenco and Parent 2008) and these two PCR fragments were fused together by following the instruction of the kit. Similarly, the linear vector of p313-GAL-TA was PCR amplified using the primer pair pRS313-F/pRS313-R, and the KanMX expression cassette was PCR amplified using the primer pair KanMX4-F/KanMX4-R from the plasmid pUG6 (Guldener et al.1996). The primers used in this study were listed in Table 2. All recombinant plasmids were sequenced by BGI (Shenzhen, China) to verify the insertion of the ZeoR and KanMX expression cassettes. Table 1. Plasmids and S. cerevisiae strains used in this study. Plasmids or strains Description Reference or source Plasmids pYES2/CT-GC 2μ plasmid, autotrophic URA3 gene Zhang et al. (2015) p313-GAL-TA CEN plasmid, autotrophic HIS3 gene Zhang et al. (2015) pYES2/CT-GGGCGG-GAL1-ZeoR 2μ plasmid, ZeoR used for genome editing This study pRS313-TATAAA-GAL1-KanMX4 CEN plasmid, KanMX4, used for genome editing This study Strains ScY01 Evolved thermotolerant strain, diploid Shui et al. (2015) ScY001T Osmotolerant strain, diploid This study ScY033T Thermotolerant strain, diploid This study Plasmids or strains Description Reference or source Plasmids pYES2/CT-GC 2μ plasmid, autotrophic URA3 gene Zhang et al. (2015) p313-GAL-TA CEN plasmid, autotrophic HIS3 gene Zhang et al. (2015) pYES2/CT-GGGCGG-GAL1-ZeoR 2μ plasmid, ZeoR used for genome editing This study pRS313-TATAAA-GAL1-KanMX4 CEN plasmid, KanMX4, used for genome editing This study Strains ScY01 Evolved thermotolerant strain, diploid Shui et al. (2015) ScY001T Osmotolerant strain, diploid This study ScY033T Thermotolerant strain, diploid This study View Large Table 1. Plasmids and S. cerevisiae strains used in this study. Plasmids or strains Description Reference or source Plasmids pYES2/CT-GC 2μ plasmid, autotrophic URA3 gene Zhang et al. (2015) p313-GAL-TA CEN plasmid, autotrophic HIS3 gene Zhang et al. (2015) pYES2/CT-GGGCGG-GAL1-ZeoR 2μ plasmid, ZeoR used for genome editing This study pRS313-TATAAA-GAL1-KanMX4 CEN plasmid, KanMX4, used for genome editing This study Strains ScY01 Evolved thermotolerant strain, diploid Shui et al. (2015) ScY001T Osmotolerant strain, diploid This study ScY033T Thermotolerant strain, diploid This study Plasmids or strains Description Reference or source Plasmids pYES2/CT-GC 2μ plasmid, autotrophic URA3 gene Zhang et al. (2015) p313-GAL-TA CEN plasmid, autotrophic HIS3 gene Zhang et al. (2015) pYES2/CT-GGGCGG-GAL1-ZeoR 2μ plasmid, ZeoR used for genome editing This study pRS313-TATAAA-GAL1-KanMX4 CEN plasmid, KanMX4, used for genome editing This study Strains ScY01 Evolved thermotolerant strain, diploid Shui et al. (2015) ScY001T Osmotolerant strain, diploid This study ScY033T Thermotolerant strain, diploid This study View Large Table 2. Primers used in this study. Name Sequence (5΄ → 3΄) Application pYES2-F ATCCACATGTGTTTTTAGTA Amplifying the linear vector of pYES2/CT-GC without URA3 expression cassette pYES2-R ATCTTGACTGATTTTTCCAT Zeocin-F AAAATCAGTCAAGATCATATGCCCACACACCATAG Amplifying the ZeoR expression cassette Zeocin-R AAAACACATGTGGATGATATCAGCTTGCAAATTAA pRS313-F CCGGGCACGGATTAGAAGCC Amplifying the linear vector of p313-GAL-TA without HIS3 expression cassette pRS313-R GGGATCCACTAGTTCTAGAG KanMX4-F GAACTAGTGGATCCCCCCGGGCGTACGCTGCAGGT Amplifying the KanMX expression cassette KanMX4-R CTAATCCGTGCCCGGCGGCCGATCGATGAATTCGA Name Sequence (5΄ → 3΄) Application pYES2-F ATCCACATGTGTTTTTAGTA Amplifying the linear vector of pYES2/CT-GC without URA3 expression cassette pYES2-R ATCTTGACTGATTTTTCCAT Zeocin-F AAAATCAGTCAAGATCATATGCCCACACACCATAG Amplifying the ZeoR expression cassette Zeocin-R AAAACACATGTGGATGATATCAGCTTGCAAATTAA pRS313-F CCGGGCACGGATTAGAAGCC Amplifying the linear vector of p313-GAL-TA without HIS3 expression cassette pRS313-R GGGATCCACTAGTTCTAGAG KanMX4-F GAACTAGTGGATCCCCCCGGGCGTACGCTGCAGGT Amplifying the KanMX expression cassette KanMX4-R CTAATCCGTGCCCGGCGGCCGATCGATGAATTCGA View Large Table 2. Primers used in this study. Name Sequence (5΄ → 3΄) Application pYES2-F ATCCACATGTGTTTTTAGTA Amplifying the linear vector of pYES2/CT-GC without URA3 expression cassette pYES2-R ATCTTGACTGATTTTTCCAT Zeocin-F AAAATCAGTCAAGATCATATGCCCACACACCATAG Amplifying the ZeoR expression cassette Zeocin-R AAAACACATGTGGATGATATCAGCTTGCAAATTAA pRS313-F CCGGGCACGGATTAGAAGCC Amplifying the linear vector of p313-GAL-TA without HIS3 expression cassette pRS313-R GGGATCCACTAGTTCTAGAG KanMX4-F GAACTAGTGGATCCCCCCGGGCGTACGCTGCAGGT Amplifying the KanMX expression cassette KanMX4-R CTAATCCGTGCCCGGCGGCCGATCGATGAATTCGA Name Sequence (5΄ → 3΄) Application pYES2-F ATCCACATGTGTTTTTAGTA Amplifying the linear vector of pYES2/CT-GC without URA3 expression cassette pYES2-R ATCTTGACTGATTTTTCCAT Zeocin-F AAAATCAGTCAAGATCATATGCCCACACACCATAG Amplifying the ZeoR expression cassette Zeocin-R AAAACACATGTGGATGATATCAGCTTGCAAATTAA pRS313-F CCGGGCACGGATTAGAAGCC Amplifying the linear vector of p313-GAL-TA without HIS3 expression cassette pRS313-R GGGATCCACTAGTTCTAGAG KanMX4-F GAACTAGTGGATCCCCCCGGGCGTACGCTGCAGGT Amplifying the KanMX expression cassette KanMX4-R CTAATCCGTGCCCGGCGGCCGATCGATGAATTCGA View Large Generation and screening of mutant strains by TALENs-mediated multiplex genome editing The procedure to generate and screen stress-tolerant mutants of industrial S. cerevisiae strain with TAME was described in Fig. 1A. To facilitate improving stress tolerances of S. cerevisiae industrial strain by TALENs-mediated multiplex genome editing, the plasmids pYES2/CT-GGGCGG-GAL1-ZeoR and pRS313-TATAAA-GAL1-KanMX4 expressing the TALENs that were designed to recognise the GGGCGG and TATAAA sequence, respectively (Zhang et al.2015) were used to induce multiplex editing in the yeast genome. These two TALENs-expressing plasmids were first transformed into the previously evolved thermotolerant industrial stain ScY01 in our lab (Shui et al.2015), which is able to grow and ferment well at 40°C. The positive transformants were selected on YPD plates (per litre, 10 g yeast extract, 20 g peptone, 20 g glucose and 20 g agar) containing 70 μg mL−1 zeocin and 400 μg mL−1 geneticin G418. To induce TALENs, the transformants with TALENs-expressing plasmids were inoculated into 20 mL YP medium (per litre, 10 g yeast extract, 20 g peptone) containing 10 g L−1 galactose, 70 μg mL−1 zeocin and 200 μg mL−1 geneticin G418 in 50-mL flasks with an initial OD600 of 0.2, cultured at 200 rpm at 30°C for 24 h, and subcultured for serial transfers every 24 h for 7 days, thus allowing accumulation of TALENs-mediated multiplex genomic mutations. The resulting mutant library was further subcultured in liquid YPD medium for serial transfers every 12 h for 10 days to lose the TALENs-expressing plasmids, which could eliminate disturbing effects caused by the bindings of TALENs on the genome. To screen osmotolerant mutants, one aliquot of the mutant library was plated on YP plates with 400 g L−1 glucose producing hyperosmotic stress and grown at 40°C for 3 days until colonies appeared. Total 218 colonies were separately inoculated into 200 μL liquid YP medium containing 400 g L−1 glucose on 96-well plates and grown at 40°C for 48 h. The colony showing the best cell growth in contrast to the parent strain ScY01 and any other colonies from our experiment at hyperosmotic condition was named ScY001T (Table 1). To screen thermotolerant mutants, the other aliquot of the mutant library was plated on YP plates containing 200 g L−1 glucose and grown at 42°C for 4 days. Total 671 colonies were separately inoculated into 200 μL liquid YP medium containing 200 g L−1 glucose on 96-well plates and grown at elevated temperature of 42°C for 24 h. The colony showing the best cell growth in contrast to the parent strain ScY01 and any other colonies from our experiment at thermal stress condition was named ScY033T (Table 1). These two more stress-tolerant industrial S. cerevisiae strains were assessed for subsequent physiological characterisation and functional genomic elucidation. Figure 1. View largeDownload slide Screening of industrial S. cerevisiae by TALENs-assisted multiplex editing (TAME) for improving stress tolerance. (A) Diagram of TAME for improvement of industrial yeast stress tolerance. (B) Screening of the mutant strains for improved osmotolerance at 40°C using YP media with 400 g L−1 glucose. The cell growth of the parent strain ScY01 reached an OD600 of 1.13. (C) Screening of the mutant strains for improved thermotolerance at 42°C using YP media with 200 g L−1 glucose. The cell growth of the parent strain ScY01 reached an OD600 of 0.75. The relative cell growths in (B) and (C) were calculated in contrast to the parent strain ScY01 at the certain stress condition. The red lines in (B) and (C) represent the same cell growth to the parent strain ScY01 at the same test conditions. Figure 1. View largeDownload slide Screening of industrial S. cerevisiae by TALENs-assisted multiplex editing (TAME) for improving stress tolerance. (A) Diagram of TAME for improvement of industrial yeast stress tolerance. (B) Screening of the mutant strains for improved osmotolerance at 40°C using YP media with 400 g L−1 glucose. The cell growth of the parent strain ScY01 reached an OD600 of 1.13. (C) Screening of the mutant strains for improved thermotolerance at 42°C using YP media with 200 g L−1 glucose. The cell growth of the parent strain ScY01 reached an OD600 of 0.75. The relative cell growths in (B) and (C) were calculated in contrast to the parent strain ScY01 at the certain stress condition. The red lines in (B) and (C) represent the same cell growth to the parent strain ScY01 at the same test conditions. Characterisation of fermentation capacity To characterise fermentation capacities of stress-tolerant industrial S. cerevisiae strains, yeast cells on YPD plates were grown in 100-mL flasks containing 50 mL YP media with 200 g L−1 glucose at 30°C overnight (∼15 h). Cells were harvested by centrifugation and then inoculated into fermentation media. The initial OD600 used for fermentations was 0.5. YP medium containing 400 g L−1 glucose was used for testing fermentation capacity at hyperosmotic conditions, and fermentations were performed at 40°C. Fermentations at thermal stress conditions were carried out at 42°C using YP medium containing 200 g L−1 glucose. Fermentations at moderate stress conditions were conducted at thermal stress condition of 40°C using 200 g L−1 glucose, or 37°C using 400 g L−1 glucose, or at normal temperature of 30°C using 400 g L−1 glucose. YP medium containing 200 g L−1 glucose and supplemented with 2%–5% (v/v) or 3%-8% (v/v) ethanol as specified in the text was used for testing fermentation capacity at ethanol stress conditions, and fermentations were conducted at 30°C and 40°C, respectively. Cell growth was monitored by measuring OD600. Concentrations of glucose and ethanol were measured by high-performance liquid chromatography on an Agilent 1260 system (Agilent, USA) equipped with a refractive index detector and a Phenomenex RFQ fast acid column (100 mm × 7.8 mm ID) (Phenomenex Inc., Torrance, CA, USA). The column was eluted with 0.01 N H2SO4 at a flow rate of 0.6 mL min−1 at 55°C. Spot assay for heat-shock survival Heat-shock assays and spot assays of cell survival were performed as described previously with some modifications (Gibney et al.2013; Jarolim et al.2013). Yeast colonies were grown in 10 mL tubes containing 3 mL YP medium with 20 g L−1 glucose at 30°C with shaking at 200 rpm overnight (∼15 h). Cells were harvested by centrifugation, and inoculated into 25 mL YP medium with 200 g L−1 glucose in 50 mL flasks, to achieve an initial OD600 value of 0.2. The cell cultures were grown to early log phase at 30°C with shaking at 220 rpm for 5 h. Two aliquots containing an appreciate amount of cells were harvested and resuspended in 1 mL supernatants to obtain cell suspensions with an OD600 of 5.0. One aliquot of cell suspension was placed on ice as a pre-heat shock control. The other aliquots of cell suspension were transformed to 10 mL tube, and incubated at 50°C for 30 min with shaking at 200 rpm, and immediately chilled on ice for 5 min. Both pre-heat shock and heat-shock aliquots of cell suspensions with an OD600 of 5.0 were diluted to OD600 of 1.0, 0.3, 0.1, 0.03, 0.01, and 5 μL samples at each dilution spotted on YPD plates (YP medium with 20 g L−1 glucose) which were incubated at 30°C for 48 h. Genome resequencing and data analysis Genomic DNA isolation and the sequencing libraries of the osmotolerant strain ScY001T, the thermotolerant strain ScY033T and the parent strain ScY01 were constructed and sequenced on Illumina HiSeq 4000 using 150-bp paired-end sequencing by GeneDenovo Co. (Guangzhou, Guangdong, China). A mean of 21.2 million 150-bp clean reads was generated for each library. The S. cerevisiae S288c genome as a reference was downloaded from RefSeq at the NCBI (sequence assembly version R64, RefSeq assembly accession: GCF_000146045.2) including 16 chromosomes and the mitochondrial genome. The Genome Analysis Toolkit (GATK v3.5) Best Practices pipeline was used to detect single nucleotide polymorphisms (SNPs) and insertion/deletion (InDels) (McKenna et al.2010; DePristo et al.2011). The cleaned reads were mapped to the reference genome using the mapping tools BWA-mem (version 0.7.13) (Li and Durbin 2009), providing an average sequencing depth of ×185 and 98% sequencing coverage for each library. And variants were then called using GATK HaplotypeCaller. For genome analysis, reads with mapping quality ≥30 were included. Initially called SNPs were filtered with a minimum read depth of 10 and a quality score threshold of 20 (Clevenger et al.2015). Variants annotation was performed using the package ANNOVAR (Wang, Li and Hakonarson 2010). The affected genes of detected genic non-synonymous and intergenic variants (SNPs and InDels) in comparing ScY001T with ScY01 as well as ScY033T with ScY01 are available in supplementary material (Supplementary files 1–4), and were further analysed for enrichment in Gene Ontology (GO) terms using DAVID Bioinformatics Resources 6.8 with a P-value cut-off of less than 0.05 and a Benjamini false discovery rate cut-off of less than 0.5 (Huang da, Sherman and Lempicki 2009) (Table S1, Supporting Information). Copy number variants (CNV) corresponding to potential large-scale chromosomal duplications and loss were examined by using CNV-seq package in R (Xie and Tammi 2009). The results were plotted as log-ratio plots with calculated P-values for genomic and mitochondrial genome, respectively. The genome sequencing raw data were deposited in the NCBI Sequence Read Archive under the accession number SRP127218. Determination of cell wall structure and cell membrane integration Cells were grown in 100-mL flasks containing 50 mL YP media with 200 g L−1 glucose at 30°C overnight (∼15 h), and then inoculated into media with an initial OD600 of 0.5 after centrifugation. For treatments at hyperosmotic stress condition, cells were cultured at 400 g L−1 glucose and 40°C for 48 h. For treatments at thermal stress condition, cells were cultured at 200 g L−1 glucose and 42°C for 36 h. As controls, cells were correspondingly cultured at normal condition using 200 g L−1 glucose at 30°C for 48 or 36 h. Cells were then harvested by centrifugation and washed twice with 0.1 M phosphate buffered solution (pH 7.4). Cell pellets were resuspended in phosphate buffered solution. One aliquot of cell suspension was used for monitoring cell wall structure by electron microscope. The other aliquots were used for monitoring membrane integration by flow cytometry. For scanning electron microscopy (SEM) analysis, samples were fixed and treated as previously described (Niu et al.2017). Inspection and photomicrographs were performed with a scanning electron microscope (SU8010, Hitachi, Ltd., Japan) operating at a voltage of 1.0 kV. For transmission electron microscopy (TEM) analysis, samples were fixed and processed as previously reported (Guan et al.2012). After contrasting with uranyl acetate and lead citrate, the samples were examined with transmission electron microscope (HT7700, Hitachi, Ltd., Japan). Determination of membrane integration by flow cytometry analysis was conducted according to the previously reported method (Khan et al.2011). Briefly, cells were stained using propidium iodide (PI), which penetrates cells with severe membrane lesions only, and make the cell be coloured with red by binding the nucleic acid. After staining for 30 min, monoparametric detection of PI fluorescence was performed using a MoFlo XDP flow cytometer (Beckman Coulter, Brea, CA, USA) with a wavelength of 488/620 nm. RESULTS AND DISCUSSION Screening of stress-tolerant industrial S. cerevisiae by TAME The TALENs recognizing the critical and conserved GC box (GGGCGG) and TATA box (TATAAA) was previously designed to target 66 potential TALENs-mediated modification sites in the genome of S. cerevisiae and probably influence 98 genes, thus allowing to induce genomic modification at multiple sites (Zhang et al.2015). By applying this pair of TALENs, a TAME toolbox was established and used to successfully improve ethanol tolerance of S. cerevisiae laboratory strain BY4741 in a short amount of time (Zhang et al.2015). Here, this TAME toolbox was harnessed to improve tolerances of S. cerevisiae industrial strain ScY01 to hyperosmotic and high-temperature stresses. After TAME treatment for seven rounds, a mutant library of the industrial strain ScY01 was generated and subjected to screen stress-tolerant mutants at hyperosmotic and high-temperature stress conditions, respectively (Fig. 1). On the one hand, among total 218 colonies screened at hyperosmotic stress condition, 12.4% of cells showed more than 1.2-fold increase of cell growth in contrast to the parent strain ScY01, whereas 9.6% showed less than 0.8-fold decrease of cell growth than ScY01 (Fig. 1B). The colony with the best cell growth showed a 1.3-fold increase in contrast to the initial strain ScY01, and named ScY001T for further evaluation of osmotolerance and other stress characterisation as well as genomic analysis. On the other hand, among total 671 colonies screened at high-temperature condition, 6.6% of cells showed more than 1.2-fold increase of cell growth than the parent strain ScY01, whereas 33.4% showed less than 0.8-fold decrease of cell growth than ScY01 (Fig. 1C). The colony with the best cell growth showed a 1.4-fold increase in contrast to the parent strain ScY01, and named ScY033T for further evaluation of thermotolerance and other stress characterisation as well as genomic analysis. Physiological characterisation of stress-tolerant industrial S. cerevisiae To evaluate whether fermentation capacities of the selected stress-tolerant mutants were also improved at stress conditions, the selected osmotolerant strain ScY001T and thermotolerant strain ScY033T were compared with the parent strain ScY01 at hyperosmotic and high-temperature conditions, respectively. At hyperosmotic condition caused by high concentration of 400 g L−1 glucose, ScY001T consumed 144.3 ± 1.0 g L−1 glucose in 84 h, which was 1.3 times as much as ScY01 (Fig. 2A). Meanwhile, ScY001T produced 61.9 ± 0.3 g L−1 ethanol in 84 h, which was 1.2 times more than that of ScY01. On the other hand, ScY033T consumed 108.9 ± 2.6 g L−1 glucose in 84 h at 42°C, which was 1.2 times as much as ScY01. Meanwhile, ScY033T produced 46.3 ± 0.2 g L−1 ethanol in 84 h, which was 1.3 times more than that of ScY01 (Fig. 2B). These results indicated that the selected stress-tolerant strains could also obtain enhanced fermentation capacities at stress conditions. However, glucose consumption of both ScY001T and ScY01 got slow after fermenting for 72 h at hyperosmotic condition, leading to more than half of the residual glucose left in the media. Both ScY033T and ScY01 consumed only a little more than half of the glucose at higher temperature (42°C). These results might be due to the combination inhibitory effects of high-temperature and hyperosmotic stresses, since fermentations at hyperosmotic condition were performed at 40°C instead of normal temperature of 30°C and fermentations at thermal stress condition were conducted using a relatively high concentration of glucose of 200 g L−1 glucose and a higher temperature of 42°C. To verify this assumption, we compared the fermentation capacities of the stress-tolerant and parent strains at moderate stress conditions (Fig. S1, Supporting Information). The parent strain ScY01 has been previously evolved to grow and ferment well at 40°C (Shui et al.2015). Thus, when fermentations were conducted using 200 g L−1 glucose at 40°C, the stress-tolerant strains ScY001T and ScY033T showed similar fermentation capacities to the parent strain ScY01 and completely consumed glucose within 48 h (Fig. S1A, Supporting Information). When fermentations at hyperosmotic condition were performed at a moderate temperature of 37°C and a normal temperature of 30°C (Fig. S1B and C, Supporting Information), much more glucose was consumed in contrast to 42°C (Figs 2A and 3B). Furthermore, the osmotolerant strain ScY001T obviously consumed glucose more quickly than ScY01 and ScY033T at 37°C (Fig. S1B, Supporting Information). By contrast, at 30°C, ScY001T consumed glucose slightly faster at the beginning of fermentation but eventually slower than ScY01 and ScY033T (Fig. S1C, Supporting Information). Therefore, these results confirmed the combination inhibitory effects of high-temperature and hyperosmotic stresses. Figure 2. View largeDownload slide Physiological characterisation of industrial S. cerevisiae mutants with improved stress tolerance. (A) Characterisation of osmotolerant industrial S. cerevisiae mutants. Fermentations of the osmotolerant strain ScY001T and parent strain ScY01 were performed at 40°C using YP media with 400 g L−1 glucose. (B) Characterisation of thermotolerant industrial S. cerevisiae mutants. Fermentations of the thermotolerant strain ScY033T and parent strain ScY01 were performed at 42°C using YP media with 200 g L−1 glucose. Initial OD600 of 0.5 was used for all the fermentations. Data represent the mean and standard error of duplicate cultures at each condition. Figure 2. View largeDownload slide Physiological characterisation of industrial S. cerevisiae mutants with improved stress tolerance. (A) Characterisation of osmotolerant industrial S. cerevisiae mutants. Fermentations of the osmotolerant strain ScY001T and parent strain ScY01 were performed at 40°C using YP media with 400 g L−1 glucose. (B) Characterisation of thermotolerant industrial S. cerevisiae mutants. Fermentations of the thermotolerant strain ScY033T and parent strain ScY01 were performed at 42°C using YP media with 200 g L−1 glucose. Initial OD600 of 0.5 was used for all the fermentations. Data represent the mean and standard error of duplicate cultures at each condition. Figure 3. View largeDownload slide Physiological characterisation of the osmotolerant strain ScY001T and the thermotolerant strain ScY033T at other stress conditions. (A) Fermentation profiles of the osmotolerant strain ScY001T and parent strain ScY01 at thermal stress condition of 42°C using YP media with 200 g L−1 glucose. (B) Fermentation profiles of the thermotolerant strain ScY033T and parent strain ScY01 at hyperosmotic condition at 40°C using YP media with 400 g L−1 glucose. Initial OD600 of 0.5 was used for all the fermentations. Data represent the mean and standard error of duplicate cultures at each condition. Figure 3. View largeDownload slide Physiological characterisation of the osmotolerant strain ScY001T and the thermotolerant strain ScY033T at other stress conditions. (A) Fermentation profiles of the osmotolerant strain ScY001T and parent strain ScY01 at thermal stress condition of 42°C using YP media with 200 g L−1 glucose. (B) Fermentation profiles of the thermotolerant strain ScY033T and parent strain ScY01 at hyperosmotic condition at 40°C using YP media with 400 g L−1 glucose. Initial OD600 of 0.5 was used for all the fermentations. Data represent the mean and standard error of duplicate cultures at each condition. During the industrial production process, S. cerevisiae often suffers from multiple stresses. Thus, strains resistant to multiple stresses would be desirable for their industrial applications. To test the physiological response of the selected stress-tolerant strains to other stress conditions, fermentation capacity of the selected osmotolerant strain ScY001T was evaluated at thermal stress condition, while fermentation capacity of the selected thermotolerant strain ScY033T was detected at hyperosmotic condition. Furthermore, both ScY001T and ScY033T were evaluated for fermentation capacities at other stress conditions. At thermal stress condition, osmotolerant strain ScY001T consumed 144.3 ± 1.0 g L−1 glucose and produced 39.5 ± 0.2 g L−1 ethanol in 60 h at 42°C, which were 1.1 times as much as that of ScY01 (Fig. 3A). At hyperosmotic stress condition, thermotolerant strain ScY033T consumed 117.3 ± 4.1 g L−1 glucose and produced 43.6 ± 2.6 g L−1 ethanol in 84 h, which were 1.1 times as much as that of ScY01 (Fig. 3B). These results suggested that the selected osmotolerant strain might acquire slightly increased thermotolerance and the selected thermotolerant strain might also acquire slightly increased osmotolerance. To evaluate the ethanol tolerance of ScY001T and ScY033T, we performed fermentations using media supplemented with 0% (v/v), 3%(v/v), 6% (v/v) or 8% (v/v) ethanol concentration at 30°C (Fig. S2, Supporting Information) and with 0% (v/v), 2% (v/v), 3% (v/v) or 5% (v/v) ethanol concentration at 40°C (Fig. S3, Supporting Information). At normal condition using 200 g L−1 glucose at 30°C, glucose was depleted within 18 h, and ScY001T and ScY033T exhibited similar fermentation capacities to ScY01 (Fig. S2A, Supporting Information). With the increasing concentration of ethanol supplemented in media, fermentation capacities of these strains were gradually hampered at either 30°C or 40°C (Figs S2 and S3, Supporting Information), and completely inhibited at the condition of 5% (v/v) ethanol at 40°C. Overall, both ScY001T and ScY033T did not exhibit greatly different fermentation capacities from parent ScY01 at ethanol stress condition (Figs S2 and S3, Supporting Information), indicating that the screening process of both osmotolerant and thermotolerant mutants might have no significant impact on enrichment of ethanol-tolerant mutants. Furthermore, the physiological responses of the selected stress-tolerant strains to short-term heat shock were monitored by measuring cell viability through spot assay before and after heat shock treatment. As for cells without heat shock treatment, both the selected stress-tolerant strains ScY001T and ScY033T showed similar cell viabilities to the parent strain ScY01 (Fig. 4A). Inherently, both ScY001T and ScY033T acquired increased long-term thermotolerance at prolonged stress conditions (Figs 2 and 3), but they both showed dramatically decreased cell viabilities after heat shock treatment in contrast to ScY01 (Fig. 4B). This result implicated that there might be a trade-off mechanism for acquiring long-term stress tolerance and maintaining short-term stress tolerance, such as heat shock response. The heat shock response is appropriately considered to be evolutionarily selected to prevent damage caused by an anticipated stress rather than to promote recovery from an existing insult (Verghese et al.2012). In contrast to potential lethal damages caused by short-term severe stresses, long-term moderate stresses lead to sustained but nonlethal impacts on physiological activities of cells, which might provide cells a chance to be adapted to stress conditions. It has been reported that adaptively evolved thermotolerant S. cerevisiae strain show trade-off at ancestral temperatures and preadaptation to other stresses (Caspeta and Nielsen 2015), and even expressed heat stress response at normal temperature of 30°C (Caspeta, Chen and Nielsen 2016). Furthermore, S. cerevisiae has distinct regulatory mechanism of thermotolerance at long-term thermal stress from the heat shock response at short-term thermal stress (Shui et al.2015). Figure 4. View largeDownload slide The physiological response of the osmotolerant strain ScY001T and the thermotolerant strain ScY033T to heat shock treatment. (A) Spot assay of cells before heat shock treatment. (B) Spot assay of cells after heat shock treatment. Strains ScY001T, ScY033T and ScY01 were grown in biological duplicates to early-log phase in YP medium containing 20 g L−1 glucose. Cells were harvested and resuspended to 5.0 OD600. One aliquot of the 5.0 OD600 cell suspension without heat shock treatment and the other aliquot after treated at 50°C for 30 min were serially diluted to the OD600 indicated and spotted on YPD plates after. The plates were incubated for 2 days at 30°C and imaged. Figure 4. View largeDownload slide The physiological response of the osmotolerant strain ScY001T and the thermotolerant strain ScY033T to heat shock treatment. (A) Spot assay of cells before heat shock treatment. (B) Spot assay of cells after heat shock treatment. Strains ScY001T, ScY033T and ScY01 were grown in biological duplicates to early-log phase in YP medium containing 20 g L−1 glucose. Cells were harvested and resuspended to 5.0 OD600. One aliquot of the 5.0 OD600 cell suspension without heat shock treatment and the other aliquot after treated at 50°C for 30 min were serially diluted to the OD600 indicated and spotted on YPD plates after. The plates were incubated for 2 days at 30°C and imaged. Taken together, aiming to further improve fermentation capacities at multiple stress conditions, more efforts of strain breeding remained to be undertaken. TAME toolbox would be a promising approach to accelerate the breeding of multiple stress-tolerant strains, especially by performing several rounds of TAME treatment and switching the screening conditions between different stresses. Functional genomic elucidation of stress-tolerant industrial S. cerevisiae To uncover the genomic alterations behind the increased stress tolerances of the selected strains, both mutant strains ScY001T and ScY033T as well as the parent strain ScY01 were subjected to whole-genome resequencing. By comparing with the reference genome of strain S228c, the common and different SNPs and InDels were detected in ScY001T versus ScY01 as well as ScY033T versus ScY01 (Table 3). Among the different SNPs and InDels between ScY001T and ScY01, total 200 genes with non-synonymous variants were annotated to result in mutant or non-functional proteins (Supplementary file 1); meanwhile, total 709 genes with intergenic variants were predicted to be modulated in terms of gene expressions in the osmotolerant mutant ScY001T (Table 3, Supplementary file 2). Among the different SNPs and InDels between ScY033T and ScY01, total 205 genes with non-synonymous variants were annotated to result in mutant or non-functional proteins (Supplementary file 3); meanwhile, total 707 genes with intergenic variants were predicted to be altered in terms of gene expressions in the thermotolerant mutant ScY033T (Table 3, Supplementary file 4). Remarkably, a majority of these variants were classified as heterozygous in either or both of stress-tolerant strains and the parent strain, where more than one allelic variant was genotyped by genome sequencing (Supplementary file 1–4). This might be due to the diploid nature of these strains (Shui et al.2015). Similarly, a large portion of differential variants were also observed to be heterozygous in another industrial diploid strain of S. cerevisiae strain Ethanol Red and its evolved strain ISO12 (Wallace-Salinas et al.2015). However, the difference was that the percentage of heterozygous variants slightly increased in the previously reported evolved strain ISO12 (Wallace-Salinas et al.2015), but obviously decreased in the stress-tolerant strains ScY001T and ScY033T in this study (Supplementary file 1–4). It seemed that more alleles became homozygous during the TAME treatment, which might be beneficial to eliminate the deleterious effects of heterozygous mutations causing haploinsufficiency in diploid strains (Deutschbauer et al.2005). For instance, Ste50, which is an adaptor protein for osmosensory signalling pathway (Jansen et al.2001), was detected to have a heterozygous mutation of D109G in the parent strain ScY01 and become homozygous wild-type Ste50 in the osmotolerant strain ScY001T (Supplementary file 1). It has been reported that mutations in the SAM domain of Ste50 at position 30 to 104 cause signalling defects in the pathways for mating, filamentous growth and osmotolerance (Jansen et al.2001). Here, the mutation D109G is localised near the SAM domain, implicating its adverse effect on osmotolerance, which was eliminated because of the homozygous wild-type Ste50 in ScY001T. Furthermore, compared with the stress-tolerant mutants derived from industrial diploid strains, the evolved mutants derived from laboratory haploid strains were not reported to have heterozygous mutations (Caspeta et al.2014; Satomura et al.2016). Thus, the identified point mutations affecting the functions of ERG3 and CDC25 could be easily verified for their positive effects on stress tolerance by reconstructing them in the parent strains (Caspeta et al.2014; Satomura et al.2016). By contrast, it would be difficult to verify the effects of the identified genomic mutations on stress tolerance in industrial diploid strains due to their homo-/heterozygosity, but worthwhile to make further efforts in the future. Table 3. Results of the variant calling and analysis. Variants (SNPs & InDels) Count ScY001T vs. ScY01 ScY033T vs. ScY01 General statistics Total number (vs. S288c) 69487 & 7010 69529 & 6991 Background variants 67782 & 6496 67811 & 6493 Mutant variants 1705 & 514 1718 & 498 Location of mutant variants Intergenic variants 812 & 410 838 & 387 Coding region variants 861 & 94 846 & 102 Unknown 32 & 10 34 & 9 Effect of coding region variants Synonymous effects 480 & 0 466 & 0 Non-synonymous effects 381 & 94 380 & 102 Genes with non-synonymous effects 177 & 48 a 178 & 49 b Genes with intergenic variants 380 & 485 c 401 & 474 d Variants (SNPs & InDels) Count ScY001T vs. ScY01 ScY033T vs. ScY01 General statistics Total number (vs. S288c) 69487 & 7010 69529 & 6991 Background variants 67782 & 6496 67811 & 6493 Mutant variants 1705 & 514 1718 & 498 Location of mutant variants Intergenic variants 812 & 410 838 & 387 Coding region variants 861 & 94 846 & 102 Unknown 32 & 10 34 & 9 Effect of coding region variants Synonymous effects 480 & 0 466 & 0 Non-synonymous effects 381 & 94 380 & 102 Genes with non-synonymous effects 177 & 48 a 178 & 49 b Genes with intergenic variants 380 & 485 c 401 & 474 d a aThe total number of genes with non-synonymous SNPs or/and InDels in ScY001T vs. ScY01 is 200. b bThe total number of genes with non-synonymous SNPs or/and InDels in ScY033T vs. ScY01 is 205. c cThe total number of genes with intergenic SNPs or/and InDels in ScY001T vs. ScY01 is 709. d dThe total number of genes with intergenic SNPs or/and InDels in ScY033T vs. ScY01 is 707. View Large Table 3. Results of the variant calling and analysis. Variants (SNPs & InDels) Count ScY001T vs. ScY01 ScY033T vs. ScY01 General statistics Total number (vs. S288c) 69487 & 7010 69529 & 6991 Background variants 67782 & 6496 67811 & 6493 Mutant variants 1705 & 514 1718 & 498 Location of mutant variants Intergenic variants 812 & 410 838 & 387 Coding region variants 861 & 94 846 & 102 Unknown 32 & 10 34 & 9 Effect of coding region variants Synonymous effects 480 & 0 466 & 0 Non-synonymous effects 381 & 94 380 & 102 Genes with non-synonymous effects 177 & 48 a 178 & 49 b Genes with intergenic variants 380 & 485 c 401 & 474 d Variants (SNPs & InDels) Count ScY001T vs. ScY01 ScY033T vs. ScY01 General statistics Total number (vs. S288c) 69487 & 7010 69529 & 6991 Background variants 67782 & 6496 67811 & 6493 Mutant variants 1705 & 514 1718 & 498 Location of mutant variants Intergenic variants 812 & 410 838 & 387 Coding region variants 861 & 94 846 & 102 Unknown 32 & 10 34 & 9 Effect of coding region variants Synonymous effects 480 & 0 466 & 0 Non-synonymous effects 381 & 94 380 & 102 Genes with non-synonymous effects 177 & 48 a 178 & 49 b Genes with intergenic variants 380 & 485 c 401 & 474 d a aThe total number of genes with non-synonymous SNPs or/and InDels in ScY001T vs. ScY01 is 200. b bThe total number of genes with non-synonymous SNPs or/and InDels in ScY033T vs. ScY01 is 205. c cThe total number of genes with intergenic SNPs or/and InDels in ScY001T vs. ScY01 is 709. d dThe total number of genes with intergenic SNPs or/and InDels in ScY033T vs. ScY01 is 707. View Large Based on GO Cellular Component enrichment analysis (Huang da, Sherman and Lempicki 2009), the affected genes of non-synonymous variants in both ScY001T and ScY033T showed similar enrichments, and were mostly enriched for integral component of membrane and cell wall-related proteins (Fig. 5A, Table S1, Supporting Information). Besides the enrichment for cell wall and membrane-related components, some affected genes were enriched for vacuole. GO Molecular Function enrichment analysis indicated that the affected genes of non-synonymous variants in ScY001T were specifically enriched for structural constituent of cell wall, while those in ScY033T were specifically enriched for receptor activity proteins (Fig. 5A, Table S1, Supporting Information). Furthermore, according to GO Biological Process enrichment analysis, only the affected genes of non-synonymous variants in ScY001T were found to be enriched in biological processes including fungal-type cell wall organisation, transmembrane transport and flocculation-related processes proteins (Fig. 5A; Table S1, Supporting Information). Membrane-associated stress proteins include not only chaperones but also other proteins (Horvath et al.2008). Wallace-Salinas et al. (2015) reported that non-synonymous variants were significantly enriched in GO terms related to cell periphery, membranes and cell wall during the adaptive evolution of an industrial S. cerevisiae strain to combined heat and hydrolysate stress. Our observations were consisted with the previous findings, suggesting that cell wall and membrane-related proteins might be major genomic targets behind improved tolerances of industrial S. cerevisiae strain to different stresses. Figure 5. View largeDownload slide Enriched Gene Ontology (GO) terms of genes influenced by genic nonsynonymous (A) and intergenic (B) variants in mutant strains ScY001T (blue line) and ScY033T (red line). The affected genes of detected genic non-synonymous and intergenic variants (SNPs and InDels) in comparing ScY001T with ScY01 as well as ScY03T with ScY01 were analysed for enrichment in GO terms using DAVID Bioinformatics Resources 6.8 (Huang da et al.2009). The percentage associated with each GO terms including GO Biological Process (BP), GO Cellular Component (CC), GO Molecular Function (MF) was calculated as the percentage of genes involved in the corresponding GO terms among the pool of genes that were influenced by genic non-synonymous or intergenic variants. Figure 5. View largeDownload slide Enriched Gene Ontology (GO) terms of genes influenced by genic nonsynonymous (A) and intergenic (B) variants in mutant strains ScY001T (blue line) and ScY033T (red line). The affected genes of detected genic non-synonymous and intergenic variants (SNPs and InDels) in comparing ScY001T with ScY01 as well as ScY03T with ScY01 were analysed for enrichment in GO terms using DAVID Bioinformatics Resources 6.8 (Huang da et al.2009). The percentage associated with each GO terms including GO Biological Process (BP), GO Cellular Component (CC), GO Molecular Function (MF) was calculated as the percentage of genes involved in the corresponding GO terms among the pool of genes that were influenced by genic non-synonymous or intergenic variants. Furthermore, variations in non-coding regions have been investigated to have an important impact on phenotypic diversity given their influence on gene expression level (Connelly et al.2013). Based on GO term enrichment analysis, the affected genes of intergenic variants in ScY033T and ScY001T were also found to have enrichments for cell wall-related cellular components, molecular functions and biological processes (Fig. 5B, Table S1, Supporting Inormation), suggesting that gene expression of cell wall-related proteins might be reshaped in the stress-tolerant strains due to these intergenic variants. Remarkably, both the affected genes of intergenic variants in ScY001T and ScY033T were enriched for the biological process of response to stress, and those in ScY001T had a specific enrichment for cellular components of cytoplasmic stress granule. These results implicated that altering gene expression of stress response-related genes might be another genomic targets behind improved tolerances of industrial S. cerevisiae strain to different stresses. Satomura et al. (2016) also reported that evolved thermotolerant strains showed highly upregulated gene expression involved in response to stress and heat, due to a CDC25 point mutation that led to the downregulation of the cAMP-dependent protein kinase A signalling pathway. However, the effects of intergenic variants were too complicated to be verified. DNA CNV—amplification or deletion of DNA segments—is also an important source of genetic variation, changing the original number of DNA copies that could lead to phenotypic variations. Thus, CNVs of the selected stress-tolerant strains were assessed for not only nuclear genomic DNA but also mitochondrial genomic DNA (mtDNA) (Figs 6 and 7). Amplification-type CNVs seemed to be prevalent on mtDNA of ScY001T (Fig. 6A). Distribution histograms of amplification-type (blue in Fig. 6B) and deletion-type (red in Fig. 6B) CNVs clearly displayed that amplification-type and deletion-type CNVs were equally distributed across nuclear chromosomes while the majority of CNVs were amplification-type on mtDNA (Fig. 6B). Similarly, amplification-type CNVs of ScY033T were also observed on mtDNA instead of nuclear chromosomes (Fig. 7). These observations implicated that the stress-tolerant mutants ScY001T and ScY033T might have more mtDNA copies or more mitochondria than the parent strain ScY01. Mitochondria are the sites for producing ATP through respiration, and also seen to regulate nuclear gene expression and cellular functions (Whelan and Zuckerbraun 2013). Previous study reported that ATP is an important factor for yeast cells to maintain normal physiological levels at stress conditions (Postmus et al.2011). The mechanism underlying mtDNA amplifications in the stress-tolerant strains remained to be discovered. Figure 6. View largeDownload slide Copy number variation (CNV) between ScY001T and ScY01 in nuclear and mitochondrial genomes. (A) A log2 ratio plot for CNV in mitochondrial genome generated by CNV-seq. The red and blue coloured plots represent amplification-type and deletion-type CNVs, respectively. The red and blue colour gradients represent the P-value calculated on each of the ratios where an increase in brightness shows a decrease in P-value. (B) Log2 ratio distribution for CNVs on nuclear genome (left) and mitochondrial genome (right). Figure 6. View largeDownload slide Copy number variation (CNV) between ScY001T and ScY01 in nuclear and mitochondrial genomes. (A) A log2 ratio plot for CNV in mitochondrial genome generated by CNV-seq. The red and blue coloured plots represent amplification-type and deletion-type CNVs, respectively. The red and blue colour gradients represent the P-value calculated on each of the ratios where an increase in brightness shows a decrease in P-value. (B) Log2 ratio distribution for CNVs on nuclear genome (left) and mitochondrial genome (right). Figure 7. View largeDownload slide Copy number variation (CNV) between ScY033T and ScY01 in nuclear and mitochondrial genomes. (A) A log2 ratio plot for CNV in mitochondrial genome generated by CNV-seq. The red and blue coloured plots represent amplification-type and deletion-type CNVs, respectively. The red and blue colour gradients represent the P-value calculated on each of the ratios where an increase in brightness shows a decrease in P-value. (B) Log2 ratio distribution for CNVs on nuclear genome (left) and mitochondrial genome (right). Figure 7. View largeDownload slide Copy number variation (CNV) between ScY033T and ScY01 in nuclear and mitochondrial genomes. (A) A log2 ratio plot for CNV in mitochondrial genome generated by CNV-seq. The red and blue coloured plots represent amplification-type and deletion-type CNVs, respectively. The red and blue colour gradients represent the P-value calculated on each of the ratios where an increase in brightness shows a decrease in P-value. (B) Log2 ratio distribution for CNVs on nuclear genome (left) and mitochondrial genome (right). In our previous study (Zhang et al.2015), ethanol tolerance of a laboratory strain BY4741 was rapidly improved by coupling TAME-accelerated genome evolution with the screening method of using ethanol as selective stress. However, when applying the same strategy to the industrial strain ScY01, we did not obtain any mutations with enhanced ethanol tolerance (data not shown). These differential results implicated that the consequence of TALENs-mediated genome editing in the industrial strain might be different from that in the laboratory strain. Furthermore, genome sequencing of the ethanol-tolerant laboratory strain showed that 7 sites (10.6%) of a total of 66 potential TALENs modification sites contained the TALENs-induced InDels (Zhang et al.2015). Unexpectedly, in this study, none of potential predesigned TALENs modification sites were found to have any genomic variants in sequenced genomes of ScY001T and ScY033T. Thus, the improved stress tolerances of TAME-treated cells might be due to non-target genomic modifications (Zhang et al.2015) and indirect impacts of genome editing via TALENs rather than introducing genomic variants at potential predesigned target sites. Compared with S. cerevisiae laboratory strains, the genome editing via CRISPR/Cas9 in an industrial strain is relatively difficult to be established because of low transformation and genome editing efficiencies (Stovicek, Borodina and Forster 2015). Similar difficulties might exist in genome editing via TALENs in an industrial strain. Furthermore, it has been reported that oligonucleotide-directed gene editing activates damage response pathway and replication fork stress in mammalian cells (Bonner et al.2012). Although there were no genomic variants introduced into industrial strains by genome editing via TALENs, DSBs generated by TALENs might induce DNA replication stress thereby triggering stress responses of yeast cells. Alteration of cell wall structure and cell membrane integration in stress-tolerant strains When cells are exposed to various environmental stresses, cell wall and membrane act as the first barrier against external stresses. Some recent demonstrated that cell wall remodelling and alteration of membrane composition and structure seem to be the primary mechanisms required for protection against cell damage in stress-tolerant strains (Caspeta et al.2014; Wallace-Salinas et al.2015; Kitichantaropas et al.2016). Furthermore, genomic variants of stress-tolerant strains in this study converged on genes related to cell wall, cell periphery and cell membrane (Fig. 5, Supplementary file 5). Thus, we determined and compared cell wall structure and cell membrane integration of stress-tolerant strains ScY001T and ScY033T with the parent strain ScY01 at hyperosmotic and thermal stress conditions using electron microscope and flow cytometry. Based on the inspection by SEM, the stress-tolerant strains, especially ScY033T, had more rough surfaces than ScY01 at hyperosmotic stress condition (Fig. 8A), and TEM analysis further recognised fimbriate cell surfaces of ScY001T and ScY033T in comparison with ScY01 (Fig. S4, Supporting Information). On the other hand, at thermal stress condition, ScY001T and ScY033T showed slightly more surfaces than ScY01 according to SEM analysis (Fig. 8A), and their relatively short hairy surfaces were observed by using TEM analysis (Fig. S4, Supporting Information). These results suggested that cell walls of the stress-tolerant strains ScY001T and ScY033T might be remodeled in comparison with the parent strain ScY01, leading to a more robust wall to protect cells against hyperosmotic or thermal stresses. Furthermore, genomic variants related to cell wall might be responsible for cell wall remodelling of the stress-tolerant strains. For instance, compared with ScY01, both ScY001T and ScY033T had non-synonymous variants related to cellular surface properties such as adhesion (AGA1, FIG2), and biofilm and flocculation (FLO1, FLO5, FLO9) (Supplementary file 5), which were also observed to be mutant in previously reported thermotolerant strain (Wallace-Salinas et al.2015). Besides, SED1 encoding major stress-induced structural GPI-cell wall glycoprotein also showed non-synonymous mutations in ScY001T and ScY033T (Supplementary file 5). Interestingly, 16 and 14 of 24 PAU genes encoding structural constituent of cell wall, which might be differentially induced and possess specific roles for the adaptation of S. cerevisiae to certain environmental stresses (Luo and van Vuuren 2009), were found to have intergenic mutations in ScY001T and ScY033T, respectively (Supplementary file 5). This observation further confirmed that cell wall remodelling of ScY001T and ScY033T might be due to these genomic variants. In addition, the same stress-tolerant strain showed more rough cell surface at hyperosmotic stress condition than at thermal stress condition, indicating more activated cell wall remodelling at hyperosmotic stress condition. Figure 8. View largeDownload slide Scanning electronic microscope images (A) and percentage of PI-stained cells (B) of the stress-tolerant (ScY001T and ScY033T) and parent (ScY01) strains at different conditions. Hyperosmotic stress condition was performed using 400 g L−1 glucose at 40°C, and cells were cultured for 48 h. Thermal stress condition was conducted at 42°C using 200 g L−1 glucose, and cells were cultured for 36 h. As controls, cells were correspondingly cultured at normal condition using 200 g L−1 glucose at 30°C for 48 or 36 h. The percentage of PI-stained cells was analysed and calculated by flow cytometer. Data represent the mean and standard error of duplicate cultures at each condition. Figure 8. View largeDownload slide Scanning electronic microscope images (A) and percentage of PI-stained cells (B) of the stress-tolerant (ScY001T and ScY033T) and parent (ScY01) strains at different conditions. Hyperosmotic stress condition was performed using 400 g L−1 glucose at 40°C, and cells were cultured for 48 h. Thermal stress condition was conducted at 42°C using 200 g L−1 glucose, and cells were cultured for 36 h. As controls, cells were correspondingly cultured at normal condition using 200 g L−1 glucose at 30°C for 48 or 36 h. The percentage of PI-stained cells was analysed and calculated by flow cytometer. Data represent the mean and standard error of duplicate cultures at each condition. Besides cell wall remodelling, cell membrane integration might be also modified due to genomic variants in ScY001T and ScY033T (Supplementary file 5). Flow cytometric monitoring of PI uptake has been well established to inspect cell membrane integration (Davey and Hexley 2011). Under control conditions without stress, percentages of PI-stained cells were lower than 5% for both the stress-tolerant strains ScY001T and ScY033T and the parent strain ScY01 (Fig. 8B). By contrast, percentage of PI-stained cells of ScY01 increased to 16.6 ± 0.5% at hyperosmotic stress condition and 19.6 ± 0.2% at thermal stress condition, respectively. Compared with ScY01, ScY001T showed slightly lower percentage of PI-stained cells at hyperosmotic stress condition and apparently lower percentage (10.9 ± 0.1%) of PI-stained cells at thermal stress condition. On the other hand, ScY033T showed significantly lower percentage (6.0 ± 0.1%) of PI-stained cells at hyperosmotic stress condition but similar percentage of PI-stained cells at thermal stress condition in contrast to ScY01. These results suggested that the stress-tolerant strains might develop more robust cell membranes to different extent. Meanwhile, many genes (Supplementary file 5), which encode membrane-associated proteins, membrane transporters, integral component of membrane, etc., were found to have non-synonymous variants in the stress-tolerant strains ScY001T and ScY033T, which might be beneficial to cell membrane integration at stress conditions. Overall, the stress-tolerant strains ScY001T and ScY033T seemed to remodel cell wall and alter cell membrane integration to different extent by developing related genomic variants, thus protecting cells from adverse stresses. Further efforts, however, are required to clarify the precise molecular mechanisms underlying stress tolerance of these strains. In the future, further improvement of transformation and genome editing efficiencies of S. cerevisiae industrial strains would be beneficial to the application of the TAME toolbox in the breeding of multiple stress-tolerant strains. TAME-introduced genomic mutations were generated and accumulated when the TALENs-induced DSBs were being fixed through the NHEJ (nonhomologous end joining) pathway of DSB repair. In yeast, two major competing pathways including homologous recombination (HR) and NHEJ are involved in DSB repair (Aylon and Kupiec 2004), but the NHEJ efficiency is pretty low (Li et al.2011). It was recently reviewed that inhibiting critical NHEJ proteins, such as Ligase IV, a serine/threonine protein kinase DNA-PK responsible for initiating the NHEJ pathway and the heterodimeric Ku complex for binding DSB ends, could enhance HR-mediated genome editing (Pawelczak et al.2018). On the contrary, inhibition of HR activity would increase the NHEJ efficiency, thereby providing a promising approach to improve the TALENs-mediated genome editing efficiency and increase the rate of TALENs-introduced genomic mutations in industrial S. cerevisiae strains. In addition, population genomics studies reported that phenotypic variation of S. cerevisiae isolates correlates with genomic variation (Liti et al.2009). Therefore, comparative genome analysis between S. cerevisiae industrial and lab strains would identify the genomic variants that might determine the differential consequence of TAME treatment in these strains, thus providing clues to further improve the TAME efficiency in industrial strains by modulating these genomic variants. SUPPLEMENTARY DATA Supplementary data are available at FEMSYR online. Acknowledgements We thank Dr Guoqiang Zhang for constructing the TALENs-expressing plasmids with drug-resistant marker genes. We thank Lixian Wang and Huanhuan Zhai (Technical Support Center, Tianjin Institute of Industrial Biotechnology, Chinese Academy of Sciences) for technical assistance in flow cytometry and electron microscope, respectively. FUNDING This work was supported by the National Science Foundation of China (31470214 and 31700077), the National Science Foundation of Tianjin (16JCYBJC43100) and the Science and Technology Support Program of Tianjin, China (15PTCYSY00020), and funding from the Science and Technology Foundation for Selected Overseas Chinese Scholar of Tianjin to Yuping Lin. Conflict of interest. None declared. REFERENCES Abdel-Banat BM , Hoshida H , Ano A et al. High-temperature fermentation: how can processes for ethanol production at high temperatures become superior to the traditional process using mesophilic yeast? Appl Microbiol Biot 2010 ; 85 : 861 – 7 . Google Scholar CrossRef Search ADS Alexander WG . A history of genome editing in Saccharomyces cerevisiae . Yeast 2018 ; 35 : 1 – 6 . Google Scholar CrossRef Search ADS Aylon Y , Kupiec M . DSB repair: the yeast paradigm . DNA Repair (Amst) 2004 ; 3 : 797 – 815 . Google Scholar CrossRef Search ADS PubMed Bonner M , Strouse B , Applegate M et al. DNA damage response pathway and replication fork stress during oligonucleotide directed gene editing . Mol Ther- Nucl Acids 2012 ; 1 : e18 . Google Scholar CrossRef Search ADS Caspeta L , Chen Y , Ghiaci P et al. Altered sterol composition renders yeast thermotolerant . Science 2014 ; 346 : 75 – 8 . Google Scholar CrossRef Search ADS PubMed Caspeta L , Nielsen J . Thermotolerant yeast strains adapted by laboratory evolution show trade-off at ancestral temperatures and preadaptation to other stresses . mBio 2015 ; 6 : e00431 . Google Scholar CrossRef Search ADS PubMed Caspeta L , Chen Y , Nielsen J . Thermotolerant yeasts selected by adaptive evolution express heat stress response at 30°C . Sci Rep 2016 ; 6 : 27003 . Google Scholar CrossRef Search ADS PubMed Clevenger J , Chavarro C , Pearl SA et al. Single nucleotide polymorphism identification in polyploids: a review, example, and recommendations . Mol Plant 2015 ; 8 : 831 – 46 . Google Scholar CrossRef Search ADS PubMed Connelly CF , Skelly DA , Dunham MJ et al. Population genomics and transcriptional consequences of regulatory motif variation in globally diverse Saccharomyces cerevisiae strains . Mol Biol Evol 2013 ; 30 : 1605 – 13 . Google Scholar CrossRef Search ADS PubMed Davey HM , Hexley P . Red but not dead? Membranes of stressed Saccharomyces cerevisiae are permeable to propidium iodide . Environ Microbiol 2011 ; 13 : 163 – 71 . Google Scholar CrossRef Search ADS PubMed de Nadal E , Ammerer G , Posas F . Controlling gene expression in response to stress . Nat Rev Genet 2011 ; 12 : 833 – 45 . Google Scholar CrossRef Search ADS PubMed Deparis Q , Claes A , Foulquie-Moreno MR et al. Engineering tolerance to industrially relevant stress factors in yeast cell factories . FEMS Yeast Res 2017 ; 17 : DOI: 10.1093/femsyr/fox036 . DePristo MA , Banks E , Poplin R et al. A framework for variation discovery and genotyping using next-generation DNA sequencing data . Nat Genet 2011 ; 43 : 491 – 8 . Google Scholar CrossRef Search ADS PubMed Deutschbauer AM , Jaramillo DF , Proctor M et al. Mechanisms of haploinsufficiency revealed by genome-wide profiling in yeast . Genetics 2005 ; 169 : 1915 – 25 . Google Scholar CrossRef Search ADS PubMed Dragosits M , Mattanovich D . Adaptive laboratory evolution—principles and applications for biotechnology . Microb Cell Fact 2013 ; 12 : 64 . Google Scholar CrossRef Search ADS PubMed Gibney PA , Lu C , Caudy AA et al. Yeast metabolic and signaling genes are required for heat-shock survival and have little overlap with the heat-induced genes . Proc Natl Acad Sci USA 2013 ; 110 : E4393 – 402 . Google Scholar CrossRef Search ADS PubMed Guan B , Lei J , Su S et al. Absence of Yps7p, a putative glycosylphosphatidylinositol-linked aspartyl protease in Pichia pastoris, results in aberrant cell wall composition and increased osmotic stress resistance . FEMS Yeast Res 2012 ; 12 : 969 – 79 . Google Scholar CrossRef Search ADS PubMed Guldener U , Heck S , Fiedler T et al. A new efficient gene disruption cassette for repeated use in budding yeast . Nucleic Acids Res 1996 ; 24 : 2519 – 24 . Google Scholar CrossRef Search ADS PubMed Horvath I , Multhoff G , Sonnleitner A et al. Membrane-associated stress proteins: more than simply chaperones . BBA-Biomembranes 2008 ; 1778 : 1653 – 64 . Google Scholar CrossRef Search ADS PubMed Huang da W , Sherman BT , Lempicki RA . Systematic and integrative analysis of large gene lists using DAVID bioinformatics resources . Nat Protoc 2009 ; 4 : 44 – 57 . Google Scholar CrossRef Search ADS PubMed Jarolim S , Ayer A , Pillay B et al. Saccharomyces cerevisiae genes involved in survival of heat shock . G3 2013 ; 3 : 2321 – 33 . Google Scholar CrossRef Search ADS PubMed Jansen G , Buhring F , Hollenberg CP et al. Mutations in the SAM domain of STE50 differentially influence the MAPK-mediated pathways for mating, filamentous growth and osmotolerance in Saccharomyces cerevisiae . Mol Genet Genomics 2001 ; 265 : 102 – 17 . Google Scholar CrossRef Search ADS PubMed Kavscek M , Strazar M , Curk T et al. Yeast as a cell factory: current state and perspectives . Microb Cell Fact 2015 ; 14 : 94 . Google Scholar CrossRef Search ADS PubMed Khan A , Ahmad A , Akhtar F et al. Induction of oxidative stress as a possible mechanism of the antifungal action of three phenylpropanoids . FEMS Yeast Res 2011 ; 11 : 114 – 22 . Google Scholar CrossRef Search ADS PubMed Kitichantaropas Y , Boonchird C , Sugiyama M et al. Cellular mechanisms contributing to multiple stress tolerance in Saccharomyces cerevisiae strains with potential use in high-temperature ethanol fermentation . AMB Expr 2016 ; 6 : 107 . Google Scholar CrossRef Search ADS Kumari R , Pramanik K . Improvement of multiple stress tolerance in yeast strain by sequential mutagenesis for enhanced bioethanol production . J Biosci Bioeng 2012 ; 114 : 622 – 9 . Google Scholar CrossRef Search ADS PubMed Li H , Durbin R . Fast and accurate short read alignment with Burrows-Wheeler transform . Bioinformatics 2009 ; 25 : 1754 – 60 . Google Scholar CrossRef Search ADS PubMed Li T , Huang S , Zhao X et al. Modularly assembled designer TAL effector nucleases for targeted gene knockout and gene replacement in eukaryotes . Nucleic Acids Res 2011 ; 39 : 6315 – 25 . Google Scholar CrossRef Search ADS PubMed Liti G , Carter DM , Moses AM et al. Population genomics of domestic and wild yeasts . Nature 2009 ; 458 : 337 – 41 . Google Scholar CrossRef Search ADS PubMed Luo Z , van Vuuren HJ . Functional analyses of PAU genes in Saccharomyces cerevisiae . Microbiology 2009 ; 155 : 4036 – 49 . Google Scholar CrossRef Search ADS PubMed McKenna A , Hanna M , Banks E et al. The genome analysis toolkit: a MapReduce framework for analyzing next-generation DNA sequencing data . Genome Res 2010 ; 20 : 1297 – 303 . Google Scholar CrossRef Search ADS PubMed Miller JC , Tan S , Qiao G et al. A TALE nuclease architecture for efficient genome editing . Nat Biotechnol 2011 ; 29 : 143 – 8 . Google Scholar CrossRef Search ADS PubMed Niu L , Nomura K , Iwahashi H et al. Petit -High Pressure Carbon Dioxide stress increases synthesis of S -Adenosylmethionine and phosphatidylcholine in yeast Saccharomyces cerevisiae . Biophys Chem 2017 ; 231 : 79 – 86 . Google Scholar CrossRef Search ADS PubMed Pawelczak KS , Gavande NS , VanderVere-Carozza PS et al. Modulating DNA repair pathways to improve precision genome engineering . ACS Chem Biol 2018 ; 13 : 389 – 96 . Google Scholar CrossRef Search ADS PubMed Postmus J , Tuzun I , Bekker M et al. Dynamic regulation of mitochondrial respiratory chain efficiency in Saccharomyces cerevisiae . Microbiology 2011 ; 157 : 3500 – 11 . Google Scholar CrossRef Search ADS PubMed Sadowski I , Lourenco P , Parent J . Dominant marker vectors for selecting yeast mating products . Yeast 2008 ; 25 : 595 – 9 . Google Scholar CrossRef Search ADS PubMed Satomura A , Miura N , Kuroda K et al. Reconstruction of thermotolerant yeast by one-point mutation identified through whole-genome analyses of adaptively-evolved strains . Sci Rep 2016 ; 6 : 23157 . Google Scholar CrossRef Search ADS PubMed Shi DJ , Wang CL , Wang KM . Genome shuffling to improve thermotolerance, ethanol tolerance and ethanol productivity of Saccharomyces cerevisiae . J Ind Microb Biotech 2009 ; 36 : 139 – 47 . Google Scholar CrossRef Search ADS Shui W , Xiong Y , Xiao W et al. Understanding the mechanism of thermotolerance distinct from heat shock response through proteomic analysis of industrial strains of Saccharomyces cerevisiae . Mol Cell Proteomics 2015 ; 14 : 1885 – 97 . Google Scholar CrossRef Search ADS PubMed Snoek T , Picca Nicolino M , Van den Bremt S et al. Large-scale robot-assisted genome shuffling yields industrial Saccharomyces cerevisiae yeasts with increased ethanol tolerance . Biotechnol Biofuels 2015 ; 8 : 32 . Google Scholar CrossRef Search ADS PubMed Stovicek V , Borodina I , Forster J . CRISPR–Cas system enables fast and simple genome editing of industrial Saccharomyces cerevisiae strains . Metab Eng Commun 2015 ; 2 : 13 – 22 . Google Scholar CrossRef Search ADS Tsai SQ , Wyvekens N , Khayter C et al. Dimeric CRISPR RNA-guided FokI nucleases for highly specific genome editing . Nat Biotechnol 2014 ; 32 : 569 – 76 . Google Scholar CrossRef Search ADS PubMed Verghese J , Abrams J , Wang Y et al. Biology of the heat shock response and protein chaperones: budding yeast (Saccharomyces cerevisiae) as a model system . Microbiol Mol Biol R 2012 ; 76 : 115 – 58 . Google Scholar CrossRef Search ADS Voordeckers K , Kominek J , Das A et al. Adaptation to high ethanol reveals complex evolutionary pathways . PLoS Genet 2015 ; 11 : e1005635 . Google Scholar CrossRef Search ADS PubMed Wallace-Salinas V , Brink DP , Ahren D et al. Cell periphery-related proteins as major genomic targets behind the adaptive evolution of an industrial Saccharomyces cerevisiae strain to combined heat and hydrolysate stress . BMC Genomics 2015 ; 16 : 514 . Google Scholar CrossRef Search ADS PubMed Wang K , Li M , Hakonarson H . ANNOVAR: functional annotation of genetic variants from high-throughput sequencing data . Nucleic Acids Res 2010 ; 38 : e164- . Google Scholar CrossRef Search ADS PubMed Whelan SP , Zuckerbraun BS . Mitochondrial signaling: forwards, backwards, and in between . Oxid Med Cell Longev 2013 ; 2013 : 1 – 10 . Google Scholar CrossRef Search ADS Xie C , Tammi MT . CNV-seq, a new method to detect copy number variation using high-throughput sequencing . BMC Bioinformatics 2009 ; 10 : 80 . Google Scholar CrossRef Search ADS PubMed Zhang G , Lin Y , Qi X et al. TALENs-assisted multiplex editing for accelerated genome evolution to improve yeast phenotypes . ACS Synth Biol 2015 ; 4 : 1101 – 11 . Google Scholar CrossRef Search ADS PubMed © FEMS 2018. This article is published and distributed under the terms of the Oxford University Press, Standard Journals Publication Model (https://academic.oup.com/journals/pages/about_us/legal/notices)

Journal

FEMS Yeast ResearchOxford University Press

Published: Apr 17, 2018

There are no references for this article.

You’re reading a free preview. Subscribe to read the entire article.


DeepDyve is your
personal research library

It’s your single place to instantly
discover and read the research
that matters to you.

Enjoy affordable access to
over 18 million articles from more than
15,000 peer-reviewed journals.

All for just $49/month

Explore the DeepDyve Library

Search

Query the DeepDyve database, plus search all of PubMed and Google Scholar seamlessly

Organize

Save any article or search result from DeepDyve, PubMed, and Google Scholar... all in one place.

Access

Get unlimited, online access to over 18 million full-text articles from more than 15,000 scientific journals.

Your journals are on DeepDyve

Read from thousands of the leading scholarly journals from SpringerNature, Elsevier, Wiley-Blackwell, Oxford University Press and more.

All the latest content is available, no embargo periods.

See the journals in your area

DeepDyve

Freelancer

DeepDyve

Pro

Price

FREE

$49/month
$360/year

Save searches from
Google Scholar,
PubMed

Create lists to
organize your research

Export lists, citations

Read DeepDyve articles

Abstract access only

Unlimited access to over
18 million full-text articles

Print

20 pages / month

PDF Discount

20% off